How much longer can your corporate network compete?

Like most people I have been watching the “Cloud Services” develop and participated in some of the discussions surround the space. These are a collection of the best articles I have found that have shaped my thinking heavily.

Tearing down the walls that limit business

A series of articles on designing Open Networks – Jericho Forum

http://www.blackhat.com/presentations/bh-usa-04/bh-us-04-simmonds.pdf

http://www.opengroup.org/jericho/

TechEd Australia ‘08 Locknote

This is the future of IT over the next 10 years as predicted by Microsoft’s chief navel gazer. I gotta say, I think he’s right.

http://technet.microsoft.com/en-gb/dd819085.aspx
You want the session by Miha Kralj
How IT will change over the next 10 years and why you should care

http://www.craigbailey.net/live/post/2008/09/07/TECHED-Lock-note-ndash3b-Predicting-the-next-10-years-in-IT.aspx

http://www.crn.com.au/News/84240,opinion-navel-gazing.aspx

Cheaper Servers

Why Commodity Data Centres are cheaper than your server room, directly from the people building them.

http://loosebolts.wordpress.com

http://loosebolts.wordpress.com/2008/12/02/our-vision-for-generation-4-modular-data-centers-one-way-of-getting-it-just-right/

Article
http://unthrottled.spaces.live.com/blog/cns!3B07BABB3D3318AA!638.entry?wa=wsignin1.0&sa=860819746

Rebuttal
http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9080738&pageNumber=1

Response to Rebuttal
http://loosebolts.wordpress.com/2008/10/20/out-of-the-box-paradox-manifested-aka-chicago-area-data-center-begins-its-journey/

StorageMojo’s Take

http://storagemojo.com/2009/02/18/clouds-over-berkeley-the-radlab-reviews-cloud-computing-pt-1/ (read the overview and original article)

http://storagemojo.com/2009/02/21/clouds-over-berkeley-the-radlab-reviews-cloud-computing-pt-2/ (read the overview and original article)

http://storagemojo.com/2009/03/04/the-amazon-keynote-at-fast-09/

http://storagemojo.com/2009/03/04/belts-suspenders-and-scale/

Physics and a great laptop cooler

I recently bought a fanless laptop mat for use with my Dell e4300. When sitting in the lounge with the laptop on my lap it gets a bit warmish for my comfort. The fact that my clothes block it’s cooling vents certainly doesn’t help.

image

These new thingies work on a very cool principal of thermodynamics. Normally when you pump heat into a material it’s temperature increases. If however that material is at a point where it’s phase changes (solid – liquid or liquid – gas), then until the phase change is complete, all the energy you put in will not increase the temperature of the material. This is known as the “Enthalpy of Fusion” and the “Enthalpy of Vaporization”.

The really nifty thing is that with the right material, it can take huge amounts of energy to change it’s state compared to simply changing it’s temperature.

Quoting from Wikipedia

The high heat storage capacity in the phase change from solid to liquid, and the advantageous phase change temperature of 32 degrees Celsius (90 degrees Fahrenheit) makes this material especially appropriate for storing low grade solar heat for later release in space heating applications. In some applications the material is incorporated into thermal tiles that are placed in an attic space while in other applications the salt is incorporated into cells surrounded by solar–heated water. The phase change allows a substantial reduction in the mass of the material required for effective heat storage (83 calories per gram stored across the phase change, versus one calorie per gram per degree Celsius using only water), with the further advantage of a consistency of temperature as long as sufficient material in the appropriate phase is available.

As such – it literally soaks up the heat without getting significantly warmer itself. There is a caveat though – once it’s adsorbed all the heat it can – it will start getting hotter along with the laptop. For this reason you can only use it for a certain number of hours before letting it cool down and “reset”. The manufacturer quotes 8 hours, however this would vary significantly depending on the power of your laptop, and the ambient temperature. Mine works fine for 3-4 hours. As my laptop only gets hot on one side, I can just turn it around to get twice as much use out of it.

They also have a dimpled surface to allow some air to the laptop, and mesh underneath to help keep some distance. They roll up, but aren’t super light. I wouldn’t bother travelling with mine as I don’t tend to sit with my laptop on my lap very often when away. I could see it being useful watching movies in bed where the sheets stop the ventilation.

The only issue I see is they do limit airflow somewhat, possibly making the laptop run hotter.

You can buy them online at the link below.

http://www.crazysales.com.au/revolutionary-laptop-notebook-fan-less-thermal-cooling-mat-black_p7279.html

Recommended – 4/5 as I love thermodynamics and this is cool!

Greenpower – what exactly am I paying for?

It may appear obvious what you receive for your Greenpower dollar, but I quickly found it wasn’t. I buy Greenpower and recommend it, but I now have a much better understanding of what I am getting for my money.

I buy my Greenpower from Origin Energy as 100% Wind in Queensland Australia. I do this so that the electricity I use comes from a renewable carbon neutral source. The thought that occurs to me is to ask where this “green” power is sourced from and  how that relates to the photovoltaic (PV) panels on my roof.

The answer it turns out is a little confusing, but consistent.

Power producers generate power and sell that power to the grid retailers at electricity market prices. It doesn’t matter about the source, the market price is the same. If you have a home PV system, you are just a power producer, and get to sell into the grid at some agreed rate.

Registered GreenPower producers also get to sell REC’s into the REC market. These REC’s represent every MW-Hr of power they sell. When you pay extra for Greenpower, you are buying these RECS. This link means that for every KW-Hr you buy, a KW-Hr of GreenPower is generated by a registered generator. The requirements for registered generators are fairly strict to ensure they are carbon neutral sources.

Under the Greenpower Program Rules REC’s can be obtained from a number of sources. These are large registered generators, and also from pooled REC’s from home PV systems that are grid connected. If you sell (wreck) your REC’s, then they may be included in the Greenpower scheme.

Home PV systems are treated a little differently to large scale generators, where the REC’s are issued for the "deemed" life of the system (15 year deeming period), rather than created yearly. This means that the Greenpower you are buying now may not be generated yet – obviously not ideal, but these systems make up less than 5% of the total REC’s sourced, so it’s not a critical issue.

Solar Hot Water system REC’s are not eligible to be used in the Greenpower scheme, so you are actually paying for real power produced, not power saved.

The electricity your home PV system exports out to the grid is just electricity. It’s not classed as Greenpower, despite the premium you are paid for it. (The premium is paid by Energex, with Origin only paying $0.07 of the $0.50 paid). The Greenpower relates to any REC’s you sell for the deemed life of the system, not to the actual amount of power going out the meter in that period.

If you sell your REC’s, this means that you have sold all the carbon savings from your system to someone else. Any power you use from the system on your roof is no longer “green” (yes – that’s a little confusing). When you pay for Greenpower, you are buying some of the REC’s back. If anything, this is where the “hole” in the scheme lies. When you buy Greenpower (REC’s) from most generators, every one of the units is exported to the grid. If the REC’s you are buying are sourced from home PV systems, then that power may never actually be exported onto the grid. I think it’s a flaw, but it’s a small enough percentage to not bother me too much.

How all this fits into the MRET (Mandatory Renewable Energy Target) and the Australian Governments new Solar Credits Scheme is yet to be legislated.

Simple hey, clear as mud. No wonder people get confused with this stuff. It is a consistent system though, and does ensure that when I leave the lights on, it’s coming from a carbon neutral source and audited to prove it.

What I pay for Electricity with Origin

Power 100% Wind Buy Green Total
15.455 4.999 20.455

What Origin pays me for Electricity

Sell State Govt Sell Total
6.1 44.0 50.1

I was given some more info by the guys at www.localpower.net.au about why Energex who runs the grid benefits from the distributed power solution.

