µCurrent GOLD on Kickstarter

[PedroDaGr8]

Nothing interesting in it except for the whole "Average Joe can't test that crazy Aussie bloke's claims of accuracy"-part.

Correct me if I am wrong but I think there were trying to say that while its entirely open source that if you build it you don't have the ability to check if the device you built is as accurate as the ones that dave is making.

[EEVblog]

Oh NO! Go for Gold Dave. If you try to match the red to the soldermask red your head will explode.

I thought about some different colour solder masks, but I don't really have enough time to spin those to check what they look like before listing them.

[con-f-use]

One could read it that way, but I think what's meant was "to test the design", not their copy. Both is possible,

[alexanderhiam]

Kickstarter is refusing to accept my US credit/debit card, even after calling my bank and having them "open the card" for overseas payments through the funding period. Is anyone else having this problem?? Am I going to have to live with just my plain old non-gold uCurrent?!?!

[Taucher]

@µCurrent as µVoltage
I'm considering to misuse the µCurrent as generic small signal preamp ...(It's simply a nuisance that so often when one wants to use some precision sensor it needs a preamp to be useable)... maybe by bodging in a switch instead of R12 (270R) in front of the opamp...  But looking at the board-image I really worry about the tiny pads down there.....
In short: I'd love to see a nice, small, affordable precision preamp-board be spun - considering your outreach into the community the volume should be enough to bring down the initial costs to affordable levels... and by adding some general filtering capabilities such a product would surely simplify many hobbyists and engineers lifes ... µVoltage µPreAmp µFilter µAmp µBBB (bread board breakout)..

@Battery/Consumption:
How much current does the µCurrent Gold take from it's coin-cell while running?
50h+ runtime on ~225 mAh ... something around 4.5 mA? (or did I forget to consider the low V dropout point)

I've got a few CR123A 3V Li batteries here... like used in modern LED flashlights... 1600 mAh + ... is there enough room in the case to fit one of those there? (That would easily give some 14 days of continuous use and even be cheaper on the money/mAh scale) 

[Taucher]

Kickstarter is refusing to accept my US credit/debit card, even after calling my bank and having them "open the card" for overseas payments through the funding period. Is anyone else having this problem?? Am I going to have to live with just my plain old non-gold uCurrent?!?!

JS filter (Noscript)?
I had to allow a sh**load of JS sites for the CC# verification step...

[alexanderhiam]

Kickstarter is refusing to accept my US credit/debit card, even after calling my bank and having them "open the card" for overseas payments through the funding period. Is anyone else having this problem?? Am I going to have to live with just my plain old non-gold uCurrent?!?!

JS filter (Noscript)?
I had to allow a sh**load of JS sites for the CC# verification step...

No, I'm not blocking anything, I've tried on Chrome and Firefox. it seems to make it through trying to authorize because it tells me that it's declined  (and yes, the card does work elsewhere)

[Taucher]

Kickstarter is refusing to accept my US credit/debit card, even after calling my bank and having them "open the card" for overseas payments through the funding period. Is anyone else having this problem?? Am I going to have to live with just my plain old non-gold uCurrent?!?!

JS filter (Noscript)?
I had to allow a sh**load of JS sites for the CC# verification step...

No, I'm not blocking anything, I've tried on Chrome and Firefox. it seems to make it through trying to authorize because it tells me that it's declined  (and yes, the card does work elsewhere)

Hm, maybe you did try ordering 2500 AUD version and hit your CC-Limit? (just joking) ...
Well, my German Visa CC went OK (some hours ago) so my second best guess is that you're not really from the US but from North Korea, Afghanistan, Iran or Syria and the NSA did put your (or a similar) name on a top secret no-buy-stuff-from-aussie-guys list... 
... but my only suggestion left is that you get in touch with Kickstarter's support on that issue - they probably know best what's happening.

PS: Backers went from 609 to 611 while typing this ... so I guess the problem does not affect everybody

[EEVblog]

In short: I'd love to see a nice, small, affordable precision preamp-board be spun - considering your outreach into the community the volume should be enough to bring down the initial costs to affordable levels... and by adding some general filtering capabilities such a product would surely simplify many hobbyists and engineers lifes ... µVoltage µPreAmp µFilter µAmp µBBB (bread board breakout)..

I do have these in mind.

How much current does the µCurrent Gold take from it's coin-cell while running?

2.5mA or something like that. I'm using a new higher efficiency led.

I've got a few CR123A 3V Li batteries here... like used in modern LED flashlights... 1600 mAh + ... is there enough room in the case to fit one of those there? (That would easily give some 14 days of continuous use and even be cheaper on the money/mAh scale) 

Not sure, don't have one to hand to test it.

[Taucher]

-snip-...µVoltage µPreAmp µFilter µAmp µBBB (bread board breakout)..

