JTM45/2204 combo with reverb and FX loop schematic review please

[RoadShow]

Hello,
After a year and a half delay due to moving from Chicago to Nashville, retirement, and other problems I’m back on my project combining a selectable 2204 and JTM45 front end that will incorporate the Metropolus 2L-FX and reverb.  I’ve used a lot of Marshall schematic references for the design and parts selection and followed a lot of schematic design input developed by Luckey.  I previously built the Valvestorm JTM45 based off of the Metropolus design and am following a lot from that including the assembly process as it works flawlessly and is extremely low noise.  Here’s a link to my post from last year:
https://el34world.com/Forum/index.php?topic=27115.msg297723#msg297723

When I bring the amp up (aside from regular check out procedures) I plan to bypass the FX and reverb, check out the 2204 and JTM45 circuits individually, implement the selection switch, and then bring in FX and reverb one at a time. Perhaps after that implement PPIMV.
What I’m requesting is if you have the time to perform a schematic review and offer suggestions or point out mistakes/issues.  I would really appreciate it. Schematics attached.

Thanks in advance!!

Bill

[sluckey]

Bias caps are backwards.

[RoadShow]

Bias caps are backwards.

Oops, yep, fixed.

Anything else?

[PRR]

Reverb driver pot cap probbaly wants to be bigger than 1pF.

Reverb output cap: is 0.01uF a typical value with the pot connected that way?

Power stage, 47k and 100r NFB network does nothing. Smells like values copied from two different designs.

There is a short right AT the line cord?

[RoadShow]

Reverb driver pot cap probbaly wants to be bigger than 1pF.

Reverb output cap: is 0.01uF a typical value with the pot connected that way?

Power stage, 47k and 100r NFB network does nothing. Smells like values copied from two different designs.

There is a short right AT the line cord?

Thx,
Not sure what you meant by a short right at the line cord?
Accidently left out the presence and have fixed, please see attached regarding the power stage.
Regarding the cap in the reverb section, all of the one tube reverb designs I found including Luckey's have that.

[PRR]

> Not sure what you meant by a short right at the line cord?

[tubenit]

Almost all of the  "one tube reverbs" I have done have had .001 cap into the dwell pot instead of .0001 (1pf) into the dwell pot.  I don't remember ever seeing 1pf cap there?


With respect, Tubenit

[RoadShow]

> Not sure what you meant by a short right at the line cord?

Thx, drew wire correctly now, see pic...

[RoadShow]

Almost all of the  "one tube reverbs" I have done have had .001 cap into the dwell pot instead of .0001 (1pf) into the dwell pot.  I don't remember ever seeing 1pf cap there?


With respect, Tubenit

Wrong conversion, meant to use "n" instead of "p". Labeled conventionally now...

[pdf64]

The heater circuit is floating, it’ll cause buzz to contaminate the signal path. I suggest heater elevation.

The bias supply seems to have typical EL34 values, may be inadequate for KT66.

When overdriven, to mitigate significant over dissipation, KT66 screen grids greatly benefit from adding a shared 1k resistor in the feed to the screen grids, eg R32 of https://mhuss.com/amps/JTM45/JTM45Classic2b.pdf

The PT noted has a rather low resistance HT winding, around 50ohms, as opposed to the 120-140ohms typical of the JTM45 PT. That may result in the peak anode current being drawn to charge the reservoir cap exceeding the 750mA limit. Hence I suggest to add some resistance somewhere in that circuit, eg between each winding out and and anode, or between the cathode and reservoir cap.
https://tubedata.altanatubes.com.br/sheets/129/g/GZ34_Mullard.pdf

The standby switch is hot switching. I suggest to use a more benign arrangement, or to omit it.
http://www.valvewizard.co.uk/standby.html

[RoadShow]

The heater circuit is floating, it’ll cause buzz to contaminate the signal path. I suggest heater elevation.

The bias supply seems to have typical EL34 values, may be inadequate for KT66.

When overdriven, to mitigate significant over dissipation, KT66 screen grids greatly benefit from adding a shared 1k resistor in the feed to the screen grids, eg R32 of https://mhuss.com/amps/JTM45/JTM45Classic2b.pdf

The PT noted has a rather low resistance HT winding, around 50ohms, as opposed to the 120-140ohms typical of the JTM45 PT. That may result in the peak anode current being drawn to charge the reservoir cap exceeding the 750mA limit. Hence I suggest to add some resistance somewhere in that circuit, eg between each winding out and and anode, or between the cathode and reservoir cap.

- Grounded heater circuit center tap

- Copied bias supply directly from the aforementioned Valvestorm JTM45 build, works great

- Added shared screen grid resistor  (missed that), used same value from Valvestorm JTM45 build

- I'll look into what you mention regarding the PT, will circle back on this...

