• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Posted new P-P power amp design

and check what power and setup I can get from my 6JN6 stack.

The 21JV6 tube is pin compatible with the 6JN6, a different heater V obviously (foil cuts for htr). It's rated 18 Watts Pdiss, but has a noticeably bigger plate than the 6JN6. Same plate size as the 24 Watt 6HJ5 used in the monoblock version. It has higher gm than the 6JN6 too, so should be easy to drive to the higher power using boosted B+, possibly without changing the input tubes.
 

Attachments

  • 21JV6_6JN6.jpg
    21JV6_6JN6.jpg
    29.6 KB · Views: 713
Last edited:
Member
Joined 2016
Paid Member
>I think someone would have said something by now if this was a problem.

Hello Jeff,

I came as well to the conclusion that the amplifier just works fine :) Some more measures and impressions...

This is the engineer amp and I'm an engineer (or so I was told in 1988 before leaving electronics for IT ;)), so I measured a few more values, and tried to calculate if I'm still in the max dissipation of the tubes. I did some more tests with 1KhZ sinus out of my Tensa generator.

With an input voltage of 4.5V peak-peak, I get 36V peak-peak at the output, which if I'm correct translates to about 13V RMS, into my 8 ohms speakers I'm about at 20W when it starts clipping, as expected. At that power-level, the cathode current is at about 60 mA, B+ down to 380 (from 390 when idle). In a dark room, no sign of red plates.

I had two swap two input tubes to be able to balance the drivers. But I have no reliable way to measure distortion, my Siglent Scope has some FFT but it's not really usable (or I'm too dumb). I shall build something based on an audio card as some other have done. Until then, I used the voltage check on xDP1/2.

I think the tubes are still breaking in (maybe 5 hours playing now), the bias current is slowly growing, right now after one hour playing, it went from 40 to 41.6. As other have written it shall settle down at some point, or I hope so.

I dropped the cheap Denon plastic speakers and connected my Magnat Quantum 508 towers. They are specified at 93db/W/m. At full volume, my two kids came down from their room to ask for some peace... and the only other way to get them downstairs is food :). Amazingly enough, the sound quality (source is a Squeezebox 3 playing from a NAS) is excellent even at full-volume... but you suddenly notice you can't talk to each other anymore. Amazing experience.

Next is finishing the box. Waiting on the laser cut of the panels.

Charles
 
Last edited:
After a few days, today I went back in and redid all the biasing and balancing, and took some more careful measurements. B+ is 364V ( a bit higher because my power transformers is rated at 300-0-300, so as expected). C+ is -62V, and the Screen voltage is spot on at 150V.

I tried to measure the actual power and by injecting a 1 Khz sinusoidal wave, I drove the amp to clipping (speakers were connected and are nominally 8 Ohms). This occurred when the output at the speaker reached 14.8 RMS, thus giving me approx 27W RMS. This is a bit higher than spec. because I have a bit higher B+ and my output transformers are a bit lower in Impedance than the one suggested in the standard BOM. Mine are 6.6K vs. the 8k that Pete specified.

With no imput the amp is QUIET. No hum, no nothing ( my speaker are efficient but not terribly so). All in all, I am delighted.

Do I dare to tackle the 50 W monoblock? I'd need to make two... :)

Hello Everyone: Now that the amp is running and playing great, the itch to upgrade is now in full force. Currently I have an Edcor XPWR205 (300-0-300, with a 50V tap). I am getting a B+ of 364 V, and the output is about 27W at clipping. My opt are already upgraded (50W CXPP50 at 6.6K), so the only thing should be to increase the B+ another 30V or so. Antek seems the preferred choice for this, so I am thinking of getting an AS-3T325 (AS-3T325 - 300VA 325V Transformer - AnTek Products Corp) plus a 50V or so CT transformer for the Bias supply (Hammond 187D56 187D56 Hammond Manufacturing | Mouser).

What do you think?

