Seymour Duncan Convertible 60W combo - AAAARRGGHH !

[Kid Squid]

Hwyl lads,

First of all, I got to say, I am not by a long shot, a Guitar amp expert, so go easy on me...

Had a bit of spare time today (very rare !), and decided to have a look at that bloody SD convertable combo amp, and have another go.
this time with the speaker connected, instead of a load resistor.

voltages, Tubes and the O/P transformer seem ok.
Bias is a fixed negative voltage applied at point E, and is around 60 odd Volts

Plugged a guitar in, and -Well, it's loud as hell, but it seems to have turned into a 'ring modulator' of sorts :shock: and the problem seems to be coming from around the driver stage.

if fed with a 1KHz tone, from the instrument input - at the driver grid 1 - signal is clean, but the output on anode 1 is 'modulated.
After putting a scope on the amp , getting a real wierd waveform across the speaker, looking like the B+ is somehow, getting onto the output.

i have tried with the output valves removed,(2x EL34's) but the problem is still there ( modulated driver output on plate 1)

big version is here
http://www.diyguitarist.com/Schematics/60wSDC-Schematic.gif
Output drawing is here
http://www.diyguitarist.com/Schematics/SDC60-PowerSchem.pdf
It is a differential driver amp (ECC83 - 12AX7), with one side being fed off the fx return, and the other grid is fed off the feedback tap, off the O/P transformer (16 ohm output). now what i cant seem to grasp, is, why the diode is in the cathode circuit - what is it doing ? ( it is a CIL-41301 - a current regulating diode- I've never come across one of these before !)
voltage on the anode of the diode, is 1.88 V, whereas the cathode side , is around -50V.

CR5, CR12, JMP-1, JMP-2 are not fitted in this amp (240V operation)

I'm scratching my head a bit with this.....any help is really appreciated.
Steve :thumb:

 

[PRR]

> why the diode is in the cathode circuit - what is it doing ?

It is a basic long-tail amp. Anything fed to one grid gets to both plates. The nearest in the usual way, the other by forcing the cathode node. Customarily the tail is 220K to a -300V supply. Fender used a cathode-bias standing on a 22K resistor so he did not need a negative supply (but did need grid-caps). A simple analysis suggests that the tail resistor should be infinite. Hence you often find a constant current source. If you can push point "E" to various voltages, you will/should see no change in plate voltages, confirming that the CCS diode rejects change in point "E". It also rejects change in the cathode voltage, forcing equal/opposite signal current to both triodes.

> turned into a 'ring modulator'

Ring Mod is two inputs. One input is your 1KHz signal. What is your other input???

> speaker, looking like the B+ is somehow, getting onto the output.

"B+" should be Pure DC. Which can't exist on the speaker side. What are you really seeing???

If you see 120Hz ripple, and the ring-mod has a 120Hz base.... the amplifier is 20 years old. The filter caps are failing. Most old guitar amps, you replace filter caps at the first sign of most any problem.

 

[mad.ax]

Kid Squid wrote: "Bias is a fixed negative voltage applied at point E, and is around 60 odd Volts"

What do you mean by: ODD volts?
-60 Volts is your polarization voltage, should be pure DC, dead flat. If it's not the case, check CR1, C4, & C5 on the PSU....

 

[bcarso]

In this context "odd" means "more or less", or "approximately".

 

[mad.ax]

Thanks bcarso, I wasn't familliar with that use of the word!
In this context, that seems much less "odd" to me...

 

[Kid Squid]

Hwyl Lads,

Firstly, I've got to apologise for my use of slang !,

around 60 odd Volts

and secondly, I wrote the first mail, after a slim stoly and tonic .... hic, the end of a long day !!!.
Anyway,
Thanks for the info PRR , - regarding the CCSdiode,
to check the operation basically, i need to find a way of varying point 'E', while measuring the plate voltages for changes ? if i understood correctly.

