Zen V4 Smoking Q5

[Diomedian]

Hi All,

First of all, I'd like to thank everyone in this forum - I've been reading it for over a year, and everyone who's asked questions, or answered them has been a great help!

So I've finally finished my Zen V4 (pictures to come when it's working again , and since the variac I'd bought on ebay turned out to be faulty, I just had to plug it in and pray. Well... Imagine my surprise when it didn't blow up! Anyway, I checked as many voltages as I could, and they seemed pretty close. 3.8V or so across Q1 and Q2, though there may have been 45V (48V!? I don't quite remember) coming out of Q5... There was a bit of a hum, but I was willing to live with that for a week until I could look at it again.

Anyway, I slowly adjusted R4 until I got 23V coming from Q1. I see now in the circuit diagram that it was supposed to be 22V, but the text later said "half of the regulated voltage plus about two volts". It also said that I should readjust half an hour later, but I didn't. So this worked for about two days, and in that time I'd been listening for several hours. Then on the third day, the amps (dual mono in the same case - separate everthing) had been on for around half an hour, and there was a bit of crackling through the speakers, and Q5 in one channel started smoking! Seconds later, before I could get to the power switches, Q5 in the other channel started smoking as well! That is, it sure looked like it was the Q5s that were smoking - I didn't leave them on long enough to make sure. Even while it was smoking, the music just kept on playing - there was just a mild crackling coming through the speakers. Maybe the amp just doesn't like Spoon.

The heatsinks were pretty warm - they're rated 0.29C/W (Fischer Elecktronic SK56 at 20cmx30cmx4cm) - maybe a bit too hot to touch, but I didn't put a thermometer to them. I used the silicon pads included with the PassDIY Q-packs, and thermal grease on the sink-mounted transistors. ("OCZ Ultra 5+ Ultra High Density Polysynthetic Silver Thermal Compound" - for computer chips.)

Does anyone have any ideas? I did a few searches, and I couldn't find anyone else with similar problems. I haven't turned the amps on again since - should I just replace the Q5s? All the transistors? Can I do any diagnostics before then?

thanks for reading,
Jeremy

 

[Nelson Pass]

If Q5 smoked, then either there is too much current going
through it, or too much voltage across it. You will want to
measure both.

 

[Panelhead]

Heat sinking

Without knowing how much current was running and the voltage drop across Q5 it is hard to say. This transitor should be dropping less than Q1.
One thought, are you using the silicon pads and grease? This may be the problem. Use the grease with mica pads. Use the silicon alone.
Thermal coupling may have killer them. If using a common supply one going down would stress the remaining channel due to increased input voltage.
They should take it fine. Mine run higher voltage drop and lower current than the recommended. Input voltage operating was 57.5 volts. The current 1.25 amps. Q5 was almost as hot as Q1. This works out as around 17 watts disapation.

George

 

[Diomedian]

Thanks for your replies, guys.

I'm a newbie, as you might have gathered... Should I replace the Q5s before trying to take measurements from them again? Also, any guesses as to why they would only smoke after two days?

I'm not quite sure why thermal grease wouldn't go well with silicon pads? Panelhead - both of the channels have completely separate supplies - right down to the IEC connectors.

thanks again,
Jeremy

 

[Panelhead]

Trust me

Use the silicon pads dry. Use grease with the mica.
Make sure it is Q5 that was smoking. Cannot imagine it is not damaged if it got that hot. That smoke was something inside leaking out.

George

 

[PeteM]

Hi

Silver thermal compound is electrically conductive IMO unsuitable where isolation is required (pc processors do not need electrical isolation just good thermal contact)

I found this out after smoking an IRF350 in a power follower that also ran ok for a few days untill the inevitable,the compound must eventually creap creating a short or changes resistance with heat/time.

Peter

 

[Diomedian]

Ah, ok, so you guys were right. Drain of Q5 has a 5 ohm resistance to ground (or 13 ohm on the other channel) when it was mounted on the heatsink, and a very high resistance to ground when removed from the sink.

After looking at the silicon pad, I could see scorch marks right around the hole for the bolt. The bolt is connected right to ground since I'm using a wooden chassis - the sink is the case ground. So this suggests that the thermal grease (despite it being advertised as non-conductive) started conducting from the drain (the metal back of the transistor) to the bolt. Maybe I'll get some nylon screws this time - they're specced to withstand 260 degrees...

I've checked the other four transistors, and they've got no connection to ground through the drain (yet), so it looks like I got lucky. The paste under Q1 however, has already dried up and hardened! I'm going to go see if I can't get my money back for this paste (Cdn$11 or so). I'm trying to clean the paste off of the backs of the Q1 and Q2. I've got new Q5s and six new silicon pads on order. Hopefully this will help things.

Ok, thanks agian for your replies!
Jeremy