Looking for ideas to DIY uA Meter (shunt amplifier)

[001]

Hi

What the best way to build amplifier x10 x100 for shunt?

Thanx

[001]

Nobody build some similar unit? 

[TheUnnamedNewbie]

If you want to make a very low current meter, or a meter with very low burden voltage, usually a TIA (transimpedance amplifier) is used.

For a shunt meter - what is wrong with a standard opamp in 10x or 100x feedback?

[001]

For a shunt meter - what is wrong with a standard opamp in 10x or 100x feedback?

I understand main idea of basic opamp usage
But I dont know how to choose values range and opamp type too
Can You link some similar project? Thanx

[Gyro]

If you want to make a very low current meter, or a meter with very low burden voltage, usually a TIA (transimpedance amplifier) is used.

Yes, start with a Picoammeter design and degrade it. 

https://www.eevblog.com/forum/projects/picoammeter-design/

[001]

If you want to make a very low current meter, or a meter with very low burden voltage, usually a TIA (transimpedance amplifier) is used.

Yes, start with a Picoammeter design and degrade it. 

https://www.eevblog.com/forum/projects/picoammeter-design/

the idea is not about femtoamperes but about low burden voltage

[Gyro]

A TIA circuit gives theoretically zero voltage burden. By "degrade it" I meant use a lower value feedback resistor.

If you're happy with low burden voltage across a shunt then copy Dave's uCurrent circuit. That's exactly what it does.

[001]

If you're happy with low burden voltage across a shunt then copy Dave's uCurrent circuit. That's exactly what it does.

Yea
I know about it
but components look too exotical and hum is too high isnt it?

[SiliconWizard]

If you're happy with low burden voltage across a shunt then copy Dave's uCurrent circuit. That's exactly what it does.

but components look too exotical and hum is too high isnt it?

How exotic is too exotic? I guess the parts he used should be available through the usual distributors (Mouser, Digikey, Farnell, Arrow, ... )
Regarding "hum", what do you mean? Output noise? I've found the uCurrent to be slightly noisy for dynamic measurements on a scope, but its accuracy is pretty good when used with a voltmeter.

So, we don't know what your requirements are. What current range? Max. common mode voltage? Max. shunt resistance (thus burden voltage)? Max. error? Bandwidth? (Is it for purely static or dynamic measurements?)

Anyway, you can start looking for "current shunt monitors"/"current sense amplifiers". There are many out there. Maybe you'll find one that fits your requirements and is available from whatever distributor you're going to use. One example would be the INA210: http://www.ti.com/product/ina210/description

[001]

I can`t understood idea and complete schematic http://www.kerrywong.com/2013/01/17/current-adapter-for-low-current-measurement-using-ts1001/
Can You told me how this circuit work?
Where I must put current terminals?


Thanx

[Gyro]

Yes, that's a TIA circuit as previously advised.

The output of the opamp swings until the current through the feedback resistor balances the current coming from the source (at the inverting input).

Google Transimpedance amplifier.

[Kleinstein]

The circuit from that last link is the TIA circuit.  The current is flowing through the resistor in the feedback path of the OP.  So there is no voltage amplification, but the OP works as a buffer and suppresses the burden voltage, so that one can use a larger resistor. The input terminals are the far left of the 1 st. circuit part. However this only the very basic circuit, without protection and it may oscillate with a capacitive source. So the pA meter thread is probably the better starting point - there should be protection parts included.
If just the µA range and no need to worry about pA's the choice of OP is relaxed quite a bit.

The shunt + amplifier circuit is good for high currents like 1 mA and up. Though it still works for lower currents, down to maybe 10 pA resolution.
The TIA circuit can work down to very low currents, but is limited at higher current like more than 1-10 mA, as the amplifier must provide the current and due to self heating at that higher current.

While the shunt circuit kind of needs a low offset (e.g. AZ type) amplifier, the TIA type circuit can get away with a more normal OP, as the offset is relative to the output voltage range, so something like 10 or 100 times less important.

[001]

The circuit from that last link is the TIA circuit.  The current is flowing through the resistor in the feedback path of the OP.  So there is no voltage amplification, but the OP works as a buffer and suppresses the burden voltage, so that one can use a larger resistor. The input terminals are the far left of the 1 st. circuit part. However this only the very basic circuit, without protection and it may oscillate with a capacitive source. So the pA meter thread is probably the better starting point - there should be protection parts included.
If just the µA range and no need to worry about pA's the choice of OP is relaxed quite a bit.

The shunt + amplifier circuit is good for high currents like 1 mA and up. Though it still works for lower currents, down to maybe 10 pA resolution.
The TIA circuit can work down to very low currents, but is limited at higher current like more than 1-10 mA, as the amplifier must provide the current and due to self heating at that higher current.

While the shunt circuit kind of needs a low offset (e.g. AZ type) amplifier, the TIA type circuit can get away with a more normal OP, as the offset is relative to the output voltage range, so something like 10 or 100 times less important.

Thank You!