Power Costs
43% energy (generation) cost
47% network (grid) cost
10% retail cost

http://www.qca.org.au/files/ER-NEP910-QCA-DraftDecBRCI-Report-1208.PDF

Toyota Landcruiser 100 Series ‘98-‘04 Temp Gauge Mod

 

Centred

It’s fairly common knowledge that the OEM Toyota temperature gauge has a large “dead spot” in the centre of it’s range. This spot is deliberately engineered to reduce the apparent fluctuations and make the car appear to run at a constant temperature unless there is a significant problem. This works fine for most, but those of us that like to know what’s going on sooner rather than later, demand a little more detail. Many people fit an aftermarket gauge somewhere in the car, I figure, if the factory gauge is already there and can be made accurate, use it.

One of the clever guys over at ih8mud figured out the circuit and how to modify it in his 80 series. He deserves full credit for the original article and all the work behind it. There is also a version for the Toyota Surf and Hilux. 

The gauge circuit in my ‘98 HZJ105R was a little different to the earlier 80 Series, so I had to re-do the calibration to suit.
I also had the opportunity to see the inside of a post ‘04 update dash, and unfortunately, it’s quite different. Someone will need to do their own testing and research on that one.

I bench tested various setups and found the following as a simple description.

test rig     
Bench testing w/ Digital Thermometer

temp test rig
Heating the sender unit in vegetable oil

wiring
Mmm, wiring.

components
Resistors and Diode

We do not change the 15 Ohm resistor.

There are 2 components we replace, a resistor and a diode. The diode is what makes the gauge “non-linear”. Rather than explaining what they do in a wheatstone bridge, I’ll explain their effect on the gauge.

The gauge with no input actually sits in the middle of the scale. The 75 Ohm resistor we change “sets” what temperature the middle of the scale is to be. A lower value resistor sets it higher, a higher value resistor sets the scale lower. I found the following:

  • 100 Ohms = 90c
  • 120Ohms = 85c
  • 82 Ohms = 95c

centering test
Gauge w/ no input – centres on scale.

The small glass diode gives a non-linear (dead spot) in the needle’s range. We replace this diode with a resistor to make the gauge react “normally”. The value of this resistor determines the “range” of the gauge. A lower value resistor gives large movement for small temp changes, a high value resistor gives less movement. Using no resistor with a 90C centre means the gauge hits the red at 94.5C – a little too low.

centering - no damping
No Damping Resistor – large deflection

I found a value of 82 Ohms gives a good range with 115C touching the Red, 125C middle of the red and 65C touching the Cold. Properly mixed coolant boils at approximately 125C – 128C at 14PSI, and I’m not interested in below 65C, as the engine is not yet at operating temp.

This combination gives me the best combination of “”operating near the middle” and “enough movement to see what’s happening”. With the above detail you can adjust your own numbers if you wish.

The 100 Ohm resistor gets hotter as engine temps increase and will possibly exceed 1 watt. I recommend a 2 watt resistor. 5 Watt is very large and may not fit or be too heavy. The temp sender resistance decreases with heat, increasing current through the resistor.
The 82 Ohm resistor dissipates less than 1/4 watt, but I used a 1/2 watt to be safe.

 

  Original Modified
130 a130 b130
120 b120 b120
110 a110 b110
100 a100 b100
90 a90 b90
80 a80 b80
70 a70 b70
60 a60 b60
50 a50 b55

 

You will need the following

  • OEM ‘98-‘04 Landcruiser Dash
  • 2 watt, 100 Ohm resistor
  • 1/2 watt, 82 Ohm resistor. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Remove the gauge pod from car and disassemble. Be careful removing it from the car – there are 4 screws and 2 bolts. The bolts are captive and hold the plug connections, but need to be unwound a lot to release the plugs. P4040008
Remove the temperature / oil pressure gauge assembly. P4040009
Remove the 75 Ohm resistor and Diode. components
Replace the 75 Ohm resistor with a 100Ohm and the Diode with a 82 Ohm. P4040001

 

Detailed pics below (5C steps)

  Original Modified
130 a130 b130
125 a125 b125
120 b120 b120
115 b110 b115
110 a110 b110
105 a105 b105
100 a100 b100
95 a95 b95
90 a90 b90
85 a85 b85
80 a80 b80
75 a75 b75
70 a70 b70
65 a65 b65
60 a60 b60
55 a55 b55
50 a50

Tuning and Understanding your Toyota Viscous Fan Clutch

Landcruiser or Hilux overheating? Your factory fan clutch is probably under-filled and incorrectly set from factory. Fix this first and you may save a lot of time chasing issues.

(4 Runner or Tacoma in the USA)

P3280006 - Copy

The stock Toyota cooling system can sometimes be somewhat marginal. The suspicion for this falls on every component and modification in the system.

  • Radiator (Size / Efficiency)
  • Thermostat (Brand / Effectiveness)
  • Water Pump (Flow, Cavitation)
  • Radiator Cap (Quality, Pressure, Leaks, Recovery)
  • Coolant (Freezing / Boiling points, Specific heat, Anti-corrosion)
  • Hoses (Restriction)
  • Engine Type (Diesel / Turbo / Petrol)
  • Engine Load / Modifications (Diving style, load on vehicle, Mods)
  • Gearbox (Auto Cooling, Slipping)
  • Airflow (Obstructions / Restrictions In / Out, Forced / Natural)
  • Ambient Operating Environment (Temp, Altitude, Terrain)
  • Shrouds (Closeness to Fan, Leaks, gaps between radiators)
  • Fan (Size / Pitch / Airflow)
  • Fan Clutch (Lockup Temp / Stages / % Slip)
  • Temperature Gauge (Damping / Accuracy)
  • Bullbars / Winches / Lights / Antenna’s / Plates / Screens

Ask anyone and they’ll start listing random items from the list above that they have seen before or are suspicious of. It would appear that the issue is simply that the system is marginal in certain areas, and several small changes may be enough to tip it over the limit.

The end goal of a cooling system is to transfer heat to the surrounding air. All the other components are only there to allow this transfer to occur in some improved fashion. There are plenty of air cooled motors in existence that do not have these complexities, and they too may be subject to overheating.

It would appear that Mr Toyota VERY closely engineers his vehicles, with many parts sharing multiple purposes, and many many tradeoffs being made. This is good engineering, but it means that small changes may have many unintended impacts. Despite this, it appears the Landcruiser and Hilux are intended to be frequently modified. There are many attachment points, and the OEM design has many dealer supported aftermarket options that are not from the Toyota factory.

If all the basic checks have been performed on the cooling system – no leaks, nothing obviously blocked, quick warm up, infrequent overheating except under specific circumstances, then it is a fair bet that the overall system is simply marginal. In this case, a dramatic increase in specific areas may yield a significant benefit.

In my case the overheating was limited to situations with a pre-turbo EGT in excess of 550C. This equated to High Load or High Speed driving. Despite expectations, off-road steep terrain (sand excluded) does not yield high EGT’s. Mountain Ranges, Large Trailers, Roof Racks, High Speed or Deep Sand all would yield high EGT’s and therefore problems.

I have measured many temperature points around the engine bay, and spent some time listening to the engagement and disengagement of the fan. All this yielded much confusion rather than understanding.

I replaced most components, some twice. It was during this that I had time to closely examine and understand the Toyota Viscous Fan Clutch. Possibly more than any other component, this is the key item in the cooling system. It is this that creates the airflow, not vehicle forward speed. Without airflow, the radiator is not effective. My experience was very similar  in a Toyota Surf I had owned previously. It is common knowledge that additional Silicon Fluid will often improve these units. What is not common knowledge is:

  • Brand new OEM clutches appear to be under-filled
  • They can be adjusted where they engage
  • There are 4 separate engagement stages
  • Testing cannot be done one the bench. The device requires centrifugal force to operate.

Credit goes to Frank for his guide on how to split and refill the fan clutch. I am just explaining the operation in further detail.

It must be remembered that these types of fluid couplings always have some slip. They may slip by 98% (free spin) or 5% (coupled), but there is always slip. It is difficult to test the slip in any simple manner, and impossible to bench test. Therefore a fan that appears to be engaging and disengaging successfully, may in fact be slipping at 50%, significantly reducing maximum airflow. Worse, the slip will be only happen at high RPM and maximum load.

The key points are that there are 4 operating stages, and that there is not enough fluid to couple the system adequately.