I do have these in mind.

*where's the "in love" smiley?* 
I'd be glad to contribute once your µCurrent stress settles

CR123A 3V

Not sure, don't have one to hand to test it.

Battery dimensions: 16,5mm x 33,6mm + holder
= 18 mm above PCB, 42mm length

Example: http://www.memoryprotectiondevices.com/datasheets/BH2-3A-SM-datasheet.pdf

[ivaylo]

Kickstarter is refusing to accept my US credit/debit card

Kickstarter does not charge the card now, only after the project is funded. They may be doing an Auth or some sort of fraud check now but I dubt it. Even more being an US company dealing with an US CC. I'd call them.

[Dr. Frank]

Hello Dave,

did you already publish the schematic and BOM of the new design somewhere?
You stated, that it's an open source project, I think...

Your article for the old version led to some confusion also on the German heise online site (Heise also publishes the - here - famous c't magazine).

The worse specification from the old article is assumed valid  for  the new design also.

Publishing the correct / updated informations would perhaps clarify everything.

Thanks & Regards Frank

PS: Just found out, that you were using a CSM2512 metal foil precision shunt, from Vishay Precision. (about 6AU$ @ 1000EA)
Wow, that's really a big improvement over the old design. Quite sophisticated.

But I'm still curious, which ChopAmp you've used, or which circuitry (u1/U4), to  achieve that 300kHz BW @ gain = 100...
Must be a 30MHz UGBW type.

[Psi]

Would be tempting to spray the case with conductive paint and then electroplate gold onto that 
Sadly the plastic is probably too flexible.

[EEVblog]

did you already publish the schematic and BOM of the new design somewhere?

The info is on my project page:
http://www.eevblog.com/projects/ucurrent/

Your article for the old version led to some confusion also on the German heise online site (Heise also publishes the - here - famous c't magazine).
The worse specification from the old article is assumed valid  for  the new design also.
Publishing the correct / updated informations would perhaps clarify everything.

I have no plans to update the article, as the point is it's the\ original silicon chip article. The new specs are pretty clear on the project page and Kickstarter.

PS: Just found out, that you were using a CSM2512 metal foil precision shunt, from Vishay Precision. (about 6AU$ @ 1000EA)
Wow, that's really a big improvement over the old design. Quite sophisticated.

Yes, not cheap.

But I'm still curious, which ChopAmp you've used, or which circuitry, to  achieve that 300kHz BW @ gain = 100...
Must be a 30MHz UGBW type.

See the schematic. Cascaded MAX4239's.

[Kean]

I was going to suggest (if it isn't too late for a board rev) having optional mounting pads for SMD fuse holders like Mike suggested (e.g. Littelfuse NANO2).
Would be great upgrade that people could do by cutting a couple of (otherwise fusible) traces and mounting these between R1 & SWA-1A, and between R1 & R9.
BTW, is R9 connected to the correct terminal of R1?  Isn't that putting the uA range through the Kelvin sense terminal?

[mikeselectricstuff]

What is the purpose of the Short position while switched on ? 
Would it not make more sense to have the short in place when in "off" mode (as well)?

Might putting a cap across R7 make it  a little less sensitive to noise in the virtual ground ?

[EEVblog]

Alternatively, if the aim is to protect an expensive/hard to get precision shunt resistor, maybe add a small sacrificial low-value 1206/0805  resistor to protect it - cheaper than a fuse, and easy enough to replace.

For the beefier mA shunt instead of my fusible track, you'd have to use a smaller size for it to blow first. So say 0603 which a typical 0 ohm jumper version has a typical 1A rated current at up to 50mOhms. So possibly up to 5 times higher burden voltage than the shunt.
http://www.yageo.com/documents/recent/PYu-RC0603_51_RoHS_L_4.pdf
0805 is better on max current, but same "<50miliohms" specs.
http://media.digikey.com/pdf/Data%20Sheets/Yageo%20PDFs/RC0805%20Pb%20Free.pdf

Not really an option for the uA range either which uses an 0805 10ohm resistor. So if you lowed the fusible resistor size to 0603 (0402 possibly takes you into different PnP machine territory), then you'd need it to be the same value or more to blow before the precision shunt does. So you've just doubled your burden voltage at least.

So any way with any type of fusing there are trade-offs.
There have been very few reports of blown resistors over the years.

[Dr. Frank]

I was going to suggest (if it isn't too late for a board rev) having optional mounting pads for SMD fuse holders like Mike suggested (e.g. Littelfuse NANO2).
Would be great upgrade that people could do by cutting a couple of (otherwise fusible) traces and mounting these between R1 & SWA-1A, and between R1 & R9.
BTW, is R9 connected to the correct terminal of R1?  Isn't that putting the uA range through the Kelvin sense terminal?

a fuse makes no sense:
1) The ultimate scope of this design is low burden, i.e. now 20mOhm compared to 70 mOhm of the old design.
2) R1 does not need a fuse, normally. It is rated for nominal 10A (1W) , 31A max, see Vishay specification.
But I estimate, it will desolder itself when > 500mW / > 7A are applied.
I doubt that the switch will survive such high currents.

