I have put together an Amp with 2 EL34s, one is a triode and the the other
is normal (Antitriode). Cathodes are connected to CCSs and are capacitively coupled. The g1 of the undriven EL34 is connected to deck via a low R as in the serial I/P EL34. One channel is driven by an EF80 and the other by an ECF82 in cascode. EF80 performs better with richer sound and more gain.
HT on PA stage is only 250V Ia is 75mA.
The problem is that the O/P power is lower than my SE EL84 amp and the gain
is lower too (half ECC81/EL84) Before it distorts the sound is excellent with very good bass and highs, doubtless due to to the O/P toroids, (current balanced to within 1mA.
What level of power output should I expect from this setup? 2 Watts or thereabouts seems a bit low!!!!
Cheers
Boiss
is normal (Antitriode). Cathodes are connected to CCSs and are capacitively coupled. The g1 of the undriven EL34 is connected to deck via a low R as in the serial I/P EL34. One channel is driven by an EF80 and the other by an ECF82 in cascode. EF80 performs better with richer sound and more gain.
HT on PA stage is only 250V Ia is 75mA.
The problem is that the O/P power is lower than my SE EL84 amp and the gain
is lower too (half ECC81/EL84) Before it distorts the sound is excellent with very good bass and highs, doubtless due to to the O/P toroids, (current balanced to within 1mA.
What level of power output should I expect from this setup? 2 Watts or thereabouts seems a bit low!!!!
Cheers
Boiss
Hey Boiss,
Hard to judge before you have presented a schematic(with trafodata). I sincerely hope you run at least 75mA/tube? With 250V B+(or is it Ua?) I would run it a lot higher.
To get G2 current balance all the way I suggest you try connecting the first EL34 as Schade instead of triode. Both tubes will then work in identical DC conditions. You will also get more power out of the amp.
Normally works great driven by a good pentode or cascode.
Osvaldo,
Use "search" and you will find quite a lot about antitriode.
Hard to judge before you have presented a schematic(with trafodata). I sincerely hope you run at least 75mA/tube? With 250V B+(or is it Ua?) I would run it a lot higher.
To get G2 current balance all the way I suggest you try connecting the first EL34 as Schade instead of triode. Both tubes will then work in identical DC conditions. You will also get more power out of the amp.
Normally works great driven by a good pentode or cascode.
Osvaldo,
Use "search" and you will find quite a lot about antitriode.
This started off with feedback from the anode of the first EL34 to the grid with both EL34s as pentodes, however the overall gain was very low. This is why the driver stage (EF80) was modded and the first EL34 trioded.
One channel was then changed to use an ECF82 on the other channel.
Current was adjusted through the EL34s to 75ma CH1 and 76mA CH2.
the difference within each pair is less than 1mA.
One channel was then changed to use an ECF82 on the other channel.
Current was adjusted through the EL34s to 75ma CH1 and 76mA CH2.
the difference within each pair is less than 1mA.
Attachments
I see two electrolytic caps from the output tube cathodes to ground. No AC signal cross feeding can occur between the tubes that way, so only one tube is driving the output. For real "anti-triode" mode, one side would also need to have a much higher gm than the other, but a triode and pentode combo will still give some SE effect in class A operation. DC balance in the OT may be an issue too, due to the screen current on one side. You should be able to roughly fix that by adjusting the two cathode current sources slightly differently to compensate (slightly higher CCS current on the pentode side to allow for average screen current loss). There are other methods as well.
Last edited:
Thanks for the fb. I will remove the 22uF caps and I noticed on something from another source re the LM337s that a small 0.1uF can be put across the R. At the moment the currents in the 2 CCS are well balanced. I can measure the g2 current, however because I have 150VA power toroids the current needs to be balanced. I guess that i could put a resistive load on the I/P side anode with another 337 to simulate the other g2 current. If the removal of the caps makes a big difference I will try the schaded approach again.
