output transformer for 4X 6V6 ?

[Hepcat]

Hello, this is my first post/question and a search didn't find this covered recently. I am building a 50 watt class amp and really like the sound of 6V6 output tubes. I checked the tube spec sheets from JJ and found that the plate impedance for 2 6V6s is 10K. 2 6L6s is 5k so I assumed (you know what that means) that a transformer pair for a 6L6 amp would work well. When finished, I could only get 25 watts out of a the four 6V6s. Above that level, it would oscillate or basically freak out. I was testing into an 8 ohm load resistor and used a scope on X10 to avoid any loading errors. Listening with a speaker and a guitar for source showed it to be clear up to a certain power level and then a very awful sounding distortion riding on the signal. After many hours of troubleshooting, I pulled the 6V6s and put in two 6L6s. The amp worked perfectly and gave me 56 watts (60v peak to peak at 8 ohms) at the onset of clipping! It finally occurred to me that I'm loading down the 6V6s too heavily. I switched to using the 4 ohm tap and the problems completely went away but I only get about 36 watts. The tubes are new JJs, matched and running at 450 volts with about 20 Ma of bias. I think that I have established in my mind that I have the wrong output transformer but am not sure what i should use. I have a Heybour that I got from Weber and it has a primary impedance of 4K. Any suggestions what transformer impedance will give me full power output a la Jim Kelley? Thanks in advance!

[brewdude]

I have not yet built a 4 x 6V6 amp.

However, I have been planning a 4 x 6V6 Dumble style build and was considering this OT.
http://www.triodeelectronics.com/surebareupho.html

I am not really all that well informed on the technical peculiarities of this configuration. I figured it would be somewhere around 40 watts. I don't have the ability to measure it.

I'm interested in hearing more details about your build.

[tictac]

I wouldn't split hair worrying over much about OT impedance. A Deluxe Reverb uses a 6K6 OT for a pair of 6V6's and a Princton uses an 8k OT both way off your 10K number; any OT made to operate a pair of 6L6's would be fine... somewhere around 4k.

Remember Jim Kelly amps are running those 6V's at 485 to 490V on the plates to achieve the claimed 60W output. Do you really want to do that?

With 420V on the plates your 20mA is just right (70% plate dissapation) you could bump it up to 23mA each but I wouldn't go beyond that....

TT

[Hepcat]

Been wondering about those Magnetic Component transformers. Wondering if anyone's had any experience with them. Good prices. TT, I have a good sound with the amp as it is using the 4 ohm tap to give me a clean 36 watts. I don't want to build something that would possible blow up during a gig as I use the amps that I have built 2-3 times a week at gigs or jam sessions. I have a 25 watt model with two 6V6s that works just fine using a Deluxe Reverb transformer set. I used a Weber Signature 12B speaker and tuned the tone stack and reset the gain structure for my taste. I'll post a picture of that sometime. Thanks for the advice on the Jim Kelley stretching the laws of physics!
Ken

[tubeswell]

If you are running the 6V6s at around 460 volts, use JJ 6V6S (because they will easily handle it), and you want a reflected load of around 3k2 - 4k for 4 of them in PP fixed bias with the screen at or near the plate voltage (which really implies a CLC filter between the plate and screen supply nodes and 470R screen grid resistors). Make the OT a 50Wer and it should be sweet.

[rdjones]

What transformer are you using ?

6V6 data sheets and established schematics indicate a range of 6.6k to 11k for a pair, so anywhere from 3.3k to 5.5k should work for a quad.
The higher numbers more commonly used with cathode bias circuits.
This pretty much overlaps the typical range of 3.8k to 6.6k spec for a pair of 6L6.

I would have expected your 4x6V6 arrangement to at least work in a usable manner with maybe just a bit lower power than optimum.
Do the 6L6s work the same in all sockets ?
Bad 6V6, socket wiring issue, feedback loop problem ?
Your description of the problem seems like more than just a relatively small mismatch.

reddog

[ToneMerc]

In my last two builds have been quad 6V6's, on both I ran 3.4K 50W Marshall iron. This one runs around 405V on plates and the one I just finished up runs about 420V with the Tung Sol.

https://tubeamparchive.com/viewtopic.ph ... t=quad+6v6

https://tubeamparchive.com/viewtopic.php?t=16468



TM

[Firestorm]

Hello, this is my first post/question and a search didn't find this covered recently. I am building a 50 watt class amp and really like the sound of 6V6 output tubes. I checked the tube spec sheets from JJ and found that the plate impedance for 2 6V6s is 10K. 2 6L6s is 5k so I assumed (you know what that means) that a transformer pair for a 6L6 amp would work well. When finished, I could only get 25 watts out of a the four 6V6s. Above that level, it would oscillate or basically freak out. I was testing into an 8 ohm load resistor and used a scope on X10 to avoid any loading errors. Listening with a speaker and a guitar for source showed it to be clear up to a certain power level and then a very awful sounding distortion riding on the signal. After many hours of troubleshooting, I pulled the 6V6s and put in two 6L6s. The amp worked perfectly and gave me 56 watts (60v peak to peak at 8 ohms) at the onset of clipping! It finally occurred to me that I'm loading down the 6V6s too heavily. I switched to using the 4 ohm tap and the problems completely went away but I only get about 36 watts. The tubes are new JJs, matched and running at 450 volts with about 20 Ma of bias. I think that I have established in my mind that I have the wrong output transformer but am not sure what i should use. I have a Heybour that I got from Weber and it has a primary impedance of 4K. Any suggestions what transformer impedance will give me full power output a la Jim Kelley? Thanks in advance!

