I can partially answer some of the questions.
The SPP uses 4x EL84's, 2 per channel in push-pull, and 2x 12AT7's, 1 per channel as gain amplifier/phase inverter.
For PT rating -- each EL84 will bias at around 40mA with the default cathode resistor (270ohms). 40mA*4 + 10mA*2 = 180mA.
So yes, 175-180mA should be the minimum... there is not too much against over-sizing this except cost and size potentially.
I know that the Hammond series EI 27X/37X work fine up until the entire rating, not sure how about toroidals.
I've used the 270HX a few times which is rated at 230mA.
For output modes, you can read any generic text about UL/Pentode/Triode, it is not specific for the SPP.
Generally, triode = less power less distortion, pentode = more power more distortion
UL = somewhere in the middle. Almost as much power as pentode with less distortion.
UL requires the screen grid taps on the output transformer.
Not sure I understand the B-side question. Which connections/materials are you referring to?
The SPP uses 4x EL84's, 2 per channel in push-pull, and 2x 12AT7's, 1 per channel as gain amplifier/phase inverter.
For PT rating -- each EL84 will bias at around 40mA with the default cathode resistor (270ohms). 40mA*4 + 10mA*2 = 180mA.
So yes, 175-180mA should be the minimum... there is not too much against over-sizing this except cost and size potentially.
I know that the Hammond series EI 27X/37X work fine up until the entire rating, not sure how about toroidals.
I've used the 270HX a few times which is rated at 230mA.
For output modes, you can read any generic text about UL/Pentode/Triode, it is not specific for the SPP.
Generally, triode = less power less distortion, pentode = more power more distortion
UL = somewhere in the middle. Almost as much power as pentode with less distortion.
UL requires the screen grid taps on the output transformer.
Not sure I understand the B-side question. Which connections/materials are you referring to?
Jerin,
Re: B-side. Are you thinking of mounting not mounting the tube sockets on the board and running wires from the board to the sockets? Can't say I'd recommend it if that's what you mean. If that's what you want to do I'd build it with tag boards a la some guitar amps.
As far a power trans goes I'd up-rate to one with more than 200mA available, say a Hammond 272HX or JX (374 HX or JX if you need 240 volt primaries). Note: If you are not using a tube rectifier a 275-0-275 volt secondary would be fine otherwise a 300-0-300 for a tube rectifier. Make sure a toroid would a 5 volt winding if you are using tube rectification. Not really necessary to go with a custom trans, lots of choice with off the shelf iron.
There's an SPP thread on the main tubes/valves forum where there have been ongoing transformer discussions.
https://www.diyaudio.com/community/threads/easy-diy-tube-amp.384618/page-3#post-6987763
Re: B-side. Are you thinking of mounting not mounting the tube sockets on the board and running wires from the board to the sockets? Can't say I'd recommend it if that's what you mean. If that's what you want to do I'd build it with tag boards a la some guitar amps.
As far a power trans goes I'd up-rate to one with more than 200mA available, say a Hammond 272HX or JX (374 HX or JX if you need 240 volt primaries). Note: If you are not using a tube rectifier a 275-0-275 volt secondary would be fine otherwise a 300-0-300 for a tube rectifier. Make sure a toroid would a 5 volt winding if you are using tube rectification. Not really necessary to go with a custom trans, lots of choice with off the shelf iron.
There's an SPP thread on the main tubes/valves forum where there have been ongoing transformer discussions.
https://www.diyaudio.com/community/threads/easy-diy-tube-amp.384618/page-3#post-6987763
Hi Everybody, novice builder here.
I built the Tubelab SPP and love it, but I need help getting my voltages down. The DC voltages seem especially high.
I followed Steve's build guide on Wall of Sound.
I'm using the Heavy option, Hammond 272JX transformer: 125v black tap.
Using Tung-Sol tubes
Measuring between connection points:
8 and 9= 6.5 AC
21 and 22=5.1 AC
12 and 18= 350 DC
Cathode voltages are all ~14.0 (are supposed to be ~11 or 12) DC
I’m using 360 ohm resistors
The coupling caps .1uF /1,000v
I used it for a couple months, despite the tubes running hotter than they should, and one of my tubes did red-plate the other week. So I swapped out the 330 ohm resistors for 360s but it hasn’t made any real difference.
