Starting a 2204 build

[glpg80]

Your B+ is being pulled down from 437V to 284V

Your bias is only -23V, is that as low as it goes? What’s your cathode current? Do you have 1 ohm resistors installed?

 

[ledvedder]

Your B+ is being pulled down from 437V to 284V

Your bias is only -23V, is that as low as it goes? What’s your cathode current? Do you have 1 ohm resistors installed?

I do have 1 ohm resistors. Around 12mV is as high as I can get the bias on the power tubes.

 

[glpg80]

I had already done the full variac power up with no tubes installed, using the procedure here, https://robrobinette.com/Tube_Amp_Startup.htm.

Here are the voltage readings on both power tube socket pins. Top is without power tubes, bottom is with them. All preamp tubes were installed for both readings. View attachment 132433

You should have voltage on both pin 2 and pin 7 which are each side of the heaters. Pin 6 should be a no connection pin so I’m not sure why voltage is there. 3 is anode and 4 is screen grid.

Check that your heater wiring is correct. Somethings up with pin 7 being no voltage but pin 2 is.

 

[ledvedder]

You should have voltage on both pin 2 and pin 7 which are each side of the heaters. Pin 6 should be a no connection pin so I’m not sure why voltage is there. 3 is anode and 4 is screen grid.

Check that your heater wiring is correct. Somethings up with pin 7 being no voltage but pin 2 is.

I'm sorry, I should have noted it. Pin 7 is getting roughly the same voltage as pin 2.

Isn't pin 6 getting voltage from B+ (yellow wire connected to bottom of 10k 3w resistor)?

 

[glpg80]

Pin 6 isn’t used in EL34s/6L6s. What they did was supply it B+ that then feeds a 1k to pin 4 which is the screen grid, so that’s why you see the 1k screen grid resistor across 6 to 4. I personally do pin 4 to 1k to pin 3 since it eliminates two extra B+ wires but that’s just me.

Technically as that drawing is given you have 11k screen grid resistors and not 1k if you do a KVL loop starting at the anode on pin 3. Remember with KVLs, caps go open. The only advantage of doing it that way in the drawing is that the screen sees a bit more filtering than the anode does and your tubes will run a bit cooler.

 

[ledvedder]

Pin 6 isn’t used in EL34s/6L6s. What they did was supply it B+ that then feeds a 1k to pin 4 which is the screen grid, so that’s why you see the 1k screen grid resistor across 6 to 4. I personally do pin 4 to 1k to pin 3 since it eliminates two extra B+ wires but that’s just me.

Technically as that drawing is given you have 11k screen grid resistors and not 1k if you do a KVL loop starting at the anode on pin 3. Remember with KVLs, caps go open. The only advantage of doing it that way in the drawing is that the screen sees a bit more filtering than the anode does and your tubes will run a bit cooler.

Hmm, that's how I've always seen it done. Ceriatone's layouts are the same. Should I change something?

 

[glpg80]

No it’s fine - you’ll get better tube life leaving it.

You have bigger fish to find lol. Tubes shouldn’t drop B+ from 440 to 280V with only 12mV of idle current.

With 4x 6550s my B+ only dropped from 480 to 440 at 46mA idle cathode current.

 

[ledvedder]

No it’s fine - you’ll get better tube life leaving it.

You have bigger fish to find lol. Tubes shouldn’t drop B+ from 440 to 280V with only 12mV of idle current.

With 4x 6550s my B+ only dropped from 480 to 440 at 46mA idle cathode current.

I'll have to find time this afternoon to draw up my power connections, since it's kind of a hybrid between Modulus' plexi layout and Ceriatone's 2204 layout.

 

[glpg80]

I'll have to find time this afternoon to draw up my power connections, since it's kind of a hybrid between Modulus' plexi layout and Ceriatone's 2204 layout.

Yeah I think the next step is to start documenting schematic wise exactly what you’ve got. There’s a small chance by design you just have too many 10k dropping resistors or you’re tapping B+ for the power tubes from the wrong location simply due to the hybrid design you’ve got.

The preamp parts we can concentrate on later.

Just out of curiosity what is your B+ voltage with no tubes at all?

Also look for loose solder balls or something that can short out B+ to chassis ground. Small details like that can cause Havock.

 

[ledvedder]

This is what I have going on. Upon 2nd inspection, the heater center tap is connected to a chassis ground lug, not the can cap ground lug. Apologies for the chicken scratch.

 

[ledvedder]

I'm guessing the light bulb limiter was causing issues with my voltages. Now I'm able to get power tubes biased to 32 and 36mV respectively. Plate voltage is then 432v.

Now I'm getting a super annoying hum and a faint high pitched squeal through the speakers that I'll have to figure out.

 

[mcstinger]

All gain/volume pots to zero. I would check OT leads to power tubes, whether they are not swapped. In such a case NFB becomes positive.

 

[ledvedder]

All gain/volume pots to zero. I would check OT leads to power tubes, whether they are not swapped. In such a case NFB becomes positive.

How do you know if the OT leads are wired to the correct power tubes?

 

[FourT6and2]

It's the light bulb limiter. I don't use those things, they just cause problems. Throw it away and start up the amp and your voltages will be ok.

