low current limit

On a sunny day (Sat, 14 Jul 2018 11:02:05 GMT) it happened "mook Jonhon" wrote in :

Series PNP from +

• -- e c ----- out b | [ ] Rlimit, beta dependent, trimpot, temperature, weather, RF, ... | ///

It is crude, but... And then there is power dissipation.

30V 20 mA 600 mW

That would work too. Need to figure out how to eliminate unit to unit trimming and I need as much of the 20mA as possible so it needs be simwhat precise.

not enough info really. The simplest way is a resistor, or 2 if you need th em for redundancy. If you don't like the large issues that causes, a transi stor current limit stage would work. If you don't like the small V drop tha t gives, a sense resistor could detect i_out, feed to opamp/comparator, fee d to FET.

NT

Accuracy? Voltage drop?

OnSemi and other have 2-terminal current limiter things, for LEDs mostly.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

He said up to 30 volts.

Rick C.

Supply rail is 30V. up to 20mA it can lose about 1V but not more than

2. in hard short current limit there will be 30V across it.

1960s solution would be two PNPs in the classic Vbe sensed limiter circuit. You'd lose a Vce-sat plus I*Re during normal usage. If you arrange the regulator dc feedback to be after the emergency limiter then that drop can be eliminated. 2n3906 + 33R + 47k + heatsunk 2n2905 would be the 1960s lineup :)

piglet

There're bunches of high side current monitors out there:

Just couple it with a regulator with on/off control input.

How does that work exactly? It would need some sort of memory or when the regulator cuts off the current drops and the regulator comes back on... Be tter might be to use the output to drive the feedback point to limit the cu rrent rather than just shutting it down. Like foldback current limiting.

Rick C.

On Saturday, July 14, 2018 at 5:17:37 PM UTC-4, snipped-for-privacy@gmail.com wro te:

ge

en

t t

e regulator cuts off the current drops and the regulator comes back on... Better might be to use the output to drive the feedback point to limit the current rather than just shutting it down. Like foldback current limiting.

That's right, he doesn't want to use the regulator. So he can use the curre nt sense output to servo a series pass MOSFET 20mA clamp. If he wants redun dancy he can use a second one to just latch the output off permanently.

That was my thought. The 'dropout' voltage from the few I've seen is kinda high.

George H.

The OnSemi limiters behave like (and probably are) depletion fets, ohmic at low voltages and then current limit. But the voltage knee is a couple of volts, so they couldn't be a precise 20 mA current limit at below 1 volt drop each.

The problem is still rather undefined.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

I think this will do what you have requested. I combined the two redundant circuits a bit to minimize the drop out which is normally below 0.75 volts but I believe they both operate separately in most failure modes. Since t he sense resistor is shared, I've drawn it using three sense resistors in s eries which should give adequate redundancy with an increase in current lim it to about 27 mA worse case.

Normal operation starts to limit the current about 16 mA by Q2 clamping lim iting to about 20 mA into a short. If either Q1 or Q2 fails shorted the rem aining device will clamp at about 24 mA. If either of the diode bias circui ts fail shorted the circuit essentially shuts down passing no current. If one fails open that transistor turns on fully and the other will limit at t he slightly higher current.

The diode bias circuits draw less than 2 mA total.

Don't know if this will do just what you want or not, but it's simple and a start.

I started off with a 1N914 and a Schottky diode, but that just didn't seem to provide enough bias for both pass transistors to turn on properly.

I used LTspice XVII for this circuit.

Rick C.

