Princeton 6G2 from Hammond L-100 (A-043) chassis

[WimWalther]

Howdy friends, first post.

I've got a Hammond L-100 chassis that I'd like to rebuild with the Princeton 6G2 circuit, using the original Hammond terminal board if possible. A member of the jazz guitar forum suggested this was the place to ask.

Rather than trying to re-invent yet another wheel, I was wondering if a custom layout or other related info already exists for this project.

Your help is appreciated, thanks in advance.

-Bill

[shooter]

before you get to "custom" you might be better off with original
https://el34world.com/charts/Schematics/files/Fender/Fender_princeton_6g2.pdf


then once it's working, you fingers have bled, the neighbors called the Police, then you'll have a better understanding of what "customs" you want to make

[sluckey]

Several members have used the AO-43 chassis for similar projects. I've used transformers for a couple projects, but I never used the chassis or board. I don't recall anyone building a 6G2 in this chassis though. Plenty of 6G2 info though. Search the forum. I'm sure you will find some interesting stuff.

What is your electronics background? Got any tube amp experience?

[WimWalther]

Hi Shooter,

before you get to "custom" you might be better off with original
https://el34world.com/charts/Schematics/files/Fender/Fender_princeton_6g2.pdf


then once it's working, you fingers have bled, the neighbors called the Police, then you'll have a better understanding of what "customs" you want to make

I think you totally misunderstood my question. I have no interest in modifying the 6G2 circuit, I'd like to build the stock circuit on the original Hammond terminal board. To this end, I'm looking for a bespoke ("custom made") layout that can accomplish this.

Otherwise, my other option would be to buy a 6G2 style board to replace the Hammond part.. which I'd really like to avoid.

[sluckey]

Hoffman sells a 6G2 board for $17. Look here...

[shooter]

misunderstood

I posted Fenders layout, strip Hammond's board, make it look Fender, OR ^^^^^ what he said

[WimWalther]

Hoffman sells a 6G2 board for $17. Look here...

Question.. are these boards a modern, rigid material like FR4, or are they the glorified cardboard used in the classic amps?

Question #2: What's the distance between the upper & lower rows of eyelets on the Fender board? The Hammond is only 1-5/8"

[WimWalther]

What is your electronics background? Got any tube amp experience?

I spent the better part of 15 years building amps, preamps and doing q/a work for a boutique hi-end manufacturer in St Paul, MN.They make these fairly ridiculous OTL designs that use up to 26 tubes per chassis in the big 200W+ units. Also took in vintage hi-fi and guitar amps for side cash.

So I at least know which end of the soldering iron to hold. ;-)

[WimWalther]

(...) strip Hammond's board, make it look Fender (...)

Oh, well if it's THAT easy, I'll just.. ;-)

Trouble is, lack of proper solder point installation to facilitate the stock layout. My hope was that someone more clever / at leisure than myself had created a bespoke layout that allowed a clean 6G2 build on the stock Hammond board.

It might not be possible.. but as these things go, someone might have figured it out.

If I could add a number of solder points, that might make it work.

Question: What's the distance between the upper & lower rows of eyelets on the Fender board? The Hammond is only 1-5/8"

[shooter]

dig around Doug's pages on designing turret boards, he has spacings, board widths, etal.  the boards are quality, designed 6 or more using his boards.

[dude]

If Doug's $17 board is "narrow" enough by all means get it, cheap and will make life easier. But the cone original bd is do-a-ble. Things are very tight in that chassis, just follow the schematic and the layout from Fender and make the changes to fit the cone Hammond bd. I have build many amps with that chassis, and it's a challenge due to the narrow width. Move the Hammond bd over, get it off the edge, use 1/4" standoffs. You'll have to cut out the el84 sockets on the chassis, cut out a rectangle and attach a plate to cover and drill the 6V6 sockets in that plate. Shouldn't be too difficult, I put a 6V6 plexi 1987 model in that chassis with the Hammond bd, not much room for error.  Lay everything out before you start cutting and drilling, you want to be able to get the pots out if one fails so move that Hammond bd keeping an eye on everything fitting. You can raise the PT on standoffs, like an inch which will give you more room.
I have several pictures of conversions I did in that chassis, if you're interested let me Know. One more thing, this chassis is not recommended, IMO, for a beginner but it can be done..?

