My uCurrent Gold Mods

[BiOzZ]

Im not sure how many other people moded they uCurrent gold but i thought i would share

i got number 258 ... just 2 off from a perfect number!

FIRST i MUST say that the uCurrent Gold is a GREAT device at a reasonable price and the fact that (the one i got) was hand tested by dave jones him self i cant ask for more than that ... i also got some freebies with it 8D

but i did have some problems ... for one i work with SMPSs without cases and a few other things close to my PC and a laptop and a few other cheep equipment so i live in a fairly noisy environment so when i hooked this to the scope i had some serious problems with quite a bit of noise, i wrapped the case in copper tape, soldered the tape together everywhere i reasonably i could, than put a tab in the case and connected it to the output ground
i ALSO cut some of the traces he used for batch testing ... just because i was in there and had my knife ...

Secondly the coin cell battery is rather useless because of all the space that it has in the case, dave was nice enough to include a 3 AAA battery holder (not sure if he did on all of you ... also some nice gold banana plugs!) so i soldered that battery holder in and glued it down ... the opamp inside can take up to 5v and i went over the circuit hard to make sure nothing was too voltage dependent on the 3v the coin produces ... i decided no and it worked! ... this does not just increase battery life and make them easy to replace but it also gives me a +/- 2.25v fullscale output per range

The binding post was really loose and quickly lost connection ... nothing a good tightening would not fix

Ok this is where i have to be a bit picky .. the amber LED is so damn bright it annoyed the holy hell out of me ... i tried filling it with hotglue and diffusing it with my fingerprints but nothing helped besides taping it over

i thought i would fix this by replacing the resistor going in to the LED with a 1K resistor ... this worked well! ... and no its not close to hitting the input binding post there ... just looks that way, i used to for atleast 3 weeks without problems but after i took this photo (a few minutes ago) i put some heat shrink over it

after all these improvements it worked very well, very well calibrated, very nice gold contacts, LED is now usable, truly usable to full bandwidth, very nice but of kit

Some things that i still dont like:
1: The binding posts are wider than standard ... this does not do any real current handling so i dont mind
2: the switches look really cheap and im curious about there longevity and there contact resistance ... i wont change them out im sure dave did his homework on there but they make be feel a bit uneased
3: i lost a screw, seriously dave you should have anticipated me losing a screw and gone far out of your way to include extras! ... KIDDING ... would have lost the spare also XP ... i had one of the exact type in my junk bin!

Things to come:
I really want to put on a 2000 count voltmeter (will probably make one of my own) and a BNC on a daughterboard for speedy measurements!

[http://i.imgur.com/hOSiNAv.jpg][/URL]

[MLXXXp]

Secondly the coin cell battery is rather useless because of all the space that it has in the case, dave was nice enough to include a 3 AAA battery holder (not sure if he did on all of you ... also some nice gold banana plugs!) so i soldered that battery holder in and glued it down ... the opamp inside can take up to 5v and i went over the circuit hard to make sure nothing was too voltage dependent on the 3v the coin produces ... i decided no and it worked! ... this does not just increase battery life and make them easy to replace but it also gives me a +/- 2.25v fullscale output per range

The problem with this is that the TPS3809L30 voltage supervisor chip will still indicate Batt OK all the way down to about 2.64V (2.58V worst case). You can only trust your uCurrent to have a full scale reading of the originally specified +/- 1.25V, without actually measuring the battery voltage (Unless you're really good and can judge it based on the LED brightness).

To be confident that you can accurately read currents that would output 1.25V to 2.25V, you would have to change the voltage monitor circuit so that the Batt OK LED goes out below 4.5V. As is, your change gives you longer battery life but not reliable extended output range.

[MLXXXp]

2: the switches look really cheap and im curious about there longevity and there contact resistance ... i wont change them out im sure dave did his homework on there but they make be feel a bit uneased

The switches are C&K L203011MS02QE. You can satisfy your curiosity by looking at the datasheet:
http://www.ck-components.com/14386/l_17may12.pdf/

[Christe4nM]

3: i lost a screw, seriously dave you should have anticipated me losing a screw and gone far out of your way to include extras! ... KIDDING ... would have lost the spare also XP ... i had one of the exact type in my junk bin!

Well, who in his right mind opens up an electronic instrument instead of just turning it on??? 