This is why so many people report success with simply adding more fluid. Adding fluid means that when the system is operating with the valve fully open, the rings are full of silicon fluid, and not partly full. The only drive is through the fluid, so insufficient fluid will reduce maximum coupling ability. There was clearly not enough fluid in the unit to fill all the rings to the depth of the final valve.

The factory engagement points are also quite high. This reduces noise and fuel consumption, but also means maximum engagement doesn’t occur until the air temp is around 95C. Engine coolant temperature will always be higher than air temperature.

This was all tested with a Digital Thermometer and a water bath on the stove.
The water was heated and cooled and the valve set points noted as it moved.

P3230007

Temperature Set Points (all at 1/2 open)

Stage Original Temp Adjusted Temp
Closed 50 40
Stage 1 55 45
Stage 2 85 75
Stage 3 95 85

Pictures of operation:

Fan Clutch P3280006 - Copy
The 2 halves opened BlueFanClutchApart[1]
The “drive disc” spins freely in the housing except for the silicon fluid. P3230019
The “drive disc” and the “front half” share these closely spaced rings. It is these rings, and the silicon fluid in the gaps between them that couple the system together.The inner ring is taller than the others.

The oil is slowly thrown to the outside of the system by centrifugal force.

P3230004 P3230011
The fluid is rated at 10000 Cst – Centistokes – a measure of viscosity P3230009
The valve that controls where the fluid flows. It operates over 4 stages:0) Closed
A) Some oil to some rings
B) Some oil to all rings
C) Maximum oil to all of rings

This is why it seems to be more than engaged / disengaged.

P3230005 - Copy - Copy
The temperature sensing Bi-Metal spring on the front face that controls the valve. BlueFanClutchthermosideup
The reservoir behind the valve disc in the front half where the fluid is stored. When operating it is held here by centrifugal force, and pumped here by the slipping “drive disc” P3230012
The “vanes” on the edge of the drive disc in the rear half that pump the fluid forward to the outer channel for return to the reservoir. Some slip is required to allow the pumping to occur.The slots in the back of the disc pump the fluid from behind the disc to the edges, and then to the channel at the front.

The rear of the disc is not used for coupling.

P3230019
P3230011
The wedge shaped guides and small holes in the front half that collect the fluid from the outer channel and push it back into the reservoir. P3230002
Adding Fluid P3230008
Adjust the valve set point by loosening the 2 screws and rotating the disc.
The outer valve should be 1/2 open at about 45C for Australia. (US quotes 35C). Air temp will always be less than engine water temp.
P3230001
Getting the fluid level right is a little difficult and involves some guesswork.The minimum amount required is enough to fill the entire outer rim past the depth of the fins in both halves, this fully couples the unit.

The maximum amount is when the reservoir in the front is full and overflows through the central hole. Not so simple though, as full is controlled by centrifugal force, so when operating it fills the “outside” of the reservoir, not the bottom.

Luckily there is a fair tolerance between the two. Overfull will couple the fan all the time.

Mine took 1.5 tubes of fluid in addition to the factory fill to stay a few mm below the level of the valve disc.

P3230014

Study on why ABS is good on a Motorcycle

The link is below.

http://www.msf-usa.org/imsc/proceedings/a-Green-ComparisonofStoppingDistance.pdf

Most interesting. My next bike will have ABS methinks. I have no delusions about how good my riding is. If I can concentrate 5% more on traffic cause I don’t have to worry about braking, that’s a good thing.

bmw_r1200gs_abs_2007

My last significant off happened emergency braking into a roundabout. I thought it was clear, then a car with right of way came out from behind a blind fence. It was far too late to wash off that much speed, I swore I was going to have him hit me. As I braked and turned, the front wheel crossed the usual oily patch where cars drop oil on the centre of the wheel track. Presto – down she went, and I know how to cadence brake. I’m not good enough however to brake 100%, unlock for 0.5m and get fully back on again when trying 100% to avoid a car. ABS would have been fast enough. It wouldn’t have avoided the situation, but it may have been enough to keep the bike upright.

Rethink your Lid

Buying motorcycle helmets used to be easy. Plastic was crap, Fibreglass was good, Kevlar / Carbon was best. If it was a Shoei, Arai or Bell it was good, everything else was only OK.

BUT – interestingly enough, every single helmet you could buy had passed the crash test standard, so the better / worse was opinion, nothing more. In Australia this is an Australian Standard (similar to the US DOT standard, and many others). Really, it’s a “minimum”, everything on the market is better than the standard, it’s just a question of how much.

There was a defacto “better” standard – Snell. This was only issued to the “best” helmets, so if you cared, you bought a helmet that also had Snell certification.

Well now there is something better. The UK govt has setup a research and testing laboratory to test and rate motorcycle helmets like they crash test cars – with a star rating. You’ll find it at http://sharp.direct.gov.uk/ 

The testing is based on the new EU standard ECE 22.05 and discussed on webbikeworld.

I just bought a new Shark RSR2 to replace my aging Arai Qantum(1996). The new Shark is comfortable, quieter than the Arai on my head, and has a 5 star rating.

I prefer the new EU standard over Snell etc after reading this debate on MotorCyclistOnline and the testing they did. The base concept is that the Snell standard is very tough, and a tough helmet is not as soft and squishy. Soft and squishy is more likely to be useful to you in the accidents you are most likely to die from. The more serious accident that the Snell rated helmet is designed to protect your head for, will leave you dead from other injuries and with brain damage anyway.

Of course nothing will protect you from those big ones that just leave you dead.

Inergen is interesting

I noticed something very strange with a new fire suppression system. There were no valve controls on the system, of the four bottles, only one was controlled. I had to look further into this.

One of the sites I work on had installed an Intergen Fire Suppression system. The basic idea is that in the event of a fire, enough oxygen is displaced from the room, that a fire cannot be sustained, but humans will remain conscious.

When I looked at the system I was mighty confused. Four high pressure cylinders indicate a high pressure non-liquefied gas. Makes sense when you look at the mix – Nitrogen, Argon, CO2 – not likely to liquefy under pressure at room temperature. A large diameter medium pressure line led to the room with no valves on it. This line was open at the room end with a simple nozzle, no valves along the way.

There was a valve on each cylinder, but no controls to any but the first one.

PB060003

 

The first cylinder below has a both an electrical and manual release attached – but what about the other three bottles?

 PB060004

 

The control valve on each slave cylinder is the interesting bit. When pressure is sensed in the line, it opens the bottle, and remains open until the bottle is empty. Basically it’s a pneumatically operated slave. The pressure in the line is created by the first bottle’s discharge, triggered electrically, and the restriction of the line and nozzles.

 PB060006

And yes, they installed vents in the room to let out the extra air. When you add another 50% of gas to a room, the displaced gas has to go somewhere, or develop an unlikely amount of pressure. They would not be the first to forget the vent.

Combining Vista Media Centre & a Virtual Windows Home Server

The time came to decommission the Home Server once I realised how much power it was pulling. My power meter debacle had concealed the 24/7 150w consumption, chewing into my solar feed in tariff at 44c in the daytime and my green power rate at 21c at night. This was costing me about $400/yr in power bills – it had to go.

I have toyed with various options, but the most obvious was using the other machine that was on 24/7 – the Vista Media Centre.

I found some info on running Windows Home Server as a virtual guest, but nothing on the impacts to the Vista Media Centre host. My host was not particularly new, a 3yo PC with the following specs:

  • MB – Asus N4L-VM-DH
  • CPU – Core 2 Duo T2400 1.8Ghz
  • RAM – 2GB Kingston
  • HDD 7200RPM 500GB WD
  • Silent Heat Pipe Video Card
  • Antec Truepower2 Power Supply (pre 80+ standards)
  • 100MBit Ethernet to router

I figured it would be a stretch for this machine to run a VM as well, but it was worth the try. It is a fairly power efficient machine, the Core 2 Duo being a laptop CPU and noted for it’s efficiency. The current video card pumps out heat 24/7 and could do with improvement, I’m waiting for the new Intel Nehalem CPU range to arrive and come down in price, by which stage on board video should be suitable – more power savings.

