Well that stinks. Just curious. Did you eliminate the protection circuitry on your F5 build?
I'm not a great advocate of protection circuitry, but this is the first power amp that I have built that has failed.
The F5 boards don't have built in protection - you have to add it as an extra.
Well, I'm going for a Bolt and Braces repair as I'm not too sure how much damage has been done.
I'm going to replace Q3, Q4. Q5, Q6, R17, R18, R13, R14 and the thermistors. I'll check all the other resistors but they look OK. R11 and R12 look roasted too.
The F5 boards don't have built in protection - you have to add it as an extra.
Well, I'm going for a Bolt and Braces repair as I'm not too sure how much damage has been done.
I'm going to replace Q3, Q4. Q5, Q6, R17, R18, R13, R14 and the thermistors. I'll check all the other resistors but they look OK. R11 and R12 look roasted too.
Have you considered replacing the coils in the drivers. I have had good results on two 10" units and one tweeter. Madisound or PE may have a kit available. I had a Carver M-1.0t bridged to power a sub, so there was a potential for up to 1,000 W. But the repairs worked.
Last edited:
As the original drivers are still available I've ordered two replacements.
I've never been too sure if replacement cones, coils, etc are exactly 100% clones of the originals, if the originals are still available you know that the character of the speaker will not be changed.
I've never been too sure if replacement cones, coils, etc are exactly 100% clones of the originals, if the originals are still available you know that the character of the speaker will not be changed.
Last edited:
K&D,
Any specific reason why the thermistor is so close to the Mosfet's drain "tab", for lack of a better term? Is it possible contact was made there between the thermistor and tab?
Just a thought.
Any specific reason why the thermistor is so close to the Mosfet's drain "tab", for lack of a better term? Is it possible contact was made there between the thermistor and tab?
Just a thought.
It's got to be close for thermal contact but there is no electrical contact. The reason that I've got to replace the thermistors is that they are epoxyed to the MOS-FETs which are being replaced.
The general consensus of Pass amplifiers is that they are so solid that no protection is necessary, that is certainly borne out with my Aleph4.
The F5 seems to be the weakling. If you are building one of these - heed my expensive experience and protect your speakers.
The F5 seems to be the weakling. If you are building one of these - heed my expensive experience and protect your speakers.
I'm referring to the high current fold back in the F5 circuit. I guess that doesn't function during a catastrophic failure.
That's odd. When I had a rail loss on one channel of my F5 it sent a few DC pulses to my speaker, but corrected just fine. I do remember being puzzled by an amplifier that continued to work without any significant issues with one rail missing. The woofer was fine, it was a Usher 8948 (on the $3000 Mini-Dancer One). I cannot remember even one report of the F5 taking out a speaker in the four years the design has been up.
I had a plenty close calls (one involving a L18 and a chipamp gone crazy - full 35VDC on the output for a few seconds, but even in that case the woofer was fine) but IME woofers are far more resilient than tweeters. I would think the crossover is fine. I would be very surprised if something happened there.
I do understand it sucks. Hopefully the repair is not too expensive. A woofer replacement should be relatively simple to do yourself, and B&W should be able to ship you a spare.
One thing though, I let the amp run a couple days without speakers connected. The F5 is quite happy to have its bias and offset set without a connected speaker.
I had a plenty close calls (one involving a L18 and a chipamp gone crazy - full 35VDC on the output for a few seconds, but even in that case the woofer was fine) but IME woofers are far more resilient than tweeters. I would think the crossover is fine. I would be very surprised if something happened there.
I do understand it sucks. Hopefully the repair is not too expensive. A woofer replacement should be relatively simple to do yourself, and B&W should be able to ship you a spare.
One thing though, I let the amp run a couple days without speakers connected. The F5 is quite happy to have its bias and offset set without a connected speaker.
Appears that you're experiencing an unlucky streak. (I'd skip a casino visit for the moment)
Your post may stimulate others to use some form of protection for their loudspeakers, at the least.
(I used to despise SSR's, but current trenchfets make a discrete solid state relay a viable option, imo)
Totally not true, it's a very sturdy design.
