Parallel SE 5881 Tweed Champ

[garytube]

Hi everyone!  I am interested in building a parallel SE 5881 in a tweed Deluxe chassis and cabinet.  I would be using an Allen TP40D transformer 320-0-320 180mA with an Edcor 25 watt 2.5k primary 8 ohm secondary.  Attached is the modified schematic.  Does it look ok?

[shooter]

couple things;

beef up your filter caps to 47uF, except maybe the 33uf if that's all the tube rect. can handle.  Would add a filter stage prior to the PA plates like you indicated.

I like each parallel tube to have it's own R, but that's a builders choice thing, just makes setup easier when you can monitor each tube individually

that Edcor is a monster    I've used 4 of the 25W'rs, great iron

[garytube]

Thanks Shooter!  With the Edcor, is the 2.5k primary impedance correct?  Or should I be going higher?  I’m really second guessing myself here.  When you say “monster”, how big is this thing?
How do I do the extra resistor?  Would I just have a 470ish ohm with the 25uF bypass on each cathode? 
Would a choke in place of the 10k dropping resistor be better than adding the extra C-R?

[shooter]

 

if memory serves they're about 8lbs

guessing 5881's are spec'd like 6V6, so ya have a handful of 370 to 550 ohm.  that way you can "dial in" the dissipation, I like ~~ 85 - 105% max plate.

caps range from 25uf to 100uf typical per tube.  that's a tweak thing once it's up and making you smile 

adding a choke is good, chassis size will be a factor with all that iron.  take your time, get dimensions of the iron, board size etc.

[PRR]

> When you say “monster”, how big is this thing?

It's 10 pounds. Almost $100. Claimed for 25W at 20Hz.

The TP40 is another 7 pounds by the time you have it bolted.

You won't get over 17W out of two 5881, and not under 80Hz-70Hz on guitar.

Edcor's XSE 15W might be ample electrically but I sure do not like the exposed tabs. Their GXSE15-2.5K may be a fair fit.

I have a hard time figuring you will get all your voltages and currents "good" on the first prototype, which means that picking a load impedance is a crap-shoot. I'd bet higher than 2.5k but probably not 5k; but with untested transformers and a hollow rectifier I think it needs to be built to figure what you need. You can omit the OT (use a 100r 10W plate resistor for expendability), bring the 5881s up to desired dissipation, measure working V and I, then pick an OT.

> guessing 5881's are spec'd like 6V6

5881 is a tough 6L6m/6L6G. Twice the heater and almost twice the Pdiss of 6V6. This is not small beer.

[tubeswell]

With 320-0-320 VAC and a 5AR4, I expect* you’ll get about 400v on the plates of a pair of cathode-bias 5881s idling at 100%. (with that shared 250R cathode resistor)


* I got similar voltages with a couple of T-bolt builds with 120mA through a shared 250R cathode resistor


Vg1-Vk is about 30v


This leaves a plate to cathode voltage of 370


Zout for 1 x 5881 at 370 in centre-bias Class A operation is:


370/(23/370) is 5k9


Halve that for 2 x 5881 in parallel


You can always lower the screen voltage a bit


It’s all ballpark stuff anyway

[garytube]

> When you say “monster”, how big is this thing?

It's 10 pounds. Almost $100. Claimed for 25W at 20Hz.

The TP40 is another 7 pounds by the time you have it bolted.

You won't get over 17W out of two 5881, and not under 80Hz-70Hz on guitar.

Edcor's XSE 15W might be ample electrically but I sure do not like the exposed tabs. Their GXSE15-2.5K may be a fair fit.

I have a hard time figuring you will get all your voltages and currents "good" on the first prototype, which means that picking a load impedance is a crap-shoot. I'd bet higher than 2.5k but probably not 5k; but with untested transformers and a hollow rectifier I think it needs to be built to figure what you need. You can omit the OT (use a 100r 10W plate resistor for expendability), bring the 5881s up to desired dissipation, measure working V and I, then pick an OT.