I grabbed a couple of 1TB WD Green drive, as my experience with them inside WD MyBooks showed them to be very quiet and efficient.

I used VMWare Workstation 6.5 as I had it, but you could use VMWare Server – it’s free. You could also use Windows Virtual Server, but I felt (possibly incorrectly)  VMWare may have slightly lower overhead and better direct hardware and USB support. Virtual PC / Server have no USB support last I checked.

The Asus MB has 3 x SATA ports, and a PATA port. Two of the SATA ports were already in use with the DVDROM and host HDD, so I ran the first HDD on the spare SATA socket expecting great results. It was awful, woefully slow, rendering the machine unable to even record TV shows. Turns out he extra SATA slot is for RAID, and due to firmware / drivers, IRQ’s went through the roof consuming 60% CPU time. A two channel PCI SATA card fixed this problem.

With that nailed down, I installed Windows Home Server w/ PP1 onto an 80GB Virtual Disk on the 1TB HDD. The host OS and TV recording was all onto the original 500GB HDD.

Virtual machine config was easy. , Setup a Virtual Machine as Server 2003 w/ 512MB RAM, an 80 GB IDE Virtual HDD and mount the Home Server CD ISO. Sound and other unnecessary things were removed.  I elected not to fully allocate the 80GB HDD, although this may impact my performance later.

Home server will install on the above with no major configuration steps. All drivers are fine. Once the install is finished, install the VMWare tools, run Windows Update, and activate your Home Server. There are instructions here – but they use a virtual SCSI HDD, greatly complicating the install for no benefit I can discern. I would use a virtual IDE.

Once mine was installed I added the extra disks. Initially I tried the disks as Direct Physical access. They were setup with a partition, but no drive letter, as per the vmware help. The VM would not even boot and there was a disk access conflict. Nothing I did could resolve this problem, so I had to settle for a couple of 900GB virtual disks (not pre-allocated) on the 2 x 1TB HDD’s (930GB formatted capacity NTFS). I tried VMWare Server 2.0, but it doesn’t support direct physical disks anymore, and also broke my Remote Desktop to the host. VMWare Server 1.08 wasn’t compatible with my VM, so I gave up.

The benefit to using Physical Disks is threefold:

  • The entire disk is allocated to Home Server, maximising space
  • The disks can be unplugged and read anywhere
  • There is no possible conflict with access to the disk.

Unfortunately it was not to be – so virtual disks it is.

image_thumb8

The initial problem I had was WOEFUL disk performance and 100% CPU usage. Task Manager showed the “System Idle Process” to be hogging the CPU. Process Monitor showed it to be 60% used by IRQ’s. Google tracked that to the HDD not being in DMA mode, but rather PIO Mode 4 due to firmware / drivers / phase of the moon. As I couldn’t fix it easily, I bought and installed a 2 port PCI – SATA card. The CPU load is normal using this card as opposed to the on board port.I have since tried storing the VM 80GB “OS” drive on both the 1TB WD 5400RPM VM dedicated disk, and the faster 7200RPM disk shared with the media centre, and couldn’t detect a difference in performance either way. I have left it on the VM disk to optimise space for TV recordings.

Next issue was awful network performance between the Host and Guest. Guest to other network computers was fine – about 4-8MB/s, but Guest – Host was shocking – about 20Kb/s. Like all good technicians today I didn’t use my brain, but hit Google again. TCP Offload seemed to be a recurring theme here. The registry keys for XP didn’t fix it, but the advanced settings for the network adapter did. TCP Offload disabled on the host and now I get the same network performance anywhere. This is not a fault with VMWare, but does seem to be a common compatibility issue with many network adapters, my onboard nic being one of them.

image1_thumbimage_thumb1

The final hurdle was again performance related. The guest Home Server would be running, but very sluggish to respond to inputs if you had left it alone for a while. It’s like VMWare let it go to sleep, and took between 1-4 minutes to assign it resources again. I made two changes here. I disabled Memory Page Trimming in the VM admin interface and disabled Page File Sharing with the line sched.mem.pshare.enable=”FALSE” in the .vmx config file.
Both of these seemed to keep the Home Server in a much more responsive state when I wanted it. It still often needs two clicks to “connect” to the console – the first fails, but network shares and backup work perfectly. As the console is not something I regularly access, I’ll ignore this issue.

There are some other tricks that can help. I did the following on the host to reduce any possible performance hits:

  • I don’t run Anti-Virus on my Media Centre, so exclusions for that weren’t necessary. If you run AV it’s recommended to exclude the VM files to reduce overhead.
  • Snapshots on the HDD used by VMWare were disabled – won’t be needing them for backup.
  • Recycle Bin disabled – don’t need that either.
  • Added a shortcut to the VM to the startup folder so it auto-starts. (VMware took away this nice feature from VMWare workstation)

With the Home Server running and all updates installed (particularly PP1) it was time to install the connector to all the PC’s in the house, and configure backups. This is mostly straightforward. There is one trick – you MUST exclude the Virtual Machine folders from the backup when you install the connector on the Vista Media Centre Host. Otherwise it will try backing up itself to itself, decide it won’t fit, and have a heart attack.

Once everything was installed I copied over all the data using Robocopy. I found that the Windows Copy with that much data to the VM wasn’t particularly reliable, although that may have been due to not having all the above tuning done first. My sequence was a learning exercise.

Vista Media Centre has an option to add remote data to it’s library. I have added music, pictures, MPG4/DivX and DVD’s stored on the Home Server. You’ll need to use the DVD library reg hack to get the latter to work.

I haven’t tried using the Home Server to store recordings of live TV, but watching movies stored on it with either DivX or a ripped DVD works just fine. Network performance is as above.

I have also setup a 1TB Mybook using USB to be the backup drive. The USB seems to work fine and is setup to automatically connect. Backups to the external drive are the usual manual deal.

Now I have access to all my data, a large file store, regular backups, a quiet media centre and a single box that only chew’s 85w. Performance is acceptable, but not amazingly snappy. I think it’s an acceptable compromise. It doesn’t really take any longer than when the old server had to spin up it’s six HDD’s. I’m not sure if the HDD’s are spinning down under VMWare, there is a few more watts to save.

I’m looking into power saving inside VMWare next, but think I’m off to a pretty good start.

There are a few things I would like to do to improve the solution:

  • Get it going on VMWare Server (and not break RDP / Remote Desktop)
  • Resolve the physical disk access issue, I assume it’s to do with running under Vista, but have no evidence to back this up. I tried disabling everything that might conflict.
  • The performance is still not snappy on the console. The host CPU, RAM and Disk are not busy, so what’s making it sluggish. Network transfers and backup are fine, it’s just the console that’s sluggish.
  • How does WMWare interact with host power saving features?

Update Oct ’09
Now running VMWare Server.
No real difference. Still can’t use native disks.
I’ll go back to pyhsical once the new Pinetrail atom is out. Although it works, I’m sick of the ultra-long boot times.
I’l make the new WHS an Atom and the VM Host media centre an i3 w/ integrated Video.

Maximising your Credit Card Interest Free period for Dummies

image

Here I am, 30 something years old, and I only just now got around to really understanding how to optimize my credit card. (At least I hope I understand it now).

Now for those of you that don’t like reading, I’ll give you the short version.

TL:DR – You MUST pay off the “Closing Balance” as close to, but before the “Due Date” ON THAT STATEMENT – and you pay no interest.
Make no other payments.
Do NOT pay the “Outstanding Balance”.
That’s it, nothing else you need to know. Do that religiously, and you’ll maximize the interest free period.

I don’t think I’m that dumb. I use Mortgage Offset accounts and leveraging the interest free period on the c/c to maximize the amount of cash in the offset account.

