Author Topic: µA Meter Sensitivity Amp  (Read 2272 times)

0 Members and 1 Guest are viewing this topic.

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
µA Meter Sensitivity Amp
« on: May 01, 2021, 07:44:10 am »
I have to replace a meter to a piece of analog test equipment in schematic that does not specify what µA range the original was or very much else.
A bunch of resistor dividers, nothing much to see, most ranges have the negative side of the meter to GND, the positive side to the junction of a 2MΩ input resistor and a 120KΩ to GND, "totem pole style". Another one a 5MΩ over a 1MΩ, etc. The only thing it says is about the meter is "range 1 V, input impedance 2 MΩ"

I've never seen the test equipment or the meter, so what did I know. I can tell you it wasn't 50µA. I'm thinking it was somewhere in the 2-5µA range.
To make it worse, it's a center zero type.

The only thing I find at the local shop which is just a local "have it today" shop from China, is a new 50-0-50. The meter barely indicates any signal.

So the question is this ( I'm tired and the answer escapes me, and it's 03:43 and my runtime is 20+ hours ) can the 5µA signal be made larger to the 50µA meter with a simple Op-Amp circuit or other IC that will be a linear translation from the original to the new range?
Where could I find such an example or does someone here know how? The new meter has a DC resistance of 2KΩ.
I wish I could  :=\ Age related insomnia..........
Thanks.
« Last Edit: May 01, 2021, 07:56:00 am by Quarlo Klobrigney »
Voltage does not flow, nor does voltage go.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14071
  • Country: de
Re: µA Meter Sensitivity Amp
« Reply #1 on: May 01, 2021, 08:01:41 am »
5 µA would be very low for a meter movement, usually only found for very small ones and those with a a light pointer. 50 µA would be pefectly normal - that was common for analog multimeters.

Amplification needes extra supply, but would be possible. It may be easier to change the resistors, so that a more normal movement would work.

Old high value resistors may fail - especially carbon film in the MOhms range tend to drift up. So it may not be the meter movement that is broken, but some of the resistors.
A simple circuit plan around the meter may help.
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 6445
  • Country: hr
Re: µA Meter Sensitivity Amp
« Reply #2 on: May 01, 2021, 08:31:06 am »
You can measure what the current was with modern DMM that has uA range. Just connect DMM in uA range instead of moving coil meter and put DUT it in, for instance, 1 V range and apply 1 V to input. It won't be accurate (because MC meter has different resistance from your DMM in uA range) but will show you order of magnitude. If you have 10V range, even better, it will have less error..

But, if you take piece of paper and draw a schematic, i would say your estimate for meter being something less than 5uA would be correct (2 Mohm resistor), provided that is correct.

« Last Edit: May 01, 2021, 08:37:35 am by 2N3055 »
 

Offline horvat.kamca

  • Contributor
  • Posts: 13
  • Country: cz
Re: µA Meter Sensitivity Amp
« Reply #3 on: May 01, 2021, 08:42:11 am »
To measure such small currents, you need precision autozero opamp. For inspiration how to convert small current to DC voltage look at uCurrent schematics. https://www.eevblog.com/files/uCurrentRev5schematic.pdf .
 

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #4 on: May 01, 2021, 08:49:51 am »
It's a German design for a stand alone test kit from wayyyyyy back in the hollow state days so it naturally has to be well designed (over complicated).
I don't want to have to utilize a separate digital meter and it has to have an analog meter for those needle movements the thickness of a bees dick.
It has a few ranges and configurations and is high impedance with the meter in a self contained box.

So an amplifier that is linear to upscale the fly farts to the new low impedance meter is what it will have to be.
Still awake :palm: so if I sound loopy, I am.

Voltage does not flow, nor does voltage go.
 

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #5 on: May 01, 2021, 08:58:04 am »
Yes thanks, but the need was last week and I don't know if it would do a 0 center for output. I'm looking for something simple, doesn't have to have 1% accuracy and can be whipped up from the junk box or standard off MY shelf components. I could even use a rail splitter https://www.ti.com/lit/gpn/tle2426 for the + and - of center if needed to keep it single supply. I just need some ideas.
This no sleep thing has left me unable to see the forest beyond the trees.
2 weeks of this now and I wish I could get back to normal.


