hi everyone.
I have bought a diy tube amp and I need to modify a couple of things.
the schematic is attached.
Thorsten loesch have given me a lot of help and mostly told me what to do. I need however guidance.
I need to change:
1) The output tubes are 6.3V heater types, directly heated.
There are a number of effects in directly heated tubes, that mean AC heating derived problem increase by the square of the Voltage.
So while AC heating the 2A3 at 2.5V can work okay especially in a PP Amp, the problems are >20dB greater with 6B4 type valves.
The sonic problems from AC include a warm, but muddy and not very detailed sound, often lacking subjective dynamics.
One solution would be to change the heaters to DC (not straightforward, bad DC on directly heater tubes sounds even worse than bad AC, but in different ways).
2) The output transformers as said, are unsuited for operation without feedback by design.
It would probably be a good idea to add some, around 10dB should help to open up the bandwidth.
Can someone guide me with those mods?
thanks
I have bought a diy tube amp and I need to modify a couple of things.
the schematic is attached.
Thorsten loesch have given me a lot of help and mostly told me what to do. I need however guidance.
I need to change:
1) The output tubes are 6.3V heater types, directly heated.
There are a number of effects in directly heated tubes, that mean AC heating derived problem increase by the square of the Voltage.
So while AC heating the 2A3 at 2.5V can work okay especially in a PP Amp, the problems are >20dB greater with 6B4 type valves.
The sonic problems from AC include a warm, but muddy and not very detailed sound, often lacking subjective dynamics.
One solution would be to change the heaters to DC (not straightforward, bad DC on directly heater tubes sounds even worse than bad AC, but in different ways).
2) The output transformers as said, are unsuited for operation without feedback by design.
It would probably be a good idea to add some, around 10dB should help to open up the bandwidth.
Can someone guide me with those mods?
thanks
Hi boy!!!
Look at this: AC heaters had been used for decades, so I don´t believe to be necessary to get DC on them. And the trouble is augmented if the same transformer is being used, because you must drop out the difference between the AC RMS voltage and the DC voltage (1.414 times greater).
And second, the amplifier mus be designed to support NFB. The output transformer mainly, can cause a phase rotation that may cause the amplifier to convert into an oscillator.If the amplifier entirely has not be designed to NFB used, the phase rotations had not being into account, and may be impossible to use with NFB. Read some data in the web relative to NFB, and you surely will agree with me.
Best regards and good luck!
Look at this: AC heaters had been used for decades, so I don´t believe to be necessary to get DC on them. And the trouble is augmented if the same transformer is being used, because you must drop out the difference between the AC RMS voltage and the DC voltage (1.414 times greater).
And second, the amplifier mus be designed to support NFB. The output transformer mainly, can cause a phase rotation that may cause the amplifier to convert into an oscillator.If the amplifier entirely has not be designed to NFB used, the phase rotations had not being into account, and may be impossible to use with NFB. Read some data in the web relative to NFB, and you surely will agree with me.
Best regards and good luck!
Is this something you are hearing or is this theoretical? I have several PP 6B4 amps and don't experience these problems.
You do have a problem with your filaments, however. You have both the output tubes and the front end tubes all connected to the same filament transformer. As such your 6V heater supply is being modulated by the output tubes through their cathode connection. You need to add a filament transformer for either the OP tubes or the front end tubes so that they are separate.
Rod Coleman who posts on this board has voltage regulator kits for DHT filaments if you think that's what you need, but at least decouple the front end tubes from the OP tubes first.
The input tubes are indirectly heated, hence, there's no problem with sharing the heaters for the input tubes with the filaments on the output tubes. The cathodes of the output tubes are bypassed by C6, hence, there will ideally be no AC component in the audio range. In reality, there will be a small AC component but it should be small enough to not interfere with the input tubes.
You can use a DC supply for the heater, but you'll find that all solutions require a higher DC input voltage than what you can get with a 6.3 V winding. Some regulators burn more power in the regulator than they deliver to the tube filaments -- something I've always found a bit silly. Hence, when I designed my DC filament regulator, I opted to use a switchmode supply. I have boards available on my website (see my signature).