Marco
 
Almost done

I know this is an old thread but I'm just completing my DCPP and wanted to share.
I'm most proud of the case which is totally custom made by me. I made the case front sides and back all from oak board. The top chassis plate is a piece of black acrylic with a copper sheet fitted on the back under the Edcore transformers that I painted black. I used a router table to "mill" the acrylic down the thickness of the copper sheet for a flush fit at the joint. I used copper foil to fully shield the underneath side of the acrylic.

I've got everything completed and ready to bring it up with the variac and make the necessary adjustments.

I'm leaving this one in it's "stock" configuration using the 6JN6's just to see how well I like the sound. Then I may decide to build the mono blocks Pete has now.

I'm currently listening to Pete's "Mighty Midget" and absolutely love it. It inspired me to continue to the DCPP and possibly beyond.
 

Attachments

  • IMG_20180814_195540.jpg
    IMG_20180814_195540.jpg
    1,004.8 KB · Views: 513
  • IMG_20180814_195658 (2).jpg
    IMG_20180814_195658 (2).jpg
    1 MB · Views: 497
Member
Joined 2016
Paid Member
Filament voltage at 6.6V

Dear colleagues,

I'm finishing the enclosure (oak+aluminium) of my Engineer's DCPP. I checked all voltages again before disconnecting things, and I found out my Edcor XPWR139-220 provides 6.6V to my filaments. My line voltage is 230V, so I guess it's normal to have 6.6V on the filaments. If I reduce the line voltage to 220V with a Variac, I have 6.4V on the filaments (measured at PC-board inputs).

The data sheets mention 6.3 (+- 0.6V) so I'm within the tolerances. Should I still try to reducing this voltage somewhat, in the hope of keeping my tubes alive for a bit longer ?

As for 6A and 0.3V to drop, I calculated a need of 0.05 ohms in serie with the filament circuit. I have 0.22 ohms 3W in stock, I could parallelize 4 of them on a small board.

Hints welcome.
Charles
 
Transformer voltages are predicated on 2 things:
1. A specific Power Line Voltage on the primary winding(s).
2. Maximum rated loads on All secondaries.
The secondary output voltages are Not rated with open secondaries, and are Not rated with different primary voltages.

A. If you do not fully load the secondaries, the primary has to deliver Less power, so the DCR of the Primary drops less voltage, because there is less primary current.

B. If you do not load one of the secondaries, there is less current in that secondary, so the DCR of that secondary drops less voltage.

c. The winding ratios are fixed. But the rest of the factors all contribute to the secondary voltages in your amp.

Simple?
 
Member
Joined 2016
Paid Member
Kodabmx, 6A3sUMMER, Tony, thanks for your explanations.
I now have 50 milli-ohms in serie, 4x3W metal film that I had in a drawer. Filament voltage is now 6.3V, measured on the tube sockets. Resistors show 53°C (127°F) on their surface, well within their specs.

About being a newbie... yes, even at 53, with a 30-years pause out of electronics, I feel like a newbie and I have a lot of respect for all the know-how I read from you guys in these forums. I wish I could retain 10% of it !

Thx again, wish you a nice Sunday evening :)

I finished the amp boxing today. Oak frame, laser-cut 3mm aluminium on top, 2mm on bottom, painted in a shop. They painted the Edcor endbells as well, because I don't really like the Edcor gentiane blue, a matter of taste. Pictures in the thread.

No volume control, as my source has it already. As many have said previously, the sound is excellent, and I look forward to move it to my living room where it fill feed two Magnat Quantum 508 (93 SPL).

I have a 2nd PCB from Pete, I ordered 3 Toroidy transformers to build the same amp again (I have two sons...) and have all the other components, so I will restart a 2nd build. This time I will try to make it an integrated amp, with 2-3 inputs, a volume control, and a remote. Not decided yet on how I will build this, I have multiple options at hand.

