I’m trying to build a Mullard designed (“Mullard Tube Circuits for Audio Amplifiers” pg. 113) 7W per channel stereo integrated amp, and I have run into a problem finding the right output transformers.
I did not notice when I first started buying parts for the amp that the output transformers called for are “20%” screened taped. When I noticed the taps I had already bought all the parts for the amp (even the output transformers), so I am trapped into continuing the build, or just wasting the money I spent on all the parts.
My question is what would happen if instead of trying to find a output that has 20% taps, if I bought one of the more common transformers that has 40% taps? I read the section in the Mullard book that takes about “distributed loading” (also called ultra-linear amplification), and I think I understand the difference between 20%, and 43%. The 20% tap offers more power output, while the 43% offers less distortion. Almost all the transformers that I see on the net are actually 40% (not the 43% given in the design). The taps on the primary in the design connects through 1K resistors to grids in the pentode section of the push-pull ECL82 output tubes. I don’t really know what would happen if I used a 40% tapped output. Would I go from less power output to less distortion? If that were the case I would just forge ahead and get a 40% screened taped output.
I found two suppliers in G.B. that have a output transformer that might work, the problem is that one does not sell to the US (why?), and the other will build a custom transformer to my specs (7W, 9K primary, 4, 8, 16 secondary, with 20% screen taps); but at what cost?
The question to the forum is, does any one know what would happen by going from 20% to 40% in a distributed load amplifier. The only thought I had but it is to simple mined to work, is to change the screen 1K ohm resistor to a 2K ohm resistor, with the idea that if I go from 20% to 40% and double the resistance of the series screen resistor it would be equivalent to having a 20% screen tap.
jerr
I did not notice when I first started buying parts for the amp that the output transformers called for are “20%” screened taped. When I noticed the taps I had already bought all the parts for the amp (even the output transformers), so I am trapped into continuing the build, or just wasting the money I spent on all the parts.
My question is what would happen if instead of trying to find a output that has 20% taps, if I bought one of the more common transformers that has 40% taps? I read the section in the Mullard book that takes about “distributed loading” (also called ultra-linear amplification), and I think I understand the difference between 20%, and 43%. The 20% tap offers more power output, while the 43% offers less distortion. Almost all the transformers that I see on the net are actually 40% (not the 43% given in the design). The taps on the primary in the design connects through 1K resistors to grids in the pentode section of the push-pull ECL82 output tubes. I don’t really know what would happen if I used a 40% tapped output. Would I go from less power output to less distortion? If that were the case I would just forge ahead and get a 40% screened taped output.
I found two suppliers in G.B. that have a output transformer that might work, the problem is that one does not sell to the US (why?), and the other will build a custom transformer to my specs (7W, 9K primary, 4, 8, 16 secondary, with 20% screen taps); but at what cost?
The question to the forum is, does any one know what would happen by going from 20% to 40% in a distributed load amplifier. The only thought I had but it is to simple mined to work, is to change the screen 1K ohm resistor to a 2K ohm resistor, with the idea that if I go from 20% to 40% and double the resistance of the series screen resistor it would be equivalent to having a 20% screen tap.
jerr
I am no expert but I doubt that you would hear the difference though you could measure it.
Another thing to consider is that if your OPTs have different input impedances, say 3.5K and 5K then the screen tap is at 40% (or whatever it is rated at) for only one of those input impedance choices.
I have OPTs that can be configured as one of three input impedances and the screen tap is stated to be 43%. However at 5K the screen tap is actually about 30% IIRC.
In any case I would say go ahead and do the build. If you decide to go with a different OPT lafter the build then changing them is is probably one of the easier mods you can do. Plus you'll then have a spare pair of transformers for your next project!
Another thing to consider is that if your OPTs have different input impedances, say 3.5K and 5K then the screen tap is at 40% (or whatever it is rated at) for only one of those input impedance choices.
I have OPTs that can be configured as one of three input impedances and the screen tap is stated to be 43%. However at 5K the screen tap is actually about 30% IIRC.
In any case I would say go ahead and do the build. If you decide to go with a different OPT lafter the build then changing them is is probably one of the easier mods you can do. Plus you'll then have a spare pair of transformers for your next project!