What values are you using for coupling caps, grid leaks and grid stoppers in your output stage? If I had to guess, I'd say your drive signal is too hot for the 6V6s at the 70% bias condition and the inputs are clamping as they start to draw grid current. You could try biasing closer to 50-55% to delay the transition to "Class B" and also experiment with the traditional fixes of reducing the size of the coupling caps and grid leaks.

When you say you switched to the 4-ohm tap and it cleaned up the problem, were you still running into the 8-ohm load? That would increase the reflected impedance.

[tubeswell]

In my last two builds have been quad 6V6's, on both I ran 3.4K 50W Marshall iron. This one runs around 405V on plates and the one I just finished up runs about 420V with the Tung Sol.

https://tubeamparchive.com/viewtopic.ph ... t=quad+6v6

https://tubeamparchive.com/viewtopic.php?t=16468



TM

Very tidy on top! Gutshot?

[Hepcat]

Yes, I did use the 4 ohm tap with an 8 ohm speaker. It increased the reflected impedance to the tubes. I suspect the PI overloading the 6V6s also.I broke the input stage to the PI and put in a voltage divider to lower the input level. It cleared up the breakup but I still only got 30 volts Peak to Peak. If I remember right, I have a high level input signal to the 6V6s. what might be a typical level at the grid to the output tubes.The coupling caps are .1 and the grid stoppers on the PI are the customary 1.5K resistors. looking at the PI output I can easily go to a square wave with a -10 dBm signal (can you tell I worked for 25 years as a radio engineer? oops!) The -10 sine wave is of similar level to my telecaster as read on the scope. The Amp began life as a Fender Blues Deluxe. I tuned to tone stack and some of the gain stages and left out the reverb and the extra preamp stage. Then i used a Weber speaker and built a 1 X 12 cabinet out of solid pine, 3/4 thick with a 1/2" speakerboard.

[rdjones]

If I remember right, I have a high level input signal to the 6V6s. what might be a typical level at the grid to the output tubes.

Use your bias voltage as a guide.
If bias is -20V you should be able to clip the output at about 40VPP.

(can you tell I worked for 25 years as a radio engineer? oops!)

You'll find others in here with similar backgrounds.
I grew up around ham radio, and was FM broadcast (First Class) engineer at WHOT-WSRD in Youngstown.

regards,
reddog Steve

[ToneMerc]

In my last two builds have been quad 6V6's, on both I ran 3.4K 50W Marshall iron. This one runs around 405V on plates and the one I just finished up runs about 420V with the Tung Sol.

https://tubeamparchive.com/viewtopic.ph ... t=quad+6v6

https://tubeamparchive.com/viewtopic.php?t=16468



TM

Very tidy on top! Gutshot?

First link page 3

TM

[Badside]

Isn't 35W for a quad of 6V6s about par?

[Hepcat]

Yes, 35 Watts is reasonable for a quad of 6V6s, I'm still having a bunch of trouble getting that. I seemed to have the amp stable and working using the 4 ohm tap with an 8 ohm speaker. Last night i was testing a guitar and the ting freaked out again. While I don't have another quad of matched fresh 6V6s laying around, I have about 10 used old tubes that will work as a test batch. I am stating to think that i have a possible tube problem. When my FM transmitters had unknown strangeness with all of the parameters seeming to be good, I'd put in a known good spare final tube. Often the problems went away and the original tube was shipped off to be rebuilt. I'll try the same in a few minutes and report what the result was. I have a 25 watts unit with almost the same circuit that works just great so I think I can make this work. Thanks for all of the suggestions so far.
ken

[Hepcat]

I put 4 old (and I mean old) ratty 6V6s that I had laying around, completely unmatched and different brands. Tung Sol, two RCA black glass, etc. I then made sure that the bias gave me all tubes in the safe territory between 10 and 20 MA. Then I measured 50 Volts Peak to peak at 8 Ohms after putting the output transformer back on the correct voice coil tap. That calculates out to 39 watts according to my calculations. Perfectly acceptable in my book! I will get a new matched quad of JJ 6V6s and stop complaining. It sounds just marvelous also! Thanks again. It appears to have been mostly damaged tubes.
Ken