Digital Voltmeter Help - Solve ATX PSU Issue

[madhatter4]

Hello all,

I made a custom power supply using a standard ATX computer PSU. I took the -12V and +12V leads to get 24V then put a LM338 on the positive side for a 1.25V to 22V adjustable ouput. Works great.

I wanted to put a digital voltmeter on the adjustable output so I would know what voltage was being supplied at the banana terminals. I bought an LED digital panel meter from ebay and wired the voltmeter supply to +5 and ground from my moded ATX PSU. When I connect the adjustable +12V and -12V leads from my PSU to the sense leads on the voltmeter my PSU shuts off. I am guessing its because of a short caused by the negative sense lead and the supply ground being connected somewhere in the voltmeter PCB. If I only connect the + lead from the adjustable output to the voltmeter I get a reading of about half what I should as its only measuring from the + side of the adjustable output.

Is there a simple way to either isolate the voltmeter power or trick the meter into thinking it is seeing an additional +12V? Would adjusting the Vref on the voltmeter do the trick?

Thanks for any help you can offer.

 

[chroot]

Try using your voltmeter to measure the voltage on a standard 1.5V AA battery first, to make sure it's functioning properly.

- Warren

 

[madhatter4]

I have. If I use a separate power supply for the volt meter and connect both sense inputs from the adjustable output the Voltmeter reads correct.

My problem is that I want to use the +5 and ground from the same power supply as the -12 and +12 that I want to measure.

Here is a pic of the meter I am using.

PSU +5 goes to meter 5V+
PSU ground goes to meter 5V-

This is a diagram of my ajustable output that I want to measure. This is from the same PSU that supplies +5 to the meter.
http://jon.simnets.com/wiki/uploads/admin/LM338K_Voltage_Regulator.jpg


PSU +12 goes to meter + IN, actually I am reading the output of the LM338
PSU -12 goes to meter - IN, When connected the PSU shuts off, when not connected the meter reads half of the actual voltage because its reference is from ground instead of -12V.

I am trying to find away around this. Could an opamp be used to combine the +12 and -12 then adjust its output to be double what the output of the LM338 is?

Im trying to explain as best I can. Sorry if I am confusing.

 

[chroot]

If you're applying +12V and -12V across the meter (24V total) across your meter, it appears you are exceeding it's range. The one you've pictured appears to have a range of only 20V.

Use a handheld ohmmeter to see if there any conductance between the negative side of the +5V power supply and the negative sense input.

- Warren

 

[madhatter4]

The total voltage is outside the meter range but I won't be using above 20V vary much if ever. The meter will display a - if it is over range.

I have not tested to see if there is conductance between the negative side of the +5V power supply and the negative sense input, but from the DIY panel meter circuits I have been looking at they are. I will test mine when I get home.

The chip in my meter is an ICL7107 and I am assuming its wired smiler to fig 15 on page 12 of the data sheet but for a 20V scale. I have tried to figure out how mine is wired but its hard to follow traces that run under the LED segments on my meter.
http://www.intersil.com/data/fn/fn3082.pdf"

 

[madhatter4]

I have tested my panel meter and there is conductance between the negative side of the +5V power supply and the negative sense input.

 

[chroot]

So the meter's negative supply is the same node as the meter's negative sense input. Don't drive it with anything and you'll be fine. If you connect only the +12V supply to the positive sense input, the meter should correctly read +12V.

- Warren