EDIT: Whoops, you figured it out haha

 

[FourT6and2]

How do you know if the OT leads are wired to the correct power tubes?

If your OT has standard Marshall color codes, that's how you'd know. Otherwise, you just have to wire it up and hope you get it right the first time. But you would know instantly if that's the case. Amp would have an ear-splitting screaming high-pitched squeal. If you're getting more of a HF oscillation that gradually gets louder, it's probably something else. Lead dress or ground issue most likely. That's why you gotta build amps with good lead dress.

Looking at your photo, it looks like you have the OT anode connections wired ok. White/red wires. White usually goes to V5 and red to V4. But that's just assuming your OT uses historical Marshall colors. I would leave them alone for now if your amp isn't instantly screaming when you flick it out of standby.

Are your power tubes red platting? Also make sure your PI output wires aren't reversed. The ones going to the grids of power tubes. You used two green wires? How do you know which side is which? That's why you should use different color wire so you don't accidentally invert the phase.

 

[ledvedder]

If your OT has standard Marshall color codes, that's how you'd know. Otherwise, you just have to wire it up and hope you get it right the first time. But you would know instantly if that's the case. Amp would have an ear-splitting screaming high-pitched squeal. If you're getting more of a HF oscillation that gradually gets louder, it's probably something else. Lead dress or ground issue most likely. That's why you gotta build amps with good lead dress.

Looking at your photo, it looks like you have the OT anode connections wired ok. White/red wires. White usually goes to V5 and red to V4. But that's just assuming your OT uses historical Marshall colors. I would leave them alone for now if your amp isn't instantly screaming when you flick it out of standby.

Are your power tubes red platting? Also make sure your PI output wires aren't reversed. The ones going to the grids of power tubes. You used two green wires? How do you know which side is which? That's why you should use different color wire so you don't accidentally invert the phase.

I was just using the wire colors in the layout, so I wouldn't lose track. I've made sure each of the wires are going to the correct pins.

 

[SpiderWars]

Keep the NFB wire away from grid wires. And try to lay the grid wires against the chassis. You can pull tubes to try and isolate where the instability is.

Think of grid wires as antennas...picking up any stray signal. Think of plate wires as transmitters...shouting at full volume in all directions. Those transmitters and antennae have a phase. If they are same phase you get some positive feedback and potential instability. If opposite phase, they would tend to cancel and potentially add stability.

 

[SpiderWars]

All this is imo, I'd love to hear from the other builders:

Regarding the hum, check your grounds. The way you have it drawn above is not ideal imo. That first filter cap after the rectifier is the dirtiest one, keep it by itself right there at the base of that cap. Then you could either ground the screen/PI cap right there at its base (as you have it) or you could move it further away from the dirty ground. Your input jacks and preamp filters and all of the V1/V2 cathode grounds and volume grounds should all ground near/at the input jacks. The Mid control and Presence (i.e, PI) ground can be grounded closer to the Presence pot along with the bias supply ground and any center taps. Ground the OT secondary at the speaker jacks and if using Marshall style jacks run the ground wire to one of the power tube grounds.

Your power tube 1 ohm cathode biasing resistors look strange in an early pic in this thread. Like it is across pins 1 and 8? The resistor should be between pin 8 and ground and pins 1 and 8 should be connected to each other. Just do each power tube at its own socket unless you have bias points like the Ceriatones. EDIT: I now think I can see pin 1 grounded which will work. The suppressor grid (pin 1) has almost no current so as long as its grounded its fine.

I'm sure I left out stuff but there's a start. Hope it helps. Also go thru the amp with a chopstick and check every solder joint.

 

[ledvedder]

All this is imo, I'd love to hear from the other builders:

Regarding the hum, check your grounds. The way you have it drawn above is not ideal imo. That first filter cap after the rectifier is the dirtiest one, keep it by itself right there at the base of that cap. Then you could either ground the screen/PI cap right there at its base (as you have it) or you could move it further away from the dirty ground. Your input jacks and preamp filters and all of the V1/V2 cathode grounds and volume grounds should all ground near/at the input jacks. The Mid control and Presence (i.e, PI) ground can be grounded closer to the Presence pot along with the bias supply ground and any center taps. Ground the OT secondary at the speaker jacks and if using Marshall style jacks run the ground wire to one of the power tube grounds.

Your power tube 1 ohm cathode biasing resistors look strange in an early pic in this thread. Like it is across pins 1 and 8? The resistor should be between pin 8 and ground and pins 1 and 8 should be connected to each other. Just do each power tube at its own socket unless you have bias points like the Ceriatones.

I'm sure I left out stuff but there's a start. Hope it helps. Also go thru the amp with a chopstick and check every solder joint.

Can you explain more about the "dirty" ground?

Regarding the biasing resistors, I've placed them as they are in the Modulus layout, with the ground wire going back to the can cap.

 

[SpiderWars]

The first main filter is filtering out the main bulk of the power supply ripple to ground and is the dirtiest ground. It's a term I read Larry use. He likes to keep it separate from all other grounds and close to the earth ground.

But if you have a known good ground scheme then stick to that. I would expect it to be good. I've seen ground schemes like that too.

I rechecked your pics after posting and edited my post. Its crazy how simple these amps are yet we all do them a million different ways.