Denon PMA-700V keeps burning

XgN

Member
2014-08-28 11:03 am
Hi guys!
New year is coming, greetings!

Got this denon PMA-700V amplifier for repair. The story goes like this. Owner connected speakers only to both R and L positive terminals, listened to music for a while and then brought the amp to me. Okey, 6/8 power transistors were burned. Also some 0.22R resistors and one driver transistor. As I was not able to get original power transistors [2SA1104/2SC2579] I replaced with more common analogues [2SB688/2SD718]. Of course, after replacing I turned on amp with light bulbs instead of one fuse on both channels. They burned. PNP [2SB688] were those that burned despite of using light bulbs. And somehow strange - only EC was shorted, BE and BC p-ns were OK. Okey, maybe some disbalance when using only one light bulb instead of on fuse or ESD for transistors or maybe defective themselves. Replaced again, put 2 light bulbs instead of fuses to one channel. Everything was working, practically no voltage drop on light bulb, managed to get the idle current in the proper range. After few more turning ons they were burned again (and more with driver transistors this time and R553 and R559 resistors). And this happens only during the turn on process.

Could this be related with transistors parameters? Seems like Vec breakdown voltage is too low, but it is the same for original PNPs, only original NPNs have more voltage compared to my analogues but they aren't which burn. Checked all around components including diodes, zeners, transistors - they are all good. Now I bought several pairs of [2SA1941/2SC5198] which withstand 140V for Vec(br). But maybe there is a thing I do not see here. Would be great if someone comes up with any idea!

Here is related schematic part:
Screenshot by Lightshot

Thanks in advance.
 
What voltage are the supply rails? You know that secondary breakdown is the key spec for an output transistor, most fail by exceeding it (reactive loads are the issue - connecting a speaker between L+ and R+ will be extremely reactive unless the channels are bridged (ie in antiphase).


The symptoms of that failed transistor are not strange, the base doesn't get massive currents through it so will be the last part of the device to cook/melt.


You may have oscillation happening due to the transistor substitutes being faster, changing output devices may require the compensation to be altered too. Try doubling the value of the compensation cap on the VAS (assuming it has standard Miller compensation, the schematic is incomplete so its hard to say)
 

XgN

Member
2014-08-28 11:03 am
What voltage are the supply rails? You know that secondary breakdown is the key spec for an output transistor, most fail by exceeding it (reactive loads are the issue - connecting a speaker between L+ and R+ will be extremely reactive unless the channels are bridged (ie in antiphase).


The symptoms of that failed transistor are not strange, the base doesn't get massive currents through it so will be the last part of the device to cook/melt.


You may have oscillation happening due to the transistor substitutes being faster, changing output devices may require the compensation to be altered too. Try doubling the value of the compensation cap on the VAS (assuming it has standard Miller compensation, the schematic is incomplete so its hard to say)

Supply rail without power amplifier load (no fuses fit) is +/- 59V.
Here is more detailed schematic of one channel:
Screenshot by Lightshot