~Tom
change the output tubes to 6a5G as this is a 2a3 indirectly heated.
and
remove r13, r14 . change r11 and r12 to 100 ohms 1W change c6 to .1 uf 250V poly
The best results with DHT I found is using a separate heater supply that is center tapped, and put the cathode circuit on the center tap. with your setup I would reuse the 1K resistors tied in parallel and bypass with the 220Uf cap and put a .22 uf poly or pio cap and maybe even another 220 uf cap just to lower the impedence of the AC ground (pathway through the capacitors). The trick is to find a good transformer that has been built nicely as the center tap has to be perfect. It easy to get there with a toroid type transformer. its about 1 out of 3 good hammonds if you want to hand pick a more traditional style transformer.
and
remove r13, r14 . change r11 and r12 to 100 ohms 1W change c6 to .1 uf 250V poly
The best results with DHT I found is using a separate heater supply that is center tapped, and put the cathode circuit on the center tap. with your setup I would reuse the 1K resistors tied in parallel and bypass with the 220Uf cap and put a .22 uf poly or pio cap and maybe even another 220 uf cap just to lower the impedence of the AC ground (pathway through the capacitors). The trick is to find a good transformer that has been built nicely as the center tap has to be perfect. It easy to get there with a toroid type transformer. its about 1 out of 3 good hammonds if you want to hand pick a more traditional style transformer.
Last edited:
The A470 output transformer is good iron. I think the comment on them needing feedback is more a function of their original use on the ST70 which used pentodes for the finals. IIRC, you can replace the EL34's on an ST70 directly with 6B4G's and disconect the feedback entirely with good results. (Someone will comment if I am all wet on that.)
This is all technically true, however, I just find that it is preferable to have the OP tubes on a separate filament transformer.
The 6A5G is really a 6B4G with the center of the filament connected to a cathode. This was done to reduce hum.
I agree.
well its something that was suggested to me. I dont like the sound of the amp so Im looking to what to do. I do need to change some key caps to better ones but I want to focus on the design flaws first....
any suggestion of which separate heater supply? I would however want to keep the 6B4 to keep the mods cost as low as possible...
I guess I shouldnt bother to add feedback then? If I apply NFB correctly shouldnt I gain something?
I'd use a 2x12 V 25 VA transformer from Antek Inc. with my regulator board... But I'm biased, obviously.
You could get a 2x6 V, 25 VA transformer and add a few windings to get to 6.3 V if you want to stick with AC heating.
~Tom
You could get a 2x6 V, 25 VA transformer and add a few windings to get to 6.3 V if you want to stick with AC heating.
~Tom
NFB is debateable. always try to build the circuit perfect. if you start clipping the signal through the path, then apply NFB.
your schematic does not have it. So what. Lets fix whats really wrong instead of lowering the gain characteristics to hide problem.
for a heater transformer I would use a hammond 166ms (5vct 3A) for a each tube. the reason to stray away from 6.3V is less tube noise on the plate and better gain characteristics over the life of the tube.
now a variant of this is to take a transformer that has class F insulation split/dual bobbin type that has two 115V primaries and two 6V secondaries 50VA. Tie the primaries in series (230V operation) but run it at 120V. now we add a layer of rfi by adding a .01UF 1000V cap from the primary's center tap to the FG or earth ground (optional). Tie the two secondaries in series (normally 12V) now you have 6VCT but with core loss, this can be an advantage as any imbalance of heater winding tap will be absorbed in the heater winding core. the reason because of the higher current rating is mainly to lower the DCR of the secondary. Another effect is the extra inductance that bucks against the heater and secondary causes a small amount of heater regulation to be established as the thermionic emission will dynamically change with the signal. This sponginess works really well. And can be increased by putting 5-15mh chokes between the heater pin and the transformer connection.
in a push pull DHT amp I always recommend using separate transformers: one for the inverting and one for the non inverting side of the output section.
your schematic does not have it. So what. Lets fix whats really wrong instead of lowering the gain characteristics to hide problem.
for a heater transformer I would use a hammond 166ms (5vct 3A) for a each tube. the reason to stray away from 6.3V is less tube noise on the plate and better gain characteristics over the life of the tube.
now a variant of this is to take a transformer that has class F insulation split/dual bobbin type that has two 115V primaries and two 6V secondaries 50VA. Tie the primaries in series (230V operation) but run it at 120V. now we add a layer of rfi by adding a .01UF 1000V cap from the primary's center tap to the FG or earth ground (optional). Tie the two secondaries in series (normally 12V) now you have 6VCT but with core loss, this can be an advantage as any imbalance of heater winding tap will be absorbed in the heater winding core. the reason because of the higher current rating is mainly to lower the DCR of the secondary. Another effect is the extra inductance that bucks against the heater and secondary causes a small amount of heater regulation to be established as the thermionic emission will dynamically change with the signal. This sponginess works really well. And can be increased by putting 5-15mh chokes between the heater pin and the transformer connection.
in a push pull DHT amp I always recommend using separate transformers: one for the inverting and one for the non inverting side of the output section.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.