I just never really sat down and thought about maximising the interest free period.
With interest calculated daily on most accounts, the more days you are using the bank’s money interest free, the more you save. BUT, and here is the gotcha – on most credit cards, if you are even one day late over the interest free period, you lose. They backdate the interest for the entire period. Just to make things worse, many (but not all), don’t give interest free on any further purchases until the closing balance has been paid. Note, this is NOT the “Minimum Payment“, OR the current “Outstanding Balance“. Ignore the minimum balance unless you are stretched that month – and it will cost you big time if you only pay the minimum, with all the interest free benefit wiped out.

The first thing you need to know. A Credit Card is not a continuously rolling credit account. If you think of it this way (as I used to) you’ll never get straight in your head.
A credit card is a continuous series of  One Month windows noted as “Statement Period” written on the statement.
Everything you spend on the account in this Statement Period gets totaled up at the end of the period and put on the statement as the “Closing Balance”. You have “x” days  to pay that Closing Balance. This date is calculated for you and put on the statement as the “Due Date”. In my case it’s 55 days from the START of the Statement Period. This is why the interest free period is described as “Up To 55 Days Interest Free”. If you make a purchase on the first day of the statement period, you’ll get 55 days (until the Due Date) to pay it off without being charged interest. If you make it on the last day of the statement period (generally 30 days later), you’ll only only get 25 days until the Due Date, before you cop interest.

The Due Date is always before the close of the next statement period, this ensures that if it’s not paid off in full, whatever remains gets tacked onto the next Closing Balance.

The next Statement Period is INDEPENDENT and has nothing to do with the previous period. It accrues on it’s own, and has it’s own closing balance and due date. The only relationship is if you don’t pay off the previous closing balance, then things roll over and impact the interest free deal.

I find that the statement information is not available on my bank’s Internet banking site. I have to look at the paper statement to find the Statement Period, Closing Balance and Due Date.

For me, the confusion came around the current “Outstanding Balance”. Most people are concerned about keeping this under the Credit Limit to ensure we don’t get charged fees for exceeding the limit. For the purposes of maximising interest free periods, the Outstanding Balance is nothing more than a distraction. It is a combination of last months Closing Balance, and whatever you have spent this month. It is NOT what has to be paid off. If you instead concentrate on the Closing Balance, you’ll be FAR better off. The only thing the “Outstanding Balance” matters for, is ensuring you don’t exceed the credit limit and incur fees. Indeed I have found that by only paying off the Closing Balance, my card now has a much higher average amount outstanding than before, whilst still not incurring interest. This means the cash that would have been sitting on the card, is now sitting in my offset account for longer. It also means I am considering upgrading my limit to support this higher Outstanding Balance. The monthly spend is still the same, in and out. It’s just that I’m only paying off what I have to, when I have to, not the entire balance.

A complication comes around when the Closing Balance is only partially paid off before the Due Date. When this happens, interest is charged on the outstanding amount. The calculation is done on a First In, First Out basis, and I’m glad it’s computer doing the numbers. You lose in a number of ways in this case. Any charges that were not paid off (earliest charges paid off first), accrue the full interest from their “Date Processed” – this is often after the “Transaction Date”. Many cards also don’t give the future charges interest free status until the previous outstanding  balance has been paid off. (There is some debate on this point).
A further complication to the above is cash advances. The simplest answer on those is DON’T. They are not interest free. I only use Cash Advance overseas as I generally find the cash exchange rate with fees and Interest to still be better than “Money Exchange” facilities. If you must do this, make your CC have a positive balance before you travel. This will basically make it a Debit account, with less overseas fees.
This is why they use computers to calculate this stuff out.

The worst case of not paying off the Closing Balance in full is the “Low Interest” transfer type deals some card issuers offer. Say you transfer a variety of other cards onto a 0% interest deal. If you then proceed to use this card for other charges, under a First In, First Out basis, you won’t be paying off any of the new charges until the entire transferred balance is paid off. This makes any interest free period other than the transfer one, essentially ineffective. Low Interest transfer deals are valuable, only if you use them for just that, transfer. As soon as you use them as a normal credit card whilst there is a transfer amount outstanding, the bank starts winning back that interest from you.

Finally, Credit Card interest rates tend to approximately average double that of Home Loan interest rates in Australia. If you are using the card to assist your home loan repayments by increasing the average balance of your offset account, and you don’t pay the Closing Balance by the Due Date, you get charged interest at a much higher rate on the card. This means that the “effective savings” of the interest are wiped out for around twice as long as if the charges were straight on your offset account without the Credit Card. Result – one missed payment takes 2-3 months to re-coup the savings, just to get back to where you were.

All of which brings us right back to the simple statement – You MUST pay off the Closing Balance as close to, but before the Due Date ON THAT STATEMENT – and you pay no interest. Make no other payments. That’s it, nothing else you need to know. Do that religiously, and you can’t do any better.

The only circumstance where you should make a payment other than the Closing Balance, is when you need to put more charges on the card that month than the Credit Limit can cover, or if you need to cash advance overseas. Exceeding your Credit Limit will result in fees being charged. In this case an additional early payment may be necessary to keep the card under the limit, but you are losing some of the interest free benefits. This may be offset by some of the insurance benefits and loyalty points programs some card issuers have. eg. Free Travel Insurance, Extended Warranties, Frequent Flyer Points etc.
Keep in mind, this early payment simply goes onto the statement as well, and will be accrued and calculated in. The rule of “Pay off the Closing Balance by the Due Date” still applies.

So what from here – go and grab a statement. Find the following key information from all the junk they print on there – it’s how you save money:

  1. Statement Period (Open – Close)
  2. Closing Balance
  3. Due Date
  4. Credit Limit

Ignore the distractions from all the other numbers.

You may want to consider moving your Statement Period so the Due Date falls somewhere just after you get your salary paid into your account. This makes it easier to manage payments – you get paid, and you pay off the card.

A note on Auto Payments. I have never done these – so can’t thoroughly comment. It would be worth checking what Auto Payment system your bank uses, and when it Auto Pays. If it pays the “Outstanding Balance” rather than the “Closing Balance”, or if it pays off the the amount much before the “Due Date” – you are losing out, and the bank makes more money.

Please keep in mind I am not a financial advisor, or financial professional. All the above is gleaned from what information I can get from my bank in Australia, your’s may be different. I have not been able to find many explanations on the Internet elsewhere for what should be common knowledge, otherwise I wouldn’t have bothered to write it.

The Australian banks are definitely not in the business of explaining how their credit products work in simple terms, or how to optimise them for the customers benefit. If I have got any of the above wrong, or it could be done better, please comment and I’ll do my best to review and incorporate it.

NAB has a reasonable explanation of Credit Card interest.
Compared to the unintelligible NAB banker speak version.

ANZ also has a unintelligible version.

Here are some additional tips I received when a professional banking friend reviewed the above:

  1. If you pay via bpay – this payment takes 2 -3 days to reach the payee – so always process 3 business days before the payment is due – this goes with all bills you are paying this way – especially important if you are paying bills like rates and rego where you get a discount for being early.
  2. Most banks offer a free service where they will set up a direct debit to the credit cards for either, entire balance (to avoid paying interest), minimum amount (to avoid late fees), set amount (to fit your budget). This payment will always be made on the due date – you just need to make sure there is always enough money in the offset / credit account to cover. This way you never have to worry about missing a payment – NB, it may take a month for the payment to be established so watch the first payment.
  3. If paying via cheque the payment is not deemed as being made until the cheque has been cleared – normally 3 business days, so again pay 3 days prior to being due.
  4. Low interest rate balance transfers – don’t use the card until the transferred balance is cleared – I recommend doing the balance transfer, then once cleared cancel the card. The less cards you have to manage the better for your budget.
  5. If you are going to clear the credit card in full every month – look for maximum interest free days / lower annual fees, not the lowest interest rate (cause you wont be paying it) – which normally has the higher annual fee.
  6. If you have a home loan – ask for the annual fee on the card to be waived, this can be done about 75% of the time.
  7. NEVER set up direct debits from your credit card, eg pay TV, internet etc. A “stop” can not be put on a credit card to stop these payments, if you need them stopped. Even if credit card is closed, the payment will reopen the credit card (for 6 months after the closure). If you use a normal account, the payments can be stopped.