> guessing 5881's are spec'd like 6V6

5881 is a tough 6L6m/6L6G. Twice the heater and almost twice the Pdiss of 6V6. This is not small beer.

Thanks PRR.  I’ve dabbled in amp building occasionally over the years, but still have lots to learn.  I haven’t tried the resistor sub for the OT.  How is that done in further detail?  Do you just put the resistor in place of the primary connections?

[garytube]

With 320-0-320 VAC and a 5AR4, I expect* you’ll get about 400v on the plates of a pair of cathode-bias 5881s idling at 100%. (with that shared 250R cathode resistor)


* I got similar voltages with a couple of T-bolt builds with 120mA through a shared 250R cathode resistor


Vg1-Vk is about 30v


This leaves a plate to cathode voltage of 370


Zout for 1 x 5881 at 370 in centre-bias Class A operation is:


370/(23/370) is 5k9


Halve that for 2 x 5881 in parallel


You can always lower the screen voltage a bit


It’s all ballpark stuff anyway

Thanks tubeswell, long time no see!  I think you are right on the money with those expected voltages. Also thanks for confirming dividing the primary by 2 when adding another parallel tube.  I wasn’t too far off the mark with 2.5k according to your calculations.  I believe Edcor has a 3k that could be the ticket...

[shooter]

5881 is a tough 6L6m/6L6G

thanks, was working outside n too sweated to look it up

I believe Edcor has a 3k 

believe for an extra $50 they'll customize what you want, knowing what you want is a different calculation 

[garytube]

5881 is a tough 6L6m/6L6G

thanks, was working outside n too sweated to look it up

I believe Edcor has a 3k 

believe for an extra $50 they'll customize what you want, knowing what you want is a different calculation 

Thats cool.  Would you happen to know if they customize for more than one secondary tap, like 8 and 16 ohm?

[shooter]

I did a quickie, clicked the custom tab, but that just "customized" the search 

so you'll have to make contact 

found this;
https://www.edcorusa.com/xse25-3k

the xSE has exposed leads that you really can't "fool-proof"
here's how I've done a couple, but a fool can still be had 

[dwinstonwood]

To give you an idea of the size of the Edcor GXSE15's, here are two I'm using in an EL34 stereo amp I built.

Keep in mind that Edcor builds to order, and it can take several weeks to get them. But, they are very reasonably priced for the quality and amount of iron (They're heavy, 2.75lbs. each. The XPWR035 power tx just visible on the left weighs 9.75lbs. A gallon of water weighs about 8.4lbs.)

[garytube]

I did a quickie, clicked the custom tab, but that just "customized" the search 

so you'll have to make contact 

found this;
https://www.edcorusa.com/xse25-3k

the xSE has exposed leads that you really can't "fool-proof"
here's how I've done a couple, but a fool can still be had 

Thanks, I might inquire to Edcor for customization.  Btw,  That’s a cool build, I like the turret/ptp combo....

[garytube]

To give you an idea of the size of the Edcor GXSE15's, here are two I'm using in an EL34 stereo amp I built.

Keep in mind that Edcor builds to order, and it can take several weeks to get them. But, they are very reasonably priced for the quality and amount of iron (They're heavy, 2.75lbs. each. The XPWR035 power tx just visible on the left weighs 9.75lbs. A gallon of water weighs about 8.4lbs.)

Holy smokes, that pt is giant!!!  Thanks for measuring the GSXE15’s... looks like they are around 3.5” x 4”...