I'm compelled to just keep increasing the resistor value, but wanted to check with you all first. Is it possible there is a problem elsewhere in the power section (choke? rectifier?). Anyway, curious for opinions and open for suggestions.
I built the Tubelab SPP and love it, but I need help getting my voltages down. The DC voltages seem especially high.
I followed Steve's build guide on Wall of Sound.
I'm using the Heavy option, Hammond 272JX transformer: 125v black tap.
Using Tung-Sol tubes
Measuring between connection points:
8 and 9= 6.5 AC
21 and 22=5.1 AC
12 and 18= 350 DC
Cathode voltages are all ~14.0 (are supposed to be ~11 or 12) DC
I’m using 360 ohm resistors
The coupling caps .1uF /1,000v
I used it for a couple months, despite the tubes running hotter than they should, and one of my tubes did red-plate the other week. So I swapped out the 330 ohm resistors for 360s but it hasn’t made any real difference.
I'm compelled to just keep increasing the resistor value, but wanted to check with you all first. Is it possible there is a problem elsewhere in the power section (choke? rectifier?). Anyway, curious for opinions and open for suggestions.
I'm not sure what 8,9 / 21,22 / 12,18 points are but I'm assuming the first is the heater and the last is your B+ ?
The cathode voltage seems fine, 14V/360ohm = 39mA which is pretty good.
Likewise 6.5 VAC should be fine for heaters.
The 350V B+ is high, it should be around 310-320V max.
A few options off the top of my head:
1) Put a CL-90 thermistor in series with the primary AC
2) Are you using solid-state or tube rectifier? If solid-state then you can try a 5AR4 or 5U4GB to drop some volts.
3) Increase resistance in your power supply. You didnt specify which choke or resistor you are using. You dont necessarily need to replace, you can just add a resistor inline with an existing choke or add another RC stage.
4) Replace the passive filtering with a HV regulator like tomchr's Neurochrome Maida Regulator.
5) If you want to slightly drop the heater voltage you can put a small resistor inline
6) Use 6P14P-EV tubes. They can take 350V on the plate
The cathode voltage seems fine, 14V/360ohm = 39mA which is pretty good.
Likewise 6.5 VAC should be fine for heaters.
The 350V B+ is high, it should be around 310-320V max.
A few options off the top of my head:
1) Put a CL-90 thermistor in series with the primary AC
2) Are you using solid-state or tube rectifier? If solid-state then you can try a 5AR4 or 5U4GB to drop some volts.
3) Increase resistance in your power supply. You didnt specify which choke or resistor you are using. You dont necessarily need to replace, you can just add a resistor inline with an existing choke or add another RC stage.
4) Replace the passive filtering with a HV regulator like tomchr's Neurochrome Maida Regulator.
5) If you want to slightly drop the heater voltage you can put a small resistor inline
6) Use 6P14P-EV tubes. They can take 350V on the plate
So the tubes are still red-plating with 360R cathode resistors? I would try 390R. That would drop the idle dissipation down from 13.1 watts to 12. You're running current production tubes? New production tubes are probably not going to last too long at 13 or 14 watts idle dissipation in class A, except maybe JJ's which have a reputation for being tough.
Or just use added resistance in the PS to drop B+ as @itsikhefez suggests.
jeff
Last edited:
Stephane:
On one of my builds using reclaimed iron I was running a high B+. I added another CR ahead of the first C on the board. 5.1 volts on the rectifier isn't too bad. As itsek says a CL-90 is a good idea, it will drop a couple of volts on the primary. Also, as he says a couple of small value resistors on the 6.3 volt line will lose some voltage. Recently I did that on a Fisher rebuild. I used some 0.025 ohm, 3 watt resistors, one on each 6.3 volt lead. If you get your B+ down to the 300 to 310 volt range your current through the EL84s should fall into line.
Steve
On one of my builds using reclaimed iron I was running a high B+. I added another CR ahead of the first C on the board. 5.1 volts on the rectifier isn't too bad. As itsek says a CL-90 is a good idea, it will drop a couple of volts on the primary. Also, as he says a couple of small value resistors on the 6.3 volt line will lose some voltage. Recently I did that on a Fisher rebuild. I used some 0.025 ohm, 3 watt resistors, one on each 6.3 volt lead. If you get your B+ down to the 300 to 310 volt range your current through the EL84s should fall into line.
Steve
Thanks a lot for your advice a few weeks ago. I haven’t had much of a chance until now to dig in. The resistors on the 6.3 leads were easy to do and made a difference. I’m getting 6.18 volts as opposed to the 6.5: Success. Thanks to you and @itsikhefez both.
As for the high B+:
The goal, getting 350v down to 300-320v
It sounds like inserting a resistor (or another rectifier?) on the line from choke leading to C1 is the best option for dialing that down.
I’m not sure how to calculate the value of the resistor needed.
Here's what's in place:
Hammond PwT. 272JX (600VCT, 288 mA)
Im using a hammond 156R choke (1.5H, 200 mA, 56ohm max resistance)
Sovtek 5AR4
I know there are formulas for Ohm’s law but I’m not sure which values go into which places, or if I’d be using it correctly (I’m good with my hands and with details, but I have a hard time with math, and a novice builder, overall).
The other thing, mentioned: adding another current rectifier? I know one can get solid state rectifiers but I'm not sure how I'd implement that. Is there another component that can rectify? A diode?
If your B+ is already high, switching to a solid state (diode) rectifier would make things worse. It has been suggested to swap the 5AR4 with a 5U4GB. I think that would be the simplest way to drop your voltage. Using a 5U4GB should get you in the low 300v.
By the way, I am using a hammond 272HX (600V C.T. @ 230ma) with a 5AR4 rectifier, and my B+ is 320V. Surprised to see 350V with the 272HX.
By the way, I am using a hammond 272HX (600V C.T. @ 230ma) with a 5AR4 rectifier, and my B+ is 320V. Surprised to see 350V with the 272HX.
Last edited:
Oh, thanks. That does sound like a easy thing to try. And it looks like the 5U4GB is an octal tube (the only other tube amp I've had used a 6CA4). I also have some 6P14P-K tubes on the way whose specs appear to be able to handle the 350 B+.
I wanted to check in and say thanks a lot for everyone's advice. The 5U4GB rectifier did the trick: B+ is down to 312, with cathodes measuring between 12.2 and 12.6. It sounds a little dryer than with the 5AR4, but, in my setup right now, with the 5U4,the instrument separation is more distinct and the central image is more cohesive. Very happy!
Hi @rongon, have you finished your SPP with 6P43P-E yet? I’m curious how it turned out. Please let us know.
What's happened is that I've come to see in simulation that there's no way to get a clean enough driver using a twin-triode with only about 275V B+.
6DJ8 is about the best one, but there's not enough gain for applying negative feedback.
12AT7 needs more voltage. With 275V B+, I can't get the DC-coupled cathodyne working without the 1st stage operating with the cathode voltage well below 1V. A 12AT7 will very likely draw too much grid current biased at only -0.7V grid-cathode.
I think that's why koda uses a 6F12P with a higher voltage B+ for the driver stage.
I guess I can re-jigger the SPP PCB so that I can use a second (small) power transformer for a driver stage B+ of 400V or so. That would clean things up.
I've since decided that the original 12AT7-EL84 tube complement makes more sense for an SPP.
Now to do something about it...
6DJ8 is about the best one, but there's not enough gain for applying negative feedback.
12AT7 needs more voltage. With 275V B+, I can't get the DC-coupled cathodyne working without the 1st stage operating with the cathode voltage well below 1V. A 12AT7 will very likely draw too much grid current biased at only -0.7V grid-cathode.
I think that's why koda uses a 6F12P with a higher voltage B+ for the driver stage.
I guess I can re-jigger the SPP PCB so that I can use a second (small) power transformer for a driver stage B+ of 400V or so. That would clean things up.
I've since decided that the original 12AT7-EL84 tube complement makes more sense for an SPP.
Now to do something about it...
Hi @rongon,
Thanks for the assessment. Sounds like your simulations basically showed that the “as published” Tubelab SPP EL86 design can’t work in a UL configuration? I assume not many people tried to build EL86 amplifiers using SPP PCB. I remember @OldHector had similar problems with either red-plating PL84s at higher B+, or insufficient drive from the 12AT7 at lower B+. IIRC @OldHector ended up with the approach you suggested; He used one power transformer, feeding both a psuedo choke input supply used for the PL84 supply, and a CRC supply for the 12AT7s.
George wrote somewhere in @OldHector’s thread: “I have used 6CW5's in the SPP with 150 volts on G2 and 340 volts on the plate into a 3300 ohm OPT to get about 30 watts per channel.” Did your simulations confirm a B+ of 340Vdc will provide sufficient voltage to the 12AT7 front end to drive the EL86 to full power? (Would you be willing to share your simulation (.ASC) file. I’m attempting to learn to run LTSpice.)
I also two Tubelab SPP PCB awaiting building into an EL86/6P43P and a 6P15P amplifier respectively. The plan to do it fixed bias, with bias and screen Voltage controlled by Dave Gilespie’s EFB circuit. I have made some progress on the SPP build intended for 6P15P output tubes using the same approach. I made minor mods to the board (basically just lifted the ends of the grid resistors from ground) to connect the -bias to g1 via the 470 KOhm grid resistors and left the cathode resistors and by-pass caps out. For these builds I bought two Antek AS-2T230 power transformers, but I found that I can’t get much over 300 Vdc under load with the PTs, even using SS rectifiers.
Thanks for the assessment. Sounds like your simulations basically showed that the “as published” Tubelab SPP EL86 design can’t work in a UL configuration? I assume not many people tried to build EL86 amplifiers using SPP PCB. I remember @OldHector had similar problems with either red-plating PL84s at higher B+, or insufficient drive from the 12AT7 at lower B+. IIRC @OldHector ended up with the approach you suggested; He used one power transformer, feeding both a psuedo choke input supply used for the PL84 supply, and a CRC supply for the 12AT7s.
George wrote somewhere in @OldHector’s thread: “I have used 6CW5's in the SPP with 150 volts on G2 and 340 volts on the plate into a 3300 ohm OPT to get about 30 watts per channel.” Did your simulations confirm a B+ of 340Vdc will provide sufficient voltage to the 12AT7 front end to drive the EL86 to full power? (Would you be willing to share your simulation (.ASC) file. I’m attempting to learn to run LTSpice.)
I also two Tubelab SPP PCB awaiting building into an EL86/6P43P and a 6P15P amplifier respectively. The plan to do it fixed bias, with bias and screen Voltage controlled by Dave Gilespie’s EFB circuit. I have made some progress on the SPP build intended for 6P15P output tubes using the same approach. I made minor mods to the board (basically just lifted the ends of the grid resistors from ground) to connect the -bias to g1 via the 470 KOhm grid resistors and left the cathode resistors and by-pass caps out. For these builds I bought two Antek AS-2T230 power transformers, but I found that I can’t get much over 300 Vdc under load with the PTs, even using SS rectifiers.
I think the problem is the EL86/PL84's limitation of max screen grid voltage = 250V. If you try to make a UL PP EL86 SPP using a single 270V B+, that means the 12AT7 has to run with a low plate voltage and therefore small grid bias, which means now you have limited voltage swing from that 12AT7, and it is likely to be running with a fair amount of grid current.Francois G said:
I tried a 6DJ8 instead but that results in not enough gain to employ global NFB around the whole amp. It also shows THD will be pretty high open loop, I guess because the first 6DJ8 voltage amp stage will be working pretty hard.
If the B+ to the 12AT7 could be up around 350V and the B+ to the EL86 output stage 270V, then that would probably work great. But that would take away quite a lot of the "Simple" from the SPP.
I'm going to say that the SPP looks best suited to EL84 outputs, and go with that. A B+ of 325V should result in the EL84s seeing a plate-to-cathode voltage of about 310V, which looks about right. Some people have posted that the SPP is one of the best sounding PP EL84 designs they've built, so I'm not inclined to tinker too much. I think George's design is solid for PP EL84, so I'll go with that.
I'll try to post the .asc after I get home from work, or this weekend.
I think 340V B+ would be fine for the 12AT7. With that B+ it should be possible to get the input stage to bias with about 1.5V at the cathode. The problem would be how to rig the SPP PCB up for 150V on the EL86 screens. I'll have to look into that. I was also hoping to make a push-pull triode amp out of this.
Attached is the .asc for the SPP using EL86-Triode and 12AT7 with 270V B+. You'll need to put in your own 12AT7 and EL86 (or 6P43P-E) models.