5E3 with SS rectifier, low voltages

[wrongtube]

Just fired up my new build – I'm a bit puzzled by my voltage readings and was hoping some of you fine folks could point me in the right direction.

Here's the schematic:

royale_5e3_v1-1.pdf

I'm using Hammond iron (290AX + 1750E). The PT has two high voltage taps, 550-0-550 and 650-0-650. Because I'm running a solid state rectifier I figured the 550 tap would get me in the ballpark of my desired B+ of 350 DC. As you can see I've added a few other mods to the circuit, mainly for my own research purposes (most mods were taken from RobRob's site):
- 10K grid stoppers for minimal HF rolloff
- "lead channel" mod on V1B with 220K plate resistor, 22nF coupling cap, and switchable bias/bypass on cathode
- 1Meg "sweet pot" variable grid stopper into the phase inverter
- 470 ohm screen resistors
- diode rectifier + 750 ohm sag resistor

Originally I had also wired up an SPDT switch to the V2A cathode for the options: bypassed cathode / unbypassed cathode / unbypassed + 56K NFB. But unfortunately the switch busted. I cleared out the unnecessary components and it's currently just a solder point for the 1.5K bias resistor. Schematic has been updated to reflect the current build.

Here are my voltages to ground:

PT high voltage = 290 VAC each side

B+ (after diodes) = 318 VDC
B+ (after sag resistor) = 274 (VDC etc)
B+ screen node = 249
B+ preamp node = 191

V3/V4 plate = 270
V3/V4 screen = 249
V3/V4 cathode = 16

V1A plate = 100
V1A cathode = 1.36

V1B plate = 104
V1B cathode = 1.87

V2A plate = 127
V2A cathode = 0.95

V2B plate = 153
V2B grid = 15
V2B cathode = 37

As you can see, my B+ and plate voltages are quite low, around 50-75 volts lower than I was expecting. The unloaded B+ (no tubes installed) is 413 DC, which seems reasonable, I just wasn't expecting to drop 100 volts with loading. Also I'm dropping a lot more across the sag resistor than I thought I would.

Initially I thought that the 6V6s might be biased too hot, pulling down the B+. But according to RobRob's bias calculator I'm at around 45% or 6.5 watts per tube which is pretty dang cool for cathode bias.

Of course, I could try hooking up the 650 AC taps instead, but it seems wrong to me – those are supposed to be the correct taps for a 5Y3 rectifier, right? Am I missing something?

I dig the sound of the amp currently, though it is relatively quiet, I guess due to the low plate voltages. Not bedroom levels, but still noticeably quieter than my SF princeton reverb.

Photos below. Thanks for your help in advance!

[sluckey]

I'm using Hammond iron (290AX + 1750E). The PT has two high voltage taps, 550-0-550 and 650-0-650

The 290AX is 275-0-275 or 325-0-325. Your voltages seem reasonable with that 750Ω sag resistor. If you want more voltage, decrease that sag resistor.

https://www.hammfg.com/files/parts/pdf/ ... 1709681208

[R.G.]

I think you're losing it in your dropping resistor. 750 ohms seems pretty big there. How did you come up with that value?

I started to get out the spreadsheets, but the dropper is where I'd look first.

You beat me to it sluckey. Drop the dropper.

[maxkracht]

The voltage on those transformers is going to drop more than you would expect under load, even without the big sag resistor. It is a champ transformer. No harm going to the higher voltage winding if decreasing the sag resistor isn’t enough for you. I built a couple PP cathode biased 6v6 amps with that transformer a while ago, expected to have way more voltage than I got.

[Stevem]

As mentioned swap over to the 325 volt windings.

That should get you 455 volts to mess with before that dropping resistor.

Two good things come from this right off the bat.

1) dropping more voltage gives better ripple rejection.

2) now your using that full amount of that copper winding the PT has to provid that 100 ma .

One thing I would like to note with you is how long your PI wires to your output tube grids are.

Runs that long open the door to oscillation issues and degraded tone.

If that build was mine I would add another terminal strip near each output tube socket and then move the PI uncoupling caps from that board right over to pin 5 and each new terminal strip.

This will lenghten each plate wire, but much more importantly shorten the grid wires down to nothing.

Grid wires act like Antennas and will pick up other signals, so the shorter the better.

This short grid wire thing applies throughout the whole amp.

[lonote]

I have not tried it myself, but have taken note that often sag resistors for SS rectified amps are inserted between the transformer & the diodes.

See Vox schematic below, R72 & R73:

VOX AC15 Sag Resistors.JPG

[R.G.]

I have not tried it myself, but have taken note that often sag resistors for SS rectified amps are inserted between the transformer & the diodes.

They are indeed, but they don't have to be. Ignoring the heating/power dissipation considerations for a moment, it makes no difference at all. In a full-wave-center-tap setup, the diodes conduct alternately, There isn't any simultaneous conduction, so the resistor(s) being before or after the diodes makes no difference to the circuit action.