[shooter]

while you're building the idea box, keep in mind the drive for the 5881 might exceed your preamps ability to push 'em hard enough to bark 

EDIT:
added size matters shots
same amp, other side. chassis is 12X8X2.5"

[garytube]

while you're building the idea box, keep in mind the drive for the 5881 might exceed your preamps ability to push 'em hard enough to bark 

EDIT:
added size matters shots
same amp, other side. chassis is 12X8X2.5"

Wow, that is a whopper of an OT!  I am starting to rethink that maybe I will use the 15 watt one.  I wonder what the size is with the 25 XSE version with the exposed terminals?  Too bad they don’t list dimensions in their specs. 
Are you thinking it might end up a “clean machine”?  I guess nothing a dimed tube screamer couldn’t solve, haha 

[tubeswell]

If you’re only going to have a champ style preamp with that pair of 5881s, then one way of getting the clipping characteristics of 6V6s, would be 4 parallel 6V6s. JJ6V6S would comfortably handle those higher voltages.


Theoretically, four parallel JJ6V6S in SE could get you about 22W output power if optimally set up. But you will have a tricky situation managing the losses from the preamp driver pushing into the lowered input impedance needed to run 4 parallel output tubes. (Although you could overcome this with an impedance buffer between the preamp driver And the output tubes)

[garytube]

I actually have a champ that I recently put a 5ar4 and single 5881, with a 15w Classictone OT, dialed in to around 95% dissipation/21.9w.  It seems to dish out the overdrive respectably, but not high gain.  Would adding the parallel 5881 change it greatly?  Really liking the sound with the single 5881, so that was the origination for the parallel idea.  Always wanting more, lol!  It seems I’m either trying to make a quiet amp loud, or a loud amp quiet, haha.
Btw, I do like those JJ 6v6s in my limited experience.  I have a pair of them.

[garytube]

If you’re only going to have a champ style preamp with that pair of 5881s, then one way of getting the clipping characteristics of 6V6s, would be 4 parallel 6V6s. JJ6V6S would comfortably handle those higher voltages.


Theoretically, four parallel JJ6V6S in SE could get you about 22W output power if optimally set up. But you will have a tricky situation managing the losses from the preamp driver pushing into the lowered input impedance needed to run 4 parallel output tubes. (Although you could overcome this with an impedance buffer between the preamp driver And the output tubes)

I re-read this again, very interesting.  So each tube that is added is halving the input impedance?  Am I thinking about this right? 

[tubeswell]

If you’re only going to have a champ style preamp with that pair of 5881s, then one way of getting the clipping characteristics of 6V6s, would be 4 parallel 6V6s. JJ6V6S would comfortably handle those higher voltages.


Theoretically, four parallel JJ6V6S in SE could get you about 22W output power if optimally set up. But you will have a tricky situation managing the losses from the preamp driver pushing into the lowered input impedance needed to run 4 parallel output tubes. (Although you could overcome this with an impedance buffer between the preamp driver And the output tubes)

I re-read this again, very interesting.  So each tube that is added is halving the input impedance?  Am I thinking about this right?

Almost right. You’ll need to be mindful of the output tubes’ maximum grid leak resistance (because you’re idling the output tubes at close to Pmax to get the most out of them). When you run tubes in parallel, you decrease the load resistances in proportion to the increase in current- all other things being equal - if you are sharing the resistors. (Grid leak resistors form part of the AC-load on the Preamp driver stage).


So running 2 tubes in parallel you’d use 1/2 grid leak resistance (if shared). 4 tubes in parallel = 1/4 of the resistance. Even if each output tube has a separate grid leak resistor, all these grid leak resistors will be in parallel as ‘seen’ by the pre-amp driver- making the impedance bridging worse. and as you lower the grid leak resistance, you increase the AC load- thereby sucking the juice out of the driver stage’s output signal.

[garytube]

As always, thank you everyone for sharing information.  Lots of experience and knowledge on this forum for sure!  Similar to Ampage long ago...