Regarding the ring mod 'effect' it is very noticable when trying to play a guitar through the amp - seems like the center frequency is around 'G'(less ringing), more so at 'C' -ringing, almost octave pedal like !?!

The reason I said about the B+ influencing the output waveform, is when i tried a scope on the output, with no input - i'm not getting a flattish line, more like a funky repeating waveform -

and previously, i scoped it, with the o/p across a 8 ohm load resistor on ch1 (top), and B+ on ch2 (bottom trace),

No Input

small input

Medium I/P

Large I/P

Excuse my camera / scope skills, cos they are not the best !!
Notice the strange B+ waveform - I thought it was halfwave rectification firstly, but i checked and replace the PSU diodes, and it's still the same.
My meter says that it is 100Hz, would it read 50Hz for half wave..just a thought.

Now i know for certain, that All of the filter caps were changed to Atoms, because it's my mates amp - and he replaced them previously - he asked me to have a look over it for him, in the idea that i could do something about it !!!
I have checked them with a cap meter, and they seem fine.

I've checked, and remade all solder joints, changed the tubes, checked the O/P transformer , which bells out as,

Across primary - 50.2 ohms
from top of pri, to CT - 27 Ohms,
from bottom of pri to CT - 23.2 ohms.
between CT and top U/L tap - 9.5 ohms,
between CT and bottom U/L tap - 10.8 ohms,
between U/L taps, - 20.2 ohms
all seems ok ?

B+ settles around 452V

Driver valve,
at point E - -51.7 V
cathode side of R39 - -45V,
anode of CCS diode / cathodes of driver - 1.73V
between cathode ( red probe) to grid ( black probe) +1.666V ( i thought this should be negative ?)

polarization voltage, should be pure DC, dead flat. If it's not the case, check CR1, C4, & C5 on the PSU....

mad.ax - I'm pretty sure, that the bias supply is very smooth , will try and get a scope pic when i can get to it. ( got my two and a half year old daughter running rings around me today, not a place for exposed B+ !!)
CR1 ,C4,C5 have also been changed.

I'm out of my depth here, but i'd love to be able to get at the bottom of it.
any suggestions, on how to proceed ?
or to put it better -

How should I have started the checks :wink:

Diolch lads (thanks)
Steve :thumb:

 

[PRR]

> Bias is a fixed negative voltage applied at point E, and is around 60 odd Volts

AND it is the only point in this amp which IS supposed to be half-wave.

Check those caps, C4 C5. Every other cap in the amp is Plus to Ground, these two MUST be Minus to ground. And when they fail, you may still read ~~60VDC, but really swinging 84V to zero 50 times a second.

And with 50Hz pumped up through driver cathodes and on EL34 grids, you really do have a Multiplier, which is the general case of Ring Modulator (but without a diode-ring).

--------------------------------------------------

> the CCSdiode, to check the operation basically

I doubt it is a problem; I was just teaching how-it-works.

If you think it may be a problem, replace it with a 22K resistor. It will work; the CCS was a frill.

> All of the filter caps were changed to Atoms, because it's my mates amp - and he replaced them

If someone else replaced caps, and the amp gives trouble, assume the cap-job may be at fault. It seems easy, but there are ways to do it wrong. Even ways that the amp will work for a while before failing.

Does it still make funny-sound when you patch a CD or iPod into the Effect Return? (Eliminating all those chips?) What do you get when you patch from Effect Out into a standard PA power amplifier and speaker, is it OK? (without EL34 "tone", of course.)

You want to go over all the DC voltages. That chart does not show them, and the links are not working today. But 60W from EL34 implies about 500V supply to OT and EL34 plates. This seems to also be the EL34 screen voltage (ouch). The grid bias is supposed to be 60-odd, but measure it again AT the grid pin. It will read a few percent less due to meter loading the grid resistor. If it is somewhat off, the EL34 may be gassy. If it is way positive, the grid capacitor is leaky.