For R1 the connection of R9 makes no difference, because R9 is 1000 times bigger than R1, and the current 1000 times smaller.
R9 should have been connected to IN- directly, but the chosen circuitry is easier concerning the switch and more precise.

Normally, R1+R9 = 10.01 Ohm gives an additional 0.1% error for the µA (10.00 Ohm) shunt, but Dave implied a "trick":
R2 = 10k is in parallel to the 10.01Ohm, thereby reducing the µA shunt again to exactly 10.00Ohm.

Frank   

[EEVblog]

What is the purpose of the Short position while switched on ? 

It allows you to switch in a short so you can change ranges (break-before-make) without disturbing your circuit. It has been a requested feature.

Would it not make more sense to have the short in place when in "off" mode (as well)?

Maybe.

Might putting a cap across R7 make it  a little less sensitive to noise in the virtual ground ?

Noise WRT what? The virtual ground is both the input and output ground reference. The PSRR takes care of any issues WRT the supply and virtual ground.

[EEVblog]

For R1 the connection of R9 makes no difference, because R9 is 1000 times bigger than R1, and the current 1000 times smaller.
R9 should have been connected to IN- directly, but the chosen circuitry is easier concerning the switch and more precise.

No, R9 needs to be were it is to compensate for the fixed R2 value in parallel.

Normally, R1+R9 = 10.01 Ohm gives an additional 0.1% error for the µA (10.00 Ohm) shunt, but Dave implied a trick:
R2  is in parallel to the 10.01Ohm, thereby reducing the µA shunt again to exactly 10.00Ohm.

Correct.
For switching limitation reasons R2 is always connected, so gives an error of 9.99 miliohms lower than nominal R9 value of 10R on the uA range. But we conveniently have R1 which is 10 miliohms - close enough!, so we put R9 and R1 in series which is then parallel with 10K R2 to give us back our precise 10R uA shunt. And this is done via the sense line, which at uA level makes no difference, but that enables another trick at the same time:
In mA mode, R9 is now switched in series with the sense line and opamp, through which no current flows, so no effect.

I was rather chuffed when I came up with this solution 

[EEVblog]

2) R1 does not need a fuse, normally. It is rated for nominal 10A (1W) , 31A max, see Vishay specification.
But I estimate, it will desolder itself when > 500mW / > 7A are applied.

The fusible trace should blow first. Haven't tested it in practice though...

[Dr. Frank]

Correct.
For switching limitation reasons R2 is always connected, so gives an error of 9.99 miliohms lower than nominal R9 value of 10R on the uA range. But we conveniently have R1 which is 10 miliohms - close enough!, so we put R9 and R1 in series which is then parallel with 10K R2 to give us back our precise 10R uA shunt. And this is done via the sense line, which at uA level makes no difference, but that enables another trick at the same time:
In mA mode, R9 is now switched in series with the sense line and opamp, through which no current flows, so no effect.

I was rather chuffed when I came up with this solution 

Well, it's a really nice design. You have really put a lot of "gray cells" into it. Plus some very high-grade components.
Before that, I did not even know low Ohm shunts with <1% tolerance.
The circuitry is really "precision" grade.

As the worst case specification would be 0.25.. 0.3% + 50µV max. error, will you match the resistors or make a precision testing of the handmade devices?

[Kean]

Dave/Frank - thanks.  Yes, clearly a lot of thought has gone into this.  I know what it is like working through design tradeoffs.  And yes, I did notice the always in circuit R2, and the trick there.
Agreed that R1 and R2 would probably be hard to damage in normal use, and looking at the PCB layout, the location of R9 means it isn't at all hard to replace if something stupid ever happens.

[zapta]

did you already publish the schematic and BOM of the new design somewhere?

The info is on my project page:
http://www.eevblog.com/projects/ucurrent/

What is the reason for having two x10 amplifiers instead of or a single x100 ?

[Jay_Diddy_B]

Hi,
The reason that you would want to use two x10 amplifiers instead of one 100x amplifier is to increase the bandwidth.

If an op-amp has a Gain Bandwidth Product, GBW of 6.5 MHz if you have a gain of 10 the bandwidth is 6.5 MHz / 10 = 650 kHz.

If you have a gain of 100 the bandwidth is 6.5 MHz / 100 = 65 kHz.

If you cascade two x10 stages you preserve the high bandwidth. This is important if you have a uP load that is in the sleep mode and comes out of the sleep mode for very short periods.


Jay_Diddy_B