Dick Burwen's Amplifier Problem

[mlucitt]

From the website:
For 28 years 17 Phase Linear 400 stereo power amplifiers drove the 5 speaker arrays. Together with the 4-way electronic crossover networks, they could produce the same sound level as a single 20,000 watt amplifier. Dick modified these amplifiers for improved frequency response, gain accuracy, and low-frequency power capability. Eventually their power transistors wore out due to thermal fatigue. It would have been OK if the amplifiers simply failed and some power transistors could be replaced. Instead they became intermittent producing a loud pop from the speaker perhaps once a month and later more often. It became too difficult to determine which transistors were failing and in 2004 all 17 units (34 channels) were replaced by QSC SRA 1222 dual 200 watt power amplifiers. Details of the system with Phase Linear amplifiers can be found in the two technical articles that appeared in Audio magazine in 1976 and 1995.

Links:
http://www.burwenaudio.com/Biography.html http://www.burwenaudio.com/Sound_System.html

Does anyone know about Dick Burwen? Apparently he knows Ed Blackwell and has been involved with Phase Linear for many years.
I am curious about his comment regarding "thermal fatigue". I only know of metal fatigue when heated beyond a certain point.
Anyone else have experience with thermal fatigue in transistors?

 

[laatsch55]

Happens everytime they heat up and cool down ..

 

[George S.]

I've run across the website a few times and wondered about that myself. Thought, yeah, I can see those tiny internal solder joints failing after many, many, years of hard usage and temperature swings. Too bad WOAD products weren't available then. That system would have been really awesome.
Really reinforces Sniff's "don't reuse that old junk" stance. I originally was going to reuse all the outputs when I rebuild that Pro 700, as they all test good, but they're date coded 0, so 1990 or 2000 manufacture date. Decided if I want it to last, new is the way to go. Was actually checking prices last night for outputs. Going to bite the bullet for new and order from Joe once he's stocked up on the 700 storage caps.
And the big question. Are those transistor internals actually soldered or micro welded? Going to cut one open next couple days and find out. I have a coffee can full so I'll do a PNP and NPN and photograph with a USB microscope.

 

[Gepetto]

I've run across the website a few times and wondered about that myself. Thought, yeah, I can see those tiny internal solder joints failing after many, many, years of hard usage and temperature swings. Too bad WOAD products weren't available then. That system would have been really awesome.
Really reinforces Sniff's "don't reuse that old junk" stance. I originally was going to reuse all the outputs when I rebuild that Pro 700, as they all test good, but they're date coded 0, so 1990 or 2000 manufacture date. Decided if I want it to last, new is the way to go. Was actually checking prices last night for outputs. Going to bite the bullet for new and order from Joe once he's stocked up on the 700 storage caps.
And the big question. Are those transistor internals actually soldered or micro welded? Going to cut one open next couple days and find out. I have a coffee can full so I'll do a PNP and NPN and photograph with a USB microscope.

welded

 

[George S.]

Copy the welded.

 

[wattsabundant]

The loud pops that are mentioned in the article are not likely to be caused by outputs. The more common issue is the TO92 transistors on the front end.

 

[mlucitt]

The more common issue is the TO92 transistors on the front end.

Because they were not matched and tie-wrapped by me and Lee...

 

[laatsch55]

LOL!!!!!!!!!!

 

[nakdoc]

I bet Burwen increased the bias in an attempt to reduce distortion. Output transistors ran hotter (even with fans, temperatures inside the TO3 case are still very high. A friend of mine made a 5 watt true class A amplifier using a tunnel heatsink I found at an electronics junk store. The transistors would last one week before making noises....luckily they were 2N3055 which were cheaper than dirt.

 

[mlucitt]

He never mentioned thermal runaway, maybe he did not let the amplifier get that hot, but many sessions of near destructive heat made them noisy.

 

[FredR]

Amazing man. Humbling.

 

[nakdoc]

I think he ran his system very loud at times. Have you seen the picture of him standing inside his wall of speakers? His De-Noiser was probably a necessity in his system.