Sherman,
Thanks for your reply, it helped me make up my mind to go ahead and build the amp. I will just have to pick up a couple of Hammond ouput transformers. My main problem was that what I did not want to happen is to fire the amp up with the 40% taps, and have the amp sound "horrible". What I had never realized till you mentioned was that if you used one of the "universal" output transformers just changing the primary taps would change the value of the transformers screen taps, so the value of the taps can not be that crictal.
jerr
Thanks for your reply, it helped me make up my mind to go ahead and build the amp. I will just have to pick up a couple of Hammond ouput transformers. My main problem was that what I did not want to happen is to fire the amp up with the 40% taps, and have the amp sound "horrible". What I had never realized till you mentioned was that if you used one of the "universal" output transformers just changing the primary taps would change the value of the transformers screen taps, so the value of the taps can not be that crictal.
jerr
Hey-Hey!!!,
You could do what I did when faced with your troubles: pick a good OPT, take it apart and have it wound with taps where ever it is possible to put them.
I happened to decide upon a 1948 Peerless design, S-265-Q. This one has 10 layers on each side of the CT on the 10k a-a primary. This leaves 10% increments to put taps at the edges of each layer.
It is also possible to use more than one tap at a time. My E-Linear designs have just this option in mind. The last run of the 265's I had made for a group build got taps at 20, 30 and 40%.
Look here if you're curious:
http://audioroundtable.com/GroupBuild/Projects/
Merlin V2 is the latest, but missing some details covered in the first one.
cheers,
Douglas
You could do what I did when faced with your troubles: pick a good OPT, take it apart and have it wound with taps where ever it is possible to put them.
I happened to decide upon a 1948 Peerless design, S-265-Q. This one has 10 layers on each side of the CT on the 10k a-a primary. This leaves 10% increments to put taps at the edges of each layer.
It is also possible to use more than one tap at a time. My E-Linear designs have just this option in mind. The last run of the 265's I had made for a group build got taps at 20, 30 and 40%.
Look here if you're curious:
http://audioroundtable.com/GroupBuild/Projects/
Merlin V2 is the latest, but missing some details covered in the first one.
cheers,
Douglas
ECL
Agree with the others; perhaps some explanation. Data is rather scarce for many tubes in UL, but all is similar, and the following for the KT88 for which I have a complete spec. will illustrate what happens:
The maximum output for this tube is constant up to 43% screen taps. The main difference is in the Zout, i.e. the internal impedance. As you may know, pentodes, as a result of their high Zout (amongst other things), are sensitive to load changes and do not like the varying loudspeaker impedances. For the KT88 (p.p.) Zout under certain conditions is given as 30K for pentode operation, 8K for 20% screen taps, and 4K for 43% taps (compared to 2.8K for triode operation). This very briefly illustrates the almost triode operation for UL but with pentode output. (Rated distortion stays about the same from 20% taps onwards as a percentage, thus actually dropping toward triode operation because the output is lower, but with obviously the disadvantage of a lower maximum output.)
You could therefore actually be better off generally with 40% taps, though we do not know at what stage the maximum output for ECL82s will start to drop. All in all not an audible difference, though loudspeaker handling ability is improved slightly with 40% taps.
Hope this will allow you to understand better.
Agree with the others; perhaps some explanation. Data is rather scarce for many tubes in UL, but all is similar, and the following for the KT88 for which I have a complete spec. will illustrate what happens:
The maximum output for this tube is constant up to 43% screen taps. The main difference is in the Zout, i.e. the internal impedance. As you may know, pentodes, as a result of their high Zout (amongst other things), are sensitive to load changes and do not like the varying loudspeaker impedances. For the KT88 (p.p.) Zout under certain conditions is given as 30K for pentode operation, 8K for 20% screen taps, and 4K for 43% taps (compared to 2.8K for triode operation). This very briefly illustrates the almost triode operation for UL but with pentode output. (Rated distortion stays about the same from 20% taps onwards as a percentage, thus actually dropping toward triode operation because the output is lower, but with obviously the disadvantage of a lower maximum output.)
