Author Topic: Latching OverVoltage Protection Circuit?  (Read 3334 times)

0 Members and 1 Guest are viewing this topic.

Offline LooseJunkHaterTopic starter

  • Frequent Contributor
  • **
  • Posts: 514
  • Country: de
Latching OverVoltage Protection Circuit?
« on: February 09, 2023, 11:32:47 pm »
I'm trying to create an adjustable OverVoltage Protection (OVP) circuit which latches during a fault. I've come across many different IC's that do OVP (ex. MAX6495, NCP346, MAX6397, can't remember the others at this point), and the MAX6495 does appear to support latching, but I'd prefer going a more DIY route to learn.

I like the design of [image 1] latching OVP circuit from [this]()

video, with the caveat of having to adjust R1, R3, & R4, depending on the *expected* voltage. How can I adjust the circuit so that the OVP trigger point is adjustable with a simple potentiometer to support a broad range of voltages (say 12v-50v)? I've tried thinking this through in my head such as by using voltage references, op-amps/comparators, but I really can't think of how to do it.
« Last Edit: February 10, 2023, 08:27:31 pm by LooseJunkHater »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22435
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Latching OverVoltage Protection Circuit?
« Reply #1 on: February 10, 2023, 01:18:38 am »
That's basically the idea, but has some rough spots:
1. Threshold depends on temperature, and whatever the trigger current is for the TRIAC.
2. TRIACs are rather inefficient as latching devices (or, SCRs are slightly simpler, but it doesn't really matter either way in this DC application); notice the max. min. holding current is 5mA.  It may stay latched down to quite low currents (even sub 1mA at high temperature?), but you can't be sure from a design standpoint.  Now, this doesn't matter here as shown, with the LEDs providing that holding current -- but it's not something you'd want on a battery-powered device for example.
3. You might want an auto-restart function, or an electronic reset signal rather than the pushbutton.

#1, and your variable threshold, are handily solved with a voltage reference and comparator.  Some can even be used as both in one.

Consider TL431, an "adjustable zener", but considered as a three-terminal device, it's really better called an op-amp with only one (inverting) input, an open-collector (only pulls down) output which is also its VCC, and a freaking massive yet suspiciously accurate 2.50V input offset voltage.  This description encompasses its full behavior: it's an amplifier, a control loop -- your warning that its impedance depends on frequency (whereas a zener is ~instantaneous for all in-circuit purposes, TL431 only gives low impedance at low frequencies), that it can oscillate if put in poorly conditioned circuits, that it mostly behaves as an integrator (as most [dominant pole compensated] op-amps do), etc.

Or put another way, you can treat TL431 as a kind of ideal BJT (A = 'E', REF = 'B', K = 'C'): very low "base" current, stable 2.50V "Vbe", "Vce(sat)" about a diode Vf below 2.5V (because there are internal diodes, one which will pull down REF through K*), and a 30V 100mA operating range.  The one catch is, "Vbe" is within tolerance for Ik > 1mA (or less for other "improved" versions e.g. TLV431).

*Actually the same is true of a BJT, since B-C is a junction just as B-E is.  It's just that, Vce is running out in general (Vce(sat) can be 10s of mV) so you can't really notice it. :P

Adjustment is simply putting a potentiometer to the REF pin, which allows you to make it 2.5V or higher; to go the other way, use a comparator and reduce the V(REF) with a pot, comparing to the input voltage.  (Or divide both to get other ranges, while respecting comparator input voltage range.)

So, instead of a thyristor, a threshold like this (alone or with a comparator as well) can be married to a flip-flop, which can be two transistors back to back (multivibrator) or actual logic (NAND or NOR gates, CD4013 or 74HC74, etc.).  You're after an R-S flip-flop, or a type D using the clear/preset inputs.  And then reset is obvious -- just pulse the other input to the flip-flop.  Which can be a time delay from the output for auto-reset, or a direct input for manual.

There are a few other catches, and of course integrating all that onto an IC helps even more for compactness and low current consumption, but that's the basics you'll want for it.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline LooseJunkHaterTopic starter

  • Frequent Contributor
  • **
  • Posts: 514
  • Country: de
Re: Latching OverVoltage Protection Circuit?
« Reply #2 on: February 10, 2023, 04:30:47 am »
I have like no idea what most of this means lol. I'm still a beginner. Would it be possible to have a rough schematic drawn out?
 