But I don`t need pA range
I`m interested of 1-100mA 0-100kHz with minimal burden voltage

[Kleinstein]

For the 1 - 100 mA range, a µA meter would not help very much   

How important is minimal burden ? Though difficult it might still be possible to use an TIA at 100 mA,  if minimal burden is really required.

The more normal way would however be a shunt with amplifier in this current range. For 100 kHz this would already be a relatively fast amplifier, probably in 2 stages.

[001]

For the 1 - 100 mA range, a µA meter would not help very much   

How important is minimal burden ? Though difficult it might still be possible to use an TIA at 100 mA,  if minimal burden is really required.

The more normal way would however be a shunt with amplifier in this current range. For 100 kHz this would already be a relatively fast amplifier, probably in 2 stages.

Sorry for misunderstood
Is it possible to keep burden voltage at shunt in regulation range of PS?
For example 5v PS that regulated 4% or 200v regulated 0.5%

[Kleinstein]

With a reasonable amplifier one could about resolve some 1 µV at the shunt. So if a 3.5 digit resolution is needed one would need some 2 mV as the minimum shunt voltage - a little more would make it easier (less critical amplifier)  and more accurate.

If the shunt is part of a power supply, the usual way is to have the voltage regulation to already compensate the drop at the shunt. So the shunt would than be more like chosen for a 100-500 mV  drop depending on the current. In addition the shunt can be part of the voltage regulation circuit.

[001]

If the shunt is part of a power supply, the usual way is to have the voltage regulation to already compensate the drop at the shunt. So the shunt would than be more like chosen for a 100-500 mV  drop depending on the current. In addition the shunt can be part of the voltage regulation circuit.

Yes
It is simple way
But I`m gonna build not a panel meter

With a reasonable amplifier one could about resolve some 1 µV at the shunt. So if a 3.5 digit resolution is needed one would need some 2 mV as the minimum shunt voltage - a little more would make it easier (less critical amplifier)  and more accurate. 

What do You mean?
It will be Ok to clone uCurrent?
Why Dave discontinued it?

[David Hess]

I do not understand the problem.  Either use a current shunt and amplifier or use a transimpedance amplifier to provide zero burden voltage.  Note that current feedback amplifiers make pretty good transimpedance amplifiers because of greater stability which can be problem with capacitive loads.  None of this is particularly difficult.

If using a current shunt, then a chopper stabilized amplifier will give the best performance however current shunts tend to have a lot of thermocouple junctions so a standard precision amplifier may be just as good.

[Kleinstein]

I would not recommend a current feedback amplifier for a TIA. They are nice and fast, but the DC performance usually kind of sucks - high bias current at the negative input and usually poor drift.  To make use of the high speed the feedback resistor is kind of fixed to a relatively low value.

There are ways around the problem with capacitive input at an TIA:  especially an RL parallel combination at the input.

Good shunts are made with a low thermal EMF material (e.g. Manganin)  and usually there are only 2 junctions at the shunt that matter. For low voltage one still has to keep the thermal layout in mind. Todays chopper stabilized OPs are not that expensive anymore - so not real problem using one of these. With a current in the mA range input bias, which can be in the 100pA range is usually not problem anymore.

There is nothing special with Daves µC current. So you are free to build your own version of shunts + amplifier - maybe just not call it µCurrent though.

[001]

Todays chopper stabilized OPs are not that expensive anymore - so not real problem using one of these

Thanx
What OP will be ok?

[Kleinstein]

If one really needs the high BW one would need a rather fast AZ-OPs, so maybe ADA4522 or OPA189 or MCP6V81.  These have quite some bias and also a little offset, that is however at least constant. For a lower BW version one could start with something like AD8551 or MCP6V27.

It also depends a lot on what to do with the result and what supply is available.

[_Wim_]

If one really needs the high BW one would need a rather fast AZ-OPs, so maybe ADA4522 or OPA189 or MCP6V81. 

If the stated 100kHz bandwith is needed also in the 100x mode, you need at least a GBW of 10MHz. The OPA189 seems than the best solution of the above

[chickenHeadKnob]

What do You mean?
It will be Ok to clone uCurrent?
Why Dave discontinued it?

Dave made it open hardware, he explicitly encourages those who want to make there own to do so.

[MiroS]

If you want to buy something that can be a good decision, they are very affordable, listed below are just exapmles , thay can be found cheaper than on  below list:

1. V640: lowest range is 150 nA (so lowest will be 5/10nA) and goes up to 1A
https://allegro.pl/oferta/multimetr-meratronik-v640-meratester-plomba-etui-7717926835

2. U726: lowest range is 1nA so lowest  value will be  ca. 20pA
https://allegro.pl/oferta/przyrzad-electronic-voltmeter-u726-meratronik-7665682869

3. U722: lowest range is 1nA so lowest  value will be ca. 20pA
https://allegro.pl/przyrzad-woltomierz-u722a-meratronik-voltmeter-i7688063688.html

Thay are very old, but usually working well , aligement is very simple, schema availbe online.

Display is analog, but they have also analaog output , for every range it will be 0-1V or somethng like that. That way you can  have digital output as well on every other multimeter capable measuring volts.

[001]

-

Thanx
But I see what burden voltage is high, isnt it?