Quote from: horvat.kamca on Today at 17:42:11
To measure such small currents, you need precision autozero opamp. For inspiration how to convert small current to DC voltage look at uCurrent schematics.
https://www.eevblog.com/files/uCurrentRev5schematic.pdf .
« Last Edit: May 01, 2021, 09:03:39 am by Quarlo Klobrigney »
Voltage does not flow, nor does voltage go.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14071
  • Country: de
Re: µA Meter Sensitivity Amp
« Reply #6 on: May 01, 2021, 09:47:29 am »
There is no need for an AZ OP, as the voltage is not small and the accuracy of a simple analog movement is quite limited. A simple low power CMOS OP is enough. Plenty of cheap types to choose from (e.g. MCP6041, MCP6401,...).  However chances are it would be just changing a few of the resistos - if the original resistors are bad amplification would not help, as dirft would likely continue.

The DMM suggestion was more for a test to see what current is really needed.

A circuit drawing of photo would really help.
 
The following users thanked this post: 2N3055

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #7 on: May 02, 2021, 02:16:07 am »
Here it is in all of its complexity, just multiplied x5 for 5 different ranges....
Edited because inline images are not working.

Quote from: Kleinstein on Yesterday at 18:47:29
What is that? 
There is no need for an AZ OP,

A circuit drawing of photo would really help
Voltage does not flow, nor does voltage go.
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 7934
  • Country: us
  • Retired, now restoring antique test equipment
Re: µA Meter Sensitivity Amp
« Reply #8 on: May 02, 2021, 03:03:06 am »
The DC resistance of M2 in your drawing will be much less than the 1 megohm resistor R2, so it is almost equivalent to R1 in series with M2.  If it really has a full-scale of +/- 1 V, then the movement would have been +/- 200 uA.  Unfortunately, good moving-coil meters are no longer common, so your only option is probably eBay.  I'm confused about inconsistencies in your posts and this diagram as to the resistor values. 
« Last Edit: May 02, 2021, 03:06:06 am by TimFox »
 

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #9 on: May 02, 2021, 05:57:29 pm »
Um... what happened to the last 2 posts that were in my email notifications?
I checked earlier and the website was offline for repair.
Did Kleinstein & 2N3055 remove their respective posts? Or did the the "repair" remove them?
Please repost if you did not remove them yourselves. Thanks.
Voltage does not flow, nor does voltage go.
 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 6445
  • Country: hr
Re: µA Meter Sensitivity Amp
« Reply #10 on: May 02, 2021, 09:02:24 pm »
Quarlo,
it got lost in repair..
I don't have a backup so here it is from memory..

Basically, Like Kleinstein said, I meant to use DMM to verify current not as replacement. Which I saw you did. In the end, it was just verification, because 5Mohm resistor cannot lie...Your ballpark figures were right.

As for amplifier, you can leave all dividers as they are, add a resistor that would act as instrument coil replacement, measure voltage on that and amplify. Also you can just leave dividers as they are and use high impedance amplifier ( to get less division) and amplify that.  Or you could put in transimpedance amp and convert current to voltage.

As for driving moving coil instrument (d'Arsonval) it is best driven with constant current.
Then can even put instrument inside diode bridge (constant current source will linearize diodes) and get absolute value meter (and AC) but that is not needed for centered scale instrument.

For examples how it's done search for "FET voltmeter" and "opamp microampermeter"...

 

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 6445
  • Country: hr
Re: µA Meter Sensitivity Amp
« Reply #11 on: May 02, 2021, 09:05:21 pm »
The DC resistance of M2 in your drawing will be much less than the 1 megohm resistor R2, so it is almost equivalent to R1 in series with M2.  If it really has a full-scale of +/- 1 V, then the movement would have been +/- 200 uA.  Unfortunately, good moving-coil meters are no longer common, so your only option is probably eBay.  I'm confused about inconsistencies in your posts and this diagram as to the resistor values.

That would actually be 200 nA...  (200e-9) ^-^

But instrument had 2MOhm/V, so 0,5uA it is....
« Last Edit: May 03, 2021, 07:23:31 am by 2N3055 »
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14071
  • Country: de
Re: µA Meter Sensitivity Amp
« Reply #12 on: May 03, 2021, 07:10:20 am »
I also don't have a copy of my reply, so here a new version:

Form the shown circuit the meter is really low current. Even in the old type that low current was rare.

