Programmable Current Limited Linear Regulator

Thread Starter

joeyd999

Joined Jun 6, 2011
5,234
I'm working on a multiple board test fixture. Each board-under-test (BUT) will be supplied with its own regulated +5VDC from the fixture.

I would like to either current limit or fold-back limit each supply independently (so, if one BUT has a short, the other BUTs will continue to test out normally, and no damage will occur on the bad BUT or the fixture).

Limiting should occur at about 50mA. I'd like the limit to be programmable, either through hardware or, preferably, via software (like SPI or IIC).

Normally, I'd just design & build my own regulator for such purposes. There are lots and lots of new-fangled integrated regulators out there now, and I am wondering if anyone knows of a part that performs such a function natively.

FYI, the fixture will have a 12V bus internally, so LDO is not necessary.
 

ErnieM

Joined Apr 24, 2011
8,377
That looks like a good choice. I like these modern regulators because they have that nice shutdown input, always a handy thing to have.

What can become quite useful is using them on digital input lines where a bad device under test may blowout a simple gate driving the input.

When doing this, since I started this back a couple of decades, was to use the LM723 as my goto device. It is a fairly simple linear regulator with all the juicy internal signals brought out so you can use it from 2 to 37 volts out. It does up to 150 ma by itself, or you can add an external pass device. While quite ancient in design it has a wide flexibility similar to the ubiquitous LM555.

Now a days these simple regulators oft do the trick. I have also used LDO types to act as a switch; the single IC was cheaper than two discrete transistors.
 

AnalogKid

Joined Aug 1, 2013
10,986
I've never been a huge 723 fan, but this application is perfect for it except for that whole programmable current limit part. Digital programming implies a D/A which implies ground-referenced current sensor output voltage, but multiple supplies implies high side current sensing. 50 mA is pretty low for a Hall sensor. Hmmm... Between LTC and Maxim there are a zillion linear regulators, and one of them probably has the shifted current sense. Also, both companies make high side current sense parts, some with shunts built in, that deliver a ground-based output voltage that could drive a current control opamp working with a D/A output voltage.

Other than that, what comes to mind is a 3-opamp regulator: voltage control, differential current sensor, current control. Not hard, but lotsa little parts. If you're not screaming for precision, one LM324 would do all of it.

ak
 

MikeML

Joined Oct 2, 2009
5,444
To convert high-side current to a ground-referenced voltage, do not overlook the parts made by Diodes Inc, typified by the ZXCT series...
 

OBW0549

Joined Mar 2, 2015
3,566
If all you need is 50 mA, I imagine a "Modified Howland Pump" circuit (with an NPN boost transistor on the opamp's output) driven by a DAC or filtered PWM would get you your programmable current limit, with the Howland circuit's output feeding a micropower 5V regulator such as the LP2951.

With load currents less than the Howland circuit's programmed limit, it's output voltage (the input to the LP2951) will simply saturate to the positive rail; once the load current attempts to exceed that value, the output voltage of the current regulator will abruptly drop down and reduce the LP2951's input voltage to maintain the load current at the programmed limit.

The ERROR output of the LP2951 will go low to give you an indication of when you're in current limit because of the resulting drop in the regulator's output voltage.

Worth considering, perhaps...
 

Thread Starter

joeyd999

Joined Jun 6, 2011
5,234
I've never been a huge 723 fan, but this application is perfect for it except for that whole programmable current limit part. Digital programming implies a D/A which implies ground-referenced current sensor output voltage, but multiple supplies implies high side current sensing. 50 mA is pretty low for a Hall sensor. Hmmm... Between LTC and Maxim there are a zillion linear regulators, and one of them probably has the shifted current sense. Also, both companies make high side current sense parts, some with shunts built in, that deliver a ground-based output voltage that could drive a current control opamp working with a D/A output voltage.

Other than that, what comes to mind is a 3-opamp regulator: voltage control, differential current sensor, current control. Not hard, but lotsa little parts. If you're not screaming for precision, one LM324 would do all of it.

ak
I'm trying to avoid building up 8 separate volage sources via discretes. The controller for the fixture will be a rather small hand-held device, so I prefer as much integration as possible. The LT part I linked to above not only does much of what I want, but also supplies a fault signal with which I can provide feedback to the operator. The part uses a fixed resistor to define the upper current limit. This is ok...but a SPI/IIC digital limit control would have been nice.
 

Thread Starter

joeyd999

Joined Jun 6, 2011
5,234
If all you need is 50 mA, I imagine a "Modified Howland Pump" circuit (with an NPN boost transistor on the opamp's output) driven by a DAC or filtered PWM would get you your programmable current limit, with the Howland circuit's output feeding a micropower 5V regulator such as the LP2951.

With load currents less than the Howland circuit's programmed limit, it's output voltage (the input to the LP2951) will simply saturate to the positive rail; once the load current attempts to exceed that value, the output voltage of the current regulator will abruptly drop down and reduce the LP2951's input voltage to maintain the load current at the programmed limit.

The ERROR output of the LP2951 will go low to give you an indication of when you're in current limit because of the resulting drop in the regulator's output voltage.

Worth considering, perhaps...
Thanks. Too many parts.
 

AnalogKid

Joined Aug 1, 2013
10,986
The LT part ... uses a fixed resistor to define the upper current limit. This is ok...but a SPI/IIC digital limit control would have been nice.
Digital pot? Digi-Key search yielded 5K, 8% tolerance, I2C - AD5112
or Intersil X9252 - worse tolerance, but only 2.8K so better adjustability, and 4 in a box.

ak
 

Thread Starter

joeyd999

Joined Jun 6, 2011
5,234
Digital pot? Digi-Key search yielded 5K, 8% tolerance, I2C - AD5112
or Intersil X9252 - worse tolerance, but only 2.8K so better adjustability, and 4 in a box.

ak
Yup. I considered that. Not worth the trouble.

The only reason I wanted digital control of the current limit was to make the fixture 'universal'. But this is not a requirement -- just something I thought to throw in if I could.

I bet, though, that if someone came up with an integrated regulator that had digital control of both output voltage and current limit, it'd be a hit. At least with me.
 

Alec_t

Joined Sep 17, 2013
14,280
For digital control, can't you just generate a PWM signal and smoothe that to give an analogue reference for a current-sense comparator?
 
Top