Definition: The Broadside

[Bigun]

I'd like to offer a definition of a Broadside.

When, like me, you have a design with all your power devices and their drivers are in a nice row across the edge of your heatsink you have the wonderful possibility of screwing up really badly (which happened last night I'm afraid) - where all the devices and a couple of resistors get pooched at once

[singa]

Hi, Could it be that you did not isolate the devices properly with mica sheet or non heatsink grease type and plastic inserts for the screws?

[Bigun]

I used plastic screws and insulating pads. I believe the issue was thermal runaway - I had just increased the rail voltage and didn't adjust the bias back to Class AB - probably had it running a good solid Class A until it melted down.

[ostripper]

Originally Posted by Bigun

I'd like to offer a definition of a Broadside.

When, like me, you have a design with all your power devices and their drivers are in a nice row across the edge of your heatsink you have the wonderful possibility of screwing up really badly (which happened last night I'm afraid) - where all the devices and a couple of resistors get pooched at once

Got "broadsided" once (actually twice).I now hook up the drivers as a small amp first. split Re in half and tie that value X2 to the NFB. In your case , a change in Vbe , you would of seen that as a rise in emitter voltages with your change in rail voltage. Was this on your bootstrap amp ???

My other "broadside" was unavoidable , drank the "grogg" ...slipped with screwdriver , fried the whole freakin' 8 device OPS. . I have become very paranoid with OPS's ... I will even just install 1 pair , let "burn" overnight ... check Re. Finished the next day.

OS

[Steve Eddy]

While I'm sorry to hear of your misfortune, I think your new definition for "broadside" and the accompanying photo is absolutely hilarious.

Thanks for the laugh and may your future efforts be broadside-free!

se

[c2cthomas]

Ahhhhhh............... Once upon a time in another life I was a US Navy Fire Controlman - in many ways I still am.

Fire Controlman’s Creed … Nixon’s Ghost

Oh yeah - I almost forgot - after that I got into audio stuff - guess what....... I still blow stuff up!!

[WuYit]

Hi,
so terribly sorry to hear of your mishap with your DIY equipment. On behalf of the entire staff and management please accept our sincerest commiseration to you, your family and friends at this sad time. We would like to take this opportunity to remind you that our non-profit-making organization's policy remains to provide as much assistance as possible. We are pleased to be able to offer a wide range of information and expertise advice containing much food for thoughts and a word of caution to minimize similar unnecessary collateral damage.
We wish you all the very best for the future.

[ontoaba]

Originally Posted by WuYit

... and expertise advice containing much food for thoughts ...
.

I'd like to buy much food too, than buying transistor, you cannot eat it even you fried it, burn it, or blow it into pieces.

I am also using 4x1N4148 in series connected to collector and base of BC547B then connect proper resistor Rbe for BC around 100~220ohm, then bond all that circuit to heatsink to eliminate thermal dynamic for Hexfet (IRF540/9540).

[effebi]

Originally Posted by Bigun

I used plastic screws and insulating pads. I believe the issue was thermal runaway - I had just increased the rail voltage and didn't adjust the bias back to Class AB - probably had it running a good solid Class A until it melted down.

While I am sincerely very sad for your transistors , I would say that I never liked plastic (I believe Nylon 66, maybe fiberglass reinforced) screws. They never seem to be tight enough and they have the very bad habit to break sometimes under thermal stress. I have seen this several times not on DIY but on expensive CPU boards for Sun servers (that run damn hot, indeed).

[Bigun]

I'm going to be replacing the nylon screws.

Even though I've replaced the output devices and drivers and VAS the thing is still broken.

A bit of searching and I found the emitter resistor on the output device was open circuit - a darn power resistor blown - I was a bit surprised and decided to bust it open and look for evidence of damage. You can see in the photo where the resistive wire melted !

Darn thing still don't work though so a bit more detective work needed yet.