I think maybe the separation of one resistor into two, one in series with each diode, may be done for heat dissipation reasons, as it spreads the heat over two devices.

[wrongtube]

Thanks for all the replies.

I'm using Hammond iron (290AX + 1750E). The PT has two high voltage taps, 550-0-550 and 650-0-650

The 290AX is 275-0-275 or 325-0-325.

Copy that, sluckey -- forgot to halve the voltages per side. Got my wires crossed so to speak!

750 ohms seems pretty big there. How did you come up with that value?

Good question. I seem to recall reading somewhere -- though I've lost the source -- that 750 ohms was as big as you'd want to go. Naturally, I was hoping for the "most sag" given the fact I'm building with diodes, but I didn't factor in the major voltage drop. Of course, it could also have been a typo on my end, as it looks like 150 is a more common value.

So, I pulled the resistor and everything is looking a lot better. AC high voltage 285, B+ 361, 6V6 plates 355, screens 327, cathodes 22. The bias calculator has me at 12.W / 85% dissipation which looks pretty good to me.

I'm curious as to what's happening on a theoretical level here. Prior to removing the resistor, my voltage off the diodes was 312, but now it's jumped up to 361. I understand the voltage drop across the resistor, but I don't follow why the voltage BEFORE the resistor would increase once the resistor is removed.

The voltage on those transformers is going to drop more than you would expect under load, even without the big sag resistor.

This is interesting. This transformer is rated for 100mA on the high voltage secondaries -- are you saying that a 200 or 300mA trafo would come closer to the ideal (AC x √2) voltage that you'd expect from a SS rectifier? If so, why is that the case?

As mentioned swap over to the 325 volt windings.

That should get you 455 volts to mess with before that dropping resistor.

Two good things come from this right off the bat.

1) dropping more voltage gives better ripple rejection.

2) now your using that full amount of that copper winding the PT has to provid that 100 ma .

Are you saying that the 275V winding doesn't deliver as much current as the 325V? I just assumed that both windings would be rated at 100mA.

One thing I would like to note with you is how long your PI wires to your output tube grids are.

This is actually the first thing I noticed when I got the board and sockets installed. In my initial layout I didn't account for the long run from the PI coupling caps to the 6V6 grids, so I was a little disappointed once I got all the components wired in.

Now that my voltages are up to spec, at high gain settings I'm getting some shrieking oscillation as expected. The run from the 2nd stage --> "sweet pot" --> PI grid seems to be the culprit, so I'll move the sweet pot to the rear of the chassis. And I like your idea of relocating the output tube coupling caps to terminal strips near the sockets.

[Stevem]

No, the higher voltage winding uses the whole copper wire that makes up that winding while delivering 100 ma , also its running cooler and in turn so will the whole PT .

As temps go up so does the resistance in a wire and voltages then drop.

In terms of your oscillation issue , first try swapping the blue and Brown wires on pin 3 of each output tube.

If the shirick gets worse then you have the output phase right and can move on what’s going on in that 2nd gain stage.

[maxkracht]

maxkracht wrote: ↑Wed Jun 04, 2025 9:33 pm
The voltage on those transformers is going to drop more than you would expect under load, even without the big sag resistor.

This is interesting. This transformer is rated for 100mA on the high voltage secondaries -- are you saying that a 200 or 300mA trafo would come closer to the ideal (AC x √2) voltage that you'd expect from a SS rectifier? If so, why is that the case?

Turns out my memory was a little off, I was using the Mojo/Hoeber version of that transformer, not Hammond. I don't have a record of what the voltages ended up being, but I remember expecting somewhere around 450v and getting significantly less... And now checking Mojotone for numbers, it appears they only have the unloaded voltage listed... Hammond is kind enough to list both and say which is which. So, just me being dumb and expecting better documentation. No, I wouldn't recommend using a transformer bigger than you need.

[B Ingram]

The voltage on those transformers is going to drop more than you would expect under load, even without the big sag resistor. It is a champ transformer. ...

It is better to think of it as a "blackface Princeton transformer" because as Hammond's version of a 125P1B, it is sized for a pair of 6V6s.

Having measured the real thing in a 1965 Vibro Champ, I can tell you the Vibro Champ "uses a too-big Princeton power transformer," and not the other way around (with "a Princeton using an under-sized Champ transformer").

I'm using Hammond iron (290AX + 1750E).

This is interesting. This transformer is rated for 100mA on the high voltage secondaries -- are you saying that a 200 or 300mA trafo would come closer to the ideal (AC x √2) voltage that you'd expect from a SS rectifier? If so, why is that the case?

The 1750E is Hammond's take on a 5E3 Deluxe OT, right? 8.5kΩ primary, 8Ω secondary, rated for "12 watts."

AC Volts on Secondary: 12w across 8Ω --> Volts = √(8Ω x 12w) = ~9.8v RMS
AC Volts on Primary: 9.8V RMS x [√(8500Ω/8Ω)] = 9.8v RMS x 32.596 = 319v
RMS Current for 12w: 12w / 319v RMS = 37.5mA RMS
----> each 6V6 needs to idle at 38mA or a bit more; allow another 4-6mA for 6V6 screens, and 4-6mA for preamp: 88mA for the B+ winding is "plenty."


By the way, most build wind up with too much voltage, and folks implement hamfisted moves to knock down volts. So you're already way ahead by having less B+ present.

No, the higher voltage winding uses the whole copper wire that makes up that winding while delivering 100 ma , also its running cooler and in turn so will the whole PT .

As temps go up so does the resistance in a wire and voltages then drop. ...

In a vacuum, it seems like this stuff would matter.

In a practical build, it does not. But the higher voltage output does matter, because as noted above, folks then try to knock down the B+, or raise cathode resistor values, etc. A cathode biased amp works better when a needlessly high B+ is avoided from the outset.

[wrongtube]

Been occupied with other projects the last week, but I managed to get back into the 5E3 and make some changes the last few days. Long story short, the high-frequency oscillation is improved but not eradicated.

Now that the going concern is no longer "low voltages," please let me know if it's better I start a new thread or continue along here.

Here are the changes I've made:
- removed the "Sweet Pot" altogether
- relocated phase inverter coupling caps (and 220K grid leaks) closer to the power tubes
- added a 22uF cathode bypass cap to V2A
- replaced the run from the input switch to each of the V1 grids with shielded wire

After this step, I no longer have oscillation when a guitar is plugged into the amplifier. However, if nothing is in the input jack, oscillation starts when the volume dial is around 7 on the modified channel. The "normal" channel doesn't exhibit this behaviour. When the channels are paralleled, I get oscillation when either channel is set past 7.

Again, these issues seem to occur only when I do NOT have anything plugged into the input jack. So, maybe good enough for rock and roll. But I'd like to figure out what's going on regardless.

In an attempt to solve it, I swapped out the grid wire into the V2B phase inverter with shielded cable. This didn't make a difference.

My next thought is to replace all of the wire to the volume and tone pots with shielded. But before I go down that rabbit hole, I'd love a second opinion. Is there anything else that jumps out at you in my design or lead dress that might cause oscillation?

One thought I had was the grid stoppers on V1 -- I used 10K here off of a suggestion I got from the 18 Watt forum: "10K is ideal for input grid resistor. Increase up to 33K to avoid radio station interference, at the potential cost of increased pickup or rectifier buzz." Of course I could swap them for 33K if you think that will help.

I did not try swapping the OT leads (Stevem's recommendation), as I was under the impression that potential problem only exists in circuits with NFB. Please let me know if I'm wrong in my assumption (this circuit does not have NFB).

Here is the updated schematic:

royale_5e3_v1-2.pdf

And photos:

IMG_0001.JPG

IMG_0002.JPG

IMG_0003.JPG

[maxkracht]

This isn't a proper solution, but if you replaced your input jack with one that shorts the input to ground like in most amps it would turn off V1 and shouldn't be able to oscillate. Probably better to actually find the problem, but if it's only happening with nothing plugged in, that should work.

[wrongtube]

I've had the build done for a month now but it's taken me that long to get this post written! Anyhow, I thought I'd drop a little update to let you all know that the amp is up and running and sounds excellent.

I did end up swapping the input jack for a shorting jack (thanks maxkracht for the suggestion). Sorted out the oscillation issue for the most part -- although I also found that when I tested with a lower-output guitar (Strat or Jag) the oscillation would come back (I had been testing primarily with a Les Paul). Borrowed an oscilloscope and signal generator from a friend, and was able to trace the gremlin back to the grid stopper on V1B -- the wire at the pin was too close to the AC heater line. Re-positioned the resistor and bent the wire up for a nice high clearance, and all is well!

At this point I had finished troubleshooting, and it was time to tweak the circuit. I found that the cathode-switching network on the "lead" input yielded very little tonal difference, and mainly sounded best at a centre-biased or slightly warm setting. No audible difference between using 1uF, 10uF, or 25uF bypass caps. So, I left V1B with the 1.5K/25uF setup and moved the switching network to V2A. Also changed some values around, so the choices are: 5K/10K/1.3K bias, 1uF/none/25uF bypass. Wow! The difference between the settings is quite pronounced. On the 10K/no bypass setting, it gets a real nice clean chime. Switching between 1uF vs 25uF is more subtle than switching the bias, but it does tighten up the low end.

Finally I added a 130R sag resistor (a much more reasonable value than the 750 I had previously) and on the lead channel a 500pF cap bypassing the volume pot. Found the lead channel too dark in comparison to the normal (because of the tone pot arrangement) but I may change this value as the lead is now a bit too bright! Still, I dig the sound with the tone backed off a little.

Anyway, thanks for the help everyone. Here's the final schematic, if you're interested.