More at the end of the day.
More at the end of the day.
Hey Boiss,
Noted your cascode is not a cascode, it is a betafollower and will not drive a Schade. The EF80 might not be the best suited tube for audio/Schade at all. Why not go for an E280F or a D3a?
Anyway the big differnce will be removing the caps as Don pointed out to make the the tube work in antitriode.
If you want to go pure SE you only need to remove the cap on the right/pentode-side together with the "horizontal" caps.
Noted your cascode is not a cascode, it is a betafollower and will not drive a Schade. The EF80 might not be the best suited tube for audio/Schade at all. Why not go for an E280F or a D3a?
Anyway the big differnce will be removing the caps as Don pointed out to make the the tube work in antitriode.
If you want to go pure SE you only need to remove the cap on the right/pentode-side together with the "horizontal" caps.
Removed the 22uF caps, Gain dropped dramatically and also distortion has risen at quite low levels, a quick check with a DVM showed strange readings on the cathodes of the EL34s 17v and 21V, (previously around 14V) suspected instability. Have spent the afternoon trying to repair my 40yr old scope. 1 channel of scope now sort of works so will try and work out what is happening tomorrow. Will also try a 1K square wave from the scope cal source.
The ECF82 driver was only connected directly. the EF80 was connected to the EL34 with a feedback resistor of 82K and without the cathode bypass cap. Further the Anode load for the EF80 was 12k before (schaded config).
Further I have used what was available to me in the way of valves (tubes), the EL34 I got a couple of years ago with the intention of making an amp along these lines ie self-splitter etc, so I have to make the best of what I have and costs etc. I have some EL84s and some wierd russian pentodes with topcaps etc. but they have higher ra than el84.
The ECF82 driver was only connected directly. the EF80 was connected to the EL34 with a feedback resistor of 82K and without the cathode bypass cap. Further the Anode load for the EF80 was 12k before (schaded config).
Further I have used what was available to me in the way of valves (tubes), the EL34 I got a couple of years ago with the intention of making an amp along these lines ie self-splitter etc, so I have to make the best of what I have and costs etc. I have some EL84s and some wierd russian pentodes with topcaps etc. but they have higher ra than el84.
if your using a matched pair surely this wouldnt be an issue? although you could bias them separately and use a single capacitor to couple them. as for the 100 ohm resistor off of the grid of the second EL34, i think that should be a 100n cap, although i may be wrong, ive never seen this configuration with these valves before.
I don´t know if you found the thread about our succesful 6B4G/6L6 anti-triode with mains toroids? Anyway we used a currentmirror to automaticly equal out the currents.
What voltage toroids do you use. Think I saw 150VA and that should mean quite low inductance.
Rod Coleman also did a great design with EL84 and servo in my ECL86 thread.
What voltage toroids do you use. Think I saw 150VA and that should mean quite low inductance.
Rod Coleman also did a great design with EL84 and servo in my ECL86 thread.
OK an update to the previous thread, everything appears to be working quite well now with around 10W O/P. Changes to the O/P stage were;- removing the 22uF in the cathode circuit. Paralleling 0.22uF caps across the current setting resistors in the LM337 CCSs. Increasing the Ia quiescent current to 120mA for each EL34 (Va to Vk is 230V).
Cathode resistor of the EF80 changed to 690 ohms and the Anode load resistor changed to 50K, Coupling cap increased to 1uF.
It now sounds pretty good with lots of low down umpf and very clear mids and highs, overall very nice. I will try and take some measurements when I get my scope sorted out
Thanks for your suggestions and help.
Boiss
Cathode resistor of the EF80 changed to 690 ohms and the Anode load resistor changed to 50K, Coupling cap increased to 1uF.
It now sounds pretty good with lots of low down umpf and very clear mids and highs, overall very nice. I will try and take some measurements when I get my scope sorted out
Thanks for your suggestions and help.
Boiss
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.