[sluckey]

Everything you need to know about that board isin the link I posted. Read the info at the top of the page and also follow the link about Fender board sizing.

[dude]

Everything you need to know about that board isin the link I posted. Read the info at the top of the page and also follow the link about Fender board sizing.

I think Doug's boards are 3 1/4" in width, won't fit in the AO-43 chassis, you'll have to use the cone bd or cut down new bd material.
This one I used a board from A&E, cut the width to fit, not much length on the legs but you want to be able to get the pots and sockets.

[sluckey]

Doug's boards are 3-1/8" wide and the distance between outer rows of eyelets is 2". It's all in the link I posted. The eyelets are centered on the width of the boards, but you could rip it down to be 2-1/2" wide by taking equal cuts on each edge. Will a 2-1/2" wide board fit that chassis?

I can do a layout on the original cone board. Can you add a couple eyelets to the existing board?

[dude]

I’m sure you can add a few eyelets, 2 & 1/2” will fit, needs to be a few eyelets longer, like you say. Would be a nice small amp.

[WimWalther]

To clear up one point, I don't presently plan to change out the EL84s for 6V6s. Personally, I much prefer the 84, at least for hifi audio use. And I'm aware that they're also used in numerous guitar amps, so I see no reason to change them.

They are more sensitive (gain-wise) than 6V6, but I'm not really sure how much trouble that might cause, if any.

[dude]

Won’t be a Princeton then..?

[WimWalther]

Won’t be a Princeton then..?

Won't really be one ever, for that matter.. just a simulacrum in any case. It doesn't have Fender iron, or caps / resistors, or cab / speaker etc etc.

The main idea is to use the Princeton circuit, with some practical-minded parts changes but retaining the original topology and features.

If it's just totally unacceptable with the el84s, I can always convert the chassis for 6V6 by installing a plate with octal sockets on short standoffs over the old el84 holes. Might also be able to upsize the original holes, if there's room.

[Dave]

It'll be just fine with EL84's. I have built plenty of Fenderish amps and used EL84 power tubes because I have hundreds of NOS ones.


Dave

[sluckey]

You should be able to install in your chassis using this drawing. If you use EL84s you will need to lower the bias voltage and possibly the tremolo oscillator signal too. Listen to Dude. His hands-on advice is worth a lot. For cleaning the cone boards I first removed all components and wires. Then I took the board outside and while heating a cone I would blow the solder out with an air compressor. This leaves a nicely tinned cone that looks brand new. Very messy, hence outside job. Coveralls are useful. Safety glasses are a must.

EDIT... Replaced attachment with final drawing in pdf format. Should print fine.

[dude]

How do I print that layout..? Can't get the print to capture the full layout..? I have a few more AO-43's with Hammond iron, might just do this conversion. I could just drill out one 9 pin el84 socket  and a 12Ax7 socket for the 6V6, no plate needed.  Also, don't have to lift the PT either. Would make a nice 6G2 head.
As far as the  "simulacrum", getting technical on a copy of any amp isn't the real thing, maybe not, but are Reissues "simulacrums".. .
Cheers, tone is in the eyes of the beholder.

[WimWalther]

You should be able to install in your chassis using this drawing.

You, sir, are a gentleman and a scholar. THANK YOU!

If you use EL84s you will need to lower the bias voltage and possibly the tremolo oscillator signal too.

So by "lower" a negative voltage, do you mean make it less negative or less positive? =) No worries.. iirc the el84 need less negative grid voltage as they have greater voltage gain than the 6v6.. maybe -20v vs -35v for 6v6.

For cleaning the cone boards (...)

Not a possibility here.. no compressor. I'll probably use the time-tested, traditional "heat-up well & slam down on edge of trashcan" method.

