Hello,
In my project, I need to use a current source delivring 40 outputs of 100mA each for 40 parallel wires. I know I can't find such device in the net (with this huge output number).
First I thought to use a single supply with a current divider, but the problem is, wires could be cutted during working, meaning current in other working wires will arise and exceed 100mA.
Does anyone know a multi-output commercial device ( current source )?
If not, do you think that I can build my own product if I gather 40 device of this ?

https://www.amazon.com/Constant-Current-Power-Supply-100mA/dp/B00E4AOMMO
https://www.amazon.com/Constant-Current-Power-Supply-100mA/dp/B00E4AOMMO
Thanks

 

I'm not great with transistors, but I guess you could make a simple "current mirror" to accomplish this.

 

I'm not great with transistors, but I guess you could make a simple "current mirror" to accomplish this.

Hi pmd34,
Do you think that this approch can lead to a clear, concrete and presice result?
Thanks

 

Hello,
In my project, I need to use a current source delivring 40 outputs of 100mA each for 40 parallel wires. I know I can't find such device in the net (with this huge output number).
First I thought to use a single supply with a current divider, but the problem is, wires could be cutted during working, meaning current in other working wires will arise and exceed 100mA.

Could you give us a better idea of what you are doing? Why do you need the 40 outputs, it’s very unusual. It might help find a better solution if you provided more information about the application and not just this tiny part.

 

You can use 40 of these and drive them with a low DC voltage.


Search term: lm317 led driver

 

Could you give us a better idea of what you are doing? Why do you need the 40 outputs, it’s very unusual. It might help find a better solution if you provided more information about the application and not just this tiny part.

Hi Yaakov,
Thanks for the reply, regarding my project, it's a test to validate electrical conduction in wires, we inject 100mA in every single wire and we mesure voltage drop on each one; The problem is wires are mecanically trained during the test and could be cutted by vibration at any time, that's why it's not a good idea to link wires in series (once a wire is cutted, current would be cancelled in other wires), same for current divider, as I mentioned , current will exceed 100 mA.
Please feel free if you have other questions.

You can use 40 of these and drive them with a low DC voltage.

View attachment 172299
Search term: lm317 led driver

Thank you DickCappels, I will see that now.

 

This will give you a lower dropout V than a LM317 -



https://www.onsemi.com/pub/Collateral/NCP1117LP-D.PDF

Dropout is the min required voltage differential, input to output to keep
regulator in regulation, in this case to keep current source working. Also
a lower dropout allows design to minimize worst case power dissipation.

Regards, Dana.

 

This will give you a lower dropout V than a LM317 -

View attachment 172302

https://www.onsemi.com/pub/Collateral/NCP1117LP-D.PDF

Dropout is the min required voltage differential, input to output to keep
regulator in regulation, in this case to keep current source working. Also
a lower dropout allows design to minimize worst case power dissipation.

Regards, Dana.

Thanks Dana for being here,
Do you mean that the NCP1117LP is more suitable to provide 100mA constant current ?

 

Hi Yaakov,
Thanks for the reply, regarding my project, it's a test to validate electrical conduction in wires, we inject 100mA in every single wire and we mesure voltage drop on each one; The problem is wires are mecanically trained during the test and could be cutted by vibration at any time, that's why it's not a good idea to link wires in series (once a wire is cutted, current would be cancelled in other wires), same for current divider, as I mentioned , current will exceed 100 mA.
Please feel free if you have other questions.

Is it a 40 conductor cable?

 

You might want to look at his sort of thing: https://sdgezheng.en.alibaba.com/pr...spm=a2700.details.pronpeci14.2.3d0c7fa3BeXEkU

Many versions available, at different currents.

You can also get bare board versions, and they can be very cheap. Search for LED drivers.

 

I'm still missing some information on what the precision requirements for the measurements are?

