Deall forum members,
after having finished my little Class-D project for the kitchen's "hifi" - I'd like to move to a new project.
I have 2 matched pairs of Ulyanov 6C33C-B (6S33S-V) tubes, so 4 tubes alltogether. I got them from my father, he's a big EL34/KT88 builder but these are some different animals and I thought I give it a try here and ask if there're ANY special considerations we need to take into account when building a stereo amp with these. I would like to make him a stereo integrated amp for Xmas 2019, a good one.
Tubes are new. If they're matched pairs REALLY - who knows, anyone can state anything on the internet, right ? Anyway, the stamps on the glasses are matching. They're here and I'd like to get some really great triode sound out of them.
I read some issues regarding these tubes about reliability, or at least - put it that way - the quality spread seems to be big. What's the truth behind that, shall we be worried ? These tubes are easy to get cheap here from the neighbouring country, Ukraine, nevertheless I'd like to employ some basic protections in the circuit, either in the design itself or using a small microcomputer (like an Arduino) to monitor current, voltages, power-on hours and setting/modifying bias automatically, or even turn-off the amp via relays if that's the case. Maybe you have some basic hints what we need to take into account with these tubes .. ?
Design considerations for now:
- we read the article with Tim Mellow's OTL design but we think we might stick to a good old classic design with OT which doesn't stress the tubes at all.
- I'd like to have a good, reliable, great sounding Class AB, 2 tubes per side in pushpull, toroidal OT (from Transformatory toroidalne - Producent transformatorow Toroidy.pl) calculated into 6+ Ohms (8 in reality but most 8 ohm drivers go reasonably below 8 Ohms in reality so we design the OT to 6 Ohms then and using the amp with nominal 8 Ohm speakers) and then it should be fine. XLR inputs, fully balanced design with good common mode noise rejection - that's all I have in mind right now.
My question rather focuses on reliability, stability and longevity but if somebody has an already-finished good schematics, go ahead..
after having finished my little Class-D project for the kitchen's "hifi" - I'd like to move to a new project.
I have 2 matched pairs of Ulyanov 6C33C-B (6S33S-V) tubes, so 4 tubes alltogether. I got them from my father, he's a big EL34/KT88 builder but these are some different animals and I thought I give it a try here and ask if there're ANY special considerations we need to take into account when building a stereo amp with these. I would like to make him a stereo integrated amp for Xmas 2019, a good one.
Tubes are new. If they're matched pairs REALLY - who knows, anyone can state anything on the internet, right ? Anyway, the stamps on the glasses are matching. They're here and I'd like to get some really great triode sound out of them.
I read some issues regarding these tubes about reliability, or at least - put it that way - the quality spread seems to be big. What's the truth behind that, shall we be worried ? These tubes are easy to get cheap here from the neighbouring country, Ukraine, nevertheless I'd like to employ some basic protections in the circuit, either in the design itself or using a small microcomputer (like an Arduino) to monitor current, voltages, power-on hours and setting/modifying bias automatically, or even turn-off the amp via relays if that's the case. Maybe you have some basic hints what we need to take into account with these tubes .. ?
Design considerations for now:
- we read the article with Tim Mellow's OTL design but we think we might stick to a good old classic design with OT which doesn't stress the tubes at all.
- I'd like to have a good, reliable, great sounding Class AB, 2 tubes per side in pushpull, toroidal OT (from Transformatory toroidalne - Producent transformatorow Toroidy.pl) calculated into 6+ Ohms (8 in reality but most 8 ohm drivers go reasonably below 8 Ohms in reality so we design the OT to 6 Ohms then and using the amp with nominal 8 Ohm speakers) and then it should be fine. XLR inputs, fully balanced design with good common mode noise rejection - that's all I have in mind right now.
My question rather focuses on reliability, stability and longevity but if somebody has an already-finished good schematics, go ahead..
Individual Self Bias Cathode Resistors (with appropriate individual bypass capacitors)
does several things:
(with proper design and calculations)
Advantage: Allows the circuit to come up gracefully at the very first power up (no multiple bias adjustments to do quickly at first power up that fixed adjustable bias requires).
Advantage: Individual self bias does not require a regulated B+.
(Disadvantage for fixed adjustable bias circuits: they require regulated B+ unless you want to do multiple back and forth Re-adjustments of the bias).
Advantage: Allows you to easily see the match of cathode currents.
Advantage: Tends to reduce the chance of thermal run-away.
Advantage: Allows you to easily check the drift and aging of the individual cathode currents over time.
Are there disadvantages to individual self bias, yes. But all the above advantages always apply.
"All Generalizations Have Exceptions" - Me
does several things:
(with proper design and calculations)
Advantage: Allows the circuit to come up gracefully at the very first power up (no multiple bias adjustments to do quickly at first power up that fixed adjustable bias requires).
Advantage: Individual self bias does not require a regulated B+.
(Disadvantage for fixed adjustable bias circuits: they require regulated B+ unless you want to do multiple back and forth Re-adjustments of the bias).
Advantage: Allows you to easily see the match of cathode currents.
Advantage: Tends to reduce the chance of thermal run-away.
Advantage: Allows you to easily check the drift and aging of the individual cathode currents over time.
Are there disadvantages to