Offline pqass

  • Super Contributor
  • ***
  • Posts: 1087
  • Country: ca
Re: Latching OverVoltage Protection Circuit?
« Reply #3 on: February 10, 2023, 06:28:49 am »
I have like no idea what most of this means lol. I'm still a beginner. Would it be possible to have a rough schematic drawn out?

Basically, the TL431 acts like a switch (conducts) when its REF input goes over 2.5V.  Your (monitored) input voltage is divided through a resistor network (or pot) to select your trip point.  When the TL431 conducts, you can use that as input to a flip-flop clock pin that gates a H on its D pin thru to Q.  Q remains H until you reset the flip-flop.

See attached for a fixed 5.3V OVP circuit.  Adjust the right margin "Voltage" slider to change the input voltage.
Change the two parallel 2.2K resistors for a pot and you can choose a different input voltage to trip on. 
Note the values of the two voltmeters as you cross the trip point.
Not shown, use the flip-flop output to control a pass transistor to remove the input voltage from the rest of your circuit. 
You can substitute the flip-flop shown with a transistorized version. Google for "transistor flip flop".
AND you shouldn't let the REF input go above the cathode voltage (+5V) via 4.3V zener clamp (not shown).

Play with it here: http://www.falstad.com/circuit/circuitjs.html?ctz=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-TIHcsVA654E4MBlzkRpzEwohzi9DYW3RFC0NhMwcBwbCrh7CC-FOatYPPJRji-CdSgABxVIDdXUO1wIQNh9m8D4PidMZbAQisqVSQM-HxEhjiwbkmAzaBJKDGSPQHejNko3j+IwQSOHqZRGLHBMfxKNiUxndMlC1VNMy4xDYDdJFKEIWZzGONyd3LRwGCpD0cACAIDDcPA2FDVMwj86AAs8HhYNC8KGWc2FXPcvTIC8kd5W-JUp0iyggLnbEhl8D1A24XApjMQNGHozjdyYKL-JFIL7xwZkUCNC8nBa2EeHajwpm60InPK2sSWquRiDqz8TNyyc-3BJAYW2OyGv1XyhiEIRHgsXwSUDerbVAqk4DICw3N8ShLE0RdIqU86sACNAuvwDA7rw6ldsORoiF2ODsqYxNfxTRxxxBwp2NW4RwdMkEobBhzmCU2KgsIQIECxvFvM0FtoJrNytDc2YMJ0FgACcRCoVCBl28AtCgEAPEgFgACVqYZgR6Y+JmqFUFa+dcFh3052nOZacA+ZYABzTnuF4enPDYaWWMEfw+FUIxeAVqWmhYAB3dWdZ4eXTdZo3tc1-nYdUVn7iQIkVemFWLGdvw+CgFgyQEWyXZAAJ3edkBCjoAAzABDABXXJsgYXIU3kIWTkpgPef9wO+GNJnODgWW0+d7PM9wFX9ap-hmcgZ32mZ83kDzo2K91pXze9mu0F4SWtE7u2QHIbBkD5qBXBAAA1AB7WOI5lugWCUEAPl1bOnaz3gaZAbIKaj2fy-bzMa5Lpnpjz8v8Gtvgz8P3UOkNi+darvgD4fi27-PivJdZjnF9Xhez4aNemYCyHmGEWRJH4CEPrrQ+SdWbZAoEYc+hwBgfxAEwQQMVWo8EaBMPSHpeBaTgIsOQsInzGBaB4euIdw7R1jqnXizstYaxQbnT+FA9J8A9kg8A7CmgiGoLwkBVMuFdQGFw3WNN4C3zEabK24ipHsJEX3B+iiX4r39sIj2rM1Yd0rs7S+psYG3x0brJurcgA
« Last Edit: February 10, 2023, 07:29:23 am by pqass »
 
The following users thanked this post: T3sl4co1l

Offline ledtester

  • Super Contributor
  • ***
  • Posts: 3838
  • Country: us
Re: Latching OverVoltage Protection Circuit?
« Reply #4 on: February 10, 2023, 07:04:07 pm »
More info in this TI publication:

Using the TL431 for Undervoltage and Overvoltage Detection
https://www.ti.com/lit/an/slva987a/slva987a.pdf
 

Offline LooseJunkHaterTopic starter

  • Frequent Contributor
  • **
  • Posts: 514
  • Country: de
Re: Latching OverVoltage Protection Circuit?
« Reply #5 on: February 10, 2023, 08:24:39 pm »
I'm still very confused by everyone's comments but have been reading into different types of gate circuits (AND, OR, NOR). I'm having a bit of trouble wrapping my head around SR latch and SR flip flop circuits, but I think that's something that will probably allow for this OVP circuit to work?

A while ago I drew up this circuit [image 2] which I think would be a functional OVP protection (and theoretically be adjustable with a simple potentiometer instead of fixed resistors), based on a voltage reference and comparator. I'm trying to imagine how I'd incorporate a latching circuit into it but I have no idea (plus, there may be even simpler OVP circuits that exist?).

Edit: I think this form of latch would be what's needed to incorporate into my circuit (except replacing the switch with the comparator?), but again, I have no idea how to incorporate it.

« Last Edit: February 10, 2023, 08:27:03 pm by LooseJunkHater »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22435
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Latching OverVoltage Protection Circuit?
« Reply #6 on: February 10, 2023, 08:35:26 pm »
The gate signal in this circuit is a logic-level (12V) output; this is compatible with CD4013, 4011, etc. gates.  You might want to pick up a few and play around, see how they work. :)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Online iMo

  • Super Contributor
  • ***
  • Posts: 6312
  • Country: sm
Re: Latching OverVoltage Protection Circuit?
« Reply #7 on: February 10, 2023, 08:42:36 pm »
The output of the comparator has to be "latched" - it means you would need a memory component to remember the state (the "fault" condition). Like the cmos logic chips or a thyristor (scr).
Without the "latch" the comparator may (or will) oscillate - as when you disconnect the load, the voltage at the comparator's input will jump up and so on..
« Last Edit: February 10, 2023, 08:53:23 pm by imo »
Readers discretion is advised..
 

Offline pqass

  • Super Contributor
  • ***
  • Posts: 1087
  • Country: ca
Re: Latching OverVoltage Protection Circuit?
« Reply #8 on: February 10, 2023, 09:35:27 pm »
I'm still very confused by everyone's comments but have been reading into different types of gate circuits (AND, OR, NOR). I'm having a bit of trouble wrapping my head around SR latch and SR flip flop circuits, but I think that's something that will probably allow for this OVP circuit to work?

A while ago I drew up this circuit [image 2] which I think would be a functional OVP protection (and theoretically be adjustable with a simple potentiometer instead of fixed resistors), based on a voltage reference and comparator. I'm trying to imagine how I'd incorporate a latching circuit into it but I have no idea (plus, there may be even simpler OVP circuits that exist?).

Edit: I think this form of latch would be what's needed to incorporate into my circuit (except replacing the switch with the comparator?), but again, I have no idea how to incorporate it.



There's a minor difference between SR latch vs flip-flop but isn't important for your purposes (ask google). Either will do.

The video shows a latch but doesn't have a reset switch. I've redrawn the circuit but added a reset.  HOWEVER, IT WON'T WORK for your purposes since the SET line needs a +5V input to engage it. 

It won't work (in your situation) because LM393 comparators have an open collector output. That is, when -Vin > +Vin the output pin is grounded but when +Vin > -Vin the output pin is left unconnected (Vcc will never be emitted).  So, I've redrawn the latch circuit but inverted the sense of the transistors.  That way, a ground on the SET will engage the latch and a +5V on the RESET will release it. Now the LM393 is compatible with the SET parts.

It's perfectly acceptable to use a comparator with a reference if you find it easier to understand (vs TL431).