Because if it's a fixed resistance limit, then a solution could be one low voltage supply of 1 to 2 volts and a resistance of 10 to 20 Ohm per measuring channel.

Overall, the above solution will use less than 10 watts, and each reference resistor may be 1/2 or 1/3 watts.

 

This https://www.analog.com/en/products/lt3092.html might be a better choice for a current source. It is reverse voltage and current protected.

To get the voltage up, you might want to add a series resistor. With some extra circuitry, you may want to have an enable for each source, You might be able to use this to detect shorts if that's important.

 

Is it a 40 conductor cable?

no, but the data acquisition system that I use to record mesurement has 40 analog input.

You might want to look at his sort of thing: https://sdgezheng.en.alibaba.com/pr...spm=a2700.details.pronpeci14.2.3d0c7fa3BeXEkU

Many versions available, at different currents.

You can also get bare board versions, and they can be very cheap. Search for LED drivers.

Thank you so much Yaakov, this may help.

 

The LM317, NCP1117 are the lowest cost, the Analog Devices LT1092 more.

What is worst case max supply V to the current sources ?


Regards, Dana.

 

The LM317, NCP1117 are the lowest cost, the Analog Devices LT1092 more.

What is worst case max supply V to the current sources ?


Regards, Dana.

I haven't indications about supply voltage, all what I know is current equal to 100mA, and the resistance wire is of millohm order

 

Hello,
In my project, I need to use a current source delivring 40 outputs of 100mA each for 40 parallel wires. I know I can't find such device in the net (with this huge output number).
First I thought to use a single supply with a current divider, but the problem is, wires could be cutted during working, meaning current in other working wires will arise and exceed 100mA.
Does anyone know a multi-output commercial device ( current source )?
If not, do you think that I can build my own product if I gather 40 device of this ?
Thanks

If the "load" is a piece wire then why can't you use a 5 Volt Power Supply and QTY 40 @ 50 Ohm Resistors = 100ma each?

 

If the "load" is a piece wire then why can't you use a 5 Volt Power Supply and QTY 40 @ 50 Ohm Resistors = 100ma each?

Hi mvas,
Thanks for the idea, but i think this doesn't work beacuse wire resistance is instable due to some crimping parametres, and if I apply voltage source, current may exceed or fall behind 100mA.

 

Hi mvas,
Thanks for the idea, but i think this doesn't work beacuse wire resistance is instable due to some crimping parametres, and if I apply voltage source, current may exceed or fall behind 100mA.

NO, I do not see how the current can ever exceed 100 ma, given the Current Limiting Resistor on EACH Channel

You stated the wire would be in the "milliohm" range.
But, you have never stated ... the working range of x milliohms of each wire.

You have never stated ... how stable the 100ma must current be.
+/-1 % ?

It would take about a 500 miliohm INCREASE to reduce the the 100ma to 99 ma ( only -1% variation )
And you would see that, as a voltage increase across your wire.

 

I haven't indications about supply voltage, all what I know is current equal to 100mA, and the resistance wire is of millohm order

You will need to learn this in order to keep from frying your current source
circuit. both from V breakdown issue and thermal issue.

Regards, Dana.

 

How about something like this? The V1 supply voltage doesn't matter much - you could use a car battery. The transistor will keep about 100mA flowing in the wire based on the value of the V2 voltage and the value of R1. The voltage on top of R1 will be a fairly steady 2.5V as long as the wire is not open or shorted. R1 needs to be rated to I^2•R = 0.01•25 = 0.25W. Use at least a 0.5W rating for safety.

Transistors and resistors are both available in array packages, so you'd need only about 4-5 of each.

This would be a lot easier if you could reduce the test current to 10mA. That would allow using an LED for quick visual feedback in series with the wire and would eliminate the issues of power ratings for the resistor and transistor.

Oh, and I put the voltage-out in the wrong spot. It should be at the bottom of R1, on top the wire.

One more thing, have you considered multiplexing?