Epiphone Galaxie 10 - Power Transformer and other issues

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

lassoharp

Well-known member
Joined
Jan 3, 2009
Messages
2,100
Location
USA
A client just brought one of these to me recently.  Before it got to me it had already burned up one OT and ate three brand new 6L6GCs.  Another tech had replaced it and the next time he used it the cathode resistor exploded.  That's when it came to me.

This looks to be partly a problem of design choice for the power transformer and partly a problem of the consistency of new production power tubes.

I'm posting details here for reference and discussion.

5W cathode resistor was in crumbs and the heat apparently blew up the bypass cap which was mounted very close in // on the pc.

Existing tube was pulling something crazy around 157ma.  Other two arced like mad.  All new production.

I replaced cathode resistor and bypass cap and put in a used vintage 6L6 I had in the drawer.

It idled at 69ma in less than 5 min.  Stock figures say 413 plate at 30.91 bias = almost 66ma = 2W = ok for 5W bias resistor.

69ma will eventually rise to 70-72ma = 3W = real hot and probably too hot for longevity.  But, that's about the best figure I got.

Client went out and bought another brand new 6L6GC 30W 500V speced.  It hummed right away and was drawing 80ma in less than 5 min. about 3.6W = cooking hot.

I managed to gather up 3 more assorted new production 6L6GCs, lightly used and all drew over 70ma.

And it's no wonder.  PT is 800V CT unloaded.  I can't see why they would even choose that high of voltage for this amp.

HV DCR is 563 ohms which charts to something like 55ma rating.  I haven't done a hand test on it yet.  You'd expect a little higher than a Champ PT so around 100ma.

Size comparison to Weber's Champ OT is interesting.  Same dimensions excepting the bell width - Weber bell is 0.25" wider.  Epiphone PT has only single filament presumably good for 1.2A.  Weber has 3A rectifier and 4A of 6.3V plus 80ma HV.  You would think the Epiphone would be closer to 100ma HV.  It may be but that high of DCR on the sec at that voltage is usually in the 50-80ma range on most of the PTs I've measured. 

Solution(s).  It didn't make sense to have the client keep buying NON-refundable new prod tubes at 17-20 a pop in hopes of finding one that biased up to Epiphone's stated values . . . . and stayed there without tearing up.  Vintage tubes might have did a little better but there were still other concerns.

A 10W cathode resistor would survive.  The original OT apparently won't.  You may need a bigger one depending on how much current the tube ends up pulling.  And, the PT might be running too hot or way too hot.

There is room to install dropping resistors between the stand by switch and spliced into the pc board lead.  Could drop about 40-50V with 750 ohms total.

You could install choke input to drop about 50V.  Might be an interesting new sound.

I decided to just buy a Champ OT from Weber for 33 bucks.  Has two HV taps at 600 and 660 plus several pri taps at 100, 120, and 125V to get something acceptable.

Now what do you do with an 800V PT with possibly only about 60ma of HV? 


The other issue is that it has high tendency to oscillate when preamp vol is wide open - fine below 8 or 9 then it squeals past that.  Seems to already be a number of complaints out there on that problem.

I had to ask why you would even want two volume controls on this amp.  The lead dress is kind of long and now you have another grid wire sitting on the bullseye.

I think you can replace the 2nd pot with fixed resistor and a shielded lead run if needed.  If 1M is still too much use 680K.

The amp is a good bargain and you could easily gut it and install tag board and start from scratch.  The cab is a keeper.
 

Attachments

  • Epiphone_Galaxie_10.jpg
    Epiphone_Galaxie_10.jpg
    80.6 KB · Views: 26
new 6L6 tubes seem to need more negative fixed bias voltage,

cathode bias is supposed to compensate for this, up to a point,

if carving a hole in the chassis is ok, you could install a tube rectifier,
this would drop the B+ a bit, depending on which tube you use,

or you could run the Masco supply circuit which uses a resistor before the OPT,
it is a pi filter with a resistor instead of a choke,

screen voltage controls plate current more than plate voltage, so maybe experiment with lower screen voltages,

you  might have ultra sonic osc going on, you can not here it but it causes tubes to red plate, check the feedback circuit polarity,

cathode resistor chews up B+ voltage, so the tube is "seeing" B+ volts minus cathode bias volts, try 500 ohms/10 watt,


 
You keep saying "OT" where I think you mean "PT"?

