Resistor / Multi meter

[peonyu]

Hey, have a question about resistors/multi meters. Some info on what im doing first - I am trying to pencil mod my Radeon 4850 card to give it some more voltage. I *know* where to pencil at on the card [ I have reference card] and how to measure voltage when the system is powered on. But my problem is with resistance.

Heres my problem...For one, the resistor I need to pencil is a 1850 kohm resistor on the back of the card. My goal is to get that down to 1k or 1.1k kohm to bring the voltage up from 1.15v to 1.35v. I have had the card out of the case MANY times now, and penciled the resistor top thoroughly [connected both bridges on it], yet whenever I measure resistance it has not changed at all. I can pencil it alittle bit, and it sits at 1850. I can pencil it black with graphite hanging ontop of it and...1850 ohm. i have also [when frustrated], penciled it a tad then popped it back into the system and tested voltage...Voltage hasnt changed. So i am stumped.

I am wondering if perhaps I am doing it wrong, or my meter is bad. Either case is possible as I am a noob when it comes to mutli-meters.

My meter is about 10 years old, its analog and the resistance settings on it are 1k ohm and 10k ohm. It doesnt have settings beyond that. I used the 10k setting to test resistance and I have calibrated it and put new batteries in etc. I have also tested resistors on a old Geforce 3 card that I have, and on that card to, I can pencil all I want and resistance does not change on the resistors. Any comments ?

 

[KGB]

I have no idea what the 4850 card looks like but taking measurements while the resistor is in circuit can give you strange readings.

First things first: with your meter set on resistance short the test leads together and see if the meter goes to zero. If it doesn't then it needs to be calibrated or repaired.

In order to properly measure the resistance of the component you should desolder and lift one end of the resistor. Never being a big fan of 'pencil mods', why not just replace the resistor with one of the desired value?

 

[peonyu]

This is the back of the card-

And the resistor in question is the one with the black pencil hovering over it-

 

[peonyu]

I am considering soldering it, I have a 100k VR to do it and it would be much more permanent and stable... But if I cant get a proper reading on a resistor for a pencil mod [which is much simpler] I dont feel im capable of doing a hard mod at this time. Thats out of my league until I figure out some of this other stuff =p.

As far as the reading goes, I am hoping its a problem with the MM. Its old and its not digital, I am mainly wondering if this is a known problem that can occur with a bad MM...Giving odd readings when it comes to resistance. The GF3 card leaves me to believe that, but as you said its best to lift the resistor out. Ill try that on the GF3 [its dead anyway] and see what happens.

 

[KGB]

What I was trying to say before is that it may not be possible to measure the resistance in circuit.

 

[peonyu]

Well I checked it again last night...And I ended up just soldering the VR on, didnt have a single problem with doing it. Core is at 825mhz now vs the 700 it was locked at yesturday .

The MM is good btw, and as far as reading goes, yea I was getting the entire circuit. I borrowed a DMM from a friend to get a correct circuit reading. My analog topped out at 10k which was throwing it off. But for voltage reading it reads fine. I dialed volts from 1.15 to 1.33v.

 

[KGB]

Good job.

:thumbsup:

 

[Paperdoc]

I see three problems here that you had to deal with:
1. Almost impossible to measure a resistance without disconnecting one end - as several others have said.
2. You are using a meter at 10,000 ohms scale to measure the difference between 1850 ohms and 1300 ohms (or more). That's a dial reading of 18% down to 13%, hard to read accurately.
3. I really question the ability of a "pencil mod" to do what you want. If I understand correctly, the idea is to place a layer of pencil "carbon" (which is only partly carbon) around the resistor so that it forms a second "resistor" in parallel, thus reducing the overall resistance. To do what you want (reduce total resistance from 1850 to 1300 ohms), the value of the new resistor created from pencil tracings would have to be around 4700 ohms. I question whether you could make a resistor of such a low value by this technique. My guess is you could get more like 100 Kohms that way, and the result would be almost negligible in terms of actual change in total resistance.