Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Looking for proper constant current circuit

Status
Not open for further replies.

Zak28

Advanced Member level 2
Joined
Aug 19, 2016
Messages
579
Helped
6
Reputation
12
Reaction score
6
Trophy points
18
Activity points
4,681
Is there a simple ~80v constant current source circuit which can source up to ~30mA but constantly changable with the turn of a pot?

Greatly appriciated.
 

Vary bias to a plain transistor through a potentiometer. Constant current goes to the load regardless of load.

pot adjust current PNP 80VDC 0-30mA.png

The pot was dialed to one extreme and back.
 

Hi,

Is temperature an issue regarding a constant-across-temperature-fluctuations current source? I do lots of temperature analyses when simulating bjt circuits to see how much the target voltage/current may change, and it can be depressing to see how e.g. 75ºC isn't the nice, hoped-for result obtained at 25ºC, occasionally often.

The Wikipedia entry for Current source may offer variations on a theme, i.e. with minimal tweaking, the Zener version improved with a temp comp. diode could be replaced with a (TL)431 used as an error amplifier or an op amp constant current source is also simple if the output voltage suits what's being driven (in this case thinking of your 80V supply and a bjt base's requirements).
 

Vary bias to a plain transistor through a potentiometer. Constant current goes to the load regardless of load.

View attachment 149296

The pot was dialed to one extreme and back.

Cannot use a 1meg pot since I cannot find 5-10 turn variants, it wont be accurate enough, highest pot I have is 100k 2w 10 turns which gives great precision. Can this circuit be made to work with a 100k pot?

I also have 50k 10k 2k variants also.

Greatly appreciated.
 

Hi,

The more precise solution is with a bjt or a mosfet controlled by an Opamp.
The Opamp won't see the 80V, thus the voltage is no problem.
Just choose an appropriate transistor.

80V x 30mA gives 2.4W of heat. Use an appropriate heatsink.

Klaus
 

Depletion mosfets lend themselves to a number of good and simple current source topologies. The simplest is this:
https://ixapps.ixys.com/DataSheet/98704.pdf

This outlines a couple more:
https://www.ixys.com/Documents/AppNotes/IXAN0063.pdf

In general these all have nearly identical enhancement (the common type) mosfet counterparts but depletion fets eliminate the necessary pull-up resistor to the 80V rail which is helpful because 80V is beyond the voltage most small resistor packages can withstand.
 

Is there a simple remedy to have this climb to ~30mA?

bla.png

max output

bla.png
 

100k 2w 10 turns which gives great precision. Can this circuit be made to work with a 100k pot?

Since it's rated 2W then it can handle 80V going through it.
Add a series bias resistor. Max current through the pot is 1-2 mA.

pot and 260k bias resis adjust current PNP 80VDC 0-30mA.png

- - - Updated - - -

Pot was dialed from one extreme to the other and back. Response appears proportional.
 

Since it's rated 2W then it can handle 80V going through it.
Add a series bias resistor. Max current through the pot is 1-2 mA.

View attachment 149301

- - - Updated - - -

Pot was dialed from one extreme to the other and back. Response appears proportional.

The mosfet cc source works well but seems to be very susceptible to noise since the oamp has to amplify millivolts to drive the gate of the fet. Im going with your circuit using a mje350 which is very adequate.
 
Last edited:

Be aware mosfet gates are prone to puncture by high voltage. This ruins the device. 80V might do that.
Check datasheets to find safe driving voltage.
 

Seems like your suggested pnp circuit demands a 1k pot not 100k after these results.

Screenshot_20180929_220719.png
 

Hi,

About schematic of post#7:

It seems you didn't do a search for proven circuits.
And it seems you don't care about typical OPAMP circuits and OPAMP operating limits.
Why don't you choose a working circuit and modify it for your needs?

There are many tutorials and design examples in the internet ... and in this forum.
For a high side current source even I posted some kind of tutorial here: https://www.edaboard.com/showthread.php?376094-Design-of-high-side-current-source
Many of the considerations apply for a low side current source, too.

Your circuit issues:
* weak power supply of 20V (voltage depends on supply_load_current
* no bulk capacitor and no fast capacitor at the power supply
* 20V input voltage a noninverting input is beyond input voltage range of usual Opamps
* no stabilisation circuit to prevent oscillations
...