Drawing you posted has 6L6 G1 and G2 grid symbols swapped. (Pin numbers are right.)

If the idle current is high or drifting up, CHECK THE G1 CIRCUIT.

Coupling caps leak. (However here the coupling cap is backed-up by the tonestack caps.)

I do not like relying on pot wiper to hold G1 down, especially on Power tube. I'd put 1Meg or 500K directly on 6L6 socket to the cathode resistor ground.

> ask why you would even want two volume controls on this amp.

Preamp distortion. This beast is all about preamp distortion. The tonestack is in the wrong place. The "poor" preamp can hardly smack the 6L6 properly with the tonestack between.

Preamp B+ is 360V. Assume peak audio (clean) is 20% of that, 72V. 6L6 has 31V of bias and first-guess needs ~~30V of drive. If tonestack loss exceeds 2.5:1, the preamp clips before the power amp. This type tonestack has (over the many different curves possible) loss of more like 5:1 or 10:1. The area around 700Hz can NOT have loss of less than 5:1.

So in the midrange the maximum output of the red-hot 6L6 is 2.5 Watts, not the expected 10 Watts. 10 Watts could happen in bass (or upper tones, if your ears can stand the shrill).

The Classics were not wrong. This is a Champ (with 6L6). AA-Champ has gain stage, tone/vol, gain stage, then output tube.
http://el34world.com/charts/Schematics/files/fender/Fender_champ_aa764_schem.pdf

Not two gain stages, tonestack, output tube.

I dunno how ugly the PCB is, or what the customer wants. But IMHO it could and probably should be reverted back to Classic topology.

> HV DCR is 563 ohms which charts to

I would not worry too much about that. Nor "chart out" when a HV PT winding can be designed many ways, depending on relative costs of copper and iron. Certainly a hand-test should have been early in the diagnosis.

> PT is 800V CT unloaded.

400VAC goes to 560V DC, youch. We expect 20+% sag, 448V DC, not absurd. Maybe more sag with the smallish 22uFd first cap. I think copper is expensive and it really is expected to sag down near VDC = VAC, around 400V DC.

Assuming schematic values are correct:

Vp-k = 382V
Ik = 0.066A
Pdiss = 25W

Approximate optimum load = 382V/0.066A= 5800 Ohms

The operating point is not wrong.

Except there is NO reason to run 385V on screen for 6L6 at 66mA. We need 132mA peak. 6L6 data says we only need 240V G2 to get that. 385V on G2 just means that much more G1 bias to keep idle down to 66mA. Guitar amps sometimes do that to delay grid-blocking, but grid-block can't be a huge problem with that tonestack choking the driver.

Can't just resistor-drop G2 because 6L6 G2 current is small and highly variable. Either you don't get much drop at idle or it collapses at high output.

Here's what I see. I really think tonestack should move left one place. Then take the B+ end of the OT off C11 and move it to C5. That gives better B+ filtering for the power stage, which can be an issue in SE amps.

The thing will never be loud enough for the stadium, nor soft enough to play around the family. So exact power output isn't critical. However it would be good to get much of the potential power of "costly" 6L6es. Yet I think 25W idle is a lot for many modern 6L6.

After moving OT B+ over, change R14 to 1K 10W. Now we get around 350V to the final, double-filtered. Pdiss drops nearer 20W. Pout drops to maybe 8 Watts, whoo-hoo.

Argh. Look at the "NFB" resistor returned to a bypass cap. This was not "designed", it "just happened".

The attached plan would be a more "Classic" small amp with a lot less stress on the 6L6.

I do not know if the tonestack rerouting is practical. Layout of this area is VERY critical on tonestacked Champs or you get squeals. I tame it by pushing wires around with a pencil; PCBs are less cooperative.

6550 would take this abuse a LOT better. But the PT needs a hard run and a hand-test before we add the extra 4 Watt of heater power; and 6550 isn't cheap. Or short.
 

