1

I'm designing a crowbar circuit to to blow a 1A fuse, when the voltage increases above 5V. I'm using an LM431 and using the design in Figure 20 in the LM431 datasheet. I designed and prototyped the following circuit, but found it didn't work as I had thought.

5V Crowbar Circuit

I used R1=39kΩ and R2=33kΩ as a voltage divider, intending to surpass the LM431 reference voltage of 2.5V when the input voltage exceeds 5.4V. When I tested it out, it didn't work as planned. With an input voltage of 5V, I measured the voltage over R2, and found it was measuring 0.8A, instead of the expected 2.3V. I took out the LM431, and found the voltage divider was functioning correctly without it in there.

So I figured my problem was with the value the value of R3. I replaced it with a 10K pot, connecting between one edge and the wiper, and tried finding a value where the voltage over R2 was greater than 2.5V. I found it stayed consistently at 0.8V, until the pot was less than 10Ω, and the voltage jumped to 4.5V, and the LM431 got very hot, before I shut it off.

I'm not sure where to go from here, I've tried replacing R1 and R2 with different values of resistors, both keeping the same ratio, and changing the ratio. As far I could tell, it didn't matter what value resistors I used, the voltage over R2 stayed at 0.8V.

I found another question on here planning on using a similar design, asking what values each of the components should be. The accepted answer has a more complicated circuit, including an extra BJT, and was wondering why that design was suggested over the more simple one from the LM431 datasheet.

Any help on the sizing of R3, or other possible problems in the circuit would be appreciated, I'm not sure what to try next.

UPDATE:

As suggested, I double checked the pinout diagram, and rather embarrassingly, noticed the pins were numbered differently to what I had set up. The pins were numbered 1-3-2, not 1-2-3. Anyway this has fixed the voltage divider issues, and now the voltage over R2 increases as expected until Vin=5.4V.

However, the fuse isn't being blown above 5.5V. Is this due to the vale of R3? I haven't tried varying this value. Does anyone know what this value should be? Or is it a problem with the triac?

UPDATE 2:

I realised that I didn't connect cathode of the LM431 to the gate of the triac when I changed the LM431 to be correct. Now that I've done this, I'm having troubles with my voltage divider not working correctly again; I get a reading of 0.8V over R2, when I expect 2.5. The voltage at the gate of the triac is 0.2V, and I'm using a SC141D triac.

It seems to me that the triac is causing some issues, as the voltage divider worked fine without it. Is there something I need to account for when designing a voltage divider?

GooberZoober
  • 72
  • 7
  • 1
    First thing that springs to mind is, well, that datasheet you've linked to shows the pinout of the TO92 package _from above_. You're sure you've got it the right way round? – peterG May 18 '20 at 10:29
  • re-test with the triac removed, that may have been skewing your readings. – Jasen Слава Україні May 18 '20 at 10:43
  • @Jasen Sorry, I should have mentioned that I was just testing with the LM431. – GooberZoober May 18 '20 at 13:10
  • 1
    @peterG Thanks for getting me to double check the pinout diagram. I finally, and rather embarrassingly, noticed that the pins are numbered 1 3 2, not 1 2 3 as with other ICs I'm used to. This has solved my problem with the voltage divider not working properly up to 5.4V, but now it still doesn't short when the voltage exceeds 5.5V. I'm about to update my question to include this information. – GooberZoober May 18 '20 at 13:11
  • What's the voltage at the triac gate at the point when you're expecting the fuse to blow? – Finbarr May 18 '20 at 14:09
  • @Finbarr voltage at the triac gate is 0.2V – GooberZoober May 19 '20 at 09:14
  • With 4.2V between the gate and the top terminal of the triac and only 0.42mA flowing through R3, there should be more than enough left to turn it on. Looks like something is wrong with the triac or connections to it. Is there any noticeable increase in supply current at this point? – Finbarr May 19 '20 at 13:39
  • @Finbarr I can try getting a new triac and seeing if that helps. Is there anyway to test it? I'm not sure if the supply current increases, as my multimeter doesn't do current, but I know it's not more than an amp, otherwise the fuse would blow – GooberZoober May 19 '20 at 13:46
  • "my multimeter doesn't do current" - it's not much of a multimeter then. – Finbarr May 19 '20 at 13:47
  • And you could always try swapping the two main terminals of the triac. – Finbarr May 19 '20 at 13:50
  • It's a good quality multimeter, if a little old. It's a hand me down Fluke 12. I have another cheapo somewhere, but I don't trust it as much as this one. – GooberZoober May 19 '20 at 13:53
  • @Finbarr I can try giving that a go tomorrow, thanks for the advice – GooberZoober May 19 '20 at 13:54
  • @peterG I bought a new LM431 and triac, and all seems to work with the new components. If you want to make your comment about checking the pinout, as the pins are numbered in an unusual order, I'll accept that as a solution. Thanks – GooberZoober Jun 02 '20 at 12:44
  • If @peterG doesn't want to write an answer thaen maybe the OP can write up an answer and accept it. Just explaining what happened. Just so the question can be closed out satisfactorily from a book-keeping perspective. Nothing wrong with writing your own answer and accepting it after the waiting period. – user57037 Jun 03 '20 at 21:40
  • @gooberzoober You're very kind but I can't in all conscience make such a minor suggestion into an answer - we've all made the same mistake! Glad you got it working. – peterG Jun 03 '20 at 21:00

0 Answers0