Point A is only 500 ohms down so the driver tube also has nearly 500V at top of the 100K. Where the 100Ks hit the plates, we expect 70%-50% of that, 350V-250V. This driver has its grids at ground, check that. The cathodes should be 1V to 4V positive of ground.

Do not blow up your probe, scope, child, or yourself checking these high voltages! You can not run ~~500V DC into a scope, and many modern scope probes are rated 350V.

The first B+, to OT, will show 10V-30V ripple. Point A has remarkably little filtering, and must be nearly as high. I'm not sure I believe that R55 is 500 ohms. 5K seems more likely. In that case point A would show a couple volts ripple. And the ripple must be double line frequency. If you got 50Hz bumps, it is half-wave, and you don't build an amp this big on a half-wave rect.

I'm totally baffled by the Full Wave / Doubler power supply. Surely they don't use the same PT primary for 115V or 230V; that would cost twice too much. And the heaters come out wrong or must be re-wired. Maybe they have two different PTs with two different HV windings, and wire the rect accordingly.

 

[Kid Squid]

Thanks a lot for the reply PRR,

Really appreciate you taking some time over this problem.

AND it is the only point in this amp which IS supposed to be half-wave.

Check those caps, C4 C5. Every other cap in the amp is Plus to Ground, these two MUST be Minus to ground. And when they fail, you may still read ~~60VDC, but really swinging 84V to zero 50 times a second.

I'll double check those, hopefully over the next day or two, I'm back in for a few shifts, so I'll try and have a check in a break.

And with 50Hz pumped up through driver cathodes and on EL34 grids, you really do have a Multiplier, which is the general case of Ring Modulator (but without a diode-ring).

this sounds like the very thing i'm hearing.

> the CCSdiode, to check the operation basically

I doubt it is a problem; I was just teaching how-it-works.

If you think it may be a problem, replace it with a 22K resistor. It will work; the CCS was a frill.

Understood.

All of the filter caps were changed to Atoms, because it's my mates amp - and he replaced them

If someone else replaced caps, and the amp gives trouble, assume the cap-job may be at fault. It seems easy, but there are ways to do it wrong. Even ways that the amp will work for a while before failing.

I will check over them again, just to be 100% positive.

Does it still make funny-sound when you patch a CD or iPod into the Effect Return? (Eliminating all those chips?) What do you get when you patch from Effect Out into a standard PA power amplifier and speaker, is it OK? (without EL34 "tone", of course.)

I'll try this as soon as i get a chance - should have thought about doing that earlier, will narrow the 'search' down. :guinness:

I took some voltages a while back, when i first started having a look at it -

** EDITED *** - Point readings were not correct - my fault they have been corrected in my last post.

I'll check the value of R55 to make sure.
also, I'll dig out the other measurements (O/P and driver stage) when i get home tomorrow.

I'm totally baffled by the Full Wave / Doubler power supply. Surely they don't use the same PT primary for 115V or 230V; that would cost twice too much. And the heaters come out wrong or must be re-wired. Maybe they have two different PTs with two different HV windings, and wire the rect accordingly.

I'm not actually sure that they ever implemented the doubler PSU - I read it somewhere , while browsing for info. I could probably find out from paul at www.diyguitarist.com, he's pretty much the only useful info i've found on the net, regarding these amps.
They can't be all bad - Jeff Beck apparently endorsed them.... :wink:

Do not blow up your probe, scope, child, or yourself checking these high voltages! You can not run ~~500V DC into a scope, and many modern scope probes are rated 350V.

I got a 100:1 , 1.2KV probe i used for checking the B+ Volts and driver volts. - thanks for the heads up tho !
Ta mate,

Diolch, nos da (good night)
Steve :thumb:

 

[PRR]

> Point D - 404V DC, 15mV ripple.

????? Point D has two 16V Zeners to ground. It should be 32V, 33V DC! I may be mis-reading the part numbers, but 404V isn't right.

OTOH, the rail is labeled A C B D, with E and I and H and G... you or the draftsperson may have got confused.

AFAICT, we only need A and E to get the tubes happy.

> Jeff Beck apparently endorsed them....

Sometimes, a traveling musician needs free gear and service. Beck could play anything, tone is in his hands.

That IS a nice muscular preamp.

What is M-1 M-2? Do different preamp and tone circuits plug in?

 

[Kid Squid]

oops

My apologies PRR,
correct readings are

Junction of R61, R55 ( B+) - 450V D.C - 25V AC ripple
Point A - 445VD.C. - 2.6V AC ripple,
Point C - 402V D.C - 70-80mV AC ripple,
Point B - 336V D.C - 15mV AC ripple.

Somehow - due to tired eyes, I had confused myself......

AFAICT, we only need A and E to get the tubes happy.

Agreed.

That IS a nice muscular preamp.

What is M-1 M-2? Do different preamp and tone circuits plug in?

Apparently, these are quite a good amp for the money - second hand on ebay all of the time.
M-1, M-2, are indeed sockets, for different modules to plug in. there are quite a range of them.

here's the two that came with the amp

Anyway, I'll call back a bit later - my job is on stop at the moment - (waiting for the riggers to do what riggers do best !!)
Thanks again PRR :guinness:

 

[Bryson]

[quote author="Kid Squid"]
They can't be all bad - Jeff Beck apparently endorsed them.... :wink:

[/quote]
Until they failed him on the road.

Jeff is an old pal of Seymour's.

The amp was full of cool ideas and sounds, but it had reliability problems which I believe led to it's demise.

 

[Kid Squid]

Hopefully, i'll have a bit of time this afternoon, to get a little further with this.

I'll write up what i find, and post it, this evening.

Until they failed him on the road.

Jeff is an old pal of Seymour's.

The amp was full of cool ideas and sounds, but it had reliability problems which I believe led to it's demise.

This is what i'm currently feeling....

Just an FYI - The power amp section of the Seymour Duncan Convertible 100 was originally designed by Bruce Kennedy and Kevin Beller (chief engineer at Seymour Duncan). Hi-fi tube guy Roger A. Modjeski, who used to make Music Reference tube hi-fi amps and sold RAM Labs tubes was hired to review the design and add his input. The preamp section was done in-house at Seymour Duncan.

just a little bit of info, i just come across...
Catch y'all in a bit,

Steve :thumb:

 

[Bryson]

Yeah, Bruce was into all the Dynaco amp mods back then (he turned me on to them)... probably what inspired the ultra linear o/p design

 

[Walrus]

Quote: Point A - 450VD.C. - 25V AC ripple,
I do hope that is peak to peak and not RMS!! :shock:

 

[Kid Squid]

Walrus,

Has been corrected.

Junction of R61, R55 ( B+) - 450V D.C - 25V AC ripple
Point A - 445VD.C. - 2.6V AC ripple,
Point C - 402V D.C - 70-80mV AC ripple,
Point B - 336V D.C - 15mV AC ripple.

Did notice something yesterday, regarding the bias voltage.
when measuring with voltmeter,at point E -voltage is fluctuating between -51.2~-51.4 VDC,
I had a look with the scope (and i must confess, i'm not 100% with using it !)
at a resolution where i'm looking at the bias volts it looks like pretty much a straight line.

however, when i 'magnify' the waveform to get a closer look - the bias waveform is jumping about vertically, at random intervals.

is this the 200mv variation ?, i would have thought the bias caps (C4,C5) would smooth this out.
So, I changed the bias caps, R62,63, and CR1 noting the correct orientation, and the amp is the same.

what should my B+ to the O/P transformer C/T look like, mine looks like this ( top trace),(bottom trace is O/P across speaker no signal in)

and here's a pic of B+ top, and speaker out ( 1/4 vol) bottom trace.

