By continuing to use this site, you agree to our use of cookies. Find out more
Forum sponsored by:
Forum sponsored by Allendale Oct 22nd

Anyone good at fault finding with amplifiers here?

All Topics | Latest Posts

Search for:  in Thread Title in  
Neil Wyatt17/08/2019 21:08:27
avatar
Moderator
16568 forum posts
687 photos
75 articles

I've got a problem with my Trace Elliot amp, it's an old AH150 based around dual mosfets (150W amp).

The problem is a simple one - there's about 4V of offset voltage on the output, it was 8V (and making the output transistors get hot) so it might be an intermittent issue. I'm pretty sure it isn't dry joints.

It's this circuit but with only two output transistors:

These are some measured voltages - the -21.5V and 5.0V on the collectors of TR203 and TR204 don't seem right.

TR201 and TR202 both have 0.4V on their bases.

Any suggestions for further fault finding welcome.

Joseph Noci 117/08/2019 21:27:20
542 forum posts
832 photos

Neil, could you measure between the bases of the the first pair (TR201/2) with your meter at lo-volt range (10v scale maybe) please, and then between each transistor base and emitter (TR201/2)

Joe

 

 

 

 

 

Edited By Joseph Noci 1 on 17/08/2019 21:32:11

Neil Wyatt17/08/2019 21:42:10
avatar
Moderator
16568 forum posts
687 photos
75 articles

Hi Joe,

Base TR202 to Base TR201: -0.02V

Base - collector TRR201: 0.63V

Base - collector TR202: 0.60V

(Obviously there's a small rounding error in there).

Gain is 100 (R207/R206) but this isn't DC coupled.

Joseph Noci 117/08/2019 21:51:11
542 forum posts
832 photos

Something fishy around TR201/2 I would think. A positive voltage on the output would tend to turn TR202 more 'off' ( via R207) and cause the voltage on TR202 collector to become more negative. It is more positive than the collector of TR201 so maybe the voltages between base and emitter of TR201/2 can show something more.

Also, C204 seems to be an electrolytic? It may have become leaky ( electrically). With no input to the amp, can you disconnevt one leg of C204 and see if the output voltage remains?

At the moment, TR203 is turned on harder than TR204, hence it's collector will be at lower potential than TR204.

Joe

Joseph Noci 117/08/2019 21:55:42
542 forum posts
832 photos

Ok, presume you meant between TR202/3 base and emitter, not collector..

What are the voltages between the base/emitter of TR203/4?

Neil Wyatt17/08/2019 22:15:40
avatar
Moderator
16568 forum posts
687 photos
75 articles
Posted by Joseph Noci 1 on 17/08/2019 21:51:11:

 

Also, C204 seems to be an electrolytic? It may have become leaky ( electrically). With no input to the amp, can you disconnect one leg of C204 and see if the output voltage remains?

Already tried - you can unplug the preamp which does this, no change.

Yes meant Base/Emitter.

As expected:

TR203: 0.84V

TR204: 0.60V

 

I'm wondering if (tomorrow) I should go through my bits boxes looking for a pair of well-matched PNP transistors to replace 201/202. They are BF423, low gain (min 50) but high voltage. Might have to buy some in.

I could pull 201/202 and test their Hfe and leakage.

Probably best I sleep on it - getting a bit too tired for playing with an amp running at 140V rail to rail...

Thanks for the thoughts so far.

Neil

Edited By Neil Wyatt on 17/08/2019 22:16:00

Simon Williams 318/08/2019 00:45:08
412 forum posts
67 photos

First suspect is dried out and leaking electrolytics, but TR203 base to emitter voltage at 0.8+ volts says it's cooked. Doesn't necessarily say why. TR 205 is also off or nearly so, 2.4 V across R211 doesn't look right.

Shouldn't need a matched pair for TR203/204 (edit correction 201/202) to get a stable and usable system, though it may be needed to get the distortion/noise spec. If you want to borrow a distortion meter I've got one.

That 21 volts negative on TR203 collector looks wrong, (80 v or so across R209) confirms that TR203 is not a happy bunny. Looks too far negative, should be nearer mid rail. But TR201 and 202 are nearly completely off, as revealed by + 67V on their collectors. Long tailed pair like this (even if upside-down) should have collectors mid rail or thereabouts. Can't hurt to pull C204 and replace it. Ditto C201 is suspect.

Keep us posted as to progress, do.

Best rgds Simon

edited for correction in italics - though the same is true of TR203/204!