[WimWalther]

How do I print this layout?

That's all down to the behavior of your print driver & software. First thing to try is switching from a landscape to portrait view. If that doesn't work, dig around in the print dialog for options like "fit to page" or "image scaling", etc.

Below is a pic of the print I was able to make by switching to portrait view.

(...) but are Reissues "simulacrums"..

"Simulacra".. ! =)

Anyway, it's not really the most accurate term either. Not exactly sure what one calls an amp like this, beyond "based upon" this-or-that.

[sluckey]

If you use EL84s you will need to lower the bias voltage and possibly the tremolo oscillator signal too.

So by "lower" a negative voltage, do you mean make it less negative or less positive? =) No worries.. iirc the el84 need less negative grid voltage as they have greater voltage gain than the 6v6.. maybe -20v vs -35v for 6v6.

EL84s want to see about -12V to -15V.

For EL84s the 1.5K grid stopper resistors shown on the octal sockets should be put on sockets between pins 8 and 2. And forget about the 470Ω screen resistors. They are not on the schematic. I just included them because it's the right thing to do.

[WimWalther]

It'll be just fine with EL84's.

Well, this has been my belief from the get-go. The el84 is a super-musical tube in my experience, among the best of the pentode & beam power types and definitely the better of the 6v6.

I have built plenty of Fenderish amps and used EL84 power tubes because I have hundreds of NOS ones.

Hmm, maybe we should talk! ;-)

[WimWalther]

So it's suggested, and I agree, that it might well be necessary to reduce the output of the trem oscillator due to the higher gain (aka greater voltage sensitivity) of the el84 vs 6v6. There are probably a few ways to go..

1) Reduce plate voltage. The oscillator runs straight off the high B+ at the 1st cap. It could be moved to the 3rd cap.

2) Increase the 3k3 cathode resistor. This adds more degenerative feedback.

3) Increase the value of the 220k from plate to Int control. Drops less signal voltage across the control.

Any other, better ideas? Thoughts on which way to go?

[sluckey]

1) No. The LFO needs big B+ and it may cause problems if you move it to the preamp node.

2) No. There is no degenerative feedback because of the 25µF bypass cap. Besides, degenerative feedback is counter-productive in a phase shift oscillator such as this. Oscillators need regenerative feedback to operate.

3) Yes, may need to increase considerably. However, build it as is before tinkering. The Intensity control may be all you need.

[WimWalther]

1) Very rough guess, the osc needs to swing about 40V p-p? Or is that way off?

2) Doh! Missed the cathode bypass cap..

3) in retrospect, should have been 1). This would be the first thing I'd try.

But yes, absolutely.. tack-in standard values until I see how it behaves. Thanks!

[sluckey]

The sinewave at the plate is typically well over 100V_pp.

What is the actual size of the board (length and width)? You already said the distance between rows of cones is 1-5/8". What is the center to center distance between adjacent cones?

My AO-43 layout shows 21 cones in each row for a total of 42 cones on the board. And the mounting brackets are between the fourth and fifth cones from either end of the board. Is this correct?

With this info I can adjust the size of the board to be actual size on my layout and repost the revised drawing. Also, post a hi-rez pic of your board as it is right now.

[WimWalther]

The sinewave at the plate is typically well over 100V_pp

That's surprising. Now why does it need to be so great? If the Vg1 of the 6v6s is -35V, wouldn't a swing of 70V p-p cover it? In kind, if Vg1 of the el84 would be closer to -20V (rough guess), then 40v p-p should cover. In either case, this swing covers a bias current from (probable) cutoff to grid current.

What am I failing to factor in?

Your description of my board is accurate. I'll post images here, later.

[sluckey]

That's surprising. Now why does it need to be so great? If the Vg1 of the 6v6s is -35V, wouldn't a swing of 70V p-p cover it? In kind, if Vg1 of the el84 would be closer to -20V (rough guess), then 40v p-p should cover. In either case, this swing covers a bias current from (probable) cutoff to grid current.