You could therefore actually be better off generally with 40% taps, though we do not know at what stage the maximum output for ECL82s will start to drop. All in all not an audible difference, though loudspeaker handling ability is improved slightly with 40% taps.
Hope this will allow you to understand better.
The guys above have given good and correct advice regards 20% and 43% (40%) Ultralinear taps. Being an unfashionable fan of ultralinear I've used both. ECL82 Amps in particular often used 20% rather than 43% becasue for this particular Triode / Pentode it fives an extra 1 watt output so the advertising guys had something to brag about.
ECL86 Amps generally used 43% as did higher power amps like EL84 based.
In general view 20% as a compromise to squeeze a smidge (technical term equivalent to 1 bee's dick) more output power from low power pentodes like in the ECL82. I've also seen 20% taps used for 6M5 Output Tubes.
Cheers,
Ian
ECL86 Amps generally used 43% as did higher power amps like EL84 based.
In general view 20% as a compromise to squeeze a smidge (technical term equivalent to 1 bee's dick) more output power from low power pentodes like in the ECL82. I've also seen 20% taps used for 6M5 Output Tubes.
Cheers,
Ian
Thanks to every one for their response to my 20% screen tap question.
I had sent out an e-mail to www.sowter.co.uk asking about output transformers, and I got a response from Brian Sowter.
Sowter makes the "exact" (9895 MULLARD 7W, ECL82 PP AMP) transformer that I was looking for.
Look at:
http://www.sowter.co.uk/acatalog/SOWTER_CLASSIC_AMPS_15.html#a303
It almost seems that they were looking at the Mullard design when they made the transformer (7W, 9K primary, 4, 8, 16 secondary, 20% screen tap).
The only problem is that they want 87 pounds ($155) plus postage for the transformer, so at 2X87 pounds it is a little to steep for my blood.
The nice thing is that if you are looking for a transformer to match the plate resistence of ECL82s (6BM8) they have a transformer for you, if you can stand the cost.
So my plan is to buy two Hammond transformers (P-T1608/P-T1609) from "tubesandmore". The P-T1608 is a 8K primary, and the P-T1609 is a 10K primary output transformer with 40% screen taps ($35).
The further question I have is when you have a choice between a 8K and a 10K primary when you are really looking for a 9K primary what would be the best choice?
Regards,
jerr
I had sent out an e-mail to www.sowter.co.uk asking about output transformers, and I got a response from Brian Sowter.
Sowter makes the "exact" (9895 MULLARD 7W, ECL82 PP AMP) transformer that I was looking for.
Look at:
http://www.sowter.co.uk/acatalog/SOWTER_CLASSIC_AMPS_15.html#a303
It almost seems that they were looking at the Mullard design when they made the transformer (7W, 9K primary, 4, 8, 16 secondary, 20% screen tap).
The only problem is that they want 87 pounds ($155) plus postage for the transformer, so at 2X87 pounds it is a little to steep for my blood.
The nice thing is that if you are looking for a transformer to match the plate resistence of ECL82s (6BM8) they have a transformer for you, if you can stand the cost.
So my plan is to buy two Hammond transformers (P-T1608/P-T1609) from "tubesandmore". The P-T1608 is a 8K primary, and the P-T1609 is a 10K primary output transformer with 40% screen taps ($35).
The further question I have is when you have a choice between a 8K and a 10K primary when you are really looking for a 9K primary what would be the best choice?
Regards,
jerr
Yep, those Sowter prices are steep. I recently had a power transformer wound here for the Dynaco ST70 (exact specs and appearance) for 27% of their price. Even the Hammonds are about 170% of manufacturing cost here (are we fortunate?). I suppose that is not an option for you?
Wrt the primary impedance it is as SY said. But if you really have to flip a coin, I would go for the 8K. Loudspeakers go more up than down - normally. There are a few anomalies on the market (and for applying that description to some upper-class loudspeakers, I will probably get .
)
Wrt the primary impedance it is as SY said. But if you really have to flip a coin, I would go for the 8K. Loudspeakers go more up than down - normally. There are a few anomalies on the market (and for applying that description to some upper-class loudspeakers, I will probably get .
)- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.