JTM45/100 under board wiring ? what's correct ?

[playloud]

- There is a yellow wire from between the two filter caps after the choke, connecting to a resistor, and to the elevated part of heater CT. I cannot find that on the "7026" schematics, although I've seen something similar on other Marshall schematics with AC standy switch. What is that?

That's a new addition for the heater elevation.

I just chose a convenient point where I was reasonably confident of the (idle) voltage. The screens see approximately the full B+ voltage (490/560V) relative to ground, so halfway between the two caps should be about half that.* The elevated voltage will then be ~ (B+/2) * (47/(150+47)), i.e. ~ 67V at 560V B+ or ~ 58V at 490V. Drop the 150K resistor to 120K if you want more elevation.

* Depending on how closely matched the two 33uf caps are in terms of actual capacitance - it doesn't have to overly precise for our purposes.

- Wouldn't it be better to connect the black heater CT (after elevation) directly to the ground lug (near the input), instead of on the board (at V1 cathod grounding)?

I don't think it makes a huge difference. I'm treating the chassis as a separate "0 node" if you will, so this is the final list of nodes/local stars:

0: Chassis
1: V1 cathodes, heater CT ground reference*, input jacks
2: V2 cathodes, volume pots, preamp filter**
3: Output jacks, presence pot, tone stack, PI filter
4: Screens filter, bias circuit
5: Mains filter, output cathodes, B+ CT

Then we're connecting those nodes via a single path in order to ensure no loops between nodes.

I don't think there's really any harm in dropping the "0th" node and using the chassis connection to make those three connections directly (plus it reduces the length of the daisy chain/implicit "bus ground"). I just think there's an elegance to having a single connection there.

* Please don't ask me to justify placing the heater ground ref. at the first node - as opposed to, say, the much more convenient fifth node. This is Larry's innovation and it seems to yield good results... but I have no idea why!

** It would be a lot more elegant if, instead of a dual section preamp filter cap, we had two separate caps for V1 and V2. Then we could assign their grounds to the respective nodes.

- There seems to be a "ground node" at the Mid pot. Wouldn't it be better to connect power supply ground together and power section (maybe also presence pot) ground together and connect those straight to ground lug (near input)?

Perhaps, but now we're mixing grounding schemes. Merlin's idea is to daisy chain the local stars together in sequence, with the final connection going to the chassis. What you're describing is moving in the direction of R.G. Keen's proposed grounding scheme, in which each local star is connected directly to a single point on the chassis. That would probably work too (and it's not hard to modify the above diagram to accommodate it).

 

[Vlad58]

Maybe I'm tired, but that diagram has the PI plate resistors fed from the choke and the screens from the following node?

This is how I wire mine, which AFAIK is how 7026 was configured.

Hi Stu, Sorry for unrelated question. I've seen your offers for some tube amp chassis around 2021. Are you still in business? I'm interested in Hiwatt DR504 chassis. If so how to get in touch with you for a quote? Not sure how to contact members anymore. Thanks.

 

[Sigs]

Hi Stu, Sorry for unrelated question. I've seen your offers for some tube amp chassis around 2021. Are you still in business? I'm interested in Hiwatt DR504 chassis. If so how to get in touch with you for a quote? Not sure how to contact members anymore. Thanks.

.

 

[BlueX]

That's a new addition for the heater elevation.

I just chose a convenient point where I was reasonably confident of the (idle) voltage. The screens see approximately the full B+ voltage (490/560V) relative to ground, so halfway between the two caps should be about half that.* The elevated voltage will then be ~ (B+/2) * (47/(150+47)), i.e. ~ 67V at 560V B+ or ~ 58V at 490V. Drop the 150K resistor to 120K if you want more elevation.

* Depending on how closely matched the two 33uf caps are in terms of actual capacitance - it doesn't have to overly precise for our purposes.



I don't think it makes a huge difference. I'm treating the chassis as a separate "0 node" if you will, so this is the final list of nodes/local stars:

0: Chassis
1: V1 cathodes, heater CT ground reference*, input jacks
2: V2 cathodes, volume pots, preamp filter**
3: Output jacks, presence pot, tone stack, PI filter
4: Screens filter, bias circuit
5: Mains filter, output cathodes, B+ CT

Then we're connecting those nodes via a single path in order to ensure no loops between nodes.

I don't think there's really any harm in dropping the "0th" node and using the chassis connection to make those three connections directly (plus it reduces the length of the daisy chain/implicit "bus ground"). I just think there's an elegance to having a single connection there.

* Please don't ask me to justify placing the heater ground ref. at the first node - as opposed to, say, the much more convenient fifth node. This is Larry's innovation and it seems to yield good results... but I have no idea why!

** It would be a lot more elegant if, instead of a dual section preamp filter cap, we had two separate caps for V1 and V2. Then we could assign their grounds to the respective nodes.



Perhaps, but now we're mixing grounding schemes. Merlin's idea is to daisy chain the local stars together in sequence, with the final connection going to the chassis. What you're describing is moving in the direction of R.G. Keen's proposed grounding scheme, in which each local star is connected directly to a single point on the chassis. That would probably work too (and it's not hard to modify the above diagram to accommodate it).

Thanks! I think I'm confusing ground lift and elevation (not native English speaking).

As I remember (can't find it now), Merlin has described ground lifted heater CT using resistor and cap (and maybe diods) but no voltage feed. In the "power supplies" book Merlin also describes heater elevation (ch 9.3.6, p240), which seems to correspond to what you did here.

Regarding grounding schemes I follow your explanation. I guess this can be done in different ways, and still work fine in most cases.

Hope I'm not confusing the OP.

 

[playloud]

Thanks! I think I'm confusing ground lift and elevation (not native English speaking).

As I remember (can't find it now), Merlin has described ground lifted heater CT using resistor and cap (and maybe diods) but no voltage feed. In the "power supplies" book Merlin also describes heater elevation (ch 9.3.6, p240), which seems to correspond to what you did here.

You're probably thinking of balancing the heaters/artificial centre taps (see previous section of the power supplies book, 9.3.5). Like heater elevation, this can be another way of reducing heater hum, so I understand the confusion!

 

[Sigs]

I have 3 minor questions re the layout, first one, I just wanted to be sure the wiring for the Mains/power switch is correct as it is , as I wired it that way.
So when you plug the lead from an amp into a wall socket, switch the wall socket on, power flows to that Mains/power switch on the amp, to me it looks like it could be wired to either side of the toggle switch, so what is the correct side ?

 

[Sigs]

The layout has no ground lugs fitted, should they be fitted ?as per ground lugs fitted image.

 

[playloud]

The layout has no ground lugs fitted, should they be fitted ?as per ground lugs fitted image.

Those ground lugs aren't necessary if you follow the rest of the diagram (note the cathodes are all ultimately connected to a ground lug near the PT).

 

[Sigs]

S

Those ground lugs aren't necessary if you follow the rest of the diagram (note the cathodes are all ultimately connected to a ground lug near the PT).

so what happens if the ground lugs are fitted and solder in as per a 1966 ? does this cause a problem ?

 

[playloud]

S

so what happens if the ground lugs are fitted and solder in as per a 1966 ? does this cause a problem ?

If you have both (i.e. the cathodes are all connected to one another, and then to the PT grounding point, plus grounded locally), you'll have many potential ground loops. Whether this is a problem in practice remains to be seen.

If you do it the original way (just ground each cathode to local lug), you'll be breaking with the "improved" grounding schemes, strictly speaking (whether Larry or single point). Honestly, I do it the original way on mine and haven't encountered any issues.

 

[Sigs]

If you have both (i.e. the cathodes are all connected to one another, and then to the PT grounding point, plus grounded locally), you'll have many potential ground loops. Whether this is a problem in practice remains to be seen.

If you do it the original way (just ground each cathode to local lug), you'll be breaking with the "improved" grounding schemes, strictly speaking (whether Larry or single point). Honestly, I do it the original way on mine and haven't encountered any issues.

So to make it look as per an original, so if wired as per "improved" grounding method and I isolate the screw and ground tag from the socket, thought of fitting an isolating washer where the tag is secured to the socket, plus ensuring the screw is not making contact with the tag, would this work ?

 

[playloud]

So to make it look as per an original, so if wired as per "improved" grounding method and I isolate the screw and ground tag from the socket, thought of fitting an isolating washer where the tag is secured to the socket, plus ensuring the screw is not making contact with the tag, would this work ?

Assuming I understand what you're asking, an isolated washer won't make a difference. You'll still have continuity between the cathodes (pin 8) and the chassis.

If the appearance thing really bothers you, I'd just ground the cathodes on the lugs and follow the improved grounding scheme (whichever you prefer) everywhere else. Removing the bus bar across the pots is likely to make a big improvement in its own right. If you really want to get into it, I suspect simply grounding the PI (i.e. presence pot) and speaker jacks in the same place makes a noticeable difference.