Play with it here: http://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWK0CcAWAHAdiwZgRrgGz4YIBMRICk11tCApgLRhgBQALiOeWj2lrkMGAbVoRmaVGVxY0KERiJ8s5ZJDBoVmBFiJFFWsClzg4IACaMAZgEMArgBtO7AE48+Y8ESGCoIGhYcO48Ip78wqJgwgFg8JDsAM4gKFR8Er7eEiCcbg6M7ADmYaLkCOnh5GoBiQBKpeCxUalU4oEMtcgI7A1p3hh+7bT+SMPQPR4tvPzEZV4S8FwgcxEdQgsgrIEyCHI0kPrykCgoW2ATZHuQSoP4uHwsWOa01vbOrg2CQjfr4PpdUZdGA9MAqP64SD8b4rCg8Ky2RwuZhORiWcDA2AcDww-DqGEwjaJEoEiwEsG1dgAdzWGTWkP4iSmXgZEKhASCIRSWlm7JmK3ZOTyBWK4DQvP4MVErPE1L+dNxguWRAqAv4wQlAW20gwvBuaXIYKIWDBZ2YMDAOBQJ3KIkECEwKpeCPeLl6IA1jU95TaHM640mIBVVFZCBQ6hlgWCTI97IetE9uAwjJeIRKieTsehFll3H00vItEGtHjcS2Or1KDAIlwmDYaHUMASmi0lEobH25S88VeiI+ouLK2IIEHeMpHkHPpHFkjevYYPUk9+k9V6jeSM4KLRGPGmndS9Eg5Mjb9dADoXzK0LWav7U5xOoMTVgX6kcSKQQT9D4efQvyhQaMN1H5PR0k2EZ-VqCZkmoH9+WDNY-xFDw8DA-gEP5EZozlVC1lHJUaQwrwEJldgAA9zDKCkwD2HgTECAQQAAZQAUQAFXI8B8BWNAkENCNMAYrw6hY1iOIovRoh-FUkBMJBIn4ESxM4ihpVGPAeLMBTmPY9hBAgbRoUGWD8WMiAMAAHSSGirNYBArJ7BIrMgZyHObKArMLNzmwScBnLlIDAmMwLSIoohRDhYgzl1IT+AAYQASzcABjBwEs4KykgACwAeypAA7KyEsKpIADcEusHKLIs-KqXSrKrLsSxrEsKylPYzK6s4ZKss4zAzhoEdM3iZ5tIASXy0rGDcJJGEyxh8tmqycpsKzURsDKkmSpLUvS6r8oAIwcTa7AO1ErM4KqkgOuaklmzaDoAT32nKAAcFqs5KcqcVFksutxPpygBbV67DcOx-uW47XuOziHgM61Aj0R98R8EBiphzbtpStLNtBtxOCSfb8pyzbsrykqsumwogA
 

Offline LooseJunkHaterTopic starter

  • Frequent Contributor
  • **
  • Posts: 514
  • Country: de
Re: Latching OverVoltage Protection Circuit?
« Reply #9 on: July 05, 2023, 12:10:44 am »
Reviving this thread just for a bit more insight/feedback.

I've redesigned my original circuit a bit, but added a latching function. https://www.falstad.com/circuit/circuitjs.html?ctz=CQAgjCAMB0l3BWcAmWDLMgZgBxmcgCxg4CcSWA7CEujQKYC0YYAUFZCIaQGwiNZkXXuD4ROkdpU7c+PQsLGEF4qKwDuIQVqFhKfbZIBmXQji7nZFnVxCYNi0Y8OsA5iHnhlHlTwlr3KzBvIL8oNQAlcH0faL4zcJlOJH8YBAcrZjDMnAlWMB4hHJkRASEhABN6IwBDAFcAGwAXRgb6CvBwiD9JACcbMptDO0w4VgAHLRwFBNwFESxEtyn5g2nwfCXJvQNdGKytJfcd-gLRTgO8vQVtEU8TiDB4CWgsFlIQCPoAZwBLb6aNQAdgBjegTAbaXL8bSLPLuaGDRFnPKESCPGIJZA4SzmCA4AA63zACCJzFJxOecCJkBpRKezygRMw9KpVPANIcc2sylxan62PMB0F-GhyTGURFjGhUpE-jR4FSUGg6Q4dhxpz4IoSqkkau15m5OqWathQluYiWmm0WRKfFtalNQgdNu0utYJhFMo1+ENizAKuVGAcBvV5jFDl9KHMyEoumxak0UfuMRcUSwkBUpHNmbskHM8s4T0SyvSHSjFYTnkq1XqzVY3y0uajGazQiLIFqDW+4KimUw5gQCcYRH8EAVhBSiRVEMyBWS3mYmzhARoi5YQgQi7lag6-fzjgdNdqjSaDfA2fOF-KRC6nZq3fBmgIKhiIpOkmfCDEl-fMU-djflef58JIUTMJe0oXGAkE7kkiolmkrDgcEChQac26geECpTi86QAPYbJ0RYEKQLzwOIsBwKo5opKwQA