WD External USB HDD’s do Spin Down

I’ve been trying to reduce the power of my Home Server and Media Centre. Since my Power Meter debacle, I am now re-testing all the equipment and getting some rude shocks.

One of the positives out of this is that my 1TB Western Digital  My Book Essential and 300GB Western Digital My Passport Essential both spin down and save power. On USB, this is a nice feature, as many of the generic external cages don’t spin the drive down.

The WD’s spin down on XP, Vista and Windows Home Server which is based on Server 2003. The timeout appears to be independent of the OS settings.

They support several power modes, and my current meter is not accurate enough to report which one they are in.

The WD 3/5″ Black is the 7200RPM desktop drive.
The WD 3.5″ Green is the 5400RPM drive generally in the MyBook
The WD 2.5″ Blue is the 5400RPM drive generally in the MyPassport

Model C’pcty Operate Idle Stdby Sleep
WD 3.5” Black 1TB 8.4 7.8 1 1
WD 3.5” Green 1TB 5.4 2.8 0.4 0.4
WD 2.5” Blue 500GB 2.5 0.85 0.25 0.1

The other thing I noted was that my new power meter (which is not rated as accurate below 10w) indicated that whilst sleeping, the 3.5″ and it’s power adapter were drawing less than 1w. The power adapter was only just warm, much better than the usual little heat wasters. Well done WD for killing the parasitic standby power.

Overall, they get my vote as low power green storage.

The Green 3.5″ and Blue 2.5″ drives are also very quiet – both operating and seek. The 2.5″ is quieter, but with half the capacity, may not stack up overall.

And you can always decode the Morse Code on the outside!

Lies, Damn Lies and Cheap Power Meters

PA050027I have had and used an Electus power meter for about 12mths since they became available in Australia. I figured it was just like the American Kill-a-watt. Diligently collecting and comparing devices to see what they used and how much I was wasting with standby power. Dan did an early review on DansData.

ATA reported some accuracy concerns early on, and stopped selling them. The indication was this was only minor, and at low power settings.

I recently purchased the newer Solarinverters (SI) replacement from ATA and found some VERY large discrepancies between what I had seen before. Unsure as to which one to trust, I fired up the clamp meter to validate the results. The results shocked me.

I then confirmed the results against two other Electus meters of the same model.

The Electus meters are TOTAL AND UTTER JUNK. The numbers that come out have NOTHING to do with reality. The behaviour does not even appear to be linearly consistent.

Device Electus SI PF (SI) Fluke

Ent System

143w

317w

0.80

1.6

Washing Machine

17w

3w

0.16

n/a

PC & Acc

85w

199w

0.71

1.1A

Kettle

2550w

2340w

1.00

9.8A

11W CF Bulb

9W

14W

0.68

n/a

Water Pump 449w 472w 0.95 2.0A

Disappointingly Jaycar is still selling them.

The below pictured numbers are all with the SAME LOAD – a computer, UPS, Router, Printer etc. All numbers were repeated three times and fluctuations were minimal at this time.

PA050021
Fluke meter validation Rig

PA050023
1.1 Amps on Clamp Meter

PA050024
0.37A on Electus Meter – kinda disagrees with the 1.1 on the clamp meter. This is where the majority of the error seems to lie.

PA050026
1.16A on SI Meter

PA050027
The voltages read the same

PA050025
85W – 0.92 PF from Electus – again, the electus can’t seem to cope with PF giving readings that are often clearly incorrect. This will significantly affect the wattage calculation.

PA050028
0.71 PF from SI – this is about what I would expect.

PA050030
199W from SI

PA050029 
278VA from SI – matches the Fluke

PA050032
Units from rear

PA050033
Units from front

Now my clamp meter hasn’t been calibrated, but it is a Fluke. It’s only got one decimal place of precision. Using my other Fluke as an inline ammeter with mains voltages was too risky (exposed terminals) for me, so I didn’t do it, depsite the extra decimal place of precision.
I don’t have an independent Power Factor meter. I could be telling lies, this is the internet. The above numbers aren’t a small % error. They are out by over 100%. This is not a minor calibration issue, this is a device that clearly misleads Australian consumers.

I’ll stick with the new one and redo the testing round my house.

The hurdles of setting up Vista Media Centre

OK – it came time to rebuild the media centre.

Here are ALL the steps I went through to getting the software install right. A Vista Media Center Build Document.

The changing of hardware, testing codecs, utilites, guides and apps had led to some long running config and stability issues that I couldn’t resolve. Application errors, crashes, codecs, screen sizes, resolutions and audio were all problematic.

After round one a while ago, I had managed to stabilise and expand the system somewhat. The stable hardware config now is

I use the VGA cable to drive the plasma – other outputs have proven problematic in the past and I can’t be bothered trying again.

The Amp is driven off the Digital Out on the M/B – a coax copper digital connection.

This current hardware setup is MUCH more stable than the previous top of the line junk I tried.

  1. Connect USB KB & Mouse (needed)
    BIOS – Change boot order – CD/HDD
    Boot to Vista DVD
    Long Delay – options page
    Select – English Australia
    Enter Serial Key
    Select – Do Not Upgrade
    Format HDD
    Create Account – Paul w/ pwd
    Set Timezone
    Approve auto-updates
    Network Location – Home
    Login
  2. Windows Update – all updates
    Disable Sidebar Autostart & Exit
    Enable Readyboost on USB stick
    IE – set Google to default search provider
    IE – set startup to blank page
    IE – Install Flash Player
    Enable Auto logon
    Start | Run | Netplwiz
    Tick the box, and enter credentials
    Set Screen Saver – Blank screen – 3 mins (to protect plasma)
    Advanced Power Settings
    10 min display off
    HDD 10 min spin down
    Never Sleep (this MB won’t resume, your’s may)
    Bluetooth KB works – no drivers req’d
    Updates finished – reboot
    Disable Welcome screen
    Set resolution to TV screen native – 1360 x 768 (1366×768 native)
    Experience Index – 4.6 (CPU 4.7)
    Activate Windows
    Check Device Manager – all devices OK from Windows Update
    Start | Network | Click “Turn on Network Discovery” on top bar | Make Network Private
  3. Install WHS Connector (if running Windows Home Server)
    Set Home Server to ignore AV errors
  4. Vista Media Centre Setup
    Custom
    Setup Signal
    Accept 2 tuners
    Use Guide
    “your postcode”
    Guide Not Available
    Scan for Services
    Vista Media Centre – Audio
    Single RCA
    5.1
    Test OK
    Vista Media Centre – Display
    Flat Panel
    VGA
    Widescreen
    Keep display resolution
    Adjust
    Check Sizing and Centering OK
    Settings – “Start Windows Media Center when windows starts”
    Enable optimization at 4am
    Set storage to leave 60gb free (400GB)
    Stop recording 4 minutes after
  5. Create restore point
  6. Ice TV
    Login
    Setup Interactive Device
    Install ICE TV Software
    Follow install instructions
    Then do a manual setup and update guide
    Leave all settings default atm
  7. Vista Media Centre – Setup Guide for VMC
    Download Guide
    Setup Channels
    Remove SD channels and doubles
    Keep HD7,HD9,HD10,HD ABC, HD SBS, ABC2
    Reset “Add Listings to channel” to get guide to update
  8. Windows Update – Reboot
    Windows Update – Reboot (Req’d)
    Windows Update – Reboot (not stated)
    Windows Update – Reboot (SP1)
  9. Install Lifextender
    Enable Automated Scanning – Midnight
    Untick “Display TV Show Info when uncommercializing”
    Set – Upon completion DELETE the original
  10. Uninstall KB950126 (it’s a known bug)
    Reboot
    Windows Update – Check for Updates
    View Available Updates
    Right Click KB950126 – Select “Hide Update”
    Install other pending updates
  11. Fix Lifeextender Auto-start
    Search – Task Scheduler
    Create Task – Lifextender
    “Run with highest privileges”
    Triggers – Begin Task – “At Logon”
    Actions – New – “C:\program files\yellow cup\lifextender\lifextender.exe”
    OK
    Delete Lifextender from Startup Folder on Start Menu
  12. Enable Terminal Services Access (if running Vista Ultimate)
    Computer – Properties – Remote Access – Remote Desktop – allow connections
  13. Enable Ripped DVD Gallery

    HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Media Center\Settings\DvdSettings
    ShowGallery -Play

    Change the Registry value data from Play to Gallery.