[BiOzZ]

The switches are C&K L203011MS02QE. You can satisfy your curiosity by looking at the datasheet:
http://www.ck-components.com/14386/l_17may12.pdf/

Thank you!

The problem with this is that the TPS3809L30 voltage supervisor chip will still indicate Batt OK all the way down to about 2.64V (2.58V worst case). You can only trust your uCurrent to have a full scale reading of the originally specified +/- 1.25V, without actually measuring the battery voltage (Unless you're really good and can judge it based on the LED brightness).

To be confident that you can accurately read currents that would output 1.25V to 2.25V, you would have to change the voltage monitor circuit so that the Batt OK LED goes out below 4.5V. As is, your change gives you longer battery life but not reliable extended output range.

im not entirely worried about battery indication ... just shove it on a too high range and see what voltage it puts out!

Well, who in his right mind opens up an electronic instrument instead of just turning it on??? 

I dont know, i occasionally have a little voice in my head telling me not to turn it on just take it appart ... i think its a platypus!

[MLXXXp]

im not entirely worried about battery indication ... just shove it on a too high range and see what voltage it puts out!

Be careful about putting too much current through on a lower range. The resulting voltage across the shunt might damage the first op amp. The shunt itself might also get too hot and be damaged. There's no built in protection for overcurrent (except the trace fuse, but that's more of a safety thing. Other components would probably sustain damage by the time it blows).

[BiOzZ]

Be careful about putting too much current through on a lower range. The resulting voltage across the shunt might damage the first op amp. The shunt itself might also get too hot and be damaged. There's no built in protection for overcurrent (except the trace fuse, but that's more of a safety thing. Other components would probably sustain damage by the time it blows).

i use it mostly with 5v, 9v and rarely 12v ... across a 10k resistor thats 2.5mw, 8.1mw and 14mw so im not worried about frying the resistor ... as for a worse case of 12v going in to the opamp i have never had any problems with over saturating an opamp and im sure it has internal protection as i have not fried it yet and normally they go quickly

[MLXXXp]

... as for a worse case of 12v going in to the opamp i have never had any problems with over saturating an opamp and im sure it has internal protection as i have not fried it yet and normally they go quickly

The absolute maximum voltage specification for the MAX4239 op amp's input pins is V_CC+0.3V and V_GND-0.3V. There's no indication of what protection is on the inputs or what the maximum input current allowed is.

The uCurrent has a 270 ohm resistor in series with the +input. If the op amp clamps inputs at 0.3V beyond the rails, at 12V you could have maybe 25mA or more into the input. It's up to you if you feel that's safe.

[jram112]

... as for a worse case of 12v going in to the opamp i have never had any problems with over saturating an opamp and im sure it has internal protection as i have not fried it yet and normally they go quickly

The absolute maximum voltage specification for the MAX4239 op amp's input pins is V_CC+0.3V and V_GND-0.3V. There's no indication of what protection is on the inputs or what the maximum input current allowed is.

The uCurrent has a 270 ohm resistor in series with the +input. If the op amp clamps inputs at 0.3V beyond the rails, at 12V you could have maybe 25mA or more into the input. It's up to you if you feel that's safe.

I just looked at at the MAX4239 datasheet and also saw the Common Mode Input Voltage range of (Vgnd - 0.1) and (Vcc - 1.3). I often need to measure small currents from a 10V - 12V source where the common mode voltage seen on the opamp would be near 10-12V.

Is this a bad idea with the uCurrent?

[MLXXXp]

jram112,
You have nothing to worry about.

It doesn't matter what the operating voltage of your source is. As long as the measured current doesn't go very far beyond the uCurrent's range, you should be fine. The first MAX4239's +input will only see the shunt burden voltage added to the virtual ground voltage.  The second MAX4239 will see the shunt burden voltage times 10 added to the virtual ground voltage.

(Changing the topic a bit)
This does bring up a point, though, that the MAX4239's maximum common mode voltage could be exceeded during normal operation of the uCurrent. This could cause inaccurate measurements when close to the specified 1250mV maximum positive output range. The worst case would be when the battery is at its minimum allowed voltage.
Calculating for this:

The second MAX4239 has a gain of 10. For an output of 1250mV its input would be at 125mV (0.125V). Both input and output are with respect to virtual ground, which is at half the (battery) supply voltage.

