Do mosfet output devices age? - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Thread Tools Search this Thread
Old 17th April 2006, 09:10 PM   #1
diyAudio Member
Join Date: Feb 2006
Location: Suffolk
Default Do mosfet output devices age?

Hi Folks,

I have a version of the Hitachi-datasheet Mosfet power-amp (the Mike Creek/ETI/R&EW version) which has been solid as a rock for the past 25 years. Unfortunately one channel has just started to become a little unstable, so I'm going to treat it to a (probably well deserved!) service and replace things like caps and maybe resistors.

Do I need to worry about replacing the output devices? I'm kinda hoping not as this would be quite expensive - each channel uses three pairs of devices (2sk134's/2sj49's).

Could I leave alone the other transistors, too - none of them seem to be readily available anymore? (though I'm sure substitutes exist).

  Reply With Quote
Old 18th April 2006, 12:26 AM   #2
poobah is offline poobah  United States
diyAudio Member
poobah's Avatar
Join Date: Nov 2005
I am sure someone here could recommend a replacement for your MOSFETS Matching might be an issue, but not as hard as you think... and not as important as some would want you to believe... all depends on the circuit.

Semiconductors, in general, die because of heat. Every 11 degrees C in operating temp cuts their life by one half. Except for some minor effects in the silicon itself; these wearout mechanism are all about heat induced stress. The die wants to rip itself from the metal base. the whisker (connecting wires) want to jump off the die etc...

Replace your caps... if you have carbon comp resistors... replace them all. See if that clears your problem before you mess with the SEMI's.

EDIT: you should replace your caps and res's on principle. But, don't forget to reflow ALL your solder joints as you go. They get weak as well from the same kinds of thermal induced stress. They are often the culprit when you encounter intermittant problems.

  Reply With Quote
Old 18th April 2006, 09:38 AM   #3
djk is offline djk
diyAudio Member
djk's Avatar
Join Date: Feb 2001
Location: USA
When the K134/J49 start going, they get low resistance from gate to source. A new pair measure infinity.

When the cap in the negative feedback loop goes, the amp will sound weak and after a while have no gain.
  Reply With Quote
Old 18th April 2006, 09:45 AM   #4
Electrons are yellow and more is better!
diyAudio Member
peranders's Avatar
Join Date: Apr 2002
Location: Göteborg, Sweden
Blog Entries: 4
I would suggest that you go over all solder joints on the pcb's. Secondly, consider change of critical caps, not everyone.

How do you mean "unstable"?

(BTW: I have the output devices for sale...)
/Per-Anders (my first name) or P-A as my friends call me
PA03 LM4780 amplifier group buy, SIGN UP HERE for the group buy 0 boards left. 118 paid.
  Reply With Quote
Old 18th April 2006, 12:59 PM   #5
ilimzn is offline ilimzn  Croatia
diyAudio Member
Join Date: Feb 2005
Location: Zagreb
Medium and high power semis (especially TO3 ones) die of heat induced stress, but usually last a VERY long time unless there is some fatal flaw. For TO3, this is normally mechanical - overstressed case when mounting, bad thermal interface (silicon grease that has turned intopowder and crumbled away').
Plastic cases in general, and older small signal TO92s and similar, have a problem with corrosion - the pins corrode cracking the plastic to metal interface, eventually, the interface to silicon gets contaminated. Heat makes it all VASTLY worse. If you see coroded pins o the small signal transistors, replace them.
Finally, there is one more thing that can end up in small signal semi death in amplifiers. Very often, the input LTP transistors do not have sufficient Vcbmax to span the supply rails that this stage is using. This is because most high gain transiostors are not high enough voltage. Under normal circumstances, this is not a problem, but in transient conditions (power up, power down, in some designs clipping or current limit) or if the amp ever had a fatal failure, such as a shorted power transistor on the output, thses can develop very odd problems - crackling noise, leakage, low gain.
One other problem that I remember that fits the description, is contamination of the flux used in the solder. In two cases I had, the cure was do thoroughly clean the PCB (in one case I figured it out only after I had completely rebuilt the whole PCB with brand new coponents )
  Reply With Quote
Old 18th April 2006, 01:32 PM   #6
poobah is offline poobah  United States
diyAudio Member
poobah's Avatar
Join Date: Nov 2005
Sorry Geesixty,

These guys are right... don't replace all your caps. Just replace all the electrolytics... your other caps should be fine.