[DummyLoad]

my take on the proposed plan... the allen amps 3KΩ:8Ω OT model TO15 is a better fit for 400V B+ with a pair of cathode biased 5881, and it weighs less than edcor part. you probably want power tube grid stoppers & at least one common screen stopper. 220KΩ grid leak is fine with 2 x 5881 - datasheets state 500KΩ for self bias is max. ckt. resistance - so 2 x 5881 would want 250K grid leak - 220KΩ is still within spec.


why the 5AR4? there's no sag to be had with a SE, so the valve rectifier has no positive attributes to add, but does have number of import negative attributes should be considered - 1) the limitation of the filtering in the plate supply for the power tubes with 2x 5881 will be need to substantial to keep PS buzz to a minimum - especially if you plan to use 10-12" speaker(s). 2) the reliability of valve rectifiers are of a magnitude lesser than that of SSR reliability. 3) again, with the limitation of the primary filter size with a valve rectifier coupled with approximately 3x the current demand of this PSE plan over that of the single 6V6, consider using a CLC filter network for the entire amp with larger filters than 32/33uF, unless one can tolerate a discernible level of PS buzz. on an aesthetic note, the SSR plan is in line with the 6V6 champ with 3 bottles - 2 large & 1 small.
 


spec'd 500V PS filters because the peak supply B+ will approach 490V during warm-up with a SSR.


--pete

[garytube]

Dummyload, those schematics are super helpful, thanks!  I’m getting excited about this project.  I overlooked TO15... never noticed that one.  Looks like a good fit.
I actually don’t mind using a solid state rectifier setup.  I used a 5ar4 in my current single 5881 build, because I had it lying around, and it got the voltages higher than the Bendix 6106 that I had tried before it.

[shooter]

Are you thinking it might end up a “clean machine”? 

most xSE operate in class A A1, clean-ish
you can OD the pre, but if you OD the PA the results are not very predictable, from experience most SE PA's sound bad hard clipping

the XSE 25w is ~~~ 3.25 x 2.25 x 2.25"

[garytube]

It is going slow on this project right now, but I ordered the SE output transformer TO15 from David Allen, since he had them in stock.
I got a response from Edcor, that they do not offer more than one secondary impedance on their SE transformers.

[garyvalve]

Garyvalve here (previously known as garytube, the OP).  Well…, talk about a lull in this thread.  I had bought all the parts for this amp in 2020, but ended up shelving the project up until last week (what, about 2 1/2 years?).  I finally built this sucker based on the solid state rectifier schematic that were offered by Pete/Dummyload.  I changed a couple parts using two 10uf filter caps instead of the 16uf’s, and used a 25uf electrolytic for the preamp bypass cap.  Sounds pretty good, and is quiet for a SE circuit, imo.  Getting 86.5% dissipation per tube with the 270 ohm cathode resistor which seems on the cool side.  It is sort of a hodge podge Frankenstein build in a tweed deluxe chassis.  Definitely was not a purist for this build ;^)
A belated thanks everyone for your help with this !  I wanted to post to let everyone know this wasn’t a wasted dead-end thread, and has a conclusion!

[kagliostro]

Very Cool Layout

did you draw it by yourself or it is from a Kit ?

if you did your own Board, can you explain what you used to build it ?

Thanks


Franco

[garyvalve]

The only kit parts were the cabinet, and chassis from a Weber 5C3 kit that I bought a long time ago. All the other parts were purchased separately from Antique Electronic Supply, Digikey, etc.  This is the third amp I have built in this chassis!
The layout was just based on the schematic that Dummyload had made, and tried to get somewhat close to a traditional 5f1 layout.  The schematics/ attachments are on one of his posts back in this thread.
The board was a hodge podge of thin U-style terminal strips (Antique Electronic Supply) screwed onto a black backer board that I had lying around. They also offer this U- terminal style on a bigger board.  The board is mounted on the chassis with nuts as spacers to make the board stand off.  Enough for the 20awg Tefzel wire to go underneath.  I also used three watt metal film resistors in place of the 1/2 watt.

[kagliostro]

Many Thanks


Franco