building a Fender "Pro-Amp" from scratch

[Rotorcraft230]

Add the 470 ohm resistors, they fix many headaches and don't change the tone much at all.
If you connect the screens to the extra OPT taps, you'll have an ultralinear output which usuallys sounds a bit more plain, And it distorts sharper, sort of like a solid state amp. It stays very clean until snapping into overdrive.
But you can instead just connect the screen supply to the plate connection if you want a more standard config, I'd suggest that.

Also, I'd recommend connecting the OPT center tap to before the power supply filter choke, and the screens after the choke. This puts much less strain on the filter choke, Also it makes the tone creamier because of some intermodulation distortion.
You can make the first filter cap a 47uF. the farther down the line the filter cap is, the lower the value should be because if you put like a 200uF filter cap on the preamp supply, if the power supply sags, it will take a long time to recover and will sound awful. That's why you should do the most filter in the early power supply stages, because there it will recover much faster even with high capacitance values

ELS, as far as the filters go I’m just going close to what Fender did(3-16uf and an 8uf in the preamp)F&T didn’t make those values so I went with 3-22uf caps and for the preamp it’s an 10uf.

I did see in other schematics Fender did split up the plates before and screens after the choke.

I was going to stay dedicated to the Tweed Pro schematic but if it makes it that much better I may do your recommendation, can you elaborate a little more in that?

Cheers
Mark

 

[ELS]

ELS, as far as the filters go I’m just going close to what Fender did(3-16uf and an 8uf in the preamp)F&T didn’t make those values so I went with 3-22uf caps and for the preamp it’s an 10uf.

I did see in other schematics Fender did split up the plates before and screens after the choke.

I was going to stay dedicated to the Tweed Pro schematic but if it makes it that much better I may do your recommendation, can you elaborate a little more in that?

Cheers
Mark

In a push pull output, the power supply ripple gets cancelled out so you don't hear hum. And the screens are much more sensitive to ripple than the plates, which too would get cancelled out.
But the headroom of the output would get modulated by the ripple, which actually is quite desirable for tone.

The reason why not to connect the plate after the choke if you want a more sterile sound, is because then there would be a lot more current flowing trough the choke, mandating a bigger choke.
So instead the screen only is connected after the choke, The plates not being nearly as sensitive to the ripple.
Also if you connect the plate supply after the choke, you will have extra sag, because the choke will have a DC resistance of around 100 ohms typically. One amp I built was unplayable because of the added sag from the choke when it was wired up like that. it would cut out.

There were some tests about power supply ripple, and they took a 5F6A bassman circuit and used a regulated power supply with no ripple, and reportedly it made it sound quite bad, and not at all like a bassman.

 

[Rotorcraft230]

In a push pull output, the power supply ripple gets cancelled out so you don't hear hum. And the screens are much more sensitive to ripple than the plates, which too would get cancelled out.
But the headroom of the output would get modulated by the ripple, which actually is quite desirable for tone.

The reason why not to connect the plate after the choke if you want a more sterile sound, is because then there would be a lot more current flowing trough the choke, mandating a bigger choke.
So instead the screen only is connected after the choke, The plates not being nearly as sensitive to the ripple.
Also if you connect the plate supply after the choke, you will have extra sag, because the choke will have a DC resistance of around 100 ohms typically. One amp I built was unplayable because of the added sag from the choke when it was wired up like that. it would cut out.

There were some tests about power supply ripple, and they took a 5F6A bassman circuit and used a regulated power supply with no ripple, and reportedly it made it sound quite bad, and not at all like a bassman.

Very nice ELS, that’s what I’ll do split the plates and the screens. I’ve been building guitar amps from kits for a while now and this one is my first scratch build. I’ll say it’s much more enjoyable.

So here is a question for anyone. Here is the mounted circuit board. It extends a little over the OT. The primary is on the PT side. On that end are the 3 22uf 500v filters. Will the caps there interfere with the OT? I can move the caps to the top and make a dog house if that will help with anything. And actually I like the doghouse idea.

 

[ELS]

Very nice ELS, that’s what I’ll do split the plates and the screens. I’ve been building guitar amps from kits for a while now and this one is my first scratch build. I’ll say it’s much more enjoyable.

So here is a question for anyone. Here is the mounted circuit board. It extends a little over the OT. The primary is on the PT side. On that end are the 3 22uf 500v filters. Will the caps there interfere with the OT? I can move the caps to the top and make a dog house if that will help with anything. And actually I like the doghouse idea.View attachment 133507

Do the caps fit between the PCB and the bottom? if so it's fine.
Make sure the leads don't short to anything on the other side of the PCB.
A dog house for the caps isn't ideal hum wise, having the caps on the PCB near the parts of the circuit they supply, sort of like a marshall layout is more ideal. Then for least hum you can use a ground bus; Have one thick wire that connects to the chassis near the input jacks. Have the preamp connections near the input side of the ground bus, next stage connections further down, and so on, Until you get to the output section which connects at the end of the ground bus.
Make sure to connect the filter caps for those stages as close to those stage ground connections as possible

Like this:


You don't actually need like a solid ground bar, you just need to keep the idea in mind when wiring, that the ground for the preamp should be near the input jacks, the ground for the 2nd stage further down, and so on until you get to the output section where it's at the end of the chassis.
Also a note, in the drawing, it shows the speaker output ground connected near the end, but actually you should connect it near the Phase Inverter ground, If you have negative feedback which most amps do have.

 

[Rotorcraft230]

Thanks ELS, the board is going to have Turrets, everything will be mounted on the side you’re looking at. I haven’t drilled it for turrets yet. I am going to make a ground buss for the grounding. Here are 2 different ways I’ve done it. Both ways are grounded by the inputs. One style is above the pots and the other is right off the circuit board, it runs the whole length right off the board.the board. I WILL not solder a buss to the backs of the pots.

 

[ELS]

Here's a modified schematic.

Notes:
The original schematic had an error where the ground for the volume pots wasnt drawn. And also early selenium and germanium diodes often were backwards from today's orientation.
The band of the diode needs to point towards the power transformer for the bias supply!
The 2 100 ohm resistors are an artificial center tap, just take a couple 100, or 220 ohm resistors, connect them from each filament supply wire to ground. This way you have much less hum than with the original layout of having one filament wire grounded. I recommend putting the resistors near the first preamp tube near the inputs.
I've numbered the grounds, the ones with the same number should be as close together as possible, and the numbers that are close to each other reasonably close (so don't put the '1' grounds a mile away from the '2' grounds)
The 2 0.0022 to 0.01 caps are optional, if you ever plug the amp into an ungrounded outlet they will reduce the hum, And if you use modern Y1 rated safety caps, they are meant for this and wont short.
The 1st main filter capacitor should be as close to the rectifier tube as possible, have short leads from the rectifier to the filter cap because they will put out a lot of hum if they're not. The ground too, the transformer center tap should be very close to the ground of the filter cap (the '5' ground)

 

[Marcomel79]

Missed this thread! Always a pleasure to see your builds @Rotorcraft230

 

[Rotorcraft230]

Missed this thread! Always a pleasure to see your builds @Rotorcraft230

Thanks man, I know it’s not a Marshall. But this is where I like to hang out.

Mark

Here's a modified schematic.
View attachment 133569
Notes:
The original schematic had an error where the ground for the volume pots wasnt drawn. And also early selenium and germanium diodes often were backwards from today's orientation.
The band of the diode needs to point towards the power transformer for the bias supply!
The 2 100 ohm resistors are an artificial center tap, just take a couple 100, or 220 ohm resistors, connect them from each filament supply wire to ground. This way you have much less hum than with the original layout of having one filament wire grounded. I recommend putting the resistors near the first preamp tube near the inputs.
I've numbered the grounds, the ones with the same number should be as close together as possible, and the numbers that are close to each other reasonably close (so don't put the '1' grounds a mile away from the '2' grounds)
The 2 0.0022 to 0.01 caps are optional, if you ever plug the amp into an ungrounded outlet they will reduce the hum, And if you use modern Y1 rated safety caps, they are meant for this and wont short.
The 1st main filter capacitor should be as close to the rectifier tube as possible, have short leads from the rectifier to the filter cap because they will put out a lot of hum if they're not. The ground too, the transformer center tap should be very close to the ground of the filter cap (the '5' ground)

so let me see if I have your suggestions right. The numbers on the schematic are the order of how they are to be connected to the ground buss, or are you saying 5 different grounds throughout the amp. I did Larry’s grounding scheme on a JTM45 and it was pretty quiet but compare that amp to some others with the star method and I didn’t think it was that much more quiet. I like using the Star method, although there are a few stray grounds on my past builds I suppose for convince. I may try a heater hum balance pot, we have one in a Twin, seems to work good (probably won’t though).

So here is another question, on the heaters CT, are we balancing voltage or the phase. When I measured the heaters on this PT both legs were 3.15, I was actually shocked. So if it’s voltage do I need a artificial CT?

Oh and the mains ground ran all the way to the inputs? That opposite ends of the amp, I’ve always done, read and been told to put it as close to the socket as possible.

 

[Marcomel79]

Thanks man, I know it’s not a Marshall. But this is where I like to hang out.

Its where the best people are at!

 

[Rotorcraft230]

I am putting it in a JTM45 chassis

 

[neikeel]

You need 6.3vac for your heaters.
If you have 3.15-3.15 and no centre tap for the heaters (ring the PT out to find one!) then you will need to raise and artificial centre tap with two resistors between the heater chain and ground. You can use anything from 220R up to a max of 100R.
If you don't do this you will likely get horrible hum.
I would not usually add a heater balance pot, rarely needed and another thing to go wrong!

Edited for 100R vs 100K

 

[Rotorcraft230]

You need 6.3vac for your heaters.
If you have 3.15-3.15 and no centre tap for the heaters (ring the PT out to find one!) then you will need to raise and artificial centre tap with two resistors between the heater chain and ground. You can use anything from 220R up to a max of 100k.
If you don't do this you will likely get horrible hum.
I would not usually add a heater balance pot, rarely needed and another thing to go wrong

You need 6.3vac for your heaters.
If you have 3.15-3.15 and no centre tap for the heaters (ring the PT out to find one!) then you will need to raise and artificial centre tap with two resistors between the heater chain and ground. You can use anything from 220R up to a max of 100k.
If you don't do this you will likely get horrible hum.
I would not usually add a heater balance pot, rarely needed and another thing to go wrong!

Now that I think about it I was measuring the heater winding, it doesn’t have a CT, I don’t know what I was thinking when I said 3.15 per leg, I do have 6.3. I am going to put the 100 ohm resistors on each leg to ground.

You need 6.3vac for your heaters.
If you have 3.15-3.15 and no centre tap for the heaters (ring the PT out to find one!) then you will need to raise and artificial centre tap with two resistors between the heater chain and ground. You can use anything from 220R up to a max of 100k.
If you don't do this you will likely get horrible hum.
I would not usually add a heater balance pot, rarely needed and another thing to go wrong!

 

[neikeel]

Oops, I put 220R at one end of value range and hit 100k by mistake, should have been 100R, edited it now

 

[ELS]

Thanks man, I know it’s not a Marshall. But this is where I like to hang out.

Mark

so let me see if I have your suggestions right. The numbers on the schematic are the order of how they are to be connected to the ground buss, or are you saying 5 different grounds throughout the amp. I did Larry’s grounding scheme on a JTM45 and it was pretty quiet but compare that amp to some others with the star method and I didn’t think it was that much more quiet. I like using the Star method, although there are a few stray grounds on my past builds I suppose for convince. I may try a heater hum balance pot, we have one in a Twin, seems to work good (probably won’t though).

So here is another question, on the heaters CT, are we balancing voltage or the phase. When I measured the heaters on this PT both legs were 3.15, I was actually shocked. So if it’s voltage do I need a artificial CT?

Oh and the mains ground ran all the way to the inputs? That opposite ends of the amp, I’ve always done, read and been told to put it as close to the socket as possible.

If there's 2 3.15V windings you can put them in series for 6.3V.
An artificial center tap is preferred but using the transformer center tap is fine too, less parts that way.
The ground connection from the power cable should be connected next to the inputs, it gives least hum and is a bit safer, in case like the whole chassis rusts away.

I checked out Larry's grounding scheme and it's exactly how I explained. He didn't use a ground but but instead the chassis, this is just by choice, some amps are easier to make a ground bus on the circuit board, others are easier connecting to the chassis, If you did the layout well, the chassis version should perform just as well as the ground bus.
With a couple exceptions tho: Ground 6- you should connect the transformer HT center tap directly to the first filter cap, no connections in between, Then you should solder another wire from the ground of the filter cap, to ground 5.
And ground 5 should be connected to ground 4, The cathodes of the output tubes should be as close to the ground for the grid leak or bias supply ground as possible.

The point of these ground busses and ground schemes is to have minimal resistance (wire length, chassis space.) between connections that have high current, For example the connection between the HT winding on the power transformer, the rectifier, and the first filter capacitor will have tons of ripple current, buzz. If you ground the first filter capacitor at the other end of the chassis, all that ripple current will go trough the whole chassis, which will make the chassis emit buzz or hum that will then be induced in any sensitive parts of the circuit.
Secondly, If there's any current going trough the ground connections regardless, you want the Cathode of a tube to be grounded at the same point where the grid leak or volume pot prior to that tube is grounded at, to have minimal distance there.
Thirly since some noise current can still be flowing trough the ground, you want to connect the earth to the inputs instead of near the outputs, this way the most sensitive parts of the circuit have the least resistance to ground and the noise gets shunted down.

So to summarize: Keep high current paths short; Connect the grid leak grounds to the cathode grounds for the same tubes; Ground the earth connection at the inputs

 

[AndyD]

If there's 2 3.15V windings you can put them in series for 6.3V.
An artificial center tap is preferred but using the transformer center tap is fine too, less parts that way.
The ground connection from the power cable should be connected next to the inputs, it gives least hum and is a bit safer, in case like the whole chassis rusts away.

I checked out Larry's grounding scheme and it's exactly how I explained. He didn't use a ground but but instead the chassis, this is just by choice, some amps are easier to make a ground bus on the circuit board, others are easier connecting to the chassis, If you did the layout well, the chassis version should perform just as well as the ground bus.
With a couple exceptions tho: Ground 6- you should connect the transformer HT center tap directly to the first filter cap, no connections in between, Then you should solder another wire from the ground of the filter cap, to ground 5.
And ground 5 should be connected to ground 4, The cathodes of the output tubes should be as close to the ground for the grid leak or bias supply ground as possible.

The point of these ground busses and ground schemes is to have minimal resistance (wire length, chassis space.) between connections that have high current, For example the connection between the HT winding on the power transformer, the rectifier, and the first filter capacitor will have tons of ripple current, buzz. If you ground the first filter capacitor at the other end of the chassis, all that ripple current will go trough the whole chassis, which will make the chassis emit buzz or hum that will then be induced in any sensitive parts of the circuit.
Secondly, If there's any current going trough the ground connections regardless, you want the Cathode of a tube to be grounded at the same point where the grid leak or volume pot prior to that tube is grounded at, to have minimal distance there.
Thirly since some noise current can still be flowing trough the ground, you want to connect the earth to the inputs instead of near the outputs, this way the most sensitive parts of the circuit have the least resistance to ground and the noise gets shunted down.

So to summarize: Keep high current paths short; Connect the grid leak grounds to the cathode grounds for the same tubes; Ground the earth connection at the inputs

Really nicely explained! Thank you.

 

[AndyD]

Beautiful build, as always. A work of art!

 

[ELS]

Really nicely explained! Thank you.

Oh I forgot to mention, the reason for the 'secondly' point. The tube stage senses the signal by comparing it to the cathode. So if the cathode voltage varies by lets say 1V, it would be the same as having the cathode not vary but the grid vary 1V.
So the cathode connection should be right next to the grid leak connection to minimize any voltage difference created between those ground connections, from hum current trough the ground.

 

[Rotorcraft230]

I’d say very well explained as well!!

I’m gonna try and follow the grounding scheme. I will probably have an exception or two.

The amp is going to be for my son and believe it or not he really doesn’t mind hum. He will probably mod it anyway, but for now it’s mine and I like them quite.

If there will be multiple grounds at one point I came up with this. I also laid out the grounds on the layout, to give me an idea of where the connections and grounds will be.

 

[Rotorcraft230]

Beautiful build, as always. A work of art!

Thanks Andy, I don’t play so I go OCD in building them.

 

[ELS]

I’d say very well explained as well!!

I’m gonna try and follow the grounding scheme. I will probably have an exception or two.

The amp is going to be for my son and believe it or not he really doesn’t mind hum. He will probably mod it anyway, but for now it’s mine and I like them quite.

If there will be multiple grounds at one point I came up with this. I also laid out the grounds on the layout, to give me an idea of where the connections and grounds will be.View attachment 133703View attachment 133704View attachment 133705

Look good.
For the ground terminals, you should create a good mechanical connection as you probably know. Bend the end of that twisted wire back trough the other hole in the terminal, then bend it around so it touches the other hole in the terminal, Then squeeze it with pliers (don't squeeze so hard you split the wire), it shouldn't move when you try to move it, even without solder.

Old Philco TV's didn't have the pot connections soldered, instead they took the stripped wire and wrapped it around the pot terminals like 8 turns or so. Those connections are still holding fine in rusty TV's 70 years later... again no solder, purely mechanical. Where there's solder, if there's a gap between the copper wires, oxides will slowly form in those gaps and isolate the connections.
For shielded cables, ground one end... Ground loops are a problem where the shield passes a lot of noise current; If the ground between those paths of the circuit has ~20mohm of resistance, and the shield in the cable is 20mohm of resistance, and you connect both ends. Half of the noise that flows trough the ground, now flows trough your shield, inducing all that noise into the audio signal flowing on the middle conductor.


Here's some things maybe of interest:
The components should have a bit of strain relief as shown here, as per nasa standard.



The power transformer wire colors are sort-of standardized, tho they aren't always following this standard.

If you want to make the amp really neat looking, look into wire Lacing. It's what was done back in the day for mil-spec electronics also often retired military electricians would do this in consumer equipment since they had knowledge of it.
The string they use is some type of string, soaked in wax preferably beeswax since that's much softer.
Here's some info on that:

401 General Requirements

Wiring and Cabling: How To Lace Cable Harnesses

AeroElectric Connection - Wire Bundle Tying Techniques

https://wiki.velocityoba.com/images/a/ab/4b-ch11_12.pdf

https://www.mathscinotes.com/wp-content/uploads/2014/01/021c99s2c08.pdf

As you can see, there's more than one way to do it, just use whatever method works best for you.

Just don't tie the output tube plate, cathode, or screen wires next to the grid wires.
Also the Preamp grid wires, next to the plate wires of the next stage, don't have those near eachother.Speaker connections should also be kept away from the preamp, And the power transformer wires should also be kept away from everything else