  Reply With Quote
Old 19th April 2006, 12:44 AM   #7
diyAudio Member
Join Date: Feb 2006
Location: Suffolk
Thanks guys for all your responses,

It sounds like it would be a good idea to replace the critical caps, though given that it won't cost much more and I'll have to dismantle the amp anyway, to also replace the other caps, too.

The resistors are all carbon film, so I'll replace these, too. In the process I'll make sure I go over the solder joints for components that don't get replaced.

A visual inspection of the trannys, output devices and the TO92's, seems to show them in ok condition with no obvious looking corrsion and the thermal grease seems ok, too, so I'll leave all of these for the moment.

However, Ilimzn, I don’t really understand the bit about voltage ratings of the input LTP. The input pair are 2SA872’s where Vcbo is -90v. The supply rails are + - 56v. Do you think this would cause a problem?

The amp actually sounds as fine as it ever did. The reason I think it maybe becoming unstable was when I connected it to a new pre-amp I've been building there was a buzzing from the speakers (sounds like 100hz but could have been 50) which I initially put down to an earthing mismatch between the two. The buzzing was quiet and would come and go. I did end up using this combo for a week or so whilst playing around with the pre-amp. Then later, the buzzing got louder and some wierd things started to happen whenever I connected the pre to the power (spkrs disconnected) like the CD player (and also the DVD) seeming to lose the ability to read discs. Perhaps more tellingly I have a touch-controlled light dimmer fitting in the room and it would go from full bright to full dim as soon as the two were connected! Possibly the new pre-amp is unstable, too, but see below.

However, on reconnecting the original pre-amp (or the other pre-amp I have, a Naim 62), the buzzing still comes and goes, which seemed odd because they’ve both always been fine before. This was the case even before leaving the new pre-amp connected for a week or so.

I borrowed an oscilloscope and, although I don't yet know how to use it properly, have made some comparative tests against the other channel (they’re monoblock amps) and also another power amp (inputs shorted on all the amps). With the probe measuring the voltage rails on the good channel (and the other pa) I see a clean triangle-type wave with the scope on 10ms/div, which I guess is the ripple voltage? On the poorly channel the wave form is about 30% lower in amplitude but very fuzzy and 'thick' like it's been drawn with a thick marker pen. On altering the scope setting to 2us/div I see an irregular wave form. The other channel doesn’t have this - unless I also power-up the poor channel in which case the irregular waveform reappears on this channel, too.

After about 10-15 mins of idle the heatsink on the good channel remains more or less cold whereas the poor channel is warm. Lastly, on the poor channel the resistor in the zobel network has started to burn out, though at the moment it still measures its original value of 10R.

  Reply With Quote
Old 19th April 2006, 02:15 AM   #8
diyAudio Member
Join Date: Dec 2005
Location: Kuala Lumpur
A cooking Zobel network resistor means that there is ultrasonic oscillation. The buzz is motorboating.

You are risking blowing your tweeter on that channel.

The oscillation is probably either caused by a dry joint or a drying electrolytic.
  Reply With Quote
Old 19th April 2006, 02:26 AM   #9
johnnyx is offline johnnyx  United Kingdom
diyAudio Member
Join Date: Jan 2004
Location: manchester
Take a look at This thread. I had a similar amp, and the zobel resistors burnt out. Read towards the end of the thread to discover why, and what I did about it. It is very likely that you have the same problem that I had.
  Reply With Quote
Old 19th April 2006, 11:48 AM   #10
diyAudio Member
Join Date: Feb 2006
Location: Suffolk
Davidsrsb, I'll definitely look at these two points - blowing tweeters would be very bad indeed

Johnnyx, thanks for this, a very interesting thread with much to digest, though I don't pretend to be able to follow the discussion on actual design details

The gate resistors, which you found to be significant in your amps, are 220R. I'm coming to the conclusion that the amp has become flaky due to simple ageing of components and solder joints rather than component values or physical layout, given that it has worked very well indeed for the past 25 years. So I guess it's not surprising that it's getting a little tired

  Reply With Quote


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

Similar Threads
Thread Thread Starter Forum Replies Last Post
Development 500 W - 2 KW @ 2 OHM Amp only with Mosfet NPN Devices Dorian Gray Solid State 57 14th March 2010 09:46 PM
Output Devices ! cloud Pass Labs 11 10th November 2008 11:37 PM
Comparison between Bipolar transistors and MOSFET as output devices raveenvijendren Solid State 4 17th November 2005 04:28 PM
More Output devices...A75 FBJ Pass Labs 14 19th March 2002 06:18 AM

New To Site? Need Help?

All times are GMT. The time now is 07:23 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2