Thanks

[pdf64]

Note the 123ohm HT winding resistance of the 290MAEX https://www.hammfg.com/files/parts/pdf/290MAEX.pdf

[RoadShow]

Note the 123ohm HT winding resistance of the 290MAEX https://www.hammfg.com/files/parts/pdf/290MAEX.pdf

Yes, I see there's quite a difference. So, I spent some time researching DCR, and per the links below a lower DCR seems to be a good thing and the higher DCR is created by trying to mimic a vintage type transformer.  Maybe I'll make a call to Hammond tech and ask them about it. Hey, something to do in retirement!!

Thanks so much for your input. Please check out these links and LMK what you think:

https://circuitdigest.com/tutorial/what-is-dcr-in-inductors-and-how-does-it-affects-your-circuit-design
https://groupdiy.com/threads/dcr-and-impedance-modern-audio-transformers.76754/
https://cettechnology.com/minimizing-power-loss-in-a-transformer/
https://www.coilcraft.com/zh-cn/edu/series/a-guide-to-forward-mode-transformers/
https://www.voltech.com/media/kwjdnbwd/testing-line-frequency-transformers.pdf

[pdf64]

Note the 123ohm HT winding resistance of the 290MAEX https://www.hammfg.com/files/parts/pdf/290MAEX.pdf

... a lower DCR seems to be a good thing and the higher DCR is created by trying to mimic a vintage type transformer.

Vintage transformers had higher HT winding resistance because they were intended for use with valve rectifiers, and there was no benefit to the equipment designer in getting the resistance down, because then additional external resistors would be required.
Such transformers necessarily had the otherwise negative characteristics of poor regulation and high copper loss, and resulting heat rise under load.

With solid state rectification, a more ideal transformer could be made without any downsides for the equipment designer.

If such transformers are to be used with valve rectification, then the valve limits require the equipment designer to take account of the shortfall in resistance.
The designer might ignore that, or proceed in ignorance, and blame subsequent shortened rectifier life / greater tendency to arcing, on bad modern valves.

[RoadShow]

Note the 123ohm HT winding resistance of the 290MAEX https://www.hammfg.com/files/parts/pdf/290MAEX.pdf

... a lower DCR seems to be a good thing and the higher DCR is created by trying to mimic a vintage type transformer.

Vintage transformers had higher HT winding resistance because they were intended for use with valve rectifiers, and there was no benefit to the equipment designer in getting the resistance down, because then additional external resistors would be required.

Ahh, gotcha. So, with the difference being 123-52=71, would a ~35ohm power resistor on each output winding be a good solution?  Some other value?

I tried to find information on Heyboers, MM, Classic, etc. but came up with nothing on resistance.

[pdf64]

... would a ~35ohm power resistor on each output winding be a good solution?  Some other value? ...

As a bare minimum that would be ok, though a pair of 47ohm would be better.
I don't understand why it's never suggested, but I think a single resistor of that value between cathode and reservoir would be fine, should be functionally equivalent. It would need twice the power rating though, as it would be conducting twice the time, ie over the full cycle. whereas the pair only conducts a 1/2 cycle each.
Whatever, a high voltage rating is needed, eg 500V. Some power resistors might only rated for lower voltage, eg 200V.

[RoadShow]

... would a ~35ohm power resistor on each output winding be a good solution?  Some other value? ...

As a bare minimum that would be ok, though a pair of 47ohm would be better.

OK, PS schematic updated, put in 47 ohm 5W.  Looking through all the schematics I've collected for this I've seen some use 100, I'll see what I have in my stash.

Also, your previous mention of the standby switch; I moved it and make it DPST. I knew about the Valvewizard point, but thought it would be a good idea for when switching between 2203 & JTM45 modes. Thoughts?

[pdf64]

… your previous mention of the standby switch; I moved it and make it DPST. I knew about the Valvewizard point, but thought it would be a good idea for when switching between 2203 & JTM45 modes. Thoughts?

The standby is still hot switching, no improvement on the previous version.
When switching preamp modes, the master volume could just be turned to minimum; it will generally need adjustment anyway.
If a standby switch is an essential user requirement, to avoid hot switching, it should be moved to some point after the reservoir cap, as per Merlin’s article.

[RoadShow]

… your previous mention of the standby switch; I moved it and make it DPST. I knew about the Valvewizard point, but thought it would be a good idea for when switching between 2203 & JTM45 modes. Thoughts?

The standby is still hot switching, no improvement on the previous version.
When switching preamp modes, the master volume could just be turned to minimum; it will generally need adjustment anyway.
If a standby switch is an essential user requirement, to avoid hot switching, it should be moved to some point after the reservoir cap, as per Merlin’s article.

I "think" I've got it now, attached...

[pdf64]

Yes, it’s no longer hot switching 
A further improvement is to add a bleed resistor across the switch, to damp down the inductive back emf, as per Merlin’s article.