Teisco Checkmate 21 Solid State amp schem + help

[Roly]

 :dbtu:

[galaxiex]

Thanks for the thumbs up Roly. 

So I got home and wired this up just like the schematic shows.

One exception... R7 at the base of Q1 is actually an 18K, not 10K.
Somehow that mistake crept in and I never noticed.
It's always been an 18K all along. (I'll fix the schem later)

Ok, so the amp is very loud and distorted like this. (BTW, the Trem still works  )
With Tone and Vol up full it squeals horribly.
Squeal (barely noticeable under the distortion) starts at about 6 on volume and gets worse as Vol increases.

At very low Vol it's "almost" clean. Any tone setting.

Can find a "sweet spot" balancing tone and vol to where it's a nice crunch but kinda muddy, lack of treble.
(But it will sustain for days...    :lmao: )

As an experiment I disconnected the buffer output after R4 10K, (so no signal on R5 at all) and jumped it to the junction of C7 and R11.

Still quite distorted (any vol except very low) and not a pleasant distortion.

My (random) thoughts/ideas/questions... no particular order... and I'm just guessing... 

1. Leaky germanium transistors can't take the good solid guitar signal from the buffer.
2. Simply too much gain somewhere for what the circuit can handle.
3. The tiny little phase transformer is saturating and not in a good way...
4. Replace all germs with silicon (one at a time and bias them appropriately? ), see what happens...
5. The above may not work because of the phase transformer?
6. Go back to 50K pots as those were probably bleeding off some signal to ground?
7. Figure out a way to eliminate the phase transformer?
8. The "power amp" section just can't handle a decent signal?

I've got small signal JFET's and BJT's and small MOSFET's etc, etc.
Even got some 2N404 germanium transistors.

Lots of questions and possibility's....

Where do I start?
Where would you start?
(besides gutting it and putting a real amp in it, still a possibility...) 

If I gut it, I still want Tremolo and Reverb to work.
I would need a preamp schem I can copy or...
Design my own? That may be over my head right now...

I do have this 60W to 100W compact power amp board kit from here...

http://www.lh-electric.net/projects/pa100ws.html

Thanks for any ideas you can throw my way.
I'm a little lost right now... Not discouraged, just not sure what to try next.

Edit; BTW I did physically remove Q2 and all associated R's and C's from the board.
Just in-case there was some parasitic signal bleed or whatever...

[phatt]

8. The "power amp" section just can't handle a decent signal?

Lots of questions and possibility's....

Where do I start?
Where would you start?
(besides gutting it and putting a real amp in it, still a possibility...) 

Thanks for any ideas you can throw my way.
I'm a little lost right now... Not discouraged, just not sure what to try next.

You start by reading;

How about you go do some reading it will save you from all the frustration.

http://sound.westhost.com/site-map.htm

Under *Articles*
Start at *amplifier basics* and work down, it will help you a lot. 8|

Also look at some of the project pages for ideas.
http://sound.westhost.com/projects-0.htm#pwr

That site is huge and bursting at the seams with well written material for the novice. :dbtu:
Cheers, Phil.

[galaxiex]

You start by reading;

How about you go do some reading it will save you from all the frustration.

Thanks Phil, I will do that.

[Roly]

With Tone and Vol up full it squeals horribly.

Instability.  Your amp is acting like an oscillator.


The conditions required for oscillation are the opposite of the conditions required for amplifier stability.

For an amplifier to oscillate there must be a feedback loop (albeit unintended) with gain, and a phase shift at the oscillation frequency which brings the feedback in phase with the output - positive feedback.  For oscillation the loop gain (forward gain less unintended feedback loop losses) must be greater than unity.

Finding this loop and killing it is your Target For Tonight.


How we prevent stages from oscillating is firstly by supply de-coupling, then spacing and shielding, keeping long runs at low impedance, careful grounding paths, etc.

You have upped the primary filter cap C24, the next important one is C16/220uF after the decoupling resistor R24.  Then there is no more decoupling.  So, add another 1k in series with the supply line to the front end (may have to cut a track; just above "vol" on cct) and add a bypassing cap somewhere close to Q1.


Fit ground shorting input jacks.  No.1 now has a very high input impedance and is wide open.  Just a few pF's of strays and she's coupled.

Where do I start?
Where would you start?

"Start"?  START!?!?  :trouble Your circuit is already at rev3.0 and counting!  You're on the third lap already.

Review this thread man, look around you, you are already in it knee-deep!  :duh  {And having a great time it seems to me}   :dbtu:

Now you have a hacking amp with some demonstrated potential (the trem), and a few problems.  Instability, particularly as you are modding a design, can be most interesting and educative, this is the wild unexplored Great Unknown for you to blaze a trail - Technician vs. Universe - who will win and bring order out of chaos?

{when I was doing industrial stuff I carried a lot of tools and equipment, but I also carried a real Magic Wand^(TM,C,P)*, and at times I have produced it and done a "nothing up my sleeve" before "pulling a rabbit out of a hat" and starting the (previously) faulty machine.  (overheard: "yeah, he's @#$%in' nuts, but he can nail jelly to a wall"  :lmao: )  *This is also my "calibrated tap" for finding microphonics preamp valves}

We reduce each problem to bite-sized lumps, then break them down into nibbles.

The power supply now seems satisfactory (if not actually magnificent).


You were having thermal runaway (in the output stage?).  After all your supercharging we need to have a look at the output stage, its biasing, its thermal compensation (lack of), and its dissipation/heatsinking; its DC and thermal conditions.  I'm pretty sure that the bias tree R25/6/7/8 requires trimming, AND it has no thermal comp (so no wonder you are getting thermal runaway) so that also needs a mod.

Then we need to look at the AC conditions, is the driver doing the required job?

Then back another stage; stable?  Sufficient?

Finally the inputs, and a fully functional amp.

And from that stable foundation you can then tinker and mod with confidence.

HTH


{Ref: cct rev3.0;
http://www.ssguitar.com/index.php?action=dlattach;topic=3535.0;attach=4841}


PS

By the way after that last post I went back and scoped the amp and the signal after the phase transformer is def clipping.
Far too much gain and hitting the rails.

Steady champ, remember this is power (E&I) drive you are looking at, and the transistor bases look a bit like a clipper to the driver, could be normal.  First the DC conditions for the new supply.


edit to add...

Before 100W-RE.gif, first I think maybe;
http://sound.westhost.com/project03.htm

or



then;

http://sound.westhost.com/project27.htm



Grouch: lookit this in isolation;



What does it tell you without the component values?

Compared with;



{"you never find just one roach in a kitchen"}

[galaxiex]

Instability.  Your amp is acting like an oscillator.

Thanks for your explanation of this. 

Target For Tonight.... I will...

1. fit a shorting jack
2. increase? C16 to 1000uF or whatever I can fit there.
Is there any benefit to adding a small 0.1uF in parallel with C16?
3. add a 1K in the supply line to Q1 and decoupling cap nearby. Value? Large? Small? Both? I'll experiment...
4. ... and other things as I think of them... but one at a time! so as not to lose the rabbit.

"Start"?  START!?!?  :trouble Your circuit is already at rev3.0 and counting!  You're on the third lap already.

Review this thread man, look around you, you are already in it knee-deep!  :duh  {And having a great time it seems to me}   :dbtu:

Ha Ha!  :lmao: Yeah, I should have said "where do I start from here"

Neck deep more like it...

Yes, having a great time!  Learning is FUN!  Esp. with such wonderful help here! 
Someday I hope to be able to nail jelly to a wall...  :lmao:

Magic Wand? Magic Wand?!!! In my trade (transmissions) I have a Magic Wrench,
I'm the guy they all bring the "tough" problems to...
I pretend to have a secret that keeps the others all in mystery and wondering "how did he fix that?"... (snicker)
If only they knew....
I'm just a little dense when it comes to this amp stuff... 

Ahhh yessss, bite sized...

The power supply now seems satisfactory (if not actually magnificent).

Magnificent! Holy carp! High praise indeed. Thank you! 

You were having thermal runaway (in the output stage?).  After all your supercharging we need to have a look at the output stage, its biasing, its thermal compensation (lack of), and its dissipation/heatsinking; its DC and thermal conditions.  I'm pretty sure that the bias tree R25/6/7/8 requires trimming, AND it has no thermal comp (so no wonder you are getting thermal runaway) so that also needs a mod.

Then we need to look at the AC conditions, is the driver doing the required job?

Then back another stage; stable?  Sufficient?

Finally the inputs, and a fully functional amp.

And from that stable foundation you can then tinker and mod with confidence.

HTH

Yes that helps immensely! But.. did you notice the output transistors?
I installed the T03 2N2148's (a few pages/posts back) they don't even get warm.
There is no more (doesn't seem to be) thermal runaway, and so far they seem to be doing ok even with the (probably) less than optimum biasing.
And Q3 and Q4 do not get warm either. Not even a little bit.

For all the rest of your above post...

Thanks! lots of good stuff there and food for thought.

I'm goin' back in swingin'... De-coupling, here I come! 

[galaxiex]

Hummm, I found a Data sheet for the 2N2148's...

It's a pdf but secured so I can't print it but I think I can link to it... 

http://alltransistors.com/pdfview.php?doc=2n2148.pdf&dire=_rca

Let's see if this works...

[galaxiex]

Ok, got a little work done.

1. Input jack shorting type
2. C16 now 1000uF
3. Added R13 1K
4. Added C9 1000uf

See schem.

Amp is "better" squeal now starts at @ 8.5 on Vol instead of @ 6.
I have been keeping the Tone on 10 as that makes it the worst.

Squeal happens even with input jack shorted to ground, so it's (obviously) the amp itself oscillating.

[galaxiex]

Just for gits and shiggles I removed C6 10uf emitter bypass cap on Q1.

I know those emitter caps maximize the signal gain of the stage.

Well... squeal is gone tho the amp lost quite a bit of volume.
Also backing the Tone down from 10 even slightly, lowers the volume quite noticeably.
Tone at 0 and the amp is barely heard and quite dull sounding.

For fun I installed a 1uF cap instead of the 10uF.
Squeal came back at 9 on Vol and amp got a bit louder again.

So my initial uneducated (as yet), thoughts... randomly...

1. The squeal originates in Q1, it just has too much gain with the better PS.
2. Q1 is ok, but too much gain overdrives the driver circuit of Q3 and Q4.
3. There is a biasing issue. Somewhere...Everywhere...?
4. The new 2N2148 output transistors bias is wrong.
5. All of the above

Thinking about it... of course lowering the gain of Q1 reduces/eliminates the squeal because turning the Vol down basically attenuates the signal and gets rid of the squeal. "sorta" the same as reducing gain.

The difference is, with C6 installed and volume down the amp is very distorted at all non-squeal settings.
With C6 gone, the amp is mostly clean with just a hint of distortion, even at full volume.
Yes, I know there is a difference between lowering gain (voltage) and attenuation. They are *not* the same thing.

Hmmm, gonna sleep on it now. It's late here and I'm tired.

[Roly]

Well another thing to consider are grounding paths.  Is there a bottleneck somewhere causing preamp returns and power amp returns to share a small common track or wire (with apparently significant resistance)?

Good grounding is single-point to the chassis (but not the mains safety earth), and radial to each of the sections from there, the sections isolated from chassis ground.  It is likely this amp has been built on the "ground everywhere" scheme, with can work, sorta.  Until somebody goes and bolts on a supercharger...    ...then moms runabout needs some chassis and suspension work.

This single ground point should also be as close to the south end of the main filter cap as possible, but not share any of its AC paths, and decoupling caps should be at the stages they decouple (the isolating resistors can be anywhere).

So the power feeds and returns should all radiate from the main filter cap; then stages don't have either supply or ground resistances in common, so they are less unintentionally coupled together via the power and ground supply.  :dbtu:

http://www.ozvalveamps.org/grounding.htm

[galaxiex]

grounding paths. 

Ahhh Ha Ha Ha ...hahaaa gaspchokesnortcoughwheeze....

Sorry... I just spewed hot coffee out my nose and all over the keyboard....

Yes, grounding paths,,, harrumph...

I actually know about star grounds etc. as I have Kevin O'Connors excellent books the TUT (The Ultimate Tone) Series 1 thru 5.

http://www.londonpower.com/audio-design-books

He writes extensively on the subject of grounding in tube amps (and other).

Too bad I read stuff without sometimes actually comprehending/understanding/retaining what I'm reading... (embarrassment)
perhaps a touch dyslexic....  :duh

In my defense, it's been a long time since I read those books and I never actually applied anything from them.
Your mention of star grounds triggered my memory. (such as it is)

Anyways... this amps grounds are probably about as random as you can get.
There are 2 sorta "L" shaped ground tracks on the board. Both originally had separate leads to ground.
One to the input jack ground, the other to one of the pots body.
I jumped the 2 tracks together so as to have one ground lead from the board.
The main filter cap ground goes to the speaker jack ground near the PS.
So the chassis is the ground path for *all* other grounds.

I will correct this before I go any further.

Thanks for the link to your site on grounding.  Great stuff!

BTW I've been reading (and actually understanding and learning ) the ESP Basics Articles on Electronics and Amplifiers.
Good stuff, easy to understand.  Learning lots and have been applying what I'm learning on breadboard circuits. 

Will report on how the ground issue/correction goes...
Stay tuned!

Oh, also BTW this amp has never had a hum issue, quite quiet actually, so I never considered it to have a ground problem.  :loco

[galaxiex]

I'm also working on a few things in parallel here...

I do make changes to the amp one at a time.

This is just my latest idea for the 3 inputs.

Normal Input is just a copy of what I already have.
High Input has a gain of 10. I lifted the circuit from an app note somewhere.
Low Input is obvious...

Will this work? Any obvious flaws?
I know the passive mixing of the three 10K resistors is not ideal...
Does the Low Input need a coupling cap?

[galaxiex]

Alright, I got the ground changes done.

1. Installed an insulated jack for the speaker. Ground of this now goes direct to main filter cap.
2. Short lead from main filter cap to chassis.
3. PS + and - now go directly to the board from main filter cap. Each go to the end of their respective traces near the PA section.
4. All 4 pot grounds now go direct to board near their respective circuitry.
5. Input jack is *not yet* insulated but I ran a jack ground wire all the way back to the main filter cap anyway.
If there is a loop, this is where it will be.
6. Input buffer is still grounded to the input jack. (mounted on it actually)

Test conditions;

C14 and C15 bypass caps on the driver transistors removed. All else just like the schem shows.

When I first fired it up I thought something was wrong, or I had made a mistake somewhere...
It was soooo darn quiet.
If I thought it was quiet before, now it's dang near silent. Barely even any hiss.

Sounds good, clean, low volume, about bedroom level. No squeal. Any settings.

Installed both bypass caps C14 and C15 (not sure how their values affect gain, if at all) used 1uF for C14 and 10uF for C15.

Test;

Squeal is back but character has changed. Due to value change on the caps?
Tone 10, Vol 2, squeal starts and is harsher and nastier than before (more high freq).

Tone 0, Vol 10, No squeal, Thick pleasant distortion. LOUD! Could gig with that sound! Crunch Rhythm only, not a good lead sound.
Bring Tone up to 2, squeal starts... nasty nasty nasty.

Next up, insulate input jack, play with C14, C15 values, see what happens...

[galaxiex]

HAH! I killed it!!!!  :dbtu:

No more squeal!!!!

More later.... 

[galaxiex]

Here are the latest changes.

1. Added insulated input jack, grounded jack on Vol pot.

2. 2N2907A's for Q1, Q3, Q4.

3. For Q1, R6-180K, R7-18K, R9-27 ohm (basically adjusted the bias to have approx 1/2 the supply Voltage on the base)

4. For Q3, R21-5.6K, R22-2K

5. For Q4, R23-1.5K C15-47uF

What killed the squeal was mostly the changes made around Q3 and Q4.
esp. R21-5.6K
I guessed that the original R21-330ohm sets the gain there,
so thought I would attempt to lower the gain.

Stuck a 10K in for R21 and got no squeal but very little volume.
Got out the R substitution box and started trying different values.
4.7K was no squeal but very loud and harsh sound.
5.6k tamed it.

Not sure how all the other changes affect things, but for reference I added the voltages measured around Q3, and 4.
What voltages should we be seeing there?
What would be considered "normal'?
Should I do something with R20-6.8K?

Amp now sounds "OK" fairly loud at Vol 10. "almost" clean at low Volume.
Tone ranges from quite thick to almost ice-pick.
All that can be tweaked as we go...

It's getting there....