18w TMB issue: dying OT?

[SleepLess]

Hi!
I'm back here for yet another amp issue. You'll see a couple of pics below and you'll instantly see that I didn't build it. The wiring is real sloppy. This amp belongs to a friend of mine who thought he could build an 18w without any prior experience. The result is that his amp didn't work at first, he even smoked the Xicon 1K5 5W on top of the 32uF F&T cap... Then he decided to rebuild it all over again. This time nothing went up in smoke but he still has an issue and asked me to solve it.
The amp works fine on the Normal channel, no issue. On the TMB channel though there is an issue with the Treble and Mids pots. If they're turned down to 0, 1 or 2 the amp works OK. But as soon as I reach 2 on either pot a high squeal shows up. It sounds like an old radio hiss/squeal. The squeal doesn't get any louder up to 10 and it's bearable but in the meantime the more I turn up the pot the less volume the amp puts out... Ain't that weird? Never seen such a thing!
My plan is to rewire the whole amp and make it clean. But if in the meantime you can point me to some tests or precise parts in the amp that'd be greatly appreciated! So far all I've done is rewiring the pots to the board. I haven't rewired the board to the sockets yet.

Good pics of a sloppy build:





Thanks!!!

[John]

The squeal may be as simple as moving some wires around, which you'll do anyway on the re-wire job. And of course check the solder joints, if one is bad that can cause some oscillation too sometimes.

Don't be so hard on your buddy. His  biggest crime was getting bit by the amp bug, that's all. His next one will be better.

[Tyrannocaster]

I don't think he did such a bad job on this one either; it's neater than I expected from your description. Fact is, a TMB is not that easy to build without issues and oscillation is a common problem; what you describe sounds exactly like oscillation. I would try using shielded cable in the signal path during the first part of the circuit at least. Between shielded cable, chop-sticking leads and shortening wires and moving them away from other wires I bet you can sort it out.

[SleepLess]

I don't think he did such a bad job on this one either; it's neater than I expected from your description. Fact is, a TMB is not that easy to build without issues and oscillation is a common problem

That's why I told him to build pedals or simpler amps first... But wait until you see the wiring made over...  
I found two bad solder joints, I managed to pull the wires out of their lugs by pulling on them with two fingers... Noise is still here... I am moving forward on the rewiring...

[Colas LeGrippa]

As I can see, coupling caps are on the board and relatively long signal wires run from these caps to the pots which can cause oscillation. By soldering the coupling caps right onto the pots and using signal  shielded wires to the next stage HELP A LOT in eliminating oscillation. The idea is to keep HV leads all the way to the pôint where you need a coupling cap, HV leads cannot be influenced by anything. Maybe you should try that before inserting ceramic caps here and there in the circuit to tame the oscillation, which cause your amp to loose it's ''MOJO'' . I have never had any oscillation problem since I use this method. I admit that soldering a cap on a pot doesn't look as neat as your beautiful layout tough, but if you use a small terminal strip to hang it on, it should look fine.

Colas

[SleepLess]

OK. Wiring made over except the grounding wires. The amp works a lot better but I still get that squeal on the treble pot above 5/10 instead of 2/10. The mids pot works fine now... On to the grounding... I hate those ground busses...

New pics:

[Tyrannocaster]

All I can say is I wish you had been around to fix all the amps I made that had problems. 

I still have my first "succesful" amp, which happened to be a TMB that I made before the idea existed. Richie Hall helped me with the circuit design (this was before he actually did the TMB that everybody builds now) and I built it point to point; that thing is the ugliest, scariest, most poorly built amp you could imagine...but it works. And it sounds so good (it actually sounds GREAT to the point that I am afraid to mess with it) that I have often wondered if Gerald Weber was right when he claimed that point to point's three dimensional aspects made amps sound better might not be right. But the fact is that the thing would be an absolute nightmare to work on and I have built all my amps since then using eyelet boards, which has helped the success ratio immensely.

Seems to me that you will soon have something as good to give back to your friend. He doesn't know how lucky he is.

[SleepLess]

Thanks! But wait a minute... Now it looks good, but I still have that damn squeal on the treble pot... Damn, I'm stumped, I don't know where to look...

[John]

Try bypassing the tone pot altogether... might be a dodgy wiper? overheated resistor/cap? If this is his first build, maybe he made some of my first-build mistakes.

I'm *ass*uming you have gridstoppers everywhere?

[Willabe]

Not saying this will fix the squeal, but it looks like you have a ground loop from the IEC socket to the PT and 2 ground lugs that are bolted to the chassis? Wht., Blk. and Grn./Ylw.?

The 2 B+ wires from the can cap are routed by/with the heater wires and also since the heaters and PT B+ secondary wires are going right by the chassis lug their CT wires should be twist together with the other 2 wires of each wind/supply.


                    Brad      

[SleepLess]

Here's the last picture of what I've done. I still have that squeal on the treble pot, the middle pot now works flawlessly.



I have three grounding spots:
1. PT bolt has the IEC ground.
2. Second PT bolt has the first filter cap ground (500uF/50V), negative side of filter cap can, HV CT (blue) and Heater CT (black).
3. OT Bolt is the buss bar ground. Pots, input jacks and board grounds go there.

 

[Willabe]

It looks like there's a Grn./Ylw. and Black. wire connected at the PT. Grn./Ylw. goes to 2'nd PT bolt. Blk. goes to IEC socket and from the IEC socket that Black. wire connects at the IEC socket to a Wht. wire that goes to the 1"st PT bolt? If this is right then as the 2 PT bolts will connect through the chassis, you now have a ground loop.

Or is the Blk. 1 end of the primary and the Grn./Ylw. the PT screen/shield connection?


                    Brad        

[Willabe]

I have three grounding spots:
1. PT bolt has the IEC ground.
2. Second PT bolt has the first filter cap ground (500uF/50V), negative side of filter cap can, HV CT (blue) and Heater CT (black).
3. OT Bolt is the buss bar ground. Pots, input jacks and board grounds go there.

You have a 4'th, speaker jack/OT ground to a chassis lug. This should ideally be grounded at/with the PP output tubes cathode ground point.


                     Brad     

[SleepLess]

Damn, thanks Willabe, I went too fast:

I have four grounding spots:
1. PT bolt has the IEC ground.
2. Second PT bolt has the first filter cap ground (500uF/50V), negative side of filter cap can, HV CT (blue) and Heater CT (black) and PT Screen wire (yellow/green).
3. OT Bolt is the buss bar ground. Pots, input jacks and board grounds go there.
4. Speaker jacks ground, near the filter can cap.

[archaos]

You have a 4'th, speaker jack/OT ground to a chassis lug. This should ideally be grounded at/with the PP output tubes cathode ground point.

Once again, both Merlin & KOC rule here.  

[SleepLess]

Or is the Blk. 1 end of the primary and the Grn./Ylw. the PT screen/shield connection?

That's it... Black one says 0V and connects at the IEC with the indicator lamp white wire. PT Grn./Ylw. is the screen.

[Willabe]

That's it... Black one says 0V and connects at the IEC with the indicator lamp white wire. PT Grn./Ylw. is the screen.

OK, then no ground loop, sorry buddy.     


                   Brad     

[SleepLess]

That's it... Black one says 0V and connects at the IEC with the indicator lamp white wire. PT Grn./Ylw. is the screen.

OK, then no ground loop, sorry buddy.     


                   Brad     

No problem Brad, any hunch is good examining at this point so thank you...
I have left the amp aside, I need to scratch my head for a while on this one... The wiring makeover got rid of the Mid pot squeal, but I still have it on the treble pot, though later on the course of the pot and not as loud as well... What the hell could it be now??? I have tested the pot and it goes from 0K to 250K so it's good... Maybe one of the two 0.022uF caps on the board are bad...?

[PRR]

Get it squealing, poke wires with a dry chopstick or wood pencil.

On one of my squealerss, putting a pencil (lead) between two wires made the squeal worse. Moving those two wires well apart was a cure.

[Fresh_Start]

Damn, thanks Willabe, I went too fast:

I have four grounding spots:
1. PT bolt has the IEC ground.
2. Second PT bolt has the first filter cap ground (500uF/50V), negative side of filter cap can, HV CT (blue) and Heater CT (black) and PT Screen wire (yellow/green).
3. OT Bolt is the buss bar ground. Pots, input jacks and board grounds go there.
4. Speaker jacks ground, near the filter can cap.

Are the power tube cathodes also grounded on the second PT bolt?  Should be with "reservoir" filter cap and looks like it is.

You have a 4'th, speaker jack/OT ground to a chassis lug. This should ideally be grounded at/with the PP output tubes cathode ground point.

Once again, both Merlin & KOC rule here. 

The TMB doesn't have negative feedback IIRC,.  However, I don't remember where I picked this up, but I have always grounded the speaker jack(s) where negative feedback is grounded.  IOW on the preamp bus. 

Also, that speaker jack ground has got a lot of current going through it and needs to be kept clear of any sensitive signal leads.

Moving away from grounding, did you use shielded wire to go from input jacks to the first triode(s)?  How about from tone stack back to recovery stage?  Those are the smallest signals in the amp and, therefore, most likely to pick up noise.

Where are the grid stoppers for the power tubes located?  If they're on the board, you might try putting them directly on the power tube sockets.

Last thought - have you tried grounding the preamp bus as close to the input jack as possible?  I always put it there, as far away from the noisy, big current grounds at the other end of the chassis.

Please don't think I'm pretending to be an expert.  Grounding drove me crazy on the first couple of builds and these comments just reflect what's worked for me.  There are many ways to skin this cat.

Hope

[SleepLess]

Hi guys!
OK, amp is made over and the squeal is gone. I think that for once the culprit wasn't a single component or wire, the squeal was heard because of a conjunction of multiple reasons among which bad soldering and bad wiring.
At first I moved the speaker jacks ground wire with a Chinese wooden stick and the squeal changed. So I soldered that wire to the buss wire, going underneath the board. The squeal was a lot quieter. Then I realized that all the yellow wires going to the sockets weren't lifted above the chassis as I usually do. Well I did it and now the squeal is totally gone!
One more lesson I've learnt: wiring really has a tremendous importance on how the amp feels and sounds. I actually never realized that before because my amps never had issues because I have perfect layouts at hand. But on this one, since I didn't have a perfect layout the smallest wire move makes a difference, it's amazing!
Watch your wiring guys and thanks for your help! Here's the last pic of a working, non-squealing 18w TMB! Never ha the chance to play one, it's actually funny how the amp seems to have a fuzz pedal in front of it past 7 on the volume knob! John Mayer's Crossroads tone without a fuzz... 

[Fresh_Start]

Way cool!
 
Thanks for the "After" pics.  I thought about saying something about vertical separation of those leads around the preamp sockets but figured it was nitpicking.  I always try to put the plate leads flat on the chassis and keep the grid leads coming straight down at the socket and away from everything else.

It's easy to forget that we have THREE dimensions to use to separate signal wires from big current sources.  I had tremolo injection into the phase inverter of two Blackface Reverb builds until I figured out that coming straight down from 1" above the board gave enough separation.

Good to know you got it worked out and, especially HOW you solved the problems.

Cheers,

Chip

[Willabe]

Yep. Good job.  

Very interesting about moving the OT's ground. 


                    Brad      

[SleepLess]

HI.
Back with this one here and I'd just like to have my diagnosis confirmed by you guys because I think that the OT has a problem here. I have 1.87V on pin 7 of V4 and 347V on pin 7 of V5. B+ is the brown wire on this one (353V) and goes to the cap can.

I took some measures because the amp works (though it is noisy) but V4 lits up a lot brighter than V5 and the sound is not as it should be. It seems that only V4 is working in the amp, sort of. But a bad OT would mean no sound at all, right?

So I measured resistance between the two primary wires (red and white, B+ is brown on this OT) and got overload (2K range). I measured resistance between red and brown (B+) and got overload. And I measured resistance between white and B+ and got 137ohms... Which tends to show that the OT is open and that the amp needs a new one, right?
Tell me what you think on this one.
Here's my full voltage chart, all heaters are at 3.2VAC:

         V1          V2         V3          V4          V5          V6
1       186        194        150          0            0           295VAC
2       0            48         0             0            0           0
3       1.53        69         1.1         8.9          8.9         353
4                                                                          43
5                                                                          43
6       139        192        242          0            0           0
7       0            48         148         1.87        347        295VAC
8       0.85        69        148           0            0           0
9                                              301         301



Here's the OT schematic btw:



Thanks!

[six el six]

Use R. Keen's method to test an OT with a small ac voltage on the secondaries.

It's on the GEOFEX tube amp debug pages.

[SleepLess]

Use R. Keen's method to test an OT with a small ac voltage on the secondaries.

It's on the GEOFEX tube amp debug pages.

I don't have a variac I can't do this test. But I've again measured the resistance across the B+ lead of the OT and the two other primary wires and I have 137ohms for the white one but overload for the red one, no matter the MM setting, from 200 ohms to 200Mohms... What surprises me is that the amp is playable... Though the sound isn't as good as it should. That's why I was hoping for a confirmation here. To me a bad OT means no sound, but here, it's playable...

[phsyconoodler]

How are you measuring the leads? You have to have no tubes installed to get your readings and all wires disconnected from anything that can cause external interferance.

[SleepLess]

How are you measuring the leads? You have to have no tubes installed to get your readings and all wires disconnected from anything that can cause external interferance.

I had the rectifier and EL84s pulled out but the wires still soldered to the sockets... Must I unsolder them?
Thanks!

Still this amp sounds... V4 shines a lot brighter than V5. But I know it's not the tube because I swapped them and V4 is the one that glows redder... 

[SleepLess]

Could someone just tell if it's normal that I should only have 1.87V on V4 pin 7 when I have 347V on the same pin on V5? I can't find any 18w TMB non master-volume voltage charts anywhere...   

If it's normal then I have a working amp, if not, then I might have a dead OT, or dying... Since the amp can be played...

[tubeswell]

Use R. Keen's method to test an OT with a small ac voltage on the secondaries.

It's on the GEOFEX tube amp debug pages.

I don't have a variac I can't do this test.

You don't need a variac - all you need is a low VAC supply on the OT secondary, which you can get from the PT's heater winding, or rectifier winding

[phsyconoodler]

Step one: make 100% sure you have a good connection when you are reading voltages.
  Your pin 7 voltage on one tube is waaaayyyyy off.
Step two: make 100% sure you don't have the center tap on the low voltage pin and one leg of the primary on the spot where the center tap should be.

 try this: drain the caps,pull the tubes.Lift the center tap wire off the board.Measure with your meter from each wire being the 'center tap' if you get a balanced reading anywhere,you may have just mixed up the center tap with one of the plate leads.Been there done that.

 If not,then I would say the OT is shot.Last resort every time,cause why is the OT shot?

[SleepLess]

Step one: make 100% sure you have a good connection when you are reading voltages.
  Your pin 7 voltage on one tube is waaaayyyyy off.

Thanks a lot, that's what I wanted to have confirmed! I'm supposed to have 347V just like on V5, right? I have taken this measure about ten times today. I never got 347V there, I have 1.87V all the time...

Step two: make 100% sure you don't have the center tap on the low voltage pin and one leg of the primary on the spot where the center tap should be.

 try this: drain the caps,pull the tubes.Lift the center tap wire off the board.Measure with your meter from each wire being the 'center tap' if you get a balanced reading anywhere,you may have just mixed up the center tap with one of the plate leads.Been there done that.

 If not,then I would say the OT is shot. Last resort every time,cause why is the OT shot?

The center tap on the OT? You mean the B+ wire, the brown one as per the schematic I posted above? I don't understand when you say I must lift it off the board... It's going to the capcan...  

[SleepLess]

In the meantime I have unsoldered the three OT primary leads, taken the rectifier and output tubes out and measured resistance again, this time they're not linked to anything, they're just dangling in the air...

MM set on 2K range
Brown/white: 137Ω
Brown/Red: overload
White/Red: overload

I'll wait for phsyconoodler to help me understand what he asked but this definitely smells like a dying OT...
Thanks.

[phsyconoodler]

Ok,the three leads.Lift the one going to the cap can.Then measure with your ohmmeter to the other leads.What is the reading?

 I'm thinking the center tap is not really the center tap and your wiring is wrong.That would explain the voltages.
  Have you ever played an 18 watt before?If you think it sounds ok as is with those readings then when it's right you will be quite surprised.

 My thinking is that what you assume is the center tap is NOT really the CENTER tap.Guessing of course but it would explain a couple of things.

[SleepLess]

Ok,the three leads.Lift the one going to the cap can.Then measure with your ohmmeter to the other leads.What is the reading?

 I'm thinking the center tap is not really the center tap and your wiring is wrong.That would explain the voltages.
  Have you ever played an 18 watt before?If you think it sounds ok as is with those readings then when it's right you will be quite surprised.

 My thinking is that what you assume is the center tap is NOT really the CENTER tap.Guessing of course but it would explain a couple of things.

Like I said:
MM set on 2K range
Brown/white: 137Ω
Brown/Red: overload
White/Red: overload

I'm not assuming anything except that the maker's schematic is correct and that in this case, as the schematic shows B+ is brown and not red, which may actually not be the case...
I have reversed the brown and red wires now. Brown (supposedly the B+) is now going to V4 pin 7 instead of the + side of the cap can and the red is now going to the capcan instead of V4 pin 7. I'll turn the amp ON tomorrow and see if it goes boom... At least I'll know why I have dead OT...

It's my third 18w build or fixing. My first TMB though, the other ones were the tremolo 18w.

Thanks!

[SleepLess]

Allright, so I fired the thing up this morning. I remind you that the B+ wire is supposed to be the brown one.

Three possible wirings:
- Brown to capcan + / white to V5 pin 7 / red to V4 pin 7: the wiring that was chosen on this amp. With this I have 1.87V on V4 pin 7, 347V on V5 pin 7. Amp is playable but V4 glows a lot redder than V5 and obviously 1.87V on pin 7 V4 is not good.

- Brown to V4 pin 7 / white to V5 pin 7 /red to capcan +: the amp emits no sound. I have 4V on both V4 and V5 pins 7. 355V on capcan +

- Brown to V4 pin 7 / red to V5 pin 7 / white to capcan +: the amp is playable but this time it's V5 that reads 2V and V4 345V. This time it's V5 that glows redder than V4. 345V on capcan +.

So can I now conclude that the OT primary's red wire is dead?
Thanks!

[sluckey]

So can I now conclude that the OT primary's red wire is dead?

Yes.

Your original voltage readings indicate half of the primary winding is open. Your ohm meter readings verify this is the case. There's no need for further testing on this type failure. Just replace it. And always wonder, what caused the OT to fail? Was the OT the culprit? Or did a shorted EL84 take it out? Or did I play it hard with no speaker connected? Or did I accidently short the plate to ground?

The OT isn't totally dead, but it is half dead. One output tube is still working. That's why you still get some sound.

[SleepLess]

Thanks a lot Sluckey! 

[phsyconoodler]

Ok,bear with me here.
 Take the three wires in your hand.Brown,red,white.Is that correct?
Take your black ohmmeter lead and put it on the red wire.Then measure the ohms to the brown wire and then the white wire.What are the readings?

-then the black ohmeter lead to the brown wire and other to the red and then the white.Get my drift?

I know it's likely shot,but humour me.

[SleepLess]

Ok,bear with me here.
 Take the three wires in your hand.Brown,red,white.Is that correct?
Take your black ohmmeter lead and put it on the red wire.Then measure the ohms to the brown wire and then the white wire.What are the readings?

-then the black ohmeter lead to the brown wire and other to the red and then the white.Get my drift?

I know it's likely shot,but humour me.

Like I said:
MM set on 2K range
Brown/white: 137Ω
Brown/Red: overload
White/Red: overload

I have ordered a new OT... 

[SleepLess]

OK. New OT came in on Thursday and I installed it yesterday and checked it first, it's good.

I fired up the amp today and it works, save for an unpleasant hiss on the TMB channel only when I turn up the Treble and Mids pots. I can turn the Bass and Volume pots all the way up I don't have any hiss, but the hiss comes in when I turn the Treb and Mid pots above 2...

I have pulled the three preamp tubes one at a time and the hiss stops when I pull V3, not before. Therefore I have changed the two 0.02uF caps and the 250pf cap that are linked to these two pots and V3, to no avail... I still get taht damn hiss...

I need you guys... Thought I'd be done with this pain of an amp...

Thanks!
(Oh and I also need your help on the 5F4 issue... See page 1!  )

[Willabe]

Could be a noisy plate load R.


           Brad     

[SleepLess]

Could be a noisy plate load R.


           Brad     

Do you mean the 100k resistor on V3 between pins 1, 6 and 7? I measured it and it's 98.5K and it's carbon film, not carbon comp... Do you suggest to change it anyway?
Thanks!

[Willabe]

Yeah from the pics I saw it was carbon film. Still could be a bad one.

Do you mean the 100k resistor on V3 between pins 1, 6 and 7?

 
Could you post a schemo for the amp? Or a schemo of the same type of PI, if that's what V3 is? Sounds like V3 is a CF?

I don't see an 18w TMB in Doug's library.


                  Brad      

[SleepLess]

There you go...

http://www.el34world.com/Forum/index.php?action=dlattach;topic=14351.0;attach=32127

[Willabe]

Thank you but I really meant a circuit schemo, not a layout schemo.

Looks like V2 is the PI and V3 is a CF for the tone stack?


               Brad     

[SleepLess]

There you go then...

http://www.el34world.com/Forum/index.php?action=dlattach;topic=14351.0;attach=32128

[Willabe]

Now that's what I'm talkin bout.    

Ok, V2/V3 are swapped in the 2 circuit and layout drawings.        That's a PITB.

So anyway, yes if you have a few 100K's laying around, I'd change the 100K plate R on the PI tube socket and if that doesn't get it then I'd try the 56K cathode R, then the 2, 100K plate R's on the PI. (Maybe I'm wrong and less chance but why couldn't a CF R be noisy just like a CC R?)

OTOH, guys have been having more and more problems with bad silver mica caps. (Although I think some are just cooking them when they solder them in.)

So if the 100K's don't fix it, maybe try changing the 250pF treb. tone cap.

These are the things I'd try 1'st.    

Did you use grid stoppers on the EL84's? I wonder if they need to be tamed a little?

                  
                           Brad        

[Willabe]

Here's a link to Douglas library on a listing amp tool.

http://www.el34world.com/Hoffman/tools.htm

That would help you track down where/which stage the noise is coming from.


             Brad         

[SleepLess]

Got it guys!!!!

Through wiring and rewiring I had done so much stuff that I tried to swap the wires going to V2 pins 1 and 6. That did the trick, the amp is not hissing on the treble and mids pots anymore!

Now I find it to be a bit noisy, but 18w are noisier than their bigger brothers if I remember correctly... 

Now on to my last issue: that damn 5F4!!!!

THANK YOU!!!!!