It's not fair to blame the design instead of one's inability to build it properly (i.e. to provide good thermal connection between the MOSFET and the heatsink).
There are F5 amps working day and night for more than 5 years without a single problem (but built by apt people).
So before you start being judgemental about the stuff you know you know little of, increase the efforts to improve your own skills - this is not the first time you declare proven design a rubbish ...
My Aleph 4 has been trouble free, a truly robust design. OK I got this one wrong but just be aware of its weaknesses. Why do you keep harping on about a poor thermal connection between the heatsink and the MOS-FET.; That IS NOT the issue. The build was over biased (my mistake)
Last edited:
KandD, Have you done any further "failure analysis"? Have you tested the failed MOSFET with an ohmmeter? Particularly, is the gate blown? Is there a short G-S? Just wondering why your gate stopper is brown colored? Could that have been an oscilation issue first, instead of a blown gate? An over voltage at the gate ussually puts a hole in the gate oxide and it's done working correctly but, may still work somewhat.
I suppose you have a D-S short right? That would be the most typical failure mode but why, from just over temp'ing for a few minutes?
When doing such tests it is possible to turn on a FET with the devices own gate leakage but, ussually the ohmeter isn't sourcing enough voltage to turn it on well at all. And, ussually your looking for a pretty zapped channel (low resistance).
Hey, what was that gonad thing?
I suppose you have a D-S short right? That would be the most typical failure mode but why, from just over temp'ing for a few minutes?
When doing such tests it is possible to turn on a FET with the devices own gate leakage but, ussually the ohmeter isn't sourcing enough voltage to turn it on well at all. And, ussually your looking for a pretty zapped channel (low resistance).
Hey, what was that gonad thing?
Last edited:
I admit I F***D up, The F5 was over biased during setting up which ultimately killed one of the MOS-FETs. An expensive mistake. The Aleph4 would take such a mistake in its stride through the sheer amount of output devices.
KandD, Have you done any further "failure analysis"? Have you tested the failed MOSFET with an ohmmeter? Particularly, is the gate blown? Is there a short G-S? Just wondering why your gate stopper is brown colored? Could that have been an oscilation issue first, instead of a blown gate? An over voltage at the gate ussually puts a hole in the gate oxide and it's done working correctly but, may still work somewhat.
I suppose you have a D-S short right? That would be the most typical failure mode by why, from just over temp'ing for a few minutes?
When doing such tests it is possible to turn on a FET with the devices own gate leakage but, ussually the ohmeter isn't sourcing enough voltage to turn it on well at all. And, ussually your looking for a pretty zapped channel (low resistance).
Hey, what was that gonad thing?
Unfortnuately the IRFP240 is fried so I can't summize what happened.
Dogs Gonads = Bo***cks
Last edited:
KandD
far be it for me to say, but from the picture on how you mounted those Mosfets, are they actually flat on the heatsink? It looks like it is on an angle from the pics. I'm just considering Juma's suggestion of improper heatsinking.
What was the voltage across the bias resistor when it was running ok?
It just seems that for the size of those heatsinks, unless you were right ontop or underneath an AC vent, 40C is a bit too cold, and perhaps on top of temporarily over biasing, heat wasn't dissipated properly.
I've over biased mine before as well.... by mistake when I rolled mosfets, up to 2V across bias resistor. I immediately turn it off and turned down the pot to, so I unless you had it really high for a while, they are pretty robust.
far be it for me to say, but from the picture on how you mounted those Mosfets, are they actually flat on the heatsink? It looks like it is on an angle from the pics. I'm just considering Juma's suggestion of improper heatsinking.
What was the voltage across the bias resistor when it was running ok?
It just seems that for the size of those heatsinks, unless you were right ontop or underneath an AC vent, 40C is a bit too cold, and perhaps on top of temporarily over biasing, heat wasn't dissipated properly.
I've over biased mine before as well.... by mistake when I rolled mosfets, up to 2V across bias resistor. I immediately turn it off and turned down the pot to, so I unless you had it really high for a while, they are pretty robust.
Fried dosent mean you can't measure them??? I do it all the time. And it will help you "sumerize" what happened.
uugh? go ask Katie about Gonads!