What am I failing to factor in?

The oscillator will produce a huge sinewave when operating properly. And for stable/reliable operation, the oscillator needs lots of B+ and a strong tube with a high gain factor. That's just the way it is. Of course the plate signal is way more than needed on down the line. That's why the signal is always run through a voltage divider before it is used. This same oscillator may be wiggling grids of output tubes (big signal), or may be wiggling the cathode of a preamp stage (little signal), of flashing a neon lamp optocoupler (big signal), or driving a LED optocoupler.

I repeat, the bias voltage for EL84s should be in the range of -12v to -15v, not -20v. This is not a guess. This is a real world measurement of a pair of EL84s operating at 70% of max plate dissipation.

Please post the measurements I asked for...

What is the actual size of the board (length and width)? ... What is the center to center distance between adjacent cones?

[WimWalther]

I repeat, the bias voltage for EL84s should be in the range of -12v to -15v, not -20v.

My bad, I missed the line in the previous comment where you said as much. Been years since I worked with an EL84 amp.

Please post the measurements I asked for...

No problem.  I ran out of time this AM, and now I'm stuck in dialysis for the next four hours.

[dude]

Worth a read:
http://carlscustomamps.com/the-history-of-the-fender-princeton-part-ii-the-brownface-princeton

[PRR]

.....why does it need to be so great? ...

It may not "need" to be, but it WILL rise as near as it can to full supply voltage, because that's what oscillators do, unless they get complicated and slow-acting (or more complicated) amplitude control.

When your stage amplifier feeds-back, and you don't fine-control it, how loud does it go? Usually FULL POWER.

> it might well be necessary to reduce the output of the trem oscillator

You have a Depth control. Use it. If you find that anything over 55% makes a bottoming-out thud, then add a resistor before the depth pot to cut the max down to 55%.

You probably do not want to tap the trem LFO down on the preamp B+ filter because it may upset the preamp. The B+ filter to the preamps is scaled to filter line buzz and low audio but not subsonic wobble.

[WimWalther]

The stock Hammond A0-43 board is 2" x 7-5/8" with eyelets 5/16" on center.

[sluckey]

The final drawing is in Reply #19.

[dude]

Thanks sluckey as always

[PRR]

Drift about Kandinsky moved to Your Other Hobbies.

[WimWalther]

Thanks sluckey as always

Like no doubt, right? What a guy.

One question about sluckey's layout. It seems he added a 470R 2W screen resistor to each power tube. These aren't on the Fender diagram..  thoughts on this?

I'm prone to omit them, mostly because I don't have any. =)

[sluckey]

Maybe you missed this...

For EL84s the 1.5K grid stopper resistors shown on the octal sockets should be put on sockets between pins 8 and 2. And forget about the 470Ω screen resistors. They are not on the schematic. I just included them because it's the right thing to do.

I might add, it's more common to see 5.6K or 8.2K grid stoppers rather than 1.5K when using EL84s.

[WimWalther]

Maybe you missed this...

Miss I did, thanks.

As for the grip stops, do you suggest that I should increase their values? Fwiw, the Hammond circuit used 1K in that location.

I don't think I've ever seen grid stops much higher than 1k5:on an el84 amp.

[sluckey]

1.5K is fine.

[WimWalther]

Coming along.. need to order parts now.

Reused a few resistors from the Hammond circuit as available. Used MF 1% parts & Illinois MPW/MWR caps in the tremolo circuit. The rest of the resistors are Digi-Key branded carbons from the.. 80s?

So what to use for caps in the audio path? Orange drop 715, more Illinois MWR, or? The lone 500pF will probably be a ceramic or maybe silver mica.

The chassis has a spare 9pin socket. Thinking of putting a magic eye tube in there.

[dude]

One thing to watch for: That Hammond board has cones that are about 3/8" long and solder easily can drip out from the "cones point" at the bottom shorting out...! Be careful and make sure you lift the board with 1/2" standoffs, give yourself room to get pots in and out, sockets too when locating the boards location.