Edited By Simon Williams 3 on 18/08/2019 00:51:42

John Olsen18/08/2019 01:34:19
988 forum posts
86 photos
1 articles

I haven't attempted to fix a direct coupled amp like this for a while now, but anyway...Because they are direct coupled and have DC feedback as well as AC, faults can manifest themselves in subtle ways. With one I worked on, replacing a cooked output transistor just lead to the immediate demise of the replacement. I ended up checking the transistors one by one with the power off to localise the culprit. That was one that had been built from a kitset by a learner, so I had to check everything. You can sometimes check them OK in circuit, but sometimes it is better to take them out so you can be certain.

Other than that I tend to agree with Simon's reasoning up above. Bear in mind that something like a leaky capacitor could put the circuit out of balance enough to possibly take out one of transistors in the early stages, so there may be more than one fault.

Also unplugging the preamp does not do anything with respect to C204 being possibly faulty. If C201 was leaky unplugging the preamp would help.

John

Joseph Noci 118/08/2019 07:42:16
542 forum posts
832 photos

Hi Neil,

As said above, unplugging the pre-amp does naught WRT C204 - not sure if you checked that cap. The other caps appear non-polarised, so unless old, paper wound types, they should not be any problem.

The key, as Simon said,is that under quiescent conditions, the voltage at TR201/2 collector should be close to mid-rail.

You should be able to check TR201/2 in circuit - the circuit is reasonably balanced in coupled impedances and comparative readings between the two should reveal large deltas. Try the usual ohms readings tween BE, BC, EC, with positive and negative polarities, and compare each transistor. ( discharge PSU caps PSC1/2 first).

If nothing reveals, you may need to remove each and check in your fancy tester..

OR - lift the leg of the mosfet end of R207 . Connect a resistor, value equal to R202, between TR202 base and ground ( midpoint of the PSU caps). On power, the voltage at the collectors of TR201/2 should now be equal and close to mid-rail.

I do the repairs for our local Hi-Fi/Sound system retailer here in Swakopmund and work on equipment of all conceivable types..From shebeens, disco's, restaurants, homes, Churches, Bands....And have followed many a garden path in the search for the fault, so what do I know...

Joe

Neil Wyatt18/08/2019 09:34:35
avatar
Moderator
16568 forum posts
687 photos
75 articles

Both transistors are in spec for gain, but don't seem to be well matched

TR201 - 80

TR202 - 64

Joseph Noci 118/08/2019 10:29:05
542 forum posts
832 photos

I think that's a good place to start Neil. I repaired a Hitachi amp - see schematic below - similar to yours , but with stages inverted, and the input long tailed pair had to be well matched - gain delta of 5 gave around hundred millivolts output offset. This amp has pots for offset adjustment, but setting the pot voltages for equal Vbe showed how close the matching had to be. Also,the input LTP in this amp were bonded together, with copper foil wrapped around - for thermal tracking.

Joe

hitachi.jpg

Joseph Noci 118/08/2019 10:32:58
542 forum posts
832 photos

If you have spares, replace them, but if you have time...Fit the originals back, verify the output offset voltage, and gently heat one of the transistors with the soldering iron - see what the offset voltage does. Try to do the same with the other transistor when the first has cooled again and see if the effect is similar - and opposite..

What is the amps age?

Joe

Neil Wyatt18/08/2019 11:02:56
avatar
Moderator
16568 forum posts
687 photos
75 articles

Wow, thanks for LOADS of advice, all of which I have carefully digested.

I then went down to replace TR201 and TR204 and try some more measurements.

I thought... if it's their gain mismatch then swapping them over should reverse the effect and make the cones displace in, not out.

This is exactly what happens!

I've ordered 10 BF423 resistors, that should let me get a better match. If that doesn't work, I've got a lot to go on, I've also ordered some BF422 so TR203/4 can be replaced if necessary.

I also noticed one of the four speakers is wired out of phase with the others (it goes out when they go in) the combo was re-speakered by the previous owner (with very nice units - Eminence B10A, cost about twice what I paid for the whole thing). This explains why I wasn't 100% happy with the bottom end on it and ended up buying a 15" extension cab...

AdrianR18/08/2019 11:09:52
272 forum posts
20 photos

Hope this is not egg sucking, and it has been 35 years since I last played with electronics.

TR205 is acting as a current mirror. This means that the current flowing down through the TR203 leg is mirrored in the TR204 leg with the ratio of R211/R208.

The voltages you have show on those resistors infer the mirror is working as they should always be similar.

If you know the value of R211 you can calculate the current in that leg. Then gestimate the resistance of the pot by its position and check if the voltage drop across the pot is correct. That will tall you if the pot is bad.

The -21V on TR203 seems to suggest the input differential pair TR201/TR202 and TR203 are working.

5V onTR204 could be about right for a 4V output, but hard to say if it is due to being driven that way or it is due to the clamp diodes D202 - D205 without a parts list.

Intermittent voltage could be the biasing pot vr201, is it an open one or sealed. Open ones are prone to get mucky and noisy. If that went open circuit nasty things would happen as all the output transistors would be biased on.

How hard they would be turned on depends on the clamp diodes.

It is unusual for mosfets to go short circuit but if all else fails, you could try popping out the output pairs, 206/208 or 207/209

Adrian

Joseph Noci 118/08/2019 11:55:12
542 forum posts
832 photos

Well, there are not many transistors in that amp so when you swap the last one out, all will work...

I second Adrian's comment on the pot - esp since you said that there was a point where the output FETs got hot - they had to be turned on to some extent for that - if the pot is open frame, clean the track with a little methylated spirits while winding the pot back and forth two or three times. ( note its position and return it to that). The output FET's standing current should vary smoothly while adjusting the pot up or down SLIGHTLY...Beware excess current in the output!

Joe

Neil Wyatt18/08/2019 13:17:23
avatar
Moderator
16568 forum posts
687 photos
75 articles

VR201 is wired with the wiper to one end of the track, so it can't go disastrously open circuit.

I noticed a lot more hum after swapping TR201/2, so I swapped them back. The polarity of the offset changed again and the hum became almost inaudible.

This suggests to me that one of the pair is definitely on the way out even though they give believable gain readings.

Changed the polarity of that rogue speaker - they are Alpha 10s not beta 10s. £100 each rather than £60. Boggled! Not actually as good for low bass as the Beta 10, meant for a PA with sub-woofers, so someone had more money than sense... should balance well with the 15" cab though.

Neil

Neil Wyatt18/08/2019 15:50:21
avatar
Moderator
16568 forum posts
687 photos
75 articles
Posted by Joseph Noci 1 on 18/08/2019 11:55:12:

Well, there are not many transistors in that amp so when you swap the last one out, all will work...

The scoundrels approach to fault finding - REPLACE EVERYTHING devil

It has occurred to me that much as these amps are venerated (a generously rated transformer and big smoothing caps mean they sag gently rather than breaking up when pushed hard), 99% of the Trace Elliot sound comes from their preamps. I could just drop in a 500W* class D amplifier module and PSU and knock a chunk off the 43kg total weight...

Neil

*~250W into the 400W/8R built in speakers or ~500W with the extension cab.

Joseph Noci 118/08/2019 17:15:58
542 forum posts
832 photos

I HATE class D amps - They are the worst to repair. Everything has to work for it to work - you cannot debug or diagnose sections, one big, miserable closed loop that makes things go bang if you open the loop at the wrong place.

And I don't consider them 'Hi-Fi'...

Joe

Mark Rand18/08/2019 22:23:51
761 forum posts

But you don't have to repair them. They are so cheap that they are disposable. I used to be able to hear 20kHz as a teenager, but I could never hear 200kHz...

Said he, who bought a Pink Triangle turntable because he didn't like the nasty ringing sound of the Lin Sondek LP12.

Neil Wyatt18/08/2019 23:06:44
avatar
Moderator
16568 forum posts
687 photos
75 articles
Posted by Mark Rand on 18/08/2019 22:23:51:

But you don't have to repair them. They are so cheap that they are disposable.

Exactly.

All Topics | Latest Posts

Please login to post a reply.

Magazine Locator

Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!

Find Model Engineer & Model Engineers' Workshop

Latest Forum Posts
Support Our Partners
Ausee.com.au
Eccentric July 5 2018
Eccentric Engineering
ChesterUK
cowbells
Warco
Meridienne oct 2019
TRANSWAVE Converters
Allendale Electronics
emcomachinetools
Subscription Offer

Latest "For Sale" Ads
Latest "Wanted" Ads
Get In Touch!

Do you want to contact the Model Engineer and Model Engineers' Workshop team?

You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.

Click THIS LINK for full contact details.

For subscription issues please see THIS LINK.

Digital Back Issues

Social Media online

'Like' us on Facebook
Follow us on Facebook

Follow us on Twitter
 Twitter Logo

Pin us on Pinterest