Thanke Pete and the community for this excellent project !
Charles
 
Member
Joined 2016
Paid Member
DC filament supply - referencing against gnd/B+

Dear all,

I'm experimenting a DC supply for my filaments. Reason are, the power tranny is a bit overloaded, mechanically vibrating, heating, etc. Had to drop voltage using resistors in serie, they heat the inside of my case, etc. Plus, I have this remaining 50 Hz hum in speakers. Well, not completely happy yet.

So I came up with this DC supply idea from the Baby Huey EL34 project, using a 12V toroid, full-wave bridge rectifier and some capacitors, then a regulated buck-converter to provide the 6.3V/6A for my 8 filaments. Schema attached below.

My question relates to the referencing of this floating DC voltage. The original CT 6.3V is connected to W14, R44/R43. I assume I have to do something similar ? Simply connect the "0V" of my 6.3 supply to W14 ?
 

Attachments

  • dc_filament_supply.jpg
    dc_filament_supply.jpg
    43 KB · Views: 518
If your DC supply is completely floating as your schematic shows, then simply connect it's negative lead to the junction of R44/R43.

I have hacked one of my boards so the the input tubes still run on the 6.3V AC transformer winding but the output tubes run from a Meanwell SMPS. The only common connection is the resistor network. This allows both heater networks to be at about 40 volts DC. I have three different Meanwell LRS-150 supplies to suit different oddball voltage sweep tubes

Warning notice for those experimenting with odd voltage TV sweep tubes:

For anyone thinking about doing this with oddball TV sweep tubes other than the recommended types, BEWARE some sweep tubes like the 35LR6 have an undocumented connection internally connecting the heater to pin 7 which is also used for the plate pin on some tubes. Make SURE that there is no internal connection to pin 7. DISCONNECT the jumper on the board that connects the plate to pin 7 if your tubes have top caps for the plate. If both the internal heater connection exists and the plate to pin 7 jumper is in place, the plate supply will be connected to the heater supply.....trust me this WILL blow stuff up, especially when you have a 650 volt 1.7 amp plate supply!
 
Member
Joined 2016
Paid Member
Georges thanks for your feedback, always learning new things! Instead of the whole setup I have thought about using a SMPS from 230VAC main to 6.3V DC, but I can't find a model that can be adjusted to 6.3V. Annoying because they are very compact and would probably fit into my existing case.

I'm experimenting a DC supply for my filaments. Reason are, the power tranny is a bit overloaded, mechanically vibrating, heating, etc. Had to drop voltage using resistors in serie, they heat the inside of my case, etc. Plus, I have this remaining 50 Hz hum in speakers.

So I have been trying to feed my filaments with DC from a distinct power supply. Neither the mechanical hum nor the 50 Hz in speakers are gone. Back to AC on filaments and search for other reasons for my 2 issues.

inputs are shorted, this signal is measured after the C17/C18/C19/C20 coupling capacitors. Period is 100 Hz, peak-peak 300-400 mV, this is most likely what I hear in my speakers after being amplified by the output stage. Now I'm not sure why I have this signal here.

It took me a while to understand where my problem was. Post #1092 by George put me on track: "Make sure that you have at least 20 volts across the regulator FET. I had only 5V... after understanding how the B+ ripple filter would work if it were functional, I found out R69 was in open circuit, so the reference voltage was at the same level than the drain of the FET... so the ripple went across the filter unchanged. After replacing R69, B+ is 376V, drain of Q2 is 398V, so 22V over Q2. Looks much better. Remaining ripple on B+ is 100 Hz, 200mV peak-peak and 150mV at ~ 5MHz... not sure what this high-frequency voltage does here, but well, that will be the subject of another debugging session.

The good news is, no more hum in speakers. The mechanical hum of the power transformer seems a bit lower, but I tightened the screws again so that might be it, I will see/hear when the amp is back on his feet in my living room. Maybe some mechanical isolation with gommets would help as other have found.

In the risk of repeating myself: this amp design is excellent, it sounds perfect to my ears and I have already read the complete thread at least 3 times, having better understanding and more fun at each lecture :)
 

Attachments

  • output_of_C18.jpg
    output_of_C18.jpg
    43.2 KB · Views: 328