The datasheet for the TPS3809L30 voltage monitor says that the threshold voltage, for the BATT OK LED to go out, could be as low as 2.58V. The virtual ground would be 2.58V / 2 = 1.29V.  For a reading at the positive high end of the range, the second MAX4239's input voltage, referenced to battery minus, would be 1.29V + 0.125V = 1.415V.

The MAX4239's specified maximum common mode input voltage is VCC - 1.3V. With a 2.58V supply (VCC) this would be 2.58V - 1.3V = 1.28V. Therefore, the input voltage exceeds the max common mode input voltage by 1.415V - 1.28V = 0.135V.

Note that this is a max/min calculation (and doesn't take into account resistor and other component tolerances which could make things better or worse). Whether this will actually cause problems in the real world depends on many factors and could vary from unit to unit. Another issue is that a 2.58V supply voltage is below the 2.7V minimum specified for the MAX4239. Plus, at low battery voltages the MAX4239 may not even be able to output the specified 1250mV (someone else can show the max/min calculation for this).

[jram112]

Thanks. I understand that the differential voltage seen by the opamp will only be the burden voltage. For the uA range, the burden voltage of 100 uA would be 1mV. That means if I'm measuring high side current from a 10V source, the (+) input would see 10V and the (-) input would see 9.9V for a difference of 1mV.

If I understand correctly, the common mode input voltage would then be close to 10V and far above the datasheet value of  VCC - 1.3V.

[con-f-use]

Did you guys see MLXXXp's uCurrent Gold Mod? Its rather nice as well.

• Change on-short to off-short [no need for it to be powered]
• Lower the LED brightness [less annoying plus less battery consumption]
• Add some input protection when of [diodes across the input when powered off]
• Increase the positive measurement range [virtual ground shifted because negative or ac not measured that often]
• Lower the case [too high for coin cell usage]

I like the idea to shield the case but I'd do it from the inside (looks less botched) and use aluminium tape, because copper looks nasty when it corrodes. Also I'd add reverse polarity protection either by low rds_on mosfet or a Schotky diode.

[MLXXXp]

If I understand correctly, the common mode input voltage would then be close to 10V and far above the datasheet value of  VCC - 1.3V.

You don't measure common mode voltage relative to the ground of your circuit. You measure it relative to the GND pin of the op amp.

The GND pin is connected to battery minus, which is only half the battery voltage lower than virtual ground (VGND). With the battery at 3V, virtual ground will be at 1.5V. For a 1mV input (which is relative to virtual ground), the common mode voltage on the input will be 1.5V + 1mV = 1.501V. With the battery at 3V the maximum common mode voltage allowed is 3V -1.3V = 1.7V.

[MLXXXp]

Did you guys see MLXXXp's uCurrent Gold Mod? Its rather nice as well.

Thanks for the compliment con-f-use.

One might also note that with my mod to increase the positive range, when using either of the two switch settings that shift the virtual ground lower, the potential problem (no pun intended ) of exceeding the MAX4239's maximum input common mode voltage (as I described a few posts back in this thread), is eliminated.

[ricardoJG]

Hello. Will get this little device soon (on its way to me now    )

One question:

If I want to measure nano-amps, should I replace my resistor with a 10K Ohm one like the one in the uCurrent?
I don't need to measure negative currents, so I probably will end up doing the "more voltage" mod.

This is my schematic and will replace the resistor with the uCurrent. In the mean time I'm doing some measurements with resistors and will compare those with the uCurrent.    <-- this little meter is nice.
 

[EEVblog]

2: the switches look really cheap and im curious about there longevity and there contact resistance

One thing the switches aren't is cheap!
Combined, they are by far the most expensive part on the board.
Quality C&K brand.

[mario]

This is my schematic and will replace the resistor with the uCurrent.

Hello.  Don't think that's the correct way to measure the current in your schematic.  Do you really need low side measurements? Why not try high side?   I'm taking an issue with those capacitors connected to the shunt (mainly the Vcap). High side will let you ground those capacitors as they should be.

I could be wrong, hopefully Dave can tell you how he connected the resistor to the Gecko in the kickstarter video. I'm pretty sure he placed the ucurrent  in parallel with the power terminals (can't really see that clearly though).