Tweedle Dee, Take two

[ChopSauce]

Within the bounds of my limited understanding: the 220k would act as voltage dividers when placed before the pot, not after.

I think you're right to suppress them - if your relays do cut the other channel - but ... maybe they can be usefull as grid resistors, too?
(so I'm raising more questions instead of solving your trouble... oops! )

[fred.violleau]

@Chopsauce thanks for clarifying about the voltage divider. I had not seen it as grid resistors but you are right that these would have an impact here as well...

Since there was none on my previous build, I will start as is and if I feel the amp gets out of control because of signal overload, I might try this easy mod.

Onto soldering!

Fred.

[pompeiisneaks]

Hey Phil, thanks for jumpin' in!

Actually, there is no 220k mixing resistor in the original 5E3. Nor is there on the Tweedle Dee.
It is a recommended when eliminating the channels interaction.
My guess was as long as my relay implementation keeps the signals separated it should act as these resistors.
These 220k resistors are not ground related, they seem to limit the interaction between the two channels, so basically big enough so the signal does not follow that path, but not too big so they don't block the signal right?

But if they act as a voltage divider, they would still divide the signal by half so that is quite big...

Still confused over these..

Fred.

Sorry I've built one and should have remembered. the 'to ground' part was usually to just 'simulate' the other channel's 'existence' when it's not in the circuit. The whole point of the mixing resistors is to eliminate cross talk between the two channels, i.e. signal backfeeding into it somehow from the other channel and creating either negative or positive feedback. Thus why they're called 'mixer' resistors, they allow a means of 'mixing' two channels passively without causing negative effects. The removal of them if you remove a channel, also increases the output of the single remaining channel and thus why it's good to keep a network in place if you want to retain the 'sound' of the original amp and just remove one of the channels.

That whole discussion is pointless if they didn't exist in the first place, (and on old tweed deluxes people often liked the 'interaction of the two channels' you'd get) and your relay's switching action completely separates the channels in a complete way. This does mean they will not 'interact' like the old tweed deluxes (or the tweedle dee here as well) do.

Hope that makes sense, and sorry for muddying the waters

~Phil

[fred.violleau]

So I finished wiring the amp.
This one proved to be a lot more complicated than what I expected. Space is really tight.

After checking all my solders, ground points and connections, I used a variac in order to power up slowly the old RCA Power transformer.
I went up to 40VAC, but smelled something weird and saw a bit of smoke.
The inrush limiter seems to get very hot (around 92 celsius) although I am not even running at 100% voltage and it smells ... like something is getting too hot. I suspect this thing did the smoke, but not sure.
So I powered off.

PT was getting hot as well, but nothing extreme. voltages seemed to check out ok though.
There was no tubes in the amp.

I will assemble an electrical schematic of what I ended up with.

Fred.

[talbany]

So I finished wiring the amp.
This one proved to be a lot more complicated than what I expected. Space is really tight.

After checking all my solders, ground points and connections, I used a variac in order to power up slowly the old RCA Power transformer.
I went up to 40VAC, but smelled something weird and saw a bit of smoke.
The inrush limiter seems to get very hot (around 92 celsius) although I am not even running at 100% voltage and it smells ... like something is getting too hot. I suspect this thing did the smoke, but not sure.
So I powered off.

PT was getting hot as well, but nothing extreme. voltages seemed to check out ok though.
There was no tubes in the amp.

I will assemble an electrical schematic of what I ended up with.

Fred.

This is why i call it the "Smoke Test"
You'll find it good luck!
Tony

[fred.violleau]

So here is the power section.
Since this is where I did most of the changes in the original design.
If you see anything wrong, please advise !
Thanks for the help!

5E3PowerSection.pdf

Fred.

[erwin_ve]

So here is the power section.
Since this is where I did most of the changes in the original design.
If you see anything wrong, please advise !
Thanks for the help!

5E3PowerSection.pdf

Fred.

You shorted the PT secondary hv windings.

Erwin

[fred.violleau]

Hey Erwin,

Thanks for your help.
So I shorted the secondary HV throught the bridge rectification right?

Fred.

[martin manning]

See below. You need to wire the rectifier as a two-phase.

[fred.violleau]

Just make sure I understand this correctly :
- No center tap on the PT, need to ground the rectification bridge in order to provide current a return path
- center Tap on PT, use 2 phase bridge rectification and usually, current's return path is throught the grounded CT

Do I get this right ?

Thanks for the help.

Fred.

[erwin_ve]

Hey Erwin,

Thanks for your help.
So I shorted the secondary HV throught the bridge rectification right?

Fred.

When using the fw bridge rectifier: no CT. In your schematic there is a path directly from one leg to the other leg. This will destroy you PT. Check a known fw bridge rectifier and compare it to yours. I'm sure you'll see the error.
Also on the CT you have B+ reducer. I'm not familiar with that circuit. Edit: Martin answered your question.

[martin manning]

Just make sure I understand this correctly :
- No center tap on the PT, need to ground the rectification bridge in order to provide current a return path

Yes, if you have a full wave bridge, you would ground the bridge through the B+ reducer. Usually there is no center tap on transformers designed to use a FWB.

- center Tap on PT, use 2 phase bridge rectification and usually, current's return path is through the grounded CT

If you are using two-phase (full wave) rectification, the CT is grounded trough the B+ reducer.

You show 374-0-374 VAC (RMS) secondary voltage, so it appears that you would want to use a two-phase arrangement. I would expect 374*1.414 = ~528VDC peak, and your B+ reducer will have to reduce that by ~126V to get to the 402V yo show on your schematic for B+1.

[fred.violleau]

Unloaded, with 120 VAC on the variac, I get around 402 VDC out of the bridge.
Will put tubes on it and see out it checks out.
So far everything looks and smells good

Fred.

[fred.violleau]

Ok, sent some 120 VAC from the variac into the amp.
Here are the charts :
B+1 : 308.9 VDC
B+2 : 224.1 VDC
B+3 : 199.8 VDC
B+4 : 188.5 VDC

V1A (12AX7)
pin1 : 99 VDC
pin2 : -
pin3 : 0.878 VDC

V1B (12AX7)
pin1 : 117.9 VDC
pin2 : -
pin3 : 0.698VDC

V2A (12AX7)
pin1 : 135.6 VDC
pin2 : -
pin3 : 0.902 VDC

V2B (12AX7)
pin1 : 165.7 VDC
pin2 : -
pin3 : 28.44 VDC

V3 (6V6GC)
pin1-8: -99.7 mV
pin3 : 290 VDC
pin4 : 216.8V

V4 (6V6GC)
pin1-8 : -99.7 mV
pin3 : 286 VDC
pin4 : 216.8V

It feel like I may not need the mosfet B+ reducer after all. When checking the voltage on the mosfet at the Source it reads -71.3V.
This thing is getting hot (mosfet reducer). If I remember correctly I had put the 91v diode in the reducer seat.
I will try to disconnect the CT from the mosfet and ground it see if it improves the voltages.

I get sound from the FX loop. So V2a sends signal downstream.
No sound from the input. I suspect the input relay since probing the voltages was audible in the speaker. Will check the wiring on this.

Fred.

[fred.violleau]

Removed the mosfet B+ reducer, grounded the center tap, and now I get 400 VDC at B+1 with 120 VAC input on the variac.
But I get only 266V at B+2 and both power tubes are red plating.

Will review the fixed / cathode bias switch and wiring as well.
Call it a night for today.

Will check these tomorrow !

Fred.