Channel switching with relay on 1974 TMB

[Warsaw]

Hi, maybe was better to continue on original 18w TMB Tube town thread but i think this is a question that can be applied also on others amp circuit. If needed, I can cut/paste this into the other thread.

I built a Tube town 18w TMB and i did some mods. Playing with this amp helped me to go deep in the sound I want.
I did a cascade mod (4 gain stages instead of 3): it can be switched manually pluggin in the normal channel and with a push pull potz.
In this way the NORM channel (one gain stage) goes into TMB channel (three gain stages) becoming quite a jcm800 with a great sound.

Now I'm trying to do the same thing in a simplier way: see the schematic attached

Mission goals:

1. Have the chance plugging in TMB input to play in the 3 stages - TMB channel (a very nice "edge of breakup sound" controlled with gtr volume), switching into 4 stages for a hard rock amazing overdriven tone.
2. Switching with a footswitch.
3. When not switched, to keep the original 18w configuration, two separate channels, TMB and NORM (sometimes i need to plug in a mic to play and sing with my daughters 😂 )

Config 1 (switch off) - ORIGINAL 18w TMB
TMB Input --> TMB 3 gain stages --> EQ section --> PI --> Power tubes
NORM input --> NORM 1 gain stage --> PI --> Power tubes

Config 2 (switch on) - 2203 config
TMB Input --> NORM 1 gain stage --> TMB 3 gain stages --> EQ section -->PI --> Power tubes
NORM Input --> disconnected

To achieve those results I surely need a relay switch:
Power section: from pt there are 6,3V AC with center tap for heaters, 5V AC unused, and 6,3v AC for EZ81 rectifier tube heater. One of those rectified and converted into 5V DC (like Hoffman board - this )
Switching: two DPDT 5v relay (about 1A 30V DC) controlled by the same remote footswitch. And a SPDT for when the footswitch isn't plugged in.

now the questions

Q1: There's a suggested reliable double DPDT relay (4 inputs / 8 outputs) instead of two DPDT to save room inside the chassis?
Q2: There's the risk of some popping of channel switching? How to avoid it?
Q3: Is the double switching before P.I. correct or redundant?
Q4: What's the best PT tap to use? Unused 5V AC?

Currently the amp works fine exactly like this scheme (obviously without the input switching). P.S: a couple of grid stoppers are missing in the schematic, it's only for the purpose of show the switching system.

Thanks in advance to anyone kind enough to help me

 

[Pete Farrington]

All that additional wiring etc at the inputs opens the door to stability and noise problems.
To mitigate pops, fit resistors to maintain DC reference paths at all times, eg 10M between V1pin7 and circuit common.

You need to understand that both positive and negative outputs of the DC relay supply must be kept isolated from circuit common. They’ll be about +/-2.5V with high current capability.
So the screen of the footswitch cable mustn’t contact the screen of signal cables.

 

[Warsaw]

All that additional wiring etc at the inputs opens the door to stability and noise problems.
To mitigate pops, fit resistors to maintain DC reference paths at all times, eg 10M between V1pin7 and circuit common.

You need to understand that both positive and negative outputs of the DC relay supply must be kept isolated from circuit common. They’ll be about +/-2.5V with high current capability.
So the screen of the footswitch cable mustn’t contact the screen of signal cables.

Thanks for the hint!

The wiring will be done with shielded cable, with ground connected on one side.

The dc ground will be also isolated from chassis with isolated jack.

I have just some doubts about placing the 68k inputs grid stoppers away from Tube V1.... they will be close to input jack, then directly in the relay board, then the signal will go with a shielded cable to V1 grids.

Currently one input grid stopper is placed close to the tube but non directly connected, the wire go forth and back to the DPDT switch of the push pull volume potz.
And the amp is very quite almost dead silent...

So I don't think there will be issues with the relay, since there will be less cable lenght.

 

[Pete Farrington]

OK, real world success trumps theoretical problems
Still, I’d fit R1 to R5 directly at the valve socket terminals, so that the switches and wiring wasn’t between them and their grids.

Have you got your head around the requirement that a grid shouldn’t have a switch in its DC path to its cathode?

 

[Warsaw]

OK, real world success trumps theoretical problems
Still, I’d fit R1 to R5 directly at the valve socket terminals, so that the switches and wiring wasn’t between them and their grids.

Have you got your head around the requirement that a grid shouldn’t have a switch in its DC path to its cathode?

I recently redesigned the switching circuit, to put the switches before grid stoppers.

But there's a doubt about the relays power supply: the PT has two heater output, 5v and 6,3v. The 6,3v is used for the tube rectifier EZ81 (+6,3 and 0 connected directly to tube socket's pins) so the 0 is common to the 5v output I need to use for the relays. Is it ok to connect the - of the relay power supply to the same negative of the rectifier?

 

[Pete Farrington]

I think that the benefit of having a separate winding for the valve rectifier’s heater may be reduced, as for safety purposes, the relay supply (and hence the winding) will need to be referenced to chassis earth.

Bear in mind that thd full HT V DC will be across the insulation between the heater and cathode of the EZ81. So if that insulation fails, there will be big problems for the relay’s power supply and the power transformer.

If you’re happy with that potential hazard, as far as I understand it, your plan should be ok.
Fuses between the HT winding and the EZ81 anodes would be a good idea.

Is it ok to connect the - of the relay power supply to the same negative of the rectifier?

Could you clarify exactly what you mean?
A schematic would be much better than a text description.

 

[Warsaw]

I think that the benefit of having a separate winding for the valve rectifier’s heater may be reduced, as for safety purposes, the relay supply (and hence the winding) will need to be referenced to chassis earth.

Bear in mind that thd full HT V DC will be across the insulation between the heater and cathode of the EZ81. So if that insulation fails, there will be big problems for the relay’s power supply and the power transformer.

If you’re happy with that potential hazard, as far as I understand it, your plan should be ok.
Fuses between the HT winding and the EZ81 anodes would be a good idea.


Could you clarify exactly what you mean?
A schematic would be much better than a text description.

With those schematics my doubt should be clearer:
Now the relay pwr supply is connected to ±3,15 VAC (6,3VAC) - in parallel with preamp and power tube heaters. All the relays (3) works fine, but i suspect that this can bring some noise in the preamp section - I have to invstigate further.

I'd like to know if it's safe to connect the relay pwr supply to the unused 5V tip, sharing the 0V tip with EZ81 heater.

note that the 0V of the relay rectifier is not grounded to the chassis.

 

[2L man]

Relay coil operate without 0VDC reference so you can use filament coil voltage and rectify and filter and even regualate it. If its current is good for extra relay coil which it should be. Bridge rectifying and filtering increase AC load 1,41 times.

Filament can also be elevated to lessen filament hum. Relay coils isolation is few hundreds volts anyway.

You can use multi wire cable between amp and FootSwitch and use one wire to connect FS metal case to Safety Earth so FS becomes safe although inside it is typical filement elevate voltage about 70VDC.

Direct heated rectifier tube filament voltage should not be used to drive relay because HV is high. EZ81 cathode is isolated from filament so its filament voltage can be used.

Remember to install a diode reverse voltage direction over the relay coil.

 

[Pete Farrington]

The relay coil goes to a footswitch?
If so, for safety, I think it must have a connection to chassis earth.
Without that, if the rectifier shorts, the relay coil circuit will be at high voltage, putting the user at risk.

 

[2L man]

The relay coil goes to a footswitch?
If so, for safety, I think it must have a connection to chassis earth.
Without that, if the rectifier shorts, the relay coil circuit will be at high voltage, putting the user at risk.

It should be safe enough when footswitch metal box is connect to the Safety Earth. Footswitch metal button is also connect to box / SE. If filament elevation is used the switch voltage inside the box come to about 60VDC.

Compare this to amp Mains and HV switches? Choose a cable which has a SE "mesh/sock" (don't know what it is called although I have seen hundreds of them and installed as well) and inside there are two wires for relay coil current.

I can't say is this allowed in "electric regulations" but there is no danger.