Well I changed the A7v&#39s voltage... downwards

[Rob Cork]

Right, I just tried the A7V mod for the third time (see my other post), and I managed to get the first leg of the 24k resistor soldered to the bottom left smd resistor. I noticed the smd resistor was at an odd angle, as the end I had soldered to had come away - there was still contact due to the solder, but wanting things to look right I *carefully* bent it back into place. It fell off. I couldn't find it (you know how small those things are). So, figuring my board was already dead, and with panic starting to set in, I pressed ahead (call it optimism, call it denial, whatever).

Of course, sod's law, my soldering skills suddenly improved tenfold, and I was able to remake the connection that the smd resistor made with solder, and then complete the volt mod soldering from there to the other smd resistor as usual. Fingers crossed, power it up - it works!!!!!

One thing - when set to 1.85V, my core voltage used to be 1.94... it now reads 1.79V. Clearly the 24k resistor wasn't quite enough to compensate for the smd resistor. My question is (doubt anyone can answer it for sure), is will soldering another 24k resistor in series with the current one, give me about 2V? If one gives an increase from 1.85 to 2.32, presumably with my core voltage of 1.79 I would get about 2.25V max - more than enough. Plus the soldering would be dead simple - just cut the wire from the 24k resistor already in place, and put another one across the ends. Any comments?

Sorry for the long post... I'm still recovering from the shock that my mobo still works at all ;D

 

[Rob Cork]

May just have answered my own question - it would probably be safe to get a 24k rheostat, and put that in series. That way I can start it on 0, and only increase it a bit at a time to make sure it stays safe (ie not 10V or anything!). Perhaps I could even turn it up veerrrry slowly with the comp on, to see how it changes the voltage read in Asus probe. Comments anyone?

 

[Eriksson]

Well I dont have the A7V board and I have therefore not studied the mod for it in details, still...

You will not get higher voltage by adding another resistor in series with the one you already got. You need less resistance not more.

However changing the resistance of the mod resistor will NOT cure the problem you are in now, no way.

The resistor you 'lost' is probably a pull up resistor for the voltage divider circuit on your board. If you start experimenting with some bizarre values for the mod resistor you will probably kill your DIE HARD Asus

You got two options if you want to save your board.

1
Find the resistor you lost and get it back in place.

2
Find out the size of the lost resistor and solder a new, same value (regular) resistor in its place.

Make sure that the lost resistor is not anywhere close to your board (It might short some connections and make probs)
Using a vacuum cleaner with new bag would both remove it from you case (board) and give you a shot towards finding it.

 

[Rob Cork]

I know normally more resistance means less volts, but the A7V mod involves soldering a 24k resistor between two points on the voltage generator, so in this case, more resistance = more volts. I can't explain it as I don't understand the circuit diagram for the voltage gen - Tom has part of it at Tom's Harware if you're curious.

I've given up hope of finding the smd resistor i lost - it's about 1mm long and exactly the same colour as my carpet. I'm pretty sure that's where it is, as I shook my mobo loads and nothing rattled

I have slightly changed the circuit layout, as my attached scribbles show, but I figure the logic is still the same for the volt mod, so it should work. As I'm not sure, I think I'll use a variable resistor, so I can start it on zero, and I know the mobo will still work. I can then increase it a tiny amount. If it increases the voltage, fine. If it decreases, then I can remove the 24k resistor, set the 24k rheostat to full (same situation as at the moment effectively), and slowly reduce it, and see if that increases it. I'm willing to risk it killing my mobo - it'll give me an excuse to get a KT7a-RAID

 

[Eriksson]

Rob Cork (Jan 11, 2001 03:43 p.m.):
I know normally more resistance means less volts, but the A7V mod involves soldering a 24k resistor between two points on the voltage generator, so in this case, more resistance = more volts. I can't explain it as I don't understand the circuit diagram for the voltage gen - Tom has part of it at Tom's Harware if you're curious.