One can bulid a simple amplifier with 1 OP to drive the current through the meter proportional to the input voltage. The circuit is relatively simple and does not need an AZ OP, as the voltages are not that small. Something like MCP6401 or MCP6041 are good enough, with many possible alternatives (finding a type in not to tiny a case may be the main difficulty).

The only slightly tricky part in the circuit in the attachment is the capacitor. The OP may need it to avoid oscillation, as most meters are inductive. Some movements tend to swing and settle slow when driven with a high impedance and may want a rather large cap, like some 100 µF to dampen the movement and mechanical resonance. Just for the OP some 100 nF should be plenty. A very slow OP like the MCP6041 could get away without the cap, but it would not hurt.
For protection it may be good to add some resistor in series to the meter, so the OP can no drive more than about 5 x the nominal current.

The OP would need a dual supply - with some types 2 x AA or similar (2x1.2 V) can be sufficient.
 
The following users thanked this post: 2N3055

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #13 on: May 11, 2021, 08:52:25 pm »
I managed to get someone to take this design, and simulate it in an :-// program.
The thing won't read as a x10 amplifier. Also the MCP6004 model won't even work.

I've yet to build the hardware version because of benchtop space and also finding the MCP6004's I bought 2 months ago.
Notice the TLE2426 is outputting 4.57V, hardly "rail splitting"
I suspect the models being at fault.

The meter I have reads 50µA each side of center zero. It measures 2077Ω, and takes 107mV to reach the 50µA mark at either polarity, making it a 51.516610496 µA meter, x2 = 103.033220992 µA full scale. So much for accuracy.

Is there any intrepid individual that can either help me in the design and/or model this in whatever is available to determine if the circuit will "multiply" ±1mV to an output of ±50mV with the MCP6004? I'm not having any luck at all.
« Last Edit: May 12, 2021, 05:45:05 am by Quarlo Klobrigney »
Voltage does not flow, nor does voltage go.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14071
  • Country: de
Re: µA Meter Sensitivity Amp
« Reply #14 on: May 12, 2021, 06:17:53 am »
For a x50 Gain one would just have to adjust one of the resistors (e.g. change 10K to 2 K). However the MCP600x are not really suitable for such small voltages. The OPs offset can be up to some 5 mV. So it would likely need a higher grade OP, unless one would add an offset trim to the circuit.

The original circuit should however give more than 1 mV out and thus no real problem with the OPs offset.

The circuit as a voltage amplifier has slightly different properties:  it is slightly more temperature sensitive as the copper coil in the meter will have a temperature dependent resistance (some -3% in the current for a 10 K temperature change). The movements are usually made to have a constant deflection for a constant current, not a constant votlage.  On the other side it does provide daming to the movement because of a low impedance drive, so no need for a large parallel capacitor.

The accuracy of the meter movements is often not that great and one would need an adjustment. AFAIK the accuracy class noted with most meter movements is for the linearity and stability and the initial full scale adjustment is extra.
 

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #15 on: May 12, 2021, 12:37:38 pm »
Yes, but we are back at suggestions for which Op-Amp to use and what values. I was never an Op-Amp guy, choosing to find other ways to do things in the digital world, so I am still clueless.
Voltage does not flow, nor does voltage go.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14071
  • Country: de
Re: µA Meter Sensitivity Amp
« Reply #16 on: May 12, 2021, 01:56:31 pm »
If it is about amplofying a 1 mV signal, the MCP6004 is not good, but the original circuit should not be in the 1 mV range but more like 100 mV. Than the MCP6004 (dual would be 6002) it OK. The 2 nd OP could be used for a virtual ground. So no more need for the TLE2426.

If it is really about very small signal in the 1 mV range, a better OP would be needed, maybe LT1013, OP202, OPA377 or MCP6V66.
 

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #17 on: May 14, 2021, 03:43:42 am »
I knew that I could get rid of the TLE rail splitter. I didn't want to have any interactions with the DUT Op-Amp. BTW the whole thing has to work from a 9 VDC battery.
The only thing I could manage to do was modify a TL082 to the new µA meter and give it a go with a multiturn pot.
What I saw was that the meter or the circuit was not linear on the positive and negative scales. There was a discrepancy of about 10mV between each for full scale deflection and along the way. :-BROKE
The circuit I am using. The full range is ± 10 mV for ±100 mV output. I will try a few different Op-Amps that I have in stock and see what's what. I don't think that it will make a difference. I'm still lost up a creek without a paddle.
« Last Edit: May 14, 2021, 04:02:41 am by Quarlo Klobrigney »
Voltage does not flow, nor does voltage go.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14071
  • Country: de
Re: µA Meter Sensitivity Amp
« Reply #18 on: May 14, 2021, 07:29:21 am »
The capacitor at the output of the OP to ground is not a good idea. OPs have a tendency to oscillate with capacitive load - especially the TL082 and related are known to be a bit sensitive and don't like more than some 200 pF. The simulation may be a little more tolerant in this respect.
There are other more tolerant OPs, but even there capacitance at the output is not a good idea and will bring the OP closer to the edge of oscillation.

9 V supply makes things more difficult, as the OP tend to need more power than. An old low power one is the TL062, but there are better ones.
If they are OK with 9 V supply (and without the capacitor at the output) different OPs should not make a big difference, mainly different supply currents and offsets.
 

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #19 on: May 17, 2021, 07:44:58 pm »
So... after not finding a suitable circuit to do as I need, I decided to look at the µCurrent front end. I only need one range. The µA range. Simulating it with the MCP6004 it performed the best from all of my in stock Op-Amps. The question is still remaining, is that 10K Ω in parallel with 10.01 Ω that I see in the schematic? Here are the muddled and the clarified schematic sections. I don't understand why such a low shunt resistance. :-//
Voltage does not flow, nor does voltage go.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14071
  • Country: de
Re: µA Meter Sensitivity Amp
« Reply #20 on: May 17, 2021, 08:02:12 pm »
The µCurrent is specially made to use a very low shunt resistor.  For the highest current range the low resistance values is kind of a money saving scheme - the amplifier may be cheaper than a larger, higher power shunt. For good accuracy the actual power dissipation should be much smaller (e.g. < 10%) than the nominal rating.
For the other current ranges the amplifier is already there.


For the original problem, there is no need to use a low shunt - just the 1 M resistor from the original circuit is good enough - so a rather high voltage to start with and thus no porblem using a cheap OP (like MCP6004) with quite some offset. The simulation usually does not include the offset. In real life the MCP6004 would not work with the µCurrent circuit (except for the virtual ground).
The MCP6004 would need a reduction in the supply voltage (maybe a white LED in series ?) when using a 9 V supply.
 

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #21 on: May 17, 2021, 08:09:05 pm »
Here is the clone simulation. I'm using a 10K Ω for the shunt. As you remember my target is 16 µA.
The last meter is set to have a resistance of 2077 Ω just as the real one does. No matter what I do I can't get the voltage to be linear with a x10 buffer circuit. I suspect lack of current drive from the 6004. However set as x1, the output is linear and proportional to the current generator. All measurements seem to be .1V low. What say you please. Edited for rushing

« Last Edit: May 17, 2021, 08:12:20 pm by Quarlo Klobrigney »
Voltage does not flow, nor does voltage go.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14071
  • Country: de
Re: µA Meter Sensitivity Amp
« Reply #22 on: May 17, 2021, 08:27:24 pm »
I don't know what the simulation does - it may be just limited resolution for display. It looks like the problem it with the simulation, not the circuit.

In the simulation it is often not a problem, but a real world MCP6004 does not like 9 V supply. Also one can not use 1 OP to regulation the supply to the other 3 in the same chip - there is only 1 pair of supply pins. This ciruit would also need a virtual ground for the amplifier, not just a supply of 1/2 of 9 V. Singel supply the meter could not output a negative voltage.
 

Offline Quarlo KlobrigneyTopic starter

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: pt
  • This Space For Rent
Re: µA Meter Sensitivity Amp
« Reply #23 on: May 19, 2021, 07:29:10 pm »
So I built the same circuit below with the TL084, but firstly the MCP6004 with a 78L05.
Split the 5V to the 6004 and it worked properly.
I wanted more swing so I just swapped out the IC with the 084, removed the 5V regulator and can run it on 9V.
Then came the tweaking to give me the 100mV to span each side of the scale.
The result is a final circuit.
It doesn't have to be a stable over 4 hours to .01%, it just has to show 5 minutes of a quick test. 0-20µA in to 100mV out, for a scale of -50 ► 0 ◄ +50 on the meter.
I can now move on with my life. Once this thing gets off the bench I'll be a happy man.

« Last Edit: May 19, 2021, 07:51:37 pm by Quarlo Klobrigney »
Voltage does not flow, nor does voltage go.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf