Help with Four channel amp with both 6v6’s and 6L6’s build?

[Jerry garrcia]

I’m bored and one build is almost complete, still working so can’t finish the Tele build due to need to have vacation so I can work at the county house and the archtop build won’t start for a month or so.
Have this idea for an amp I wanna build but lacking in experience to complete a smart schematic.
Background @printer2 and @andrewRneumann helped me to build this amazing amp. A SS rectifier, two channels grid leak biased 6SJ7 and a 12ay7, a 6SC7 PI and two 6v6’s in a 5E3 chassis.

A beautiful sounding amp and I want to take it to the next level. Want to combine that amp with the cathode biased mic and instrument channel of the Gibson GA-50.

That will not be a big hassle but I got these NOS 6L6’s from a friend, would be fun to have a higher wattage amp with the 6L6 sound.
So I want to build a 4 channel amp with the possibility to run it with both 6v6’s and be able to switch to 6L6’s.
Had this idea to connect all channels to the same PI and use a switch to activate the two different PP circuits heater filaments. Also to run the speakers, 8” and 12”, on different OT secondaries. Not that keen on the miss match option. A 6.6k OT? A 5AR4 rectifier?
Better to build two amps in one chassis with a common power supply?

https://i.imgur.com/2dGl1L7_d.webp?maxwidth=640&shape=thumb&fidelity=medium

Any creative ideas would be very appreciated on a possible setup and what kind of PT would be needed. Looking for “Yes Man” comments and solutions. Don’t wanna build me a GA-50 clone but instead to have it all.

 

[andrewRneumann]

Is that @printer2 schematic you linked to the current working idea?

Sounds ambitious, and definitely go into this knowing that you are in a somewhat "experimental" zone. Things may not work out exactly as you want.

My concerns (not problems--just areas to make sure you have thought through) would be these:
1. Noise. You have 4 inputs that are basically always active. You can turn down the volume on unused channels, but I still have my concerns that you will get more noise than you want.
2. Loss of signal at the junction of 4 channels. The amount of signal loss will depend on the position of the volume knobs and the tube that is doing the driving.
3. I assume the 1K resistor in the power supply is to drop voltage for the 6V6s. This will cause quite a bit of sag when the amp is driven. Notice on your last amp that we used a 220R. Assuming relatively similar amounts of current, you can expect almost 5x as much voltage sag with the 1K. There may be a better solution, such as using Zener diodes to drop the voltage needed for 6V6.
4. There is a schematic error in the power amp circuit. The lowest power tube is not being driven by the plate of the phase inverter.
5. Has an analysis been completed on the 270R cathode bias resistor? Does that work for 6V6 and 6L6 or do you need to break out the cathodes to individual resistors and capacitors?

 

[andrewRneumann]

Another possible concern I can see is with switching the heaters from the 6V6s to 6L6s. This would necessarily require putting the power tube heaters downstream of the preamp heaters. It is typical to heat the power tubes first, as this draws the most current and therefore causes the most noise (and the power tubes are the least sensitive to noise), then wiring the preamp heaters down the chain until the input tube is last. This minimizes current in the heaters of the input tube and theoretically should minimize noise. Since your preamp tubes will be first, followed by the power tubes, I would worry about noisy heaters in your preamp tubes. I don't have any first hand experience doing it this way, I just know what is typical in these heater chains. Power tubes-->P.I.-->Gain stages-->Input Tube(s)

Edit: Should have thought it through a little more. Now see that the preamp tube string could be branched off before the switch to the power tubes. Always have to take an extra minute to think about these things it seems.

 

[Jerry garrcia]

Is that @printer2 schematic you linked to the current working idea?

Hi again. That’s just a starting schematic for some brainstorming on how to proceed. One other idea is to have two amps in one cab with a common power supply. But that doesn’t combine it all. Two channels with 6L6 and two with 6v6’s. To few options with that configuration. Then I’ll have to build two amps and I does not get it all with just one.


Thanks for your inputs! I’ll try to analyse them tomorrow. Need to get up in 5 hours for a looong day of work.
Will also try to do some reading about the noice issue with 4 active channels.
Was planning to use a big chassis this tome so I will have more room for creative wiring than in the previous builds.

At this stage my main question is primarily if it would be possible with such a build? If so, then there is time for a solution.

@andrewRneumann you will be glad to hear that even this used up brain, have learned a bit of circuits in the last couple of month.

 

[Jerry garrcia]

Has an analysis been completed on the 270R cathode bias resistor? Does that work for 6V6 and 6L6 or do you need to break out the cathodes to individual resistors and capacitors?

No, not at this stage. A brainstorming schematic that will be, hopefully, fine tuned. If it is possible at all? Also a concern with the size of the PT and OT. Would be good to be able to carry it by hand and not on a 18 wheeler

 

[Jerry garrcia]

One thing popped up before I fell a sleep if one were to use toroidal transformers (less weight, less noise?) can’t the switch between the two power tube variants be to switch which PT that is active?

 

[Jerry garrcia]

For a tube rectangle citquit?
https://www.tube-town.net/ttstore/toroidal-56va.html only.m 0.1A
Then need to use a low voltage drop RT since B+ is 315 in a annotated GA-50 schematic I found.

Tube Type	DC mA	PIV	FIL AMPS	Max ACV	DCV	DCV DROP
Diode Rectifier					425
5AR4 / GZ-34	250	1500	1.9	450	415	10
5V4-GA	175	1400	2	375	400	25
GZ37	350	1000	2.8	450	388	37
5U4-GA	250	1550	3	450	381	44
5U4-G	225	1550	3	450	381	44
5U4-GB	275	1550	3	450	375	50
5Y3-G/GA	125	1400	2	350	365	60
5R4GYB	250	3100	2	900	362	63
5R4G/GY/GYA	250	3100	2	750	358	67

The 5AR4 looks like the best fit that purpose (got a few NOS) harder on the B+ for 5E3 goes from 350-380VAC. So a SS rectifier with a sag resistor seams like an easier path. Not as fun. Will need to find a more powerful but still light PT if
My idea is gonna work. Almost no sleep for me tonight.

 

[printer2]

Is that @printer2 schematic you linked to the current working idea?

Sounds ambitious, and definitely go into this knowing that you are in a somewhat "experimental" zone. Things may not work out exactly as you want.

My concerns (not problems--just areas to make sure you have thought through) would be these:
1. Noise. You have 4 inputs that are basically always active. You can turn down the volume on unused channels, but I still have my concerns that you will get more noise than you want.
2. Loss of signal at the junction of 4 channels. The amount of signal loss will depend on the position of the volume knobs and the tube that is doing the driving.
3. I assume the 1K resistor in the power supply is to drop voltage for the 6V6s. This will cause quite a bit of sag when the amp is driven. Notice on your last amp that we used a 220R. Assuming relatively similar amounts of current, you can expect almost 5x as much voltage sag with the 1K. There may be a better solution, such as using Zener diodes to drop the voltage needed for 6V6.
4. There is a schematic error in the power amp circuit. The lowest power tube is not being driven by the plate of the phase inverter.
5. Has an analysis been completed on the 270R cathode bias resistor? Does that work for 6V6 and 6L6 or do you need to break out the cathodes to individual resistors and capacitors?

1. yes, 2. yes. 3. yes, but need to nail down which PT and OT(s), was spitballing and a diagram was better than words 4. Ah shoot, I was trying to get it done fast to show JG along with discussing other options. 5. also agree.

I would have nailed down things with further input but I am pressed for time in other areas if life and amp going to have to bow out on this one. Other things I thought of was a common OT as shown, either 6.6k, or a multi tap secondary and to switch taps to have a 4k for the 6L6's and 8k for the 6V6. also I think there should not be any ill effects with the unused tubes connected without the heaters power. But that is not from experience. Yes, branched off heaters. Also I thought of a completely separate amp from input to output with each its own OT. The 15W 6V6 one is only two lbs and only an additional PI tube would be needed. So with a 4PTP switch you would switch both heater legs, the voltage dropping resistor and the last pole use to switch the speaker between the two output transformers. The switch will not be under too much voltage stress, the 6.3V, the output signal voltage, the voltage drop of the dropping resistor (or zeners). There will be 300V difference between the HV pole and the adjacent pole, not sure if it would be an issue internally. I would heatshrink the exterior pins of the switch if going that way.

I think the debating on the input circuits would be the first thing to discuss and whether to go separate amp route other than the PS. I do not like the voltage drop of the four resistor mixer. Mind you, this is not a high gain amp. This was a starting point.

The common OT route, I would go with the 6V6 and 6L6 each have there own cathode resistor and cap rather than combined. One less place to have problems. Some transformer options,

TT Trainwreck Express Style Output transformer 6k6 / 4-8-16 Ohm

Ausgangsübertrager für Trainwreck Express Projekte Primär: 6600 Ohm C.T. Sekundär: 4-8-16 Ohm Leistung: 35 W (22Hz -3dB) Dateien: Datenblatt

www.tube-town.net

Hammond 290CEX for Fender Power 125P26A Vibrolux

Powertransformer (Vibrolux, AB763) (Vibrolux Reverb, AA270, AA964, AB568) (Tremolux, AA763, AB763) Fender Part Number 125P26A, 022723, 036483 Pri: 240V 50Hz Sec.1: 650VCT @ 207mA Sec.2: 5 V @ 3A Sec.3: 6.3 V @ 4A Files...

www.tube-town.net

Hammond 1760JB for Fender Outputtransformer 045216 / 1846 / 1848 UPGRADE Bandmaster Tweed

Output Transformer 045216 / 1846 / 1848 Push-Pull for Bandmaster Tweed, Pro Tweed, Super Brownface, Super Tweed Pri: 6000 Ohm / CT Sec: 2.67, 4, 8 Ohm Power: 40 Watt Files: Datasheet

www.tube-town.net

Hammond 1750J for Fender Outputtransformer 125A6A / 022848 Tremolux

Output Transformer 125A6A / 022848 Push-Pull for Pro-Reverb, Tremolux, Vibrolux Pri: 4000 Ohm / CT Sec: 4 Ohm Power: 35 Watt Files: Datasheet

www.tube-town.net

Hammond 1760J for Fender Outputtransformer 125A6A / 022848 UPGRADE Vibrolux

Output Transformer 125A6A / 022848 Push-Pull for Pro-Reverb, Tremolux, Vibrolux Pri: 4000 Ohm / CT Sec: 4, 8, 16 Ohm Power: 35 Watt Files: Datasheet

www.tube-town.net

Hammond 1750JA for Fender OT 050438 Hot Rod Deluxe

Output Transformer 050438 Push-Pull for Hot Rod Deluxe Pri: 4250 Ohm / CT Sec: 4,8 Ohm Power: 40 Watt Files: Datasheet

www.tube-town.net

https://www.tube-town.net/ttstore/hammond-choke-159s.html 225 mA 5H

https://www.tube-town.net/ttstore/hammond-choke-159r.html 200mA 6H

Hammond 290EEBZ for Fender 65 Reissue Twin Reverb - 037610

Replacement transformer for Fender Part No. 037610 Universal primary windings for mains voltages of 120/230/240V, 50/60 Hz Designed for drop in replacement of original units. Physically, parts were designed to look similar to original ...

www.tube-town.net

Hammond 290EEX for Fender Power 125P7D / 022814 Bassman

Fender Style Powertransformer for (Bandmaster 6G7A, AA763, AB763, AC568, AD1269, Blackface, Blonde, Silverface) (Bassman AA165, AA270, AA864, AB165, Blackface, Silverface) (Concert 6G12A, AA763, Blackface, Brownface) Fender Item Number: 125P7D &...

www.tube-town.net

If that is possible the perfect amp will be this (just dreaming):
a four channel amp with
1) one channel cathode biased 6SJ7, like the mic channel of the GA 30/50
2) one channel cathode biased 6SJ7, like the instrument channel of the GA 30/50
3) one channel grid leak biased 6SJ7, like the previous builds
4) one channel cathode biased 12ay7, like the previous builds
5) 6SC7 PI
6) Option to run the amp on 6v6's or 6L6's for possibility to have a 12-15w or a 30-40w amp.
7) the 90% reducing circuit with a switch (or can it be done with a pot so the output can be adjusted?)
8) a spring reverb?
9) Low weight
10) All components resembling of the old specs (like only carbon resistors where it is possible and NOS tubes (have the possibility to get good tubes really cheap).

https://www.tdpri.com/attachments/ga30-pdf.977187/

I think that is mostly it.

https://www.tdpri.com/attachments/annotated-schematic_ga-50-pdf.977186/

 

[2L man]

This 74 euros toroid PP OT weights only 664g. It is HiFi rated for 25W so it would be good for 50W guitar amps. 8k is good for 6V6GT and it has been used for 6L6GC. I have test it in 5E3 and used two very different tubes trying to saturate it and there came more distortion at 100Hz when driving about 10W power than balanced tube pair did produce but sound playing guitar was not changed.

25Watt 8kOhm Push-Pull outputtransformer

Ringkern uitgangs transformator voor bijvoorbeeld 2x EL84 push-pull versterkers. Primaire impedantie : 8000 ohm Secundaire impdantie : 0 - 4 - 8 -

www.toroidal-transformer.com

Tube Town tt-t139 toroid PT weights 1890g. It has two 300V/0,15A output coils which might be too powerful for even 6L6GC PP when calculated filtered DC current RMS max comes 211mA

tt-t106 weights 1450g and has two 300/0,126A HT coils so its RMS DC turns to 177mA. 6,3V output max is 3,25A but if you don't use tube rectifier 5V/2A coil has also 6,3V/1A output to supply more pre amp tubes. When those are turned to VAs 10VA and 6,3VA are strange different and I think 6,3 is capable to output much more than 1A. I don't think they would have used thinner wire between 5V and 6,3V outputs!!!

 

[2L man]

Are you planning to use Choke for B+1 or B+2? Choke is not often used for B+1 on instrument amps. For B+2 its current rating seem way too high and inductance low so there comes unnecessary weight!

 

[printer2]

Gibson used a choke before B1 on the amp. With a smaller first capacitor it keeps the voltage down as compared to a larger capacitor. I wouldn't go with a toroid in this case, trying to get the closer to the amps in question and a EI core may be more like the original. The tubes do not have to be matched either. OK, I will pretend I didn't look.

 

[andrewRneumann]

Gibson used a choke before B1 on the amp. With a smaller first capacitor it keeps the voltage down as compared to a larger capacitor. I wouldn't go with a toroid in this case, trying to get the closer to the amps in question and a EI core may be more like the original. The tubes do not have to be matched either. OK, I will pretend I didn't look.

What plate voltages are we looking at targeting here for 6V6 and 6L6?

 

[printer2]

What plate voltages are we looking at targeting here for 6V6 and 6L6?

I am thinking on the lower end of the guitar world. Datasheet values, 360V for the 6L6's, add to that the cathode resistor bias voltage. 6V6, maybe 360V with 20V of that across the cathode resistor.

 

[andrewRneumann]

I am thinking on the lower end of the guitar world. Datasheet values, 360V for the 6L6's, add to that the cathode resistor bias voltage. 6V6, maybe 360V with 20V of that across the cathode resistor.

Not to be contrarian, but if we are focused on clean and jazzy, maybe we should consider the high end of achievable voltages.

 

[printer2]

Not to be contrarian, but if we are focused on clean and jazzy, maybe we should consider the high end of achievable voltages.

Jazz from the 30-50's. I am guessing not necessarily clean by today's standards.



Inspired by the glass of beer.



Another hour listening to amps online. Where does the time go during my day?

 

[Jerry garrcia]

I might throw in the towel for this build at this point and make it further on. To much stuff at work going on and have not got the energy needed to push through now. Also building two guitars and this damn war is not good for spare time for amp planning, reading, calculating and cabinet building. Don’t have enough knowledge at this point to do it. But soon…

Will do two much more boring projects, I think.
-A two channel, single ended, SS build in a 5F2a chassis. One grid leak biased 6sj7 and one cathode biased 6sj7 and a 6v6. Can stick that thing in an already built cabinet.
-A two channel (same preamps as above), push pull, paraphrase PI with two 6L6’s. Using a 5E3 chassis so it will also fit in another cabinet I have.
Aiming for clean tones. Do I need to bias the pentode differently than in the GA-20/30/50 circuits? Also inputs on if it is better to tweak the circuits differently than in the original schematics of the cathode biased 6sj7 in the old Gibson circuits? Some say that there are room for improvements due to some think they are not the best thought out circuits.

Hopefully I’ll gain some knowledge during those builds so It will be possible to do the dream build in this thread. A shame really but sh*t happens.

I might also build a micro amp with a grid leak biased 6SJ7 preamp and a triode as power tube. Will a 6sc7 work fort that purpose or do I have to go with a 12ax7?

So this project will probably wait.

 

[Jerry garrcia]

 

[2L man]

Aiming for clean tones. Do I need to bias the pentode differently than in the GA-20/30/50 circuits? Also inputs on if it is better to tweak the circuits differently than in the original schematics of the cathode biased 6sj7 in the old Gibson circuits? Some say that there are room for improvements due to some think they are not the best thought out circuits.

You can (first) use values from the circuits you know to produce the sound you like. If you wan't to tune tube circuits I recommend to use this link tool because it is great It is based to tube datasheets and in few minutes you can simulate circuits always when you have enough data to advance calculating currents. For preamp stages tick "resistive" but often it is a default.

For example to begin analycing circuits others have made you need anode/plate current and you need anode resistor value, its B+_ supply voltage and anode voltage Va and you subtract Var (Voltage over anode resistor). In your own circuits you just measure them. Calculated carhode current usually come different when Vc usually has less decimals.

Then set "headroom" value which is horizontal voltage peak to peak sweep which comes to anode on green. How linear it comes both side of the operating point (which is Va when it is the anode which feeds the following stage) defines this stage output distortion. When cursor is pointed to grid lines you can see what drive signal Vpp amplitude is needed for that output. Stage gain comes when output sweep is calculated using input sweep.

When green sweep is not linear the stage produce distortion which is mostly 2nd harmonic. Now you can change/tune parameters and look what loadline comes before changing resistors or tube type in your amp. Traditional method was to calculate values and draw loadline to datasheet which is slower. You can save your work, copy and post them.

Tubes have tolerances and they wear out so in actual amp voltages/currents usually come more or less different whan this tool give but is is good!!!

Universal Loadline Calculator

This calculator allows you to trace the anode characteristics graph, and to compute and tune loadline and operating point of various vacuum tubes.

www.vtadiy.com

 

[Jerry garrcia]

You can (first) use values from the circuits you know to produce the sound you like. If you wan't to tune tube circuits I recommend to use this link tool because it is great It is based to tube datasheets and in few minutes you can simulate circuits always when you have enough data to advance calculating currents. For preamp stages tick "resistive" but often it is a default.

For example to begin analycing circuits others have made you need anode/plate current and you need anode resistor value, its B+_ supply voltage and anode voltage Va and you subtract Var (Voltage over anode resistor). In your own circuits you just measure them. Calculated carhode current usually come different when Vc usually has less decimals.

Then set "headroom" value which is horizontal voltage peak to peak sweep which comes to anode on green. How linear it comes both side of the operating point (which is Va when it is the anode which feeds the following stage) defines this stage output distortion. When cursor is pointed to grid lines you can see what drive signal Vpp amplitude is needed for that output. Stage gain comes when output sweep is calculated using input sweep.

When green sweep is not linear the stage produce distortion which is mostly 2nd harmonic. Now you can change/tune parameters and look what loadline comes before changing resistors or tube type in your amp. Traditional method was to calculate values and draw loadline to datasheet which is slower. You can save your work, copy and post them.

Tubes have tolerances and they wear out so in actual amp voltages/currents usually come more or less different whan this tool give but is is good!!!

Universal Loadline Calculator

This calculator allows you to trace the anode characteristics graph, and to compute and tune loadline and operating point of various vacuum tubes.

www.vtadiy.com

I’ll just get down on my knees and kiss the ground you guys stand on in this forum. You all are just amazing!
The calculator will become much appreciated. Hopefully I’ll get time this weekend to produce a layout for the 5w pentode amp and start tuning and the building. One other question. Any way to make a temporary build before permanently solder it all in place? Is it the breadboard way that will be the best?

 

[Jerry garrcia]

Or should I build a GA-5 with a bias switch in one input and for the other input use a 5C2 kind of build with a 6SC7?