So I have ben thinking about building a beefy line amplifier with balanced output. Minimum 24dbu in 600ohm and still have acceptable distortion figures. And I also wanted to try something different. So I started looking at the ecl86 tube. It has a triode and a pentode so there is some gain if needed and I could use some feedback.
For a balanced out I opted using a transformer at the output. Galvanic Isolation is nice with tube circuits. And I like transformers. Carnhill makes this nice transformer the VTB2290 Primary 9600ohm and secondary 600 or 150. specs are in the carnhill catalog http://www.audiomaintenance.com/downloads/carnhill_design_guide.pdf
It has a gap so standing DC current should be no problem. A first draw of the circuit is this: http://www.aman-nai.be/6gw8output.pdf
So a triode input stage coupled to the output pentode wich is strapped as a triode using the 100ohm resistance the output has autobias and there is feedback from the secondary of the transformer to the cathode of the triode input stage.Gain is around 18db input to output. I did a simulation in ltspice. At an output of +4dbu in 600ohm the predicted distortion is 0.0076% At an output of +4dbu in 10k the predicted distortion is 0.0013% At an output of +24dbu in 600ohm the predicted distortion is 0.079% At an output of +24dbu in 10k the predicted distortion is 0.015% now and altough this has to be taken with some grain of salt as I don't know how much the transformer can handle.
Clipping in 600ohm is at 33dbu 3.19%THD Clipping in 10k is at +35dbu (don't touch the output it has 120v pp) 1.57%THD Now again take this with a grain of salt. at 1Khz I think this could be really possible but at 20Hz I don't think the transformer can handle that high level. Still +24dbu should be perfectly possible.
The output tube disspates a little more then 8W. It is specified for 9W so that should be ok. Supply voltage is 300v
Any ideas or comment are welcome. It is still in the design phase
For a balanced out I opted using a transformer at the output. Galvanic Isolation is nice with tube circuits. And I like transformers. Carnhill makes this nice transformer the VTB2290 Primary 9600ohm and secondary 600 or 150. specs are in the carnhill catalog http://www.audiomaintenance.com/downloads/carnhill_design_guide.pdf
It has a gap so standing DC current should be no problem. A first draw of the circuit is this: http://www.aman-nai.be/6gw8output.pdf
So a triode input stage coupled to the output pentode wich is strapped as a triode using the 100ohm resistance the output has autobias and there is feedback from the secondary of the transformer to the cathode of the triode input stage.Gain is around 18db input to output. I did a simulation in ltspice. At an output of +4dbu in 600ohm the predicted distortion is 0.0076% At an output of +4dbu in 10k the predicted distortion is 0.0013% At an output of +24dbu in 600ohm the predicted distortion is 0.079% At an output of +24dbu in 10k the predicted distortion is 0.015% now and altough this has to be taken with some grain of salt as I don't know how much the transformer can handle.
Clipping in 600ohm is at 33dbu 3.19%THD Clipping in 10k is at +35dbu (don't touch the output it has 120v pp) 1.57%THD Now again take this with a grain of salt. at 1Khz I think this could be really possible but at 20Hz I don't think the transformer can handle that high level. Still +24dbu should be perfectly possible.
The output tube disspates a little more then 8W. It is specified for 9W so that should be ok. Supply voltage is 300v
Any ideas or comment are welcome. It is still in the design phase
The idea seems good. Personally I dislike using pentodes as triodes, but it is my opinion only.
I suggest to make C30 * R72 = C31 * R75 to maintain the feedback factor constant over the whole audio range. In fact, on of this caps may be unnecessary. Also, 10KΩ as cathode bias seems too high. 2.2KΩ is more realistic for such a triode.
This is my suggestion.
I suggest to make C30 * R72 = C31 * R75 to maintain the feedback factor constant over the whole audio range. In fact, on of this caps may be unnecessary. Also, 10KΩ as cathode bias seems too high. 2.2KΩ is more realistic for such a triode.
This is my suggestion.
Attachments
Last edited:
Go for it. The line part of the Shindo Aurièges uses a similar tube (PCL86, iirc) but not a transformer. It's triode connected too (iirc, been a while since I've seen one). Somewhere on the forum someone mentioned that an ECL82/6BM8 is a much more robust tube than an ECL86. ECL82s are still being made by Electro-Harmonix. There seem to be lots of PCL86 tubes around at reasonable prices. The (slightly) different filament voltage seems to put some people off so the price isn't crazy.
Cheers, Steve
Oh, by the way, my virus nanny won't open your link. Why not just attach the schematic to your post. "Go Advanced" and you can add attachments.
Cheers, Steve
Oh, by the way, my virus nanny won't open your link. Why not just attach the schematic to your post. "Go Advanced" and you can add attachments.
I would't use ECL86 or ECL82 as they are relatively common in vintage audio gear, which increases demand. I would shoot for something less common, such as the ECL85 (6GV8). In fact, since you're wiring the pentode as a triode, why not use a dissimilar triode tube? There are plenty out there. BTW, if you're in the US there are a ton of options if you consider Compactrons and all the triode/pentode TV tubes.
I'm open for al options on different tubes It's nice to do something different. On the other hand the ECL86 is in production so I could always find an ecl86. And you are right the price 10 years ago was a lot less then now.... The thing is I would like to have a small sigal triode and an output pentode (or triode) in one glass envelope This line stage will be used in different ("modular") ways so having only one tube is nice.