This what I mean about the B+ waveform, somehow influencing the output waveform.

should the B+, be without the humps ?

Ater trying the 'magic chopstick' method, noticed that if i tapped R54, R55, i could hear the taps throught the speaker, elevating into static like bangs !?! - I'll change them when i get the replacements.

So as you can see, i got a little bit of knowledge, but as they say - that can be dangerous..... :wink:

Any Ideas ?

Best,
Steve :thumb:

 

[Kid Squid]

Hello Lads,

Got myself a bit of time, so I thought i'd have another go at the amp.

Does it still make funny-sound when you patch a CD or iPod into the Effect Return? (Eliminating all those chips?) What do you get when you patch from Effect Out into a standard PA power amplifier and speaker, is it OK? (without EL34 "tone", of course.)

If signal is patched in to the effects return - Problem is still there.
if signal is taken from the effects send, into another amp - problem is not there.

So, problem, is from driver / Output section.
Hopefully, I can get a little bit more time spent on it, over the weekend.

Any tips, on where to start ? :shock:

Steve :thumb:

 

[willtube]

Hi, I do realize this is very very old thread, but I just found it and I happen to have a seymour duncan convertible 60 with a very similar problem: loud 60Hz hum (here in the U.S.) after the volume pots (not affected by changing volume or gain knobs). I already swapped out all the caps. I did poke at C4 and C5 which did make some noises, so i re-soldered them. Then the amp worked for a while but now the problem is back.

Anyways, just wonder if you've ever found the issue or gave up? At this point I'm pretty close to giving up...

Many thanks for all the info already posted on this thread. I guess I can always reuse the cab/speaker/reverb tank...

will

 

[willtube]

Well seems like maybe the convertible 60s are slowly disappearing from the world... Anyways, I think I have found the problem and figure I post it here in case it would help another frustrated owner.

After changing the preamp, phase shifter, and power tubes, I was still able to reproduce the AC hum if I assemble the amp back into the case. The hum would come on after the amp has warmed up. I measured the temperature inside the enclosure and it was about 110F (~43C). So I figure maybe it was a capacitor or resistor that would go out of spec when the temperature is up.

I had already recapped the amp maybe six years ago, and since then I haven't played the amp much and all the electrolytics looked ok. But when I poked around the parts/board I was able to get it to buzz when I push on the bias caps (C4/C5) and sometimes the power filter caps (C7/C8? the 250V, 330uF radials). I was also able to simulate the heated environment by blowing on the board with wife's hairdryer when it was outside of the case. So I figure I'd replace the power filter caps first. So I ordered a couple of new ones and replaced them. Also fixed the PCB trace under them since they started to come off the board. Tested continuity and resistance and they seemed fine. Tuned on the amp and give it the hairdryer treatment and the buzz is back after 10/15 seconds. Hmmm...

After few days of deliberation, decided to give the board the magic chopstick test again. This time I poked on the board around C4/C5/C7 and C8, and at some point I accidentally tapped the effects return socket and got a buzz! Sure enough if I poke at that socket the buzz would come back momentarily. So I plugged in both ends of a cable into effects send and return, and wasn't able to get the amp to hum. Then I tried some contact cleaner on a 1/4 jack and exercised the effect return and send a bunch of times. Also stuck a rag inside too. Since then the amp seems to be ok.

So after all the trouble and troubleshooting it turned out the ac hum/buzz was caused by a dirty contact in the effects return socket! Well at least I learned tons about tube amps and in the process acquired a new small replacement amp (fender this time. no more seymore duncan!). Let's see how long the c60 would last this time...

will

 

[peterc]

This is also a problem with the Marshall 610 combos as well.

I eventually replaced the send & return jacks with Neutrik units, works well.

Peter

 

[willtube]

It's funny how it often comes down to the simple, physical stuff. Rust and oxidation!