Ipanel problem.

Besides the Reset problem with the Asus Ipanel which I managed to find a fix for here www.jimhearne.co.uk/ipanel there are two remaining problems with it.

When the Ipanel updates, either when you press the Mode button, or when the Ipanel display auto updates about every 30 seconds, you get a noticeable pause in what ever is running.

This is most noticeable in games or playing MP3's.

The problem occurs even in Dos applications, it is completely unrelated to operating system and so far not related to any specific hardware.
However a few people have reported they don't get this problem but there is no correlation that I've been able to work out between those people that don't get the problem and the hardware they have in the machine.

(Update: At least one of the people who supposedly weren't having this problem turns out to work for Asus !!!!!)

On the early A7M/266 motherboards the Ipanel only updates when the Mode button is pressed so you only get a pause when you are changing the mode.
But on the later A7M/266 bios there is an option to enable the Ipanel updates and then it suffers from the same problem as the CUSL2.

This mod below does not fix the pausing when you press the Mode button on the Ipanel, this delay has a different cause and I suspect cannot be fixed without a Bios update or quite possibly without hardware modifications to the Ipanel and or motherboards.

Feel free to post a link to this page or copy the details but please give me credit for working it out if you post the details somewhere else.

www.jimhearne.co.uk/ipanel

Email: Please do not email me with Questions about the Ipanel anymore, it's been removed from my pc and consigned to the junk cupboard.

At work we are no longer using Asus boards so i don't know about compatibility with newer boards though all indications from mail i've been sent are that all newer boards still suffer from the stuttering.

Results of testing on different Asus motherboards

Asus motherboard	Ipanel Problems	Tested by myself.	Feedback from others
A7V133	Does not support Ipanel despite what some early reviews said.	N/A	People asking how to connect up a Ipanel :-(
A7M/266 (bios rev 1004 or later)	Stutters if enabled in Bios.	No, but will be	Mixture, positive and negative
A7A/266	Stutters	No, won't be.	Mod does not seem to fix the problem.
A7V/266	Not known but expected to stutter.	No	None as yet.
CUSL2 series	Stutters	Yes.	Loads, Fixes stuttering in almost all cases.
TUSL2 series	Stutters.	No	1 negative, no other feedback yet.
CUV4X series	Stutters	No.	A few negative reports

Notes: Asus seem to be releasing Bios's for most Ipanel supporting motherboards that now have an option to change the time between Ipanel updates and more importantly turn of the updates completely. Though of course that makes the Ipanel a lot less useful.

The A7A/266 is a board we are not stocking at work (therefore I can't test the Ipanel on it) due to it's many reported problems after it's release. Apart from it' ability to take SDRAM and DDR memory it has no advantages over other boards.

I suspect a few of the negative reports I've had have been down to user error in doing the mod, please double check everything before emailing me.

The Cable is handed after the mod, make sure you've got the right end connected to the Ipanel.

The Anti Stutter mod.

Please note, try this at your own risk.

You need: 2 off 1 k Resistors, any wattage (suggest 1/8 or 1/4 watt)
A 0.001uf or 1nf (often labeled 102K) ceramic capacitor (any voltage)
A soldering iron.

Do the reset mod (see note below first) with one of the resistors, full details Here
Cut the wire going to pin 13 of the Ipanel connector EXTSMI, Green arrowed pin in
the picture to the right.
In between the cut ends fit the 0.001uf capacitor.
Remove the a small piece of the insulation from the wire going to pin 15 (Blue arrow) on
the Ipanel connector, this is a 5 volts supply, you don't actually need to cut the
wire, just expose enough of it to solder to. (but it might be easier to cut
and rejoin it).
Solder the second 1 k resistor between the wire going to Pin15 and the Ipanel side of the
capacitor that's in series with wire to pin 13.
Insulate all the joints and power up.

The diagram above includes the Reset mod (bottom track) as well as the Stuttering one.

Please note, Asus have now modified the Ipanel to fix the reset problem, on Ipanels from PCB 1.02 with PCBA B14

These appear to be the new plastic cased Ipanels, if you have this version or later you do not need to do the mod to pin 19

If after applying the mod the stuttering is reduced but still there try decreasing the value of the capacitor to maybe 0.47nf (471K).

If after the mod the Ipanel display does not change mode when pressing the mode button try increasing the capacitor value, try 2.2nf (222K) or even 4.7nf (472K).

But please check you've got the mod correct before changing things.

If you go too high the Stuttering will come back, if you go too low the updates will stop completely, 1nf seemed optimal on the Cusl2 series of motherboards. If you find a different capacitor value works better for you on your board please let me know.

Here's an example of the mod neatly applied to a cable kindly sent to me by Pao Nan.

Note he's made up a new cable using some 26 pin connectors rather than modding his existing Ipanel cable.

The extra pins don't hurt they just hang off the edge of the connector.

Make sure you insulate the components with some tape after doing the mod.

Some Technical stuff on the stuttering problem (I originally posted this on the CUSL2 forum which is why it reads funny)

I started with a basic system, hard drive running 98, video card, floppy.
Stuttered in 3d screen saver during Ipanel update or pressing Ipanel button.
So I went through.
Updating to latest Bios
Updating to latest drivers.
Changed PSU from cheapo 300w to high quality 300w (Sorry, haven't anything bigger *)

I then decided to see if the system still juddered in dos, I ran a old TC program that just made moving line patterns on the screen.
This also juddered when the button was pressed or when the Ipanel updated......
So, the problem is not related to OS or drivers, I continued using this dos program for testing.
Removed Hdd and tried clean dos boot from floppy.
Tried a PCI S3 video card.
Tried every setting in the bios even remotely lightly to affect anything.

Disconnected the two 5volt supply wires to the Ipanel and connected them to an external 5volt bench supply, still stutters*.

* Later Note: This was because some people on the CUSL2 board were convinced it was a PSU problem.

So, back to measuring things.

The 5 volt power feed to the Ipanel is 5.04 volts. The Ipanel draws 13 to 40ma from the 5 volts rail depending on how many display segments are on. Less than I expected.

The Ipanel pulses EXTSMI# low (3.3v logic) everytime the button is pressed or the auto update occurs.
There is then a burst of clock pulses on SMBCLK and a burst of data on SMBDTA

When the auto update occurs the EXTSMI pulse is about 50ms wide, if you press the status button the pulse is around 75ms wide.

Starting on the falling edge of EXTSMI there's a burst of activity on the SMBDTA and SMBCLK lines, 20ms on the auto update or between 20 and 600ms if you press the status button, depending on how quickly you press the button.

Interestingly pulsing EXTSMI low yourself toggles the system between normal and sleep mode.
This is what I thought EXTSMI was for anyway, a interrupt to the cpu to do with power saving, this needs some reading of Intel data sheets.

[A while later]
I think Asus may have hijacked this Interrupt to use it for updating the Ipanel display, it might explain the stuttering, if the system is momentarily going into suspend mode while the Ipanel is updated it would definatly stutter.

[A while later]
[update] This theory is supported by the fact that if you actually connect the normal SMI switch and toggle the system between suspend and normal mode the Ipanel moves on to it's next setting, exactly like it does when you press it's mode button.
[A while later]
[update 2] Just found out the EXTSMI pin on the Ipanel connector is connected directly to one side of the front panel SMI switch connector so they are the same signal.

So it's not surprising the cpu stutters, it's momentarily being suspended.......

But, why do some peoples machines not do it ?????? (allegedly)

I've tried shortening the EXTSMI pulse using an R/C circuit to a few ms and this seems to have an effect on the stuttering.

You've probably just read the results above.