For your better understanding: The pot is used for fine-adjusting the OVP trigger. The switch is used to enable/disable the latch. The whole circuit is powered by a DC40v rail, with the 5V and 10V produced by LDO's. A 25v zener+resistor will power the LDO's. The varying AC voltage is to simply *illustrate* what happens to the circuit when OVP kicks in.   

Essentially I want my circuit to be cut-off from power if the main power supply (the 40v rail) goes above 41v. If the circuit DOES go above 41v, the comparators output will be pulled low, enabling the lower P-FET.

__________

Functionally it seems like my circuit will circuit will work, but is there anything I should change in it to improve stability? Will either the P/N FET oscillate (and is that even something I need to worry about in this circuit?)? Will 3ma be fine for a trimmer pot? I'm thinking of removing the 5v LDO for a 5v resistor voltage divider (from the 10V LDO) instead, but aside from that??? 

Edit: It actually looks like I can remove the diode on the output of the P-Fet without it impacting functionality, so that diode can be disregarded.
« Last Edit: July 05, 2023, 12:18:26 am by LooseJunkHater »
 


Offline MrAl

  • Super Contributor
  • ***
  • Posts: 2019
Re: Latching OverVoltage Protection Circuit?
« Reply #11 on: July 07, 2023, 12:58:40 am »
I'm trying to create an adjustable OverVoltage Protection (OVP) circuit which latches during a fault. I've come across many different IC's that do OVP (ex. MAX6495, NCP346, MAX6397, can't remember the others at this point), and the MAX6495 does appear to support latching, but I'd prefer going a more DIY route to learn.

I like the design of [image 1] latching OVP circuit from [this]()

video, with the caveat of having to adjust R1, R3, & R4, depending on the *expected* voltage. How can I adjust the circuit so that the OVP trigger point is adjustable with a simple potentiometer to support a broad range of voltages (say 12v-50v)? I've tried thinking this through in my head such as by using voltage references, op-amps/comparators, but I really can't think of how to do it.

Hi,

Look up the ua723 voltage regulator.  The app note you want is the foldback current limit regulator.  The foldback mechanism is what holds the output off after a fault had occurred.  It is very informative in this area of circuit protection even though it is a vintage ic now.
 

Offline ledtester

  • Super Contributor
  • ***
  • Posts: 3838
  • Country: us
Re: Latching OverVoltage Protection Circuit?
« Reply #12 on: July 07, 2023, 01:05:09 pm »
You can get more design ideas / advice by looking up "crowbar circuit" - e.g. "overvoltage crowbar circuit". For instance, this discussion has some practical advice for designing SCR-based crowbars:

https://electronics.stackexchange.com/questions/453906/how-to-design-over-voltage-protection-crowbar-circuit

How can I adjust the circuit so that the OVP trigger point is adjustable with a simple potentiometer to support a broad range of voltages (say 12v-50v)? I've tried thinking this through in my head such as by using voltage references, op-amps/comparators, but I really can't think of how to do it.

Perhaps you'd be interested in the crowbar circuit presented in the TL431 datasheet:

https://www.ti.com/lit/ds/symlink/tl431.pdf

(Figure 10-8, page 27)

Essentially the TL431 is just an op amp, voltage reference and transistor (see Figure 9-1 on page 20 of the datasheet).

And here is an idea for using the TL431 crowbar circuit to trigger at high (> 36V) voltages:

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/810375/tl431-overvoltage-crowbar
« Last Edit: July 07, 2023, 01:34:28 pm by ledtester »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf