Author Topic: Sinewave guru's? [amp repair] FIXED  (Read 7466 times)

0 Members and 1 Guest are viewing this topic.

Offline stev.dkTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: dk
Sinewave guru's? [amp repair] FIXED
« on: May 12, 2013, 08:31:10 pm »
I'm trying to combine my two favorite hobby's. Playing music, and playing electrician... last one not as skillful though. Im trying to repair a 1970's vintage guitar amplifier. Everything seems to work great but the clean channel. The clean channel is only true clean, when input and output volume is really low. The distortion channel works like a charm.

I started out changing all the electrolytes, which took care of some "wobbeling" in the sound, caused by bad power supply smoothing caps.

It still distort on the clean channel way to soon. If i set the output volume to the same as the acoustic sound from my electric guitar, it starts distorting (pot at around 1-2 o'clock)... it shouldn't distort before pot at around 6-7 o'clock on the clean channel.

I measured the input in mV from my electric guitar, and created a sinewave at 440Hz at the same level, and feeded it in to the amp. The pot is about 2-3 o'clock, and the higher part of the sinewave go bananas. It has a huge spike on it, that has much more noise than the rest of the sinewave (might not be apparent by my poor mans mobilephone photo).


The lower signal is at the input of the amplifier, and the upper signal is at the output, connected across a power resistor at 8 ohms. If i turn the voloume i bit more down, the sinewaves at the top looks like titties  :-+

Are there any obvious thing causing this? (i've been so kind to supply the schematic for anyone interrested)

My next step is offcourse to backtrack with the oscilloscope and find a spot, where the extra peak disappeares, then try to figure out why the peaking starts after that exact part. But it would be awesome if a kind guru could help me out and point me in the right direction, instead of me woundering about in the dark  :-/O
« Last Edit: May 13, 2013, 06:53:39 pm by stev.dk »
Crank it up till' everybody's ears are bleeding, then back off just a notch.
 

Offline krivx

  • Frequent Contributor
  • **
  • Posts: 765
  • Country: ie
Re: Sinewave guru's? [amp repair]
« Reply #1 on: May 12, 2013, 09:34:57 pm »
What kind of amplifier is it? The obvious thing I would do is *carefully* probe my way through the signal path from input to power stage and determine where the distortion begins, this will narrow things down.
 

Offline stev.dkTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: dk
Re: Sinewave guru's? [amp repair]
« Reply #2 on: May 12, 2013, 09:39:40 pm »
It's a Session Rockette 30 Watt's guitar amplifier.

Yeah that was the next thing i was going to do, but my time is limited untill after summer :-/

I am hoping someone out there know a lot about these kinds of things, and know what could cause this wierd spike...
Crank it up till' everybody's ears are bleeding, then back off just a notch.
 

vlf3

  • Guest
Re: Sinewave guru's? [amp repair]
« Reply #3 on: May 12, 2013, 10:12:34 pm »
That's a very strange fault you have there...  Is the amplifier a valve design, or transistor ! the fault is not true distortion, this would have a flat on the top positive cycles, or more normally negative cycles as well; this looks more like an odd frequency filter peaking, and could be just at this 440 Htz  frequency.

You state a distorted channel that works okay, is this an actual effects input, within the amplifier !

Also, remove the 8 Ohm Load and re-connect the speaker; then do the scope test measurement again ? has problem disappeared !! if not try this.

Try another frequency, and see if you get the same result; if this is not repeatable with another frequency, then it points to a resonance of some sort being generated, within the amplifier pre or output stage.

What is happening here, is the positive sine-wave suddenly rises in transition from the normal level, to peak then return back to normal; my thinking is resonance at this frequency ? or "instability", that might require further DC power rails capacitor decoupling replacement, or tracing through the circuit ideally.

If this turns out to be a frequency resonance, it might be narrowed down to areas around any transformers, inter-stages pre-post amplification, where capacitor de-coupling are across the transformer windings... at present that's just a guess.

Yes, a circuit would help, to see the design and type of amplifier.
 

Offline David_AVD

  • Super Contributor
  • ***
  • Posts: 2797
  • Country: au
Re: Sinewave guru's? [amp repair]
« Reply #4 on: May 12, 2013, 11:04:38 pm »
Use your 'scope to look at the signal on the output of each opamp stage starting from the input.  That way you'll know where the distortion starts.

It's certainly an odd looking fault.  My first guess would be one of the opamps is bad.
 

Offline lewis

  • Frequent Contributor
  • **
  • Posts: 704
  • Country: gb
  • Nullius in verba
Re: Sinewave guru's? [amp repair]
« Reply #5 on: May 12, 2013, 11:53:46 pm »
That looks like one of two things, but it's very hard to tell without knowing the clipping behaviour of the amp and how the problem relates to the input amplitude and frequency.

Do you have the MOSFET version or the bipolar version? There are two power amp diagrams in the pdf.

1. An op-amp fault in the pre-amp section - check the output of each op-amp with the scope with a steady signal going into the amp to see where the weird shit starts. Then just replace that op-amp.

2. (If the bipolar circuit) It might be something to do with the bootstrapped VAS in the power amp section. (VAS = voltage amplifier stage and it's the BD139 transistor). Basically C26 gets charged through the upper 3.3K resistor when the output signal goes negative. When the signal swings towards the positive rail the capacitor is placed in series with the output voltage raising the potential at the junction of the 3.3k resistors. This increases the current through the Vbe multiplier (BC183) and VAS, and gives increased base drive to the upper output device TR5. (This is called 'bootstrapping the VAS' and is a trick for getting a bit more power out of an amplifier with given supply rails but the audiophiles hate it). Hint: In other words, the bootstrap makes the top transistor turn on more on positive excursions of the output signal.

Check the BC183 and BD139 thoroughly, and check the components around them. In fact, check all the transistors including the input transistors and output devices. In fact, check everything.

3. (if the MOSFET circuit) That's a modified Hitachi application note circuit. Check the 12K resistor R18 and all the 100R resistors, these are most prone to failure. If they're OK, check everything else.

But it's probably a knackered op-amp.
I will not be pushed, filed, stamped, indexed, briefed, debriefed or numbered.
 

Offline amspire

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sinewave guru's? [amp repair]
« Reply #6 on: May 13, 2013, 01:49:59 am »
Definitely check along the signal path to see where the distortion is occurring. My guess it the output amp, but it is very easy to confirm or eliminate that by checking the signals on each stage.

Before I looked at the circuit, my guess was something wrong around the inputs differential stage of the power amp, so I will stick with that.

For example C28 leaking or R34 going high. What is the voltage at the junction of R34 and R35? It should be a bit higher then HTvoltage * 1500/23500.

So if the HT voltage is 30V, it should be something like +2.5V at the junction of R34 and R35. If it is more like +1V, you can get the kind of waveform you are seeing.

It would also help knowing the actual voltages at the input voltage the power amp and the output, along with the HT voltage. If the input voltage to the power amp is a few volts peak, then you can get this effect  just because of the design. If C29 has gone dry reducing the power amp gain, then you have to overdrive the input and your waveform is the result. To test C29, you can actually just short it out.

One way or another I think the base of the input transistor is getting above ( (voltage at the junction of R34 and R35) - 0.7V) at the peak of the sinewave. It should not do this.

The transistors are fine, so I think it is either a bad design, or caps or resistors associated with the input stage or feedback are out. That is all assuming it is the power amp. If the distortion is happening before the power amp input, then ignore me.
 

Online IanB

  • Super Contributor
  • ***
  • Posts: 11790
  • Country: us
Re: Sinewave guru's? [amp repair]
« Reply #7 on: May 13, 2013, 02:09:08 am »
Not disputing any of the learned comments above, but if I looked at that oscilloscope trace my first impression would be that there is a push-pull stage somewhere that separately amplifies the positive and negative parts of the signal, and somehow the biasing is wrong so that instead of splitting the signal around the zero level it is being split somewhere up in the positive region.
 

Offline amspire

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sinewave guru's? [amp repair]
« Reply #8 on: May 13, 2013, 02:15:47 am »
Not disputing any of the learned comments above, but if I looked at that oscilloscope trace my first impression would be that there is a push-pull stage somewhere that separately amplifies the positive and negative parts of the signal, and somehow the biasing is wrong so that instead of splitting the signal around the zero level it is being split somewhere up in the positive region.
Basically it looks to me like the gain suddenly shoots up to maximum because above a certain voltage the feedback control dies. Otherwise the feedback would be trying to counter any non-linearity at a later stage in the power amp. This can happen if the input voltage gets so high, it starts to turns off the input stage.

When the input stage is turned off due to a transient positive overvoltage, the output will go to the maximum voltage. In a properly designed amplifier, this should never happen at normal output, so either it is a flawed design, or some resistor or capacitor fault is messing up the feedback ratio or the voltages around the input stage.
« Last Edit: May 13, 2013, 02:20:58 am by amspire »
 

Offline pickle9000

  • Super Contributor
  • ***
  • Posts: 2438
  • Country: ca
Re: Sinewave guru's? [amp repair]
« Reply #9 on: May 13, 2013, 03:01:19 am »
Regardless of the why, check at the input to the power amp to start. Splitting the difference is always quickest in the end. Not to mention "thou shalt check thy power supply voltages"   :)
 

Offline David_AVD

  • Super Contributor
  • ***
  • Posts: 2797
  • Country: au
Re: Sinewave guru's? [amp repair]
« Reply #10 on: May 13, 2013, 04:30:18 am »
Regardless of the why, check at the input to the power amp to start. Splitting the difference is always quickest in the end. Not to mention "thou shalt check thy power supply voltages"   :)

So true.  Fault finding is all about divide and conquer.  Split the problem in half (preamp / power amp) first, then do the same to whatever half the fault seems to be and so on.

Also never underestimate the creative ways that bad power and dodgy connectors can totally screw with signals and general circuit operation.   ;D
 

Offline ConKbot

  • Super Contributor
  • ***
  • Posts: 1380
Re: Sinewave guru's? [amp repair]
« Reply #11 on: May 13, 2013, 05:06:46 am »
Not disputing any of the learned comments above, but if I looked at that oscilloscope trace my first impression would be that there is a push-pull stage somewhere that separately amplifies the positive and negative parts of the signal, and somehow the biasing is wrong so that instead of splitting the signal around the zero level it is being split somewhere up in the positive region.
looking at the schematic
http://www.award-session.com/pdfs/manuals/sg30-sch.pdf

Hmm, if its at such low input levels, maybe that is the point where its exiting class A operation and going to class B operation, and if its the BJT version, if R-29 went short, it would remove some of the negative feedback for TR6, but I would expect slightly increased distortion, not it looking like it went open loop, no feedback, since the output is fed back into the long tail pair for global feedback also. Though I havent designed nor troubleshooted any amps, so I could be completely wrong.

However a failed power resistor should be pretty evident even without a multimeter, and the threadstarter seems to have his wits about him, so maybe not so likely.
 

Offline KJDS

  • Super Contributor
  • ***
  • Posts: 2442
  • Country: gb
    • my website holding page
Re: Sinewave guru's? [amp repair]
« Reply #12 on: May 13, 2013, 10:17:46 am »
Looking at that trace I'm wondering if the amp is oscillating above the frequency of the scope and in such a way to turn the device on harder to give more gain.

Offline stev.dkTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: dk
Re: Sinewave guru's? [amp repair]
« Reply #13 on: May 13, 2013, 03:15:07 pm »
lewis you are the guru i was seeking :-/O :-+

It's the bipolar version.

To show how much i appreciate your help, and what a nice guy i am, here are some genuine oscilloscope titties!



Heres the in / output i used today. Input at the top, output at the buttom.


The first ic at x10 attenuation is an almost perfect replication of the original sinewave.


The second ic is not used for the clean channel.

The third ic at x1 is also a perfect sinewave, however it seems to be a bit low?


The forth ic is also pretty good... however lets have a closer look at that


It has some noise going on... but checking the schematic, it's the one connected to the spring reverb, so noise must come from there, even though the pot is turned all the way down.


I measured across R40, the first transistor at the output stage, and what do you know. Perfect, so nothing wrong at the preamp stage.


Now, this is the collector at the first transistor (BC560) in the output stage (TR4). The base is comming from R40.


Here is the emitter from TR4, looks exactly like the output, just attenuated, cause it haven't gone through the power transistors yet.


The collector is connected to a couple of resistors that is connected to BD139. The BD139 is connected via BC183 to the emitter of TR3 and TR4. So it looks like you where right on the money lewis.

Im not sure where to go from here, if i should just get some new transistors, or if i should test something more - cause im not shure how to do any further testing right now. Thansk for the help so far! :-+
« Last Edit: May 13, 2013, 03:31:38 pm by stev.dk »
Crank it up till' everybody's ears are bleeding, then back off just a notch.
 

Offline amspire

  • Super Contributor
  • ***
  • Posts: 3802
  • Country: au
Re: Sinewave guru's? [amp repair]
« Reply #14 on: May 13, 2013, 03:37:32 pm »
Now, this is the collector at the first transistor (BC560) in the output stage (TR4). The base is comming from R40.


...

Im not sure where to go from here, if i should just get some new transistors, or if i should test something more - cause im not shure how to do any further testing right now. Thansk for the help so far! :-+
If you read my posts, I suggested the input differential stage of the power amp was turning off for part of the cycle. If you look at the collector voltage of TR4, it looks like it is turning off.

It might be worth looking at my posts. It really would help if we knew the important voltages. Most of your traces have nothing to tell us any voltages.

What is the voltage at the junction of R34 and R35 with no signal in? Is it correct (See my post - we do not know the HT voltage. What is it?)

What is the peak positive voltage on the base of Tr4 when you see the problem?
 

Offline stev.dkTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: dk
Re: Sinewave guru's? [amp repair]
« Reply #15 on: May 13, 2013, 03:58:22 pm »
amspire, im sorry bout' missing out on your post.

I just got out my multimeter and measured HT +/- 34.5V - using your calculation, the R34-R35 junction should be 34.5 * 1500/23500 = 2.20. I measured it to be 2.87V with the multimeter.

Quote
What is the peak positive voltage on the base of Tr4 when you see the problem?

Im not quite sure about that, should i measure the peak voltage with the oscilloscope, at the top of the funny looking wave, or should i measure it with a multimeter?
Crank it up till' everybody's ears are bleeding, then back off just a notch.
 

Offline stev.dkTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: dk
Re: Sinewave guru's? [amp repair]
« Reply #16 on: May 13, 2013, 04:18:07 pm »
Ok so i measured the peak voltage at the base of TR4, and the peak voltage was about 30mV with a clean wave (this is at very low output volume, almost no amplification) everything above that, creates the funny looking wave.

Edit: I don't believe C29 has gone dry, cause it's brand new.
« Last Edit: May 13, 2013, 04:23:02 pm by stev.dk »
Crank it up till' everybody's ears are bleeding, then back off just a notch.
 

Offline Tube_Dude

  • Regular Contributor
  • *
  • Posts: 115
  • Country: pt
Re: Sinewave guru's? [amp repair]
« Reply #17 on: May 13, 2013, 04:39:32 pm »
Is the output without any input signal at 0 Volts DC?
« Last Edit: May 13, 2013, 04:41:17 pm by Tube_Dude »
Jorge
 

Offline stev.dkTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: dk
Re: Sinewave guru's? [amp repair]
« Reply #18 on: May 13, 2013, 04:53:58 pm »
Is the output without any input signal at 0 Volts DC?