Klaus
 

In regards to the issue with the Pot for the PNP circuit I ran the step simulation with R12 set to 1k not 100k, the circuit does work with 100k pot.

Hi,

About schematic of post#7:

It seems you didn't do a search for proven circuits.
And it seems you don't care about typical OPAMP circuits and OPAMP operating limits.
Why don't you choose a working circuit and modify it for your needs?

There are many tutorials and design examples in the internet ... and in this forum.
For a high side current source even I posted some kind of tutorial here: https://www.edaboard.com/showthread.php?376094-Design-of-high-side-current-source
Many of the considerations apply for a low side current source, too.

Your circuit issues:
* weak power supply of 20V (voltage depends on supply_load_current
* no bulk capacitor and no fast capacitor at the power supply
* 20V input voltage a noninverting input is beyond input voltage range of usual Opamps
* no stabilisation circuit to prevent oscillations
...

Klaus

I could improve and manage the fet cc source however I found oamp amplifies millivolts and its prone to amplifying noise since noise is in that range. I have the working fet cc source made yet I prefer the pnp iteration.

bla.png
 

I ran the step simulation with R12 set to 1k not 100k, the circuit does work with 100k pot.

Yes, it should be feasible to find a way to bias the transistor/ mosfet so it gives you your desired range of response. You can try it with various potentiometer values. You can change location and ohm value of resistors. You can adapt to different operating specs of the transistor/mosfet. Of course since 80V is present, you should watch how much current goes through components.
 

Hi,

and its prone to amplifying noise
The noise is well calculable..and not that high.

Noise reduces precision, because it introduces some uncertainty.
It seems you don't like this uncertainty....thus you use the fet solution.
But the fet solution introduces other errors, like unlinearity and thermal drift. I'd say this uncertainty is much higher than the Opamp circuit uncertainty.

With an Opamp the input referred noise is in the low microvolts RMS.
If you use a shunt of 1k Ohms, then this means low nanoamperes RMS.

Klaus
 

Hi,


The noise is well calculable..and not that high.

Noise reduces precision, because it introduces some uncertainty.
It seems you don't like this uncertainty....thus you use the fet solution.
But the fet solution introduces other errors, like unlinearity and thermal drift. I'd say this uncertainty is much higher than the Opamp circuit uncertainty.

With an Opamp the input referred noise is in the low microvolts RMS.
If you use a shunt of 1k Ohms, then this means low nanoamperes RMS.

Klaus

If you can show how to replace the 400Ω shunt with higher value iteration I will go with FET cc source since BJTs do suffer immensely from thermal issues.

Screenshot_20180930_133501.png

View attachment CC-Source-MOSFET.zip
 
Last edited:

Hi,

Useful shunt value depend on current range.
With 1k and 30mA there will be 30V across the shunt.

The 1k of my post#15 is just an example.
You can use 100 ohms, 10 ohms or even 1 Ohms shunt with the Opamp solution and it will stll be more precise than any solution without Opamp.
V_be will drift with about -2mV/K..... this is huge.

What noise current can you accept?
What thermal current drift can you accept?

First decide your requirements then choose the appropriate circuit.
I don't say the Opamp solution is the best solution in any case, but it has benefits in precision.

Klaus
 

Hi,

Useful shunt value depend on current range.
With 1k and 30mA there will be 30V across the shunt.

The 1k of my post#15 is just an example.
You can use 100 ohms, 10 ohms or even 1 Ohms shunt with the Opamp solution and it will stll be more precise than any solution without Opamp.
V_be will drift with about -2mV/K..... this is huge.

What noise current can you accept?
What thermal current drift can you accept?

Klaus

I am unsure how noise will alter constant current value I intend to have a circuit which will not spew out of bounds by +/- 100uA due to noise/EMI.

The 400Ω shunt might already suffice the spec however I do not know for certain if it is.
 
Last edited:

Hi,

500uA is huge and is not a meaningful vaue caused by usual noise.

Regarding EMI:
This mainly is a question of schematic and PCB layout design.

You say "Opamp is prone to pickup noise". This is a general statement without values.
Example: Opamps are used in audio circuits, even in expensive, high quality ones.
In audio signal you will be able to hear noise in the low microvolts range. The ear is very sensitive.
In audio circuits the "audio current" is in the range of 100uApeak. In quiet parts of music the current will be less than 1uA.
If Opamps were that bad you'd hardly hear music between the noise.

Klaus
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top