Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

ICL 7107 Voltmeter accuracy

Status
Not open for further replies.

bigal_scorpio

Active Member
Hi to all,

I have just made a voltmeter from this site ICL7107 / ICL7106 - Digital Voltmeter

It is working ok in a fashion, but seems to be varying in its measured results when compared to both my DMMs.

I have changed the input resistor to 10k to give 2v scale as this is all I will be measuring but low end measurements tend to be small and the upper end towards 2v seem to be too high!

The author says to adjust the voltage between pins 35 and 36 on th 7107 to 1v but when I do the readings on the display are vastly wrong, over double those from my DMMs.

If I adjust the pot to measure a battery say 1.544v to agree with both my DMMs then the readings from a dead battery say .734v on my DMMs would show as .453v on the 7107s display.

I have used 1% resistors throughout and good spec caps. Is there some likely cause to this problem that I should be looking for or are the 7107s that innacurate (seems unlikely)?

Any ideas guys?

Thanks Al
 
hi Al,
Just in case you have not got the d/s for ICL7107

EDIT:
I have changed the input resistor to 10k to give 2v scale as this is all I will be measuring but low end measurements tend to be small and the upper end towards 2v seem to be too high!

Which resistor is that.??
 

Attachments

  • 7106.pdf
    203 KB · Views: 348
Last edited:
Hi Eric,

Thanks Eric, yes I did have the sheet and have spent some time with it but not come up with any useful data for the problem.

BTW the meter is for measuring the result of another project I made from this. DIY Audio Projects - Hi-Fi Blog for DIY Audiophiles: Simple Low Resistance Measurement

The low ohms measurement is excellent from this and as I often need to find small resistances I thought why not build a meter into the box to save getting the DMM out and holding the probes on to measure all the time, so I built both into an old laptop PSU case( I use them for allsorts and always have a good few dead PSUs handy).

If I could just get the measurements to be more accurate I would be overjoyed as the unit is easy and compact.

Any thoughts mate? Al

Edit. The resistor I changed is the 1m that goes to pin 31, as it is a voltage divider I assumed changing it to 10k would give 2v scale instead of 200v and it seems to work ok. Al
 
Last edited:
hi Al,
All the component values are in line with the datasheet values.

Are you using a +5V reg and the 7660 for -5V , I assume these voltages are steady and within spec.?

Do you have a reference diode you could use as a standard to check the project, or say a TL431 or a 2.5Vref.?
 
There is something familiar about this.

There was a thread here a couple of years ago about a 7106/7 dvm which would not read correctly, and I am sure it was based on the same article as this one.

I cant remember what the outcome was.

JimB

On edit
I just did a bit of searching and found the thread here:

https://www.electro-tech-online.com/threads/icl7107-voltmeter-sensitivity.109313/

Not as old as I thought.

The thread rambles around a bit, but at a quick read through the problem seemed to revolve around a bad capacitor in the 7660 voltage converter.

JimB
 
Last edited:
Hi Eric and JimB,

I have gone through all I can check and all seems ok.

JimB I have read the thread you posted and even changed the caps on the 7660 (1044) to no avail.

Eric I have measured the 5v at 5.02 steady and the -5v at 4.72 steady and no I don't have a TL431 or anything else like it, could I use a voltage divider say 2 x 10k in series across the 5v supply that seems very stable?

Al
 
I think the problem you are having with your circuit is that it only has one calibration adjustment. I used the 7107 in a design several years ago, My circuit had 2 pots. Look at the datasheet Eric posted, page 13 upper right. That is simaler to the circuit I used. During calibration you input what would be the zero level then adjust the zero pot until the display reads zero. then input the maximum value and adjust the scale pot to display the correct reading. This needs to be repeated 3 or 4 times until both readings agree. This is done to get both ends of the measurement range to be correct.
I would post the circuit I used but I dont have any software that does this.
 
Last edited:
I built a few meters using these chips back in the 80s. This circuit is using a voltage divider on the power supply to derive the reference voltage. I always had better results using an external reference (LM285 type I think is what I used) as described in the data sheet attached above.

The resistor on pin 31 sets the input impedance, if it is too low the meter will load the circuit under test and give bad results.
Pin 35 millivolts needs to be set to 1/2 Vin full scale, so 1 volt for 2 volts full scale is correct.
 
Hi guys,

Finally sorted the accuracy problem!

I had a few 20k multi turn pots so I tried one on the input instead of the 10k that formed the high side of the voltage divider.

ended up adjusting both that and the original reference pot at the same time in small doses! Eventually got to a point where the readings match perfectly those on both my DMMs! At last!

Anyway I now have an accurate little box that measures small ohms, well down to about 0r01 anyway which is good enough for me, in Eric's words Perfect!

Just wish the designers of the 7107 had built in a decimal point system! As anything above 10ohms now shows as 1.xxx, but never mind there will only ever be me using the meter so I can cope with that. ;) I could build in a logic gate system to move the decimal but its a lot of work for little return and I'm not sure if there's room in the box. hehehe

Thanks for all the input guys. Al
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top