I am not going to argue with you but there is no doubt in my mind that you are wrong.
It is obvious, since no mod gives 1.9V , 24k gives 2.3 and 20 k gives 2,4v.

 

[Rob Cork]

Actually I think I can see your point - the reason an increase in resistance with the volt mod gives more voltage, is that the circuit is changed. After that, decreasing the resistance used gives more voltage, as you would expect. This definitely makes more sense. I think I will try instead then replacing the 24k resistor with a 24k rheostat, as per part of my earlier plan. That way I can reduce the resistance used slowly. If I do fry my board then don't worry, I'll blame it on myself, but I can't live with 1.8V max!! I'm sure you understand ;-)

I'll post with the result (whether an increase or decrease in resistance raises the voltage), but I think now you are right - I will need to decrease the resistance. I never was much good at electronics

 

[Deviant]

Will you listen to an engineer? Eriksson is RIGHT!
Think of it this way. Without the mod there is a million ohm resistor accross the 2 mod points.
Decrease this to 24k and the voltage goes up.
Decrease it some more, and you get more volts!!!.
With the 24k resistor you can go right up to 2.34 volts. Do you really need any higher? You will run the risk of driving your voltage regulator out of speck.
Trust Eriksson and myself on this one. I have even tried a 22k resistor, and the voltages went up by about 0.5 volts. So the minimum was about1.9-2.0 volts. You can really fry you chip with a 22k, so I recommend using a 24k resistor. You will get all the voltage variance you need.

 

[Rob Cork]

It's ok, I understand now. The problem I'm having is that my maximum voltage with the resistor is now 1.79V, as I screwed up (see above posts). I assume changing the resistance of the new 24k resistor that I added would vary the voltage somehow, I just wondered whether I would need to increase or decrease the resistance. Sorry if I appeared to be ignoring the advice, it's just I can't possibly leave it as it is, my duron won't even do 900 now I just wondered whether there was anyway of solving my screwup without trying to resolder an smd resistor (my soldering skills aren't that good).

 

[Cullen]

Hey Rob

Im not an eletrical guy or anything like that but I think you are putting that rheostat in the wrong spot. It seems to me that the rheostat should go in place of the resistor that you lost, this way you could find the original resistance of the smd resistor and the voltage mod will work as it was intended to.

Please look at the attachment for a clearer picture of what im proposing.

I could be wrong but this just makes sense to me.

 

[ElectricMan]

rob, I think you should try to find the resistance of the smd chip you lost and replace it with a regular resistor, I ain't no electronics expert , but I would get your motherboard back in shape before you start adding rheostats and stuff. I am a electrician , and do know how to read a ohm meter but I do not know if it would work. I was thinking about trying and giving you a number value so you could replace, If anyone knows if this would work , please say so. BTW , rob , I do soldering all the time and I thought what a *itch as I was doing the mod and almost said *ell with it, myself. I used my kids magnified glass to do it

 

[Deviant]

Is that true, how can you be an electrician and not know how to read an ohmmeter? Things are very different in your country, or you have a funny sense of humour. Don't want to put you down, but I just thought that would be an essential in your job.

Any back to Rob Cork, I had a long chat with him on ICQ, and put him onto a few things to try. The resister actually seems to be one of the 1nF capacitors, and the mod calls for soldering to the earth side of it. We came to the conclusion that he might have soldered 2 resistors togethers or somthing, but he will keep us posted when events unfold. I suggested that the capacitor missing would probable affect stability, by not filtering noise off the power sense line.

 

[Rob Cork]

Update to the situation - it's all fixed now (apart from the capacitor ;-) ), see my post in AMD cpus for the details. Thanks guys

http://forums.overclockers.ws/UltraBoard/UltraBoard.cgi?action=Read&BID=2&TID=2064&SID=71099

 

[Eriksson]

You are welcome and congrats on your result