Timeline for Op Amp with positive feedback (hysteresis) driving N-MOSFET oscillates
Current License: CC BY-SA 4.0
20 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 29, 2020 at 2:25 | vote | accept | Lt_Flash | ||
| Feb 29, 2020 at 2:25 | vote | accept | Lt_Flash | ||
| Feb 29, 2020 at 2:25 | |||||
| Feb 29, 2020 at 2:21 | vote | accept | Lt_Flash | ||
| Feb 29, 2020 at 2:25 | |||||
| Feb 29, 2020 at 0:50 | comment | added | user57037 | Something rated for 2.5 V and with rail-to-rail outputs. | |
| Feb 28, 2020 at 21:48 | answer | added | Daniele Tampieri | timeline score: 1 | |
| Feb 28, 2020 at 10:02 | history | edited | Rohat Kılıç | CC BY-SA 4.0 |
added 214 characters in body
|
| Feb 28, 2020 at 9:49 | comment | added | Lt_Flash | Could you please suggest a part to be used in this case? I'd much appreciate that as I'm a totally noob with OpAmps and didn't even realize there are special dedicated comparators to be used with feedback hysteresis, actually :( | |
| Feb 28, 2020 at 9:42 | comment | added | user57037 | I would have tried to find a comparator to use in this application, since you are just trying to do rail-to-rail switching anyway. But I suspect you can get it to stop oscillating somehow if you keep at it. | |
| Feb 28, 2020 at 9:37 | comment | added | Lt_Flash | @mkeith thanks a lot, I will try to do that tomorrow and will reply to you asap! Appreciate your help very much! | |
| Feb 28, 2020 at 9:35 | comment | added | user57037 | Most op-amps do not like to drive capacitors. The idea of the resistor is to put it AFTER the feedback loop because if it is inside the feedback loop it will have no effect. I don't want to write a real answer and I can't post a picture in a comment. But break the wire right at the gate of Q2, and put the resistor right there at Q2. One side of the resistor is connected ONLY to Q2's gate. | |
| Feb 28, 2020 at 9:19 | answer | added | user136077 | timeline score: 0 | |
| Feb 28, 2020 at 8:40 | comment | added | Lt_Flash | The whole idea behind this circuit is to use hysteresis to keep Supercap within limits while not dissipating heat continuously with saturated BJTs or MOSFETs, I want OpAmp to turn on (completely) MOSFET when Supercap voltage is above set limit PLUS 1/2 of hysteresis and turn it off when voltage drops below set voltage MINUS 1/2 of hysteresis. This would allow Rs to cool down plus I can use a SOT-23 or similar MOSFET to control the circuit as when it's fully open - its RdsOn is very low (12mOhm in case of DMN1019) and most heat dissipation would occur in Rs that can have multiple resistors. | |
| Feb 28, 2020 at 8:32 | comment | added | Lt_Flash | Thanks, @Neil_UK, I don't want to do this because in this case MOSFET (or BJT) that discharges supercap would stay partially on all the time, thus producing a lot of heat by itself, while in my case, when OpAmp turns on MOSFET and then turns it off forces Rs to dissipate most heat instead of MOSFET. Also, as this cycles are periodical, Rs has some time to cool down so it won't heat up rapidly. If I'm using a 2.2Ohm resistor circuit works just fine with currents up to 1.5-2.0A with 10F supercap and Rs temperature stays at ~65C while MOSFET temperature is much lower (RdsOn is only ~12mOhm) | |
| Feb 28, 2020 at 8:20 | comment | added | Neil_UK | The TLV431 can be used as a comparator, against its internal voltage, why not use that directly? | |
| Feb 28, 2020 at 7:54 | answer | added | D.A.S. | timeline score: 0 | |
| Feb 28, 2020 at 7:33 | comment | added | Lt_Flash | Actually, I might have done that wrong, I've put that resistor between pin 1 of Op Amp and pin 1 of Q2 but you're saying it should be AFTER feedback network, right? Basically, right to the G2 net name and pin 1 of DMN1019? | |
| Feb 28, 2020 at 7:30 | comment | added | Lt_Flash | Thanks, I've tried 10Ohm up to 100Ohm (33, 50, 75) - that didn't help, unfortunately :( Only snubber circuit did some job, I've measured frequency of oscillation before that circuit and it was ~14kHz, after snubber circuit it was working fine until I reduced Rs to ~1.1Ohm and it started to oscillate again at ~7kHz. | |
| Feb 28, 2020 at 7:25 | comment | added | user57037 | Try putting some resistance in series with the MOSFET gate. It should be after the feedback network. Between feedback network and MOSFET gate. Try 10 Ohms and if that doesn't do it move up to 22, 47, 100, 220 or as high as you can go without slowing down the turn-on time to an unacceptable level. | |
| Feb 28, 2020 at 7:20 | review | First posts | |||
| Feb 28, 2020 at 10:37 | |||||
| Feb 28, 2020 at 7:18 | history | asked | Lt_Flash | CC BY-SA 4.0 |