Attachments

  • Galaxie-8.gif
    Galaxie-8.gif
    11 KB · Views: 25
After moving OT B+ over, change R14 to 1K 10W. Now we get around 350V to the final, double-filtered. Pdiss drops nearer 20W. Pout drops to maybe 8 Watts, whoo-hoo.

I totally missed that possibility.  Plenty of room to mount a Brown Devil on a U clip.



But IMHO it could and probably should be reverted back to Classic topology.


Done and done. 


I do not know if the tonestack rerouting is practical

It was not inviting but after staring at it for about twenty minutes I went for it.


Layout of this area is VERY critical on tonestacked Champs or you get squeals.

It's not perfect yet but likely close enough.  It still squeals only at the very last bit if Treble is past 9 and Gain is maxed and vice versa.  So Treble can be maxed provided Gain doesn't go past 9-9.5, or, Gain can be maxed provided Treble doesn't go past 9.  I may still be able to move wires and get it stable but it's a very livable flaw.  The client runs tube screamers so doesn't need Gain maxed or Treble.   


edit - It may have survived without squeals had I went with 220K 6L6 Grid like AA Champ.  I went with 470K. 


when a HV PT winding can be designed many ways
 

I had wondered about that since the size and lack of heavy filament pointed to much more current capacity.  The volt/ohm chart method has worked pretty good for most of the older PTs I've measured but I'd never tried on any really recent PTs outside of Hammond "Classics", and they measure like the older ones.


 
 
How does it sound now?

Residual squeal.... you could try 47K right at 6L6 grid (the 1.5K hardly does anything). For hi-fi I sure would, it also reduces grid-blocking. However some guitarists have built a career by living on the edge of blocking, so this has to be Artist's Choice. Anyway the treble cut is so small it might make no difference.

A perhaps-better place to shave the high-highs is 100K-300K right at (series) V1B grid. 100K against the ~~100pFd of the triode grid gives a 16KHz hi-cut. Won't hurt 5KHz guitar but might take a bit of the edge off the squeal. 300K gives 5KHz 3dB down which does shave the top of the guitar's tone; depending on style and speaker this may be acceptable, or dull.

It may just have too much gain for an as-is layout. You might snip a leg on C4, cut overall gain about half, and see if that's still plenty. (With pedalboard, it should be; but some players need more-more-more in the amp even when they are drowning in gain.)
 
How does it sound now?

Good and better to my ears than the stock amp.  Seems to breathe better.


I went with a 100K on V1B grid.  The squeal is still there and seems unchanged.  I had read some reporting they cured this squeal by moving the shielded V1A input grid wire around but I've had no luck with that.

I also tried grounding V1A grid to see if it stopped the oscillation and got hum - at all gain positions instead of the expected silence.  Maybe ground loop issues? 

I noticed the other tech had the OT pri leads reversed and so put them back to see what would happen.  Maybe a slight change in amp tone but no effect on oscillation. (also noticed he put in a Valve Jr 5W OT . . . . I guess we'll see how many licks it takes!)

Since this amp apparently had the oscillation problem with the Gain control in stock configuration I'm wondering if the parasitics are coming from board traces.

The good thing is that the extra bit of rotation on the Treble pot is redundant.  You already get your max treble by about 7 and Gain control can be maxed without setting it off.


edit:

I put the brand new, new production 6L6 back in.  It hummed pretty bad in the stock amp and now only hums a tiny bit.  I think acceptable for a practice amp so it's not a total loss on that tube.  It still wants to idle very high - 32V vs 25-27V for the assorted used vintage tubes I had sitting around.  68ma seems ok for avoiding smoke.

I did a hand test on the PT at 68ma idle for 10-12min.  Only high warm - so it's running fairly cool and should be fine.
 
If the treble pot makes little difference across that last bit of throw outside of squealing, why not slug it by putting extra resistance off the top of it, so it can never actually go to '10'? 
 
emrr said:
If the treble pot makes little difference across that last bit of throw outside of squealing, why not slug it by putting extra resistance off the top of it, so it can never actually go to '10'?

I looked at doing that.  Super tight fit on pots.  There are eye holes above board and it may work.  Worry there is it making the shaft come up crooked through the chassis and stressing joints when tightening nut.  It might do better on the backside but will have to remove pot and check.
 

Latest posts

Back
Top