[WimWalther]

Currently, I'm planning to leave the board right where it is. It looks like there will be plenty of room for the pots & switches.

ETA: @sluckey - preliminary testing shows that the 30K in the bias supply needs to be more like 13K. This pulls the bias voltage to roughly -17V which is more in the desired range.

@All - Only issue is that the B+ rose to 370V due to the lighter load. I consider this  excessive.. Need to dig out a 5Y3 and see how much B+ that absorbs. My recollection is that the Y3 has a lower purveyence than the U4.

Another option is to up the value of the 64R 10W at the head of the B+ supply, but that's not exactly desirable either. That said, I still have an unused 50uF cap section. Maybe it might make sense to use it as the first pole, purely for voltage dropping.

[WimWalther]

I did a little more work on the fixed bias conversion and here's what I've found:

Switching from the stock 5U4 to a 5Y3 seems to have tamed the excess B+ nicely. B+ is now 336V. Total Ik at 5Y3 pin 8 is 63mA. This will increase slightly (~5mA) when the drivers are installed. But either way, well within limits.

EL84 Power tube voltages & currents are now:

B+ = 336V
Ep = 328V / 329V
Eg1 = -14.7V**
Eg2 = 333V
Ik = 64mA (pair) or 32mA (each)
Pt = 21W (pair) or 10.5W (each)

(** Subbing 11K for the 30K in the bias supply curcuit)

Assuming a Pmax of 12W/each, 10.5W = 88% max. Maybe a bit hot, but probably sounds nice. Comments?

[sluckey]

You could benefit from a simple adjustable fixed bias supply. Copy Hoffman's Princeton Reverb. May still need to juggle the 100K to get a suitable adjustment range that works with your PT, but once that's done it becomes very easy to set the bias...

   https://el34world.com/Hoffman/files/Hoffman_PrincetonReverb.pdf

[WimWalther]

You could benefit from a simple adjustable fixed bias supply. Copy Hoffman's Princeton Reverb. May still need to juggle the 100K to get a suitable adjustment range that works with your PT, but once that's done it becomes very easy to set the bias...

   https://el34world.com/Hoffman/files/Hoffman_PrincetonReverb.pdf

That's all planned. But right now, I can't find my stock of 10-turn pots.. no idea where that stuff went. Also need a couple 1R 1-2W 1% current viewing resistors, don't seem to have any.

Btw, that 100K does about 240mW and runs too hot to touch. Should really be a 1-2W part.

[sluckey]

Btw, that 100K does about 240mW and runs too hot to touch. Should really be a 1-2W part.

That's why I used a 3 watt resistor on the layout. Both bias resistors, both B+ dropping resistors and both screen resistors are all 3 watt metal oxide. Two watters would work but I dont have any shapes for 2W MO.

[WimWalther]

Btw, that 100K does about 240mW and runs too hot to touch. Should really be a 1-2W part.

That's why I used a 3 watt resistor on the layout. Both bias resistors, both B+ dropping resistors and both screen resistors are all 3 watt metal oxide. Two watters would work but I dont have any shapes for 2W MO.

Ah, ok. I guess that I wasn't sure if or how I should interpret the various part sizes. It would really be best to either mark them explicitly or create a  legend table of power vs. part size.

In the real world, size sometimes has little direct relationship to rating.

For instance, the Caddock MM215 parts, the size of most 1/2W carbons, are rated 1.2W.

On the subject of adjustable bias, the Hammond chassis already has a screw-adjust 10K pot that was used as a gain control. That could be a bias pot, I'd think.. need to try it, see if there's enough control resolution.

ETA: Using the extant gain pot in series with 8K25 puts us right where we want, with -15V near mid-range.  Resolution could be better, but it's what I'd call "acceptable" as-is. 1/2 of the pot track is doing only about 15mW, fine.

Think I might play around with some series-parallel stuff and see if the resolution can be improved. The pot can be left where-is and either accessed from inside or through a hole in the new front panel.