Glitching out and mosfet overheating.

More
8 months 1 week ago #10075 by daddyfredregill
Hello all, I am having some strange problem that is making my clock unsable. I have bought and built a classic rev4 clock. I am running this with 12v and a small heatsink on the regulator.

The build was pretty straight forward, I had no problems during the build process. I powered it up for the first time and all seemed normal. The clock ran for 10 min with no problems. I picked up the pcb on the edges with one hand and used the other to feel the mosfet and regulator. When I touched the metal back part of the mosfet the digits on the clock started to glitch out. I let go and it all resumed like normal. I tried it it again this time the digits went completely blank and the mosfet got warm. I unplugged the power source. I tested a few more times with the same result.

I decided to order a new irf840 mosfet, after it arrived I installed it and the problem still continues and this time after the digits went blank the mosfet got super hot and started smoking.

I think the mosfet is ruled out at this point considering i replaced it with a new one.

I am not putting any pressure on the mosfet, it is a super light touch. I am sorry for the long story but I am at a loss. There is something obviously wrong with the way I built it.

Any help would be greatly appreciated.

Please Log in or Create an account to join the conversation.

More
8 months 1 week ago #10078 by Ty_Eeberfest
Are you sure you replaced the MOSFET with an IRF840 and not an IRF840A? Others have reported that the "A" version doesn't work correctly in this circuit.

The metal tab of the MOSFET is "live" (connected internally to drain) so anything touching it may disrupt HV generator operation in unexpected ways. Is the heatsink you mentioned well clear of the MOSFET's tab? If it is touching intermittently there will be problems because the two parts' tabs are both live but at different potentials.

Look into it later when the dust is clearing off the crater.

Please Log in or Create an account to join the conversation.

More
8 months 1 week ago #10080 by daddyfredregill
As far as I can tell the mosfet is an irf840. The regulator is pointed up with a heat sink anf the mosfet is bent at a 90 degree angle. They are not touching.

So is it possible that when i am handling the PCB on one end and then touching the mosfet my body is creating some sort of interference? Like a ground loop or static discharge?

The only time when the mosfet was hot and started to smoke was when all the digits went blank. I am just curious what the mosfet was doing or not doing that would cause the extreme heat? Why would the blank digits make any difference?

Thanks

Please Log in or Create an account to join the conversation.

More
8 months 1 week ago #10081 by Ian
You have to know that the HV circuit is driven by the controller. As long as the controller is running things are more or less OK.

If for some reason the controller stops (for example, the clock circuit is dodgy) the controller will stop wiggling the gate line on the MOSFET. At that point you have two outcomes: Either the gate is off and it's all OK, or the gate is high and the MOSFET is put across the power lines. Two choices: 50% chance of shorting the supply. In either case, HV is no longer produced, and the display will blank.

In your case, you are building this into a PC aren't you? The power supplies on PCs are total beasts.

I don't have a good reason for you touching it causing it to glitch. It sounds a bit like the clock circuit, which is pretty sensitive if it is not correct.

Please Log in or Create an account to join the conversation.

More
8 months 1 week ago #10082 by Ty_Eeberfest
Okay so it's definitely not shorting between the tabs. Shorting between tabs wouldn't really explain the excessively hot MOSFET anyway (on second thought) but would have explained the blank tubes.

Touching things has the potential to cause weird behavior because of your body acting like a conductor and/or capacitor, but I don't think that necessarily applies in this case. I just tried touching and poking my fingers all over a running clock board and nothing interesting happened at all. We need to look further.

Most likely the tubes go blank when the MOSFET heats up because these are both symptoms of the HV generator having a problem. I'll bet that when the MOSFET was smoking the HV at the test point was extremely low.

Here are the things I'm thinking about, tho without scope traces it's damn hard to decide which is right:

MOSFET getting stuck on -> heat and no HV. Cause would be something to do with R7 (missing, wrong value, bad solder) or the traces that connect Pin 1 of the MOSFET, Pin 15 of the controller, and one end of R7 together.

*** At this point Ian's post appeared ***

I think Ian makes a good point about the clock circuit (crystal and 22pF caps on Pins 9 & 10 of the controller). I'll save my other thoughts for later after you've investigated the clock circuit.

Look into it later when the dust is clearing off the crater.

Please Log in or Create an account to join the conversation.

More
8 months 1 week ago #10083 by daddyfredregill
How would i go aboit checking the clock circuit. I have a good multi-meter but no scope.

Please Log in or Create an account to join the conversation.

Moderators: AccutronTy_EeberfestIan
Time to create page: 0.170 seconds

Search

Tube Suppliers

Go to top
JSN Boot template designed by JoomlaShine.com