Version 4 SHEET 1 880 680 WIRE -336 -304 -576 -304 WIRE -192 -304 -336 -304 WIRE -176 -304 -192 -304 WIRE -64 -304 -96 -304 WIRE 48 -304 16 -304 WIRE 144 -304 128 -304 WIRE 176 -304 144 -304 WIRE 320 -304 272 -304 WIRE 368 -304 320 -304 WIRE 528 -304 464 -304 WIRE 592 -304 528 -304 WIRE -576 -256 -576 -304 WIRE 592 -240 592 -304 WIRE -192 -144 -192 -304 WIRE -112 -144 -192 -144 WIRE 16 -144 -48 -144 WIRE 128 -144 80 -144 WIRE 224 -144 224 -240 WIRE 224 -144 128 -144 WIRE -576 -128 -576 -176 WIRE 224 -128 224 -144 WIRE 592 -128 592 -160 WIRE 224 -16 224 -48 WIRE -192 32 -192 -144 WIRE -112 32 -192 32 WIRE 16 32 -48 32 WIRE 352 32 80 32 WIRE 416 32 416 -240 WIRE 416 32 352 32 WIRE 416 64 416 32 WIRE 416 176 416 144 FLAG 416 176 0 FLAG -576 -128 0 FLAG 592 -128 0 FLAG -336 -304 V30 FLAG 144 -304 E1 FLAG 320 -304 C1E2 FLAG 128 -144 B1 FLAG 224 -16 0 FLAG 352 32 B2 FLAG 528 -304 C2 SYMBOL pnp 272 -240 M270 WINDOW 0 58 62 VLeft 2 WINDOW 3 96 91 VLeft 2 SYMATTR InstName Q1 SYMATTR Value 2N3906 SYMBOL res -192 -288 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 0 56 VBottom 2 SYMATTR InstName R1 SYMATTR Value 5.6 SYMBOL res 400 48 R0 SYMATTR InstName R6 SYMATTR Value 33k SYMBOL diode 16 -128 R270 WINDOW 0 40 31 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D2 SYMATTR Value 1N914 SYMATTR Description Light Emitting Diode SYMATTR Type led SYMBOL voltage -576 -272 R0 WINDOW 3 11 100 Left 2 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value PULSE(0 30 100ms 1s 1s 1s 4s) SYMBOL res 576 -256 R0 SYMATTR InstName R7 SYMATTR Value 1000 SYMBOL diode -112 -128 R270 WINDOW 0 40 34 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D1 SYMATTR Value 1N914 SYMATTR Description Light Emitting Diode SYMATTR Type led SYMBOL diode 16 48 R270 WINDOW 0 40 31 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D3 SYMATTR Value 1N914 SYMATTR Description Light Emitting Diode SYMATTR Type led SYMBOL diode -112 48 R270 WINDOW 0 40 34 VTop 2 WINDOW 3 0 32 VBottom 2 SYMATTR InstName D4 SYMATTR Value 1N914 SYMATTR Description Light Emitting Diode SYMATTR Type led SYMBOL res 208 -144 R0 SYMATTR InstName R5 SYMATTR Value 33k SYMBOL pnp 464 -240 M270 WINDOW 0 58 62 VLeft 2 WINDOW 3 96 88 VLeft 2 SYMATTR InstName Q2 SYMATTR Value 2N3906 SYMBOL res -80 -288 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 0 56 VBottom 2 SYMATTR InstName R2 SYMATTR Value 5.6 SYMBOL res 32 -288 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 0 56 VBottom 2 SYMATTR InstName R3 SYMATTR Value 5.6 TEXT -426 -64 Left 2 !.tran 4

You posted a link (a while back) to some that had a more precise current limit, but had a dropout voltage of 7(?) V*.

George H.

I have a voltage source that 30VDC. That source will power another device that will draw 0-say 17mA in normal operation and requires over

28V.

In a fault condition the power device can go short circuit for a period of time before it recovers. Under this condition if more than 40mA goes in it will perminantly damage the unit.

The team wants a 2:1 safety factor on that and dual redundant current limits so the peek current into the device when it suddendly goes short is less than 20mA.

Of course during current limt the output voltage will fall to 20mA X load resistance just like a lab powersupply in current limit mode.

Once the fault is cleared, the voltage should return so it cannot latch off like a fuse or non resetable breaker.

hiccup mode is Ok

One version of a current limit is already in place with a current sense amplifier and a series PMOS.

Looking for a somewhat efficient second one with a different technology.

The redundant second technology is in case there is a lot based issue with the PMOS, current sense amp version that allows more than 20mA out. The second technology should not use the same parts or technique bot accomplise the same 20mA current limit.

Belts and suspenders.

Overkill? yes. Bordering silly yes. but still a hard requirement :/

the DMOS version would require I boost the output voltage by the amount of the voltage drop across the DMOS or CLD. These have fairly wide tolerance as well.

Thanks! I'll take a look.

It doesn't meet your separate technology requirement. I think you are sayi ng you don't even want the same parts in the second stage so that if there is a faulty lot of components both trip circuits won't be affected. The 17 mA operation and 20 mA trip current are not so easy to meet with this circ uit. You will need to run more current through the bias diodes by making t he series resistors smaller. Your description this morning is much better, but you don't say how much current you can use to power the circuit. If y ou get that up to say, 5 mA the diodes hold their voltages better and the l imit current tightens up a lot.

Can you get by the dual technology requirement by using two different part numbers or do they have to be different technologies? Does the same rule a pply to the biasing diodes? What about the resistors???

Rick C.

The second should not use the same "technique"? Not sure what that means.

Is the load grounded or can a current sense be used in the ground leg?

Rick C.

I get it. You are working against silly requirements. That always makes a problem harder.

Redundant microprocessors. SSRs. Work up a budget for the team.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics