uCurrent Gold troubleshooting?

[osbock]

I couldn't find anyplace to buy one (which I would still prefer!), so I got some boards from OSHPark, and ordered parts from digikey and mouser, but probably due to my bad soldering, it doesn't work. I've already replaced all the opamps.

Can anyone give me some clues about how to trace where the problem(s) might be? I have a Rigol DS2072 scope.

[ja421]

I know this doesn't help much...

I made 3 of them.  The 1st worked pretty good, and I hastily checked it; and I build a 2nd  and sent it to a friend.... Arogantly, i didn't check it -- and it didn't work.

The one I had worked on all but one range; so unsoldered the parts that were relevant to that circuit.  I tested each part & resoldered. Still no go.  Since I had 3 boards, I just built the 3rd -- it totally didn't work... at all. Worse than the one i sent to my friend.
I tested each part, again, and all seemed ok. 
I then measured the switches... I think that might be where the problem is, the resistance on the switches (the 4 that I had here) all measured different resistance.

Someone suggested that it might be A/C interference, so I put it in  metal project box and it didn't help either.
My friend, who builds boards for a living, striped the board & rebuilt with new parts; still had issues.

I ordered some of the 'non critical' components (those that are not spec'd at 0.1 or better tolerance) with better tolerances to see if that helps, but the switch is still the weak point.  I am just grateful that i have one (out of 3) that works on 2 ranges... but that's not good when you consider it's 2 out 9 ranges that worked....

I never got a reply to my post...
(I found your post because I needed the dead range to work & started searching for solutions...)
Sorry I have no helpful news, but I guess "you're not alone".
I used all the parts as specified...

[LA7SJA]

I couldn't find anyplace to buy one (which I would still prefer!), so I got some boards from OSHPark, and ordered parts from digikey and mouser, but probably due to my bad soldering, it doesn't work. I've already replaced all the opamps.

Why did you replace the opamps?

You have not given any clue on what's working/not working, but if you measure the following I could try to help you:

1.) If measured from the battery negative side to output ground do you measure battery voltage/2?
2.) With power on + short do you measure less than ±1mV on the output?
3.) If output >±1mV then connect your DMM/scope with ground to output ground and check pin 1 on IC1 and IC4 and report back.
4.) If 1 and 2 is ok, power off the uCurrent and measure the resistens on the input jacks:
     in the mA/mV range it should be 0.01 Ohm (10mOhm)
     in the  uA/mV range it should be 10 Ohm
     in the  nA/mV range it should be 10 KOhm

Johan-Fredrik

[TheAmmoniacal]

Post a picture of your board, the higher res the better.

[ja421]

Post a picture of your board, the higher res the better.

Here's mine -- the board limits the image file size, it's the best I could do w/in the size restriction...

[LA7SJA]

Here's mine -- the board limits the image file size, it's the best I could do w/in the size restriction...

What's not working, have you done any measurements at all?

Johan-Fredrik

[ja421]

Thanks Johan-Fredrik.

It all works as expected, except on nano range.  On the nano amp range it shows voltage w/o anything connected to it.
If I move my hands around the input (not touch, just sort of move them around) it generates a lot of ripples when hooked to an analog meter or a scope.
If I hook a test resister up; I get about the value expected, but it's usually off a bit, which I end up either  believing it if it's what I expected or distrust the whole thing since on that nano range, it is putting out volts w/o anything connected.  I built another that no test worked as expected, no real life reading worked out.  I stripped all the parts off and tested what I could; and the were all w/in specs, so replaced opamps s I had no good way to test them at the time; rebuild and still fails.
It's almost like there is a flaw in the board itself, or, more likely to me, the switches themselves play a much  bigger roll in the whole circuit and there is variations in manufacturing that screws it up.
The board that mostly works I have only changed the resistors directly involved in the nano range so as not to do damage to the other two working ranges...

Thanks for your help!

[FrankD]

Did you clean the board thoroughly after you replaced the components?

[ja421]

yes.  Removed all solder & flux residue made sure all traces were good; and no direct shorts between pads that shouldn't have shorts...

[LA7SJA]

4.) If 1 and 2 is ok, power off the uCurrent and measure the resistens on the input jacks:
     in the mA/mV range it should be 0.01 Ohm (10mOhm)
     in the  uA/mV range it should be 10 Ohm
     in the  nA/mV range it should be 10 000 Ohm, is it?

Johan-Fredrik

[ja421]

     in the  nA/mV range it should be 10 000 Ohm, is it?

I do not recall the value (and am not home to test it) -- but I recall it being off by about 5% with a resistor test, with nothing attached it read from 2 to 7mv, or more if I put my finger close to the terminals.  Anything attached to it w/o power or an open circuit would read up to 20mv.  I ruled out interference by a/c lines & computers.
And that was the 'good one' the others were much worse.

[LA7SJA]

Genuine MAX4239AUT op-amp?
I read 0,03mV with nothing connected and 0,1 in On+Short setting.
I also read mV if I move my finger close to the positive input terminal and the reading goes down when I put my finger on the the ground terminal.
Have you checked the output on a scope?

Johan-Fredrik

[ja421]

Genuine MAX4239AUT op-amp?

Yes.
I even went so far as to replace all parts (that I could) with closer tolerances... (instead of 1% I would find 0.1%) in hopes that it might help

I read 0,03mV with nothing connected and 0,1 in On+Short setting.
I also read mV if I move my finger close to the positive input terminal and the reading goes down when I put my finger on the the ground terminal.
[/qoute]
Well; not as bad as me, but it's still off.  Should show 0 on short position...   This leads me back to thinking it's the switch is the problem, need to find a better switch.

Have you checked the output on a scope?

Sort of.  Attached is an image that is similar to mine.
The yellow lines are parts I deviated from (supplier part # wise) and blue are the part #'s I originally used.

[LA7SJA]

If I read your "sort off" scope image it look's like U1 is oscillating (or the virtual ground op-amp U2  is oscillating), can you check from battery ground to pin 4 on U2? If you connect a 1.5V battery with very short cables tru a 1.5MOhm resistor to the input in nA range what do you read on the output (if you stay away from the circuit)? (0,993377V)

Johan-Fredrik

[ludzinc]

Thanks Johan-Fredrik.

It all works as expected, except on nano range.  On the nano amp range it shows voltage w/o anything connected to it.
If I move my hands around the input (not touch, just sort of move them around) it generates a lot of ripples when hooked to an analog meter or a scope.

Not trying to be a smart ass, but what behavior were you expecting?  You have a high inpedance input (10k shunt resistor) and a system gain of 100.  You are going to pick up stray noise!

Short the input - the output (all else being correct) should read zero 0V in that case.

If I hook a test resister up; I get about the value expected, but it's usually off a bit, which I end up either  believing it if it's what I expected or distrust the whole thing since on that nano range, it is putting out volts w/o anything connected. 

How have you hooked up a test resistor?  Setting up a nano-amp load is quite difficult to do correctly... 

[ja421]

Thanks @ludzinc.. sorry for the late reply.
No offense taken -- the doc says w/o anything connected, it will read 0.  Tested the uCurrent per manual; a test resistor (actually a series of 5 of 1% resistors) yield 120-160 depending on where my fingers are. A dead short does show 0.
All the other ranges are spot on & do not waver.
And this is the only working one.  The rest (2 physical units, one completely rebuilt on same PCB) are off on all ranges.
Tried mounting one in a metal box to try to insulate out stray signals (didn't help, didn't make it worse either).
Thanks.