PDA

View Full Version : Proper way to make a 120K VR??



Kunaak
03-22-2005, 12:55 AM
its been awhile since I had to make a VR of any sort, so I am not sure if I got this right or not.
I need a 120K VR, but since they are impossible to find here, I just made one from a 100k VR and a 20K fixed.
I just need to confirm I have the right method for this, before I use it.
it's been about a year or so since I needed to do this, so I just wanna make sure I remember this right.

http://www.xtremesystems.org/forums/attachment.php?attachmentid=27086&stc=1

reject
03-22-2005, 04:54 AM
couldnt u get a larger vr and turn it down?
anyway that will work

celemine1Gig
03-22-2005, 12:14 PM
its been awhile since I had to make a VR of any sort, so I am not sure if I got this right or not.
I need a 120K VR, but since they are impossible to find here, I just made one from a 100k VR and a 20K fixed.
I just need to confirm I have the right method for this, before I use it.
it's been about a year or so since I needed to do this, so I just wanna make sure I remember this right.

http://www.xtremesystems.org/forums/attachment.php?attachmentid=27086&stc=1


Looks OK, but keep one thing in mind:

The method that you described gives you a resistance range from 120K-20K and nothing below that, as the fixed resistor will stay fixed, even when connected to the VR. So, basicaly, if you need resistances below 20K, your plan won't work.
If you need the full range from 120K-0, one way would be to just use a 20K VR instead of the 20K fixed resistor, thus making the full range adjustable.

If 120-20K is enough though, then go on. ;)

Kunaak
03-22-2005, 12:58 PM
yes I know, I measured the reisitance last night.
a 20K VR for what I need would blow up my motherboard.
120K is actually overkill for what I am doing, but I like small offset from max resistance, then big offsets. I really only need a 110K for the job, but all my fixed resistors are 20K up to 200K. so 120K is the closest I can get to my goal.

celemine1Gig
03-22-2005, 02:59 PM
yes I know, I measured the reisitance last night.
a 20K VR for what I need would blow up my motherboard.
120K is actually overkill for what I am doing, but I like small offset from max resistance, then big offsets. I really only need a 110K for the job, but all my fixed resistors are 20K up to 200K. so 120K is the closest I can get to my goal.

That's not absolutely correct. If I fully understand what you are saying, you have more than one 20K fixed resistor and 110K would be the exact resistance that you'd need, right?

Now to achieve 10K of resistance from 2x 20K resistors is very easy. Just connect them in parallel. You take two 20K resistor and connect them in parallel which gives you 10K and then you take these 10K and add them in series to the 100K potentiometer. That would give you exactly the 110K that you'd need. ;)

remember:

the formula for 2x resistors in parallel is:

R1 x R2 / R1 + R2

In your case

20K x 20K / 40K = 400K² / 40K = 10K :toast:

Edit:

BTW, you said lower than 20K resistance would cause the voltage to be too high, thus killing your mainboard.
THE BEST WAY IMHO, would be to take 2x 20K fixed resistors in paralell (as described above, resulting in a 10K fixed resistance) and connect these two ristors in series to 2x 50K VRs, which are also connected in series.
This would give you very good precision (better than one 100K VR). Because with the 100K VR, the lower resistance fields are very hard to control by the turn of the little "screwhead" on the VR. As the VRs are not linear, i.e. for example half a turn does not always equal 2K (as an example). So you can have 15 turn with 100K which gives a rough 6.666K per turn, or you can have 2x 50K VRs which give you 30 turns with 3.333K per turn. Understand what I mean?
You've got the fixed resistor to prevent the voltage to reach the extremely dangerous levels, but on the last turn before these levels, the extra precision could mean that the system is working, while it wouldn't with the 100K pot.

You see, basically there are hundreds of combinations of fixed resistor and VRs, to fit your needs. You just have to choose what you think would be best for your situation, according to the voltages that you want to achieve and those that you want to prevent from being applied.
So if even 20K would kill your board instantly, you could also take 40K fixed resistance (2x20K in series) and just add a a 20K and 50K VR in series to that. :)

phobos
03-23-2005, 07:40 AM
As the VRs are not linear, i.e. for example half a turn does not always equal 2K (as an example).

That's not right. You can buy linear and logarithmically VRs. And most high precision VRs with more than a single turn are linear. But if you do a Vmod on your board it does not mean that the voltage change is linear. ;)

I would also take a 10k fixed resistor (or 2x20k parallel) and 2x50k VRs in series. That should give you the possibility to change the voltage in very small increments.

celemine1Gig
03-23-2005, 10:17 AM
That's not right. You can buy linear and logarithmically VRs. And most high precision VRs with more than a single turn are linear. But if you do a Vmod on your board it does not mean that the voltage change is linear. ;)

I would also take a 10k fixed resistor (or 2x20k parallel) and 2x50k VRs in series. That should give you the possibility to change the voltage in very small increments.

Oh, thx for the info. I must have gotten it mixed up in my head. :D Was pretty late when I posted that yesterday.

Anyway, the outcome is the same. 2x 50k VRs would still be better than 1x 100K VR, linear or non-linear.

phobos
03-23-2005, 02:10 PM
Anyway, the outcome is the same. 2x 50k VRs would still be better than 1x 100K VR, linear or non-linear.

That's true ;)