  14. Install KLite Standard 4.1.7 – all settings default
  15. Clean up Vista SP1 removal files – Start | Search | CMD | type “vsp1cln.exe”
    This will free up between 1 and 1.5GB of disk space

Presto – you should now have a system that can play most things, be reasonably stable and do all the things it should out of the box.

It’s only 98 real steps to set it up, not including all the “next, next, finish” buttons.

Wouldn’t it be great if this was how it came from Microsoft in Australia.

Depending on your hardware – your mileage may vary. This was FAR more stable than installing all the drivers from the vendor sites.

You may optionally want to

a)  Have Terminal Services Access without interrupting your TV (haven’t done this yet)

b) Improve your HDD performance, but also have increased risk of data corruption.

Update Sept 09 – due to hardware failure I’m not running W7 and new hardware.
W7 is much better. Full review coming.
Still needs help though.

My digital home just isn’t quite there yet

I’ve been trying to have the whole digital home experience for a while now. I’ve reviewed Vista Media Centre, Home Server, Wireless N before, the experiences there are documented.

The impossible dream I am thinking of consists of seamless integration between:

  • Home Server
  • Media Centre
  • Extenders
  • Game Consoles
  • Digital Picture Frames
  • Media Players
  • Wireless
  • Broadband
  • VOIP vs VOIP vs VOIP
  • Desktop PC’s
  • Cloud Services
    • So lets look at the current state of play for these things.

    Home Server

    I reviewed home server here. It’s a great product that acts as a network store for your folders, and backs up all the PC’s in your house. It has a brilliant Disaster Recovery solution and one of the most innovative backup and storage solutions I have seen in a long time.

    Unfortunately that’s all it is, and that means 40watts at idle 24/7 – 88KwHrs / qtr, and more if I make it work hard.

    Media Centre

    This thing‘s purpose in life is to record TV, and play content. In Australia it does an average job of both. It’s another 50watts 24/7.

    Trying to get a PC to work well in a lounge room is an exercise in frustration. Noise, power, keyboards, mice and windows hiccups are just not fun. Add in the TV compatibility problems many experience trying to get TV interfaces and resolutions correctly out of a PC.

    It’s great when it works, but it’s definitely high maintenance.

    As many people have said before me, what about storing the videos on Home Server, and getting it to do the recording too. That way they noisy PC can stay in the back room, and one PC can have an excuse for staying up chewing power 2/47, not two.

    I don’t really stream my video, I watch it off a remote file share. A performance comparison of Streaming vs File Serving is here.

      Extenders

    I can’t claim to have one of these yet. The reviews are average, and every XBox 360 I hear about dies an early death.

    These seem to be the obvious choice for the lounge room. They are quieter, more aesthetically pleasing, have native TV interfaces (PAL, Component etc), and an interface that never needs a mouse to drive it. They also play games, so many people end up wit one anyway.

    The well known issues with power consumption on some of these come up, but at least it’s only when being watched, and it’s probably not more than the plasma TV anyway.

    A chipped XBox with XBox Media Centre was my weapon of choice for years. It’s an aging solution now, and doesn’t have HiDef, but as a media player for over the network content, it beat Vista Media Centre hands down.

    I can see this being the future of the lounge room.

    The current problem being the limited support for what can / cannot be streamed to the current extenders, and how to connect the things.

      Game Consoles

    The Wii seems to excel here, as  a pure games console. The best comparison i reckon is here.

    If you want all the video streaming, it’s XBox 360 or possibly, Playstation 3.

      Digital Picture Frames

    Nice toy, chews power all the time, unlike an old fashioned photos. Be good if they only displayed when they sensed movement.

    The wireless versions I have seen seen only work off Flikr and other web services. None seem to be connecting to a local PC to show photo’s from. Great for using bandwidth.

      Media Players

    Everyone has an iPod, even me. I dont’ use iTunes at all, so it’s solid MP3’s. Unfortunately Apple in their wisdom designed the thing on ID3 tags, not on files / folders. If your tags aren’t perfect, you’ll have all sorts of fun navigating the thing.

    I’m still looking for the cheap simple, web managed, wireless MP3 player I can plugin to my amp.

    An old laptop or Asus eePC is topping the list at the moment.

      Wireless

    I ran wireless G for ages – just plain G, none of the tricky variants as my laptops built in is only basic G. It’s fine for web browsing and acceptable for file copying, but coverage and streaming movies never really worked.

    I recently threw a bucket of cash a a Wireless N solution, sticking to a single vendor to keep the process smoother. Netgear is more common that Linksys in Australia.

    Well the coverage is mildly better, and it is a bit faster, and that’s all. The problem is that the speed is still very variable, so even though the average is OK, for playing movies over the LAN – it all falls down.

    Luckily my house lets me run a cable underneath to the lounge easily, so I have a solution, but it’s a disappointment.

    I will have no such luck with running a cable for the spare room with the kids TV and their (might get for XMAS) 360. The consoles also don’t support Wireless N, so that’s more kit to buy.

    So much for the wireless dream.

    Looks like some others agree with me here and here

      Broadband

      I’m got lucky here, with my house being able to get Naked (Unbundled Local Loop) ADSL2+. I get about 7 Mbit download speeds. Being ULL means I don’t have to pay Telstra any money, and that makes me happy.

      I don’t quite understand why a GB is cheaper on ADSL2+ than the same GB on ADSL, but I won’t complain seeing as I get the service.

      This is one part of the equation that works well for me.

      I do however have to run THREE devices pulling another 40Watts 24/7. A Wireless/Firewall/Router, an ADSL Modem and a VOIP adapter, all pulling their own power, each with their own old fashioned power brick radiating heat.
      Where oh were is the all in one Modem/Router/Wireless N/VOIP adapter that only pulls 5 watts.

      VOIP vs VOIP vs VOIP

      “Skype” vs “VOIP from my ISP” vs “Messenger”

      And never shall any of the above meet. Good luck all ye who enter here.

      Then we move onto Bluetooth headsets, Wireless Handsets, GSM Picocells, and Cordless phones just for laughs.

      Desktop PC’s

      These integrate acceptably into the home network. My biggest hurdles here are

      • What version of Vista
      • Remote Desktop
      • Local User logons
      • Authentication and Passwords for file / printer sharing

      I’m used to domain authentication, and sharing stuff around a home network isn’t quite as simple as I would have thought. More work to be done in this space.

      It is still common to email a file via their internet email address, that is in the same house as you! That tells me point to point file transfer is still way too hard.

      Hmm – maybe device authentication is a better idea in this space than traditional user authentication.

      Wireless Printers

      I have a HP wireless (802.11g) All In One unit. It works well enough except for one critical problem. It gets it’s IP Address via DHCP from the Router. If I don’t set a reservation, this changes sometimes due to the alignment of Venus and Mars. The software fails badly at finding it when it’s IP address changes.

      Tip for Wireless device designers, home IP addresses are very dynamic, plan on your drivers having to work with that.

      A reservation fixes the problem – until the next firmware upgrade.

      Cloud Services

      And finally we have the latest set of toys, the one’s the world’s been saying will happen for years. Well it’s not there yet, even with ADSL2+, but I agree, it’s coming.

      I do like some of the services and I’m using more and more over time.

      • Google Docs is interesting for sharing stuff with my wife.
      • Windows Live Mesh is a new toy – no comment yet
      • Newsgator / FeedDemon
      • GMail w/ Outlook IMAP (testing – it’s a hassle)
      • Domain hosted with Bluehost
      • WordPress for this

      I tried online backup of 20Gb of data, and it does work, but it’s nowhere near as practical as Home Server for me. For small quantities of data it would be very good.