The output is at -120mV DC, with no input signal.
Crank it up till' everybody's ears are bleeding, then back off just a notch.
 

Offline Tube_Dude

  • Regular Contributor
  • *
  • Posts: 115
  • Country: pt
Re: Sinewave guru's? [amp repair]
« Reply #19 on: May 13, 2013, 05:13:13 pm »
Inspect if R23/32 are reading 12 Ohms  and R33/29 reading 1,5 Ohms... ( obviously with amp disconnected .... ;)  )
Jorge
 

Offline stev.dkTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: dk
Re: Sinewave guru's? [amp repair]
« Reply #20 on: May 13, 2013, 05:33:54 pm »
Inspect if R23/32 are reading 12 Ohms  and R33/29 reading 1,5 Ohms... ( obviously with amp disconnected .... ;)  )

R23/32 is spot on. R29 is spot on. R33 is bad, reads anything from 14KOhms and rising to 100KOhms and falling.

Im afraid i cant just replace it with any 1,5ohm resister i have though. It looks like a high power wire wound or something...

Crank it up till' everybody's ears are bleeding, then back off just a notch.
 

Offline lewis

  • Frequent Contributor
  • **
  • Posts: 704
  • Country: gb
  • Nullius in verba
Re: Sinewave guru's? [amp repair]
« Reply #21 on: May 13, 2013, 05:38:53 pm »
That's very very weird.... you should see the negative half-cycle heavily clipped if this resistor is open circuit, especially with a dummy load connected to the output.
I will not be pushed, filed, stamped, indexed, briefed, debriefed or numbered.
 

Offline stev.dkTopic starter

  • Regular Contributor
  • *
  • Posts: 67
  • Country: dk
Re: Sinewave guru's? [amp repair]
« Reply #22 on: May 13, 2013, 05:58:29 pm »
I only have 1/4 watt resistors, could i try and change the faulty one with one off these (staying at really low levels offcourse!), to see if it makes a difference. Worst case i blow up the resistor right?

My other (here and now) option is to change both with a couple of 1 Ohm 3 Watts resistors.

And anyway i'll have to figure out how many watt's the resistors need to be to get some new ones, any suggestions?

I would really be happy if i can avoid to order only the two resistors (pay a lot of postage), wait a few days for it to arrive, just to find out i need to order something else to, and once again pay a lot of shipping :-/
Crank it up till' everybody's ears are bleeding, then back off just a notch.
 

Offline lewis

  • Frequent Contributor
  • **
  • Posts: 704
  • Country: gb
  • Nullius in verba
Re: Sinewave guru's? [amp repair]
« Reply #23 on: May 13, 2013, 06:04:13 pm »
Just replace them both with 1R, 3W. That will be fine...
I will not be pushed, filed, stamped, indexed, briefed, debriefed or numbered.
 

Offline Tube_Dude

  • Regular Contributor
  • *
  • Posts: 115
  • Country: pt
Re: Sinewave guru's? [amp repair]
« Reply #24 on: May 13, 2013, 06:12:21 pm »
I only have 1/4 watt resistors, could i try and change the faulty one with one off these (staying at really low levels offcourse!), to see if it makes a difference. Worst case i blow up the resistor right?
My other (here and now) option is to change both with a couple of 1 Ohm 3 Watts resistors.

The first option is only possible if you try it at a very, very, low levels, so not much useful.

The second one look better. If you have enough 1 Ohms resistors you can use 3 in series and put two strings of 3 - 1 Ohms resistors in parallel. You get 1,5 Ohms. Bingo, you have a temporary fix...

By the way 4 or 5 Watts will be Ok when you order new resistors.

Glad you found the culprit.

Anyway inspect the Darlington transistor TR6, usualy a resistor don't burn without a reason...  ;)
« Last Edit: May 13, 2013, 06:26:37 pm by Tube_Dude »
Jorge
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf