10 amp DC?

[CopperCone]

How could I make a stable 10A DC signal.

Is there some kind of transistor I can use with a voltage reference to make one?

[David Hess]

The usual simple way is with a voltage reference, precision low valued resistor, power MOSFET, and operational amplifier.

The questions are what kind of precision do you need, is the load tied to ground, power, or floating, and how much compliance voltage do you require?

[CopperCone]

Its for calibrating clamp meters so all I care about is the current being stable. I'd like 5 digits of stability if possible.

[JohnPi]

A good (e.g. Agilent) power supply with current limit might work for you. However, with 10 A, there will be a significant amount of heating (in the power supply) which will make stability harder. If you could use  an external R (or a zener, but you cant get 10 A zeners -- you'd have to build a circuit) it'd drop most of the voltage and would move the power dissipation outside.


How are you going to measure the stability -- with a DC meter ?

[HalFET]

The main issue you're going to see is that the power resistor will need some time to heat up, so you'll have to take that into account, but otherwise a simple opamp circuit as David Hess suggests ought to do the job.

A fun trick I once saw in a big fat power supply was to put a diode in series with a negative temperature coefficient which was matched to the temp co. of the resistor, and they clamped them together with some thermal compound in between.

[HighVoltage]

Just use a stable 1A source and then loop it 10 times through the clamp meter.

[rs20]

Do clamp meters exist which even come close to 5 digits worth of accuracy/repeatability?

[4CX35000]

How could I make a stable 10A DC signal.

Is there some kind of transistor I can use with a voltage reference to make one?

What voltage are you after ?

[texaspyro]

I have several Valhalla current calibrators / shunts.  You input a low DC voltage and they produce a current proportional to the voltage.  They can be used to produce DC or AC currents.  I think the 2575 goes to 100 amps.  They are basically an op amp / switchable precision current shunt resistor bank.  They are rated at 0.01% or so accuracy.

[JohnPi]

That's the best option. More than 10 turns would be even better.

[David Hess]

I am dubious that a DC current clamp can produce more than 3 significant digits but lets say I wanted to make a best effort at a reasonable cost.

The largest error contribution comes from the temperature coefficient of the resistive current shunt.  With that in mind and besides the obvious like using 4-wire sensing, the current shunt should be power derated and a lower than normal value of resistance should be used.  The increased error from offset voltage can be made up for by using a chopper stabilized amplifier.  As mentioned, the current may be lowered by winding 4 or 5 or even 10 turns through the clamp to produce 10 amps from 2.5, 2, or 1 amp.

That leaves another question.  Is absolute accuracy required or will the calibrator be itself calibrated from another source?  The best 4-wire current shunts have 10 times worse absolute accuracy than 0.1% tolerance resistors.  If no calibration is available, then standard 0.1% tolerance resistors used for the current shunt will produce better absolute accuracy.  Alternatively, low current high absolute current source could be used to calibrate the high current high stability 4-wire current shunt.  This argues for building two current sources with one used to calibrate the other.  Or if you trust your best 4-wire bench multimeter, then it can be used for this calibration by measuring the 4-wire shunt resistance directly.

What kind of budget is available?  Vishay metal foil 4-wire precision current shunts cost about $50.  Chopper stabilized amplifiers and a good enough reference are not cheap either.  I would probably plan on building two designs with the first being a less expensive model using standard 0.1% resistors.  If I was going all out, then I might try using a temperature sensor, peltier cooling, and careful mechanical design to control the temperature of the current shunt.

This challenge is a lot like the last project Bob Pease worked on:

https://www.edn.com/design/analog/4416534/Bob-Pease--A-tribute-to-his-last-challenge---What-s-all-this-voltage-reference-stability-stuff--Part-one
https://www.edn.com/design/analog/4418068/Bob-Pease--His-last-challenge--Part-two
https://www.edn.com/design/analog/4422201/Bob-Pease--His-last-challenge--Part-three-The-precision-resistor