      The whole area is changing, and local sync options are making it more interesting. Google gears enables some of this functionality and is worth looking for.

      Security and privacy are significant concerns. Google can inform you about google here.

      A single authentication solution like OpenID starts to make all of this much easier.

      Summary

      Well the seamless integration of all this stuff is still a long way off. The problems are solvable, but the maintenance is high.

      All I want is devices that chew no power, work together seamlessly, is available anywhere in the world and takes no effort from me. It’s not too much to ask for is it?

      Watch this space.

      Here’s a AD DNS Screwup I have seen firsthand

      This is why I learnt a long time ago – if you have an AD problem – it’s probably DNS.

      Creating an empty DNS zone with the same name as your internal zone can lead too all sorts of frustration – especially with the multiple locations in AD that it can end up in. You’ll find yourself knee-deep in ADSI Edit faster than you would ever want to be.

      I love DNS, but it’s gotta be right, and it’s easy to get wrong.

      Brisbane Translink Go Card Charges in Pounds Sterling

      Sheesh – at the exchange rate – no wonder public transport is expensive! Pounds Sterling – what next – Euro’s?

      image

      image

      This is of course on top of what would have to the  slowest proximity readers on the planet, displays that are impossible to read, an arcane array of “beep” codes, and an “Auto Top Up” function – that doesn’t.

      Great system guys, matches your train network perfectly.

      Winch Mount Testing and Breaking Rope

      I decided that my home made winch mount IMG_3804needed to be tested so I could trust it somewhat. I see from some of the conversations I am not the only one with doubts. The mount performed perfectly, although he synthetic winch rope broke.

      The tests were all done on the first layer of the drum to give maximum tension. This also places additional load on the mount due to the increased height adding leverage.

      Test 1 – Drag Car on Dirt (Wheels Locked)- Passed

      Test 2 – Drag Car on Dirt Up Hill (Wheels Locked) – Passed

      Test 3 – Drag Car on Bitumen (Wheels Locked) – Passed

      Test 4 – Drag 2 Cars Uphill on Dirt (Wheels Locked) – Rope Failed

      I believe the rope failed due to 3 factors

      1. The Technora fibre is wearing noticeably more than the Amsteel Blue. I have read that the high temp materials are less abrasion resistant.
      2. The fairlead internal radius is too sharp compared to the external radius. The radius should be 4x the diameter of the rope.
      3. The fairlead sits approx 3/4 up the height of the winch, giving a fair bend to the rope as it goes over the fairlead and down the bottom of the drum. Rope rated at 13,700lb loses strength when bent. The tighter the radius, the greater the strength loss. The numbers are hard, as it depends on the diameter or the rope, and it’s construction. 12 Strand is a good construction for bending.

      I am surprised a 9500lb winch can break a 13,700lb rope.

      At no point did the mount appear to bend excessively, or sustain any visible damage. There was NO permanent twist.

      The fairlead mount also suffered no damage, and the rope was spliced back together easily.

      I was surprised at the amount of spring in the rope, you can see it coiled up under the tree where it ended up.

       IMG_3798 IMG_3799 IMG_3800 IMG_3802 IMG_3803 P8030027 P8030026

      4WD Winch without the Weight

      I wanted to fit a winch P4260063to my ’98 HZJ105 without wearing a huge  weight penalty, or spending too much money. That ruled out walking into ARB and handing over the Credit Card for a new Bullbar and Warn. It meant I had to DIY this little exercise.

      I wanted to keep the weight down, so decided a 9500lb was smaller and lighter than a 12,000lb. If I needed more pulling power, I would use a pulley block.

      After reading the SA 4WD Winch Review, and getting a good deal, I bought a 9500lb Ironman Winch for $625 w/ 3yrs warranty. They are claimed to be waterproof and come with what I needed. I would have preferred a Premier winch due to the brake not being in the drum making it more suited to synthetic rope, but it was out of my price range. I would really have preferred a hydraulic Milemarker, but that was really really out of my price range, especially when I factored in questions around power steering pump flow rates.

      After using steel cable and hating it, synthetic winch rope was a requirement. I figured 100ft (28M) of 5/16″ (8mm) rated at 13,700LB breaking strain would do. Less rope on the drum keeps the winch closer to it’s rated pulling force. I specifically bought their “Hybrid” line, where the first 25ft is a Technora based rope, and the rest is Amsteel Blue. The Technora is more temperature stable for use on winches with a brake inside the drum. I bought mine from www.cseoffroad.com, along with the alloy fairlead, rope protector and safety thimble. The alloy fairlead was a free bonus at the time. I also grabbed 100ft (28M) of 3/8″ (10mm) Amsteel Blue winch extension rated at 19,600lb breaking strain. The separate extension allows the flexibility of simply extending the pull, connecting two different points, putting a winch block in between the two etc. You need the thicker rope size if you are going to use a winch block, as it will double the pulling force – 9500×2 = 19,000lb.

      The next item was a mount. Other mounts I have seen use 6mm steel plate in various arrangements. I settled for the thickest and strongest alloy channel I could get – 8.4mm x 6.8mm tempered alloy. It doesn’t have a rated “strength” in this configuration, especially after I chopped it up to get the winch to fit. I can say that in a pull to stall test, the rope snapped before the winch stalled, and the mount, while having a slight twist, suffered no permanent deformation. The winch is in an ideal position, directly between the chassis rails, very low, and as far back as possible. The mount acts to protect the radiator from sticks etc.

      The fairlead I mounted into the bullbar with 20x heavy gauge rivets. Whilst I am not that confident as to their strength, in shear the steel pins combined with the tight fit should be very strong. At a 45deg pull they will be in both tension and shear, a less desirable situation. The alloy bar mounts are particularly weak in a sideways direction, so I think I’ll be avoiding heavy angled pulls. The fairlead needed to have it’s inside edge rounded a lot, as the winch sits quite low in relation to the bar.

      Finally it was just a matter of the control box hidden inside the bar, the wiring and lockout switches. I recommend the lockout switches be waterproofed underneath with silicon, and rubber caps fitted (I used rubber “feet”). Water pulls inside them and the copper contacts stop working.The winch is wired with a switch to each AGM battery, letting me use either one, or both.

      By dropping the steel cable and roller fairlead, using an alloy mount, not using a 12,000lb winch and keeping the alloy bar I have kept nearly 75KG off the front of the car. Just as importantly the weight is as low and far back as possible, something most winch bars don’t do. Whilst I accept it’s not as strong as a steel winch bar, it has passed every test I can throw at it so far.

      Finally – it’s a whole lot cheaper than driving into ARB, and with Mickey T MTZ’s, twin Air Lockers and some lift, hopefully I won’t need to winch too often (yeah right).

      Item Weight (KG)
      Winch 21
      Controller 3
      Rope & Thimble 2
      Fairlead 2
      Mount 2.5
      Total 29.2
      Item Weight Saved (KG)
      9500 vs 12000 10
      Fairlead 4
      Rope & Hook 8.5
      Mount / Steel Bar 50
      Total 72.5KG
      The heaviest aluminium alloy channel I could find. It is also tempered (or so the guy said). 8.4mm base, 6.8mm walls. PB290026
      Trimmed to fit winch and chassis rails with lower bull-bar mounts. I should have rounded the corners more to stop fatigue. PB290027
      Winch fits nice and snug. PB290025
      Alloy hawse fairlead mounted in the factory bar. Yeah I know, I know. If it snaps the rivets then I’ll do something sttronger. The alloy is 6mm checker-plate. The bar is not that strong, nor are the bar mounts. PC080067
      Terminating the high temp Technora fibre. PC080072
      Had to grind the allen key to fit the link on the rope. PC080069
      All spooled up. Rope protector is the black bit.
      I have a strong plastic / rubber  flap that covers this gap keeping grass and mud out.
      PC080074
      Pic when fitted. Yes, I suspect it may snag something one day. P4260062a
      The lockout switches with waterproof covers. They don’t like water in them, and then don’t work. P4260062b
      Internals of the waterproof control box. It’s zip-tied inside the bullbar. PB290028