For the transformer it's indeed a pity they don't supply all info. So I will have to experiment. The transformer would probably also need an series RC circuit over the secondary to avoid ringing. Otherwise the feedback could have some trouble However I would have to test this when I have the transformer.
It's nice to do something different. On the other hand the ECL86 is in production so I could always find an ecl86. And you are right the price 10 years ago was a lot less then now.... The thing is I would like to have a small sigal triode and an output pentode (or triode) in one glass envelope This line stage will be used in different ("modular") ways so having only one tube is nice. For the transformer it's indeed a pity they don't supply all info. So I will have to experiment. The transformer would probably also need an series RC circuit over the secondary to avoid ringing. Otherwise the feedback could have some trouble However I would have to test this when I have the transformer.
If you have lots of them...
Build a power amp too, based on the Rogers Cadet of old. Sure it's only about 6 to 8 watts but might be fun.
See the "Wave 16?" thread I posted recently using 6BM8s. Not a great amp but OK. Similar to the Cadet power amp.
Steve
Build a power amp too, based on the Rogers Cadet of old. Sure it's only about 6 to 8 watts but might be fun.
See the "Wave 16?" thread I posted recently using 6BM8s. Not a great amp but OK. Similar to the Cadet power amp.
Steve
Attachments
+24dBu IN 600r is just +24dBm.
Historically, if a studio needed more than the +18dB possible from a 6J7, they moved to push-pull 6F6/6V6. Partly because they could afford it. Also because if one side of a hot-worked pair failed, signal would still get through. But also because a push-pull core is much smaller than SE, and the THD is very much less.
The way you have taken NFB there is no galvanic isolation.
Do you actually have 600r loads?
I'm open for al options on different tubes
It's nice to do something different.
On the other hand the ECL86 is in production so I could always find an ecl86.
And you are right the price 10 years ago was a lot less then now....
Really? Where?
6FM7 is dissimilar triode in one bottle. There are others too.
ECL86/6GW8 are NOT in current production - very popular tube because they were truly excellent - becoming rocking horse poo availability. Tube manufacturers - we want ECL86s brought back. ECL85/6GV8 are not in current production either but there are a heap of NOS about. Very popular in tube TVs for frame oscillator (Triode) and scan coil power drive (Pentode) Will eventually become rare.
Some Guitar Amp Guys are using them for small power amps because they could no longer rely on getting ECL86. ECL84/6DX8 are in current production and are good quality, being used for small HiFi and Guitar amps. ECL82/6BM8 are in current production too. Another good quality tube, again being used for small Guitar and HiFi Amps.. There certainly used to be an ECL80 but haven't seen one for years. Don't know if ECL81 or ECL83 ever existed. Go for ECL84 or ECL82 would be my recommendation.
Cheers,
Ian
Some Guitar Amp Guys are using them for small power amps because they could no longer rely on getting ECL86. ECL84/6DX8 are in current production and are good quality, being used for small HiFi and Guitar amps. ECL82/6BM8 are in current production too. Another good quality tube, again being used for small Guitar and HiFi Amps.. There certainly used to be an ECL80 but haven't seen one for years. Don't know if ECL81 or ECL83 ever existed. Go for ECL84 or ECL82 would be my recommendation.
Cheers,
Ian
If you go with ECL82, I've found these these 6F3P tubes are excellent and don't break the bank.
6F3P / 6BM8 / ECL82 tube - - Tubes-Store.com
6F3P / 6BM8 / ECL82 tube - - Tubes-Store.com
I repaired a gates sta-level ones and they used a pushpull 6v6. Very beefy output you could say.
I do like the simplicity of SE. The neve 1073 output is also SE with standing DC trough the primary and it doesn't sound bad you could say. Ok it's with transistors but the principle (se) is the same.
As the feedback. I could take feedback from the plate (primary). Then the transformer is not in the feedback loop. If it turns out that the transformers gives problems when in the feedback loop (= ringing/oscillation) then I could do this. I liked the idea of the transformer in the feedback loop as it lowers output impedance and extends the frequency response.
For having 600ohm loads. When you make a passive mixing bus and start adding subgroup busses the loading impedance goes down. If you use higher resistance mixing resistors the noise goes up. I mainly use neve outputs and they just drive like everything.... So I wanted a tube circuit that can do the same is SE and drive like everything
Some older gear we have is 600ohm also and some of the "cheaper" whimpy preamps just don't cut it. If I remember correctly, the original 1176 (especially the revision A "blue-stripe") has a low impedance input.
Ok so I will checkout some more tubes. I like the dissimilar triodes. Didn't know they existed.
HI Ian,
all of these three valves existed. They have in common their common cathodes, which severely resticts their usability. As the ECL80 has a 6.3V/0.3 A heater, there was no need for an extra series heated version. Series heated versions of both the other tubes existed, though: PCL81 and UCL81 were identical to ECL81, besides the heaters, of course. But be careful with that damned Mullard ECL/PCL83 - same pentode, but very dissimilar triodes (high µ in the E, low to medium µ in the P Version)!
Btw, which manufacturer still does ECL84's/6BX8's? This is a rather interesting tube, as it looks like and behaves like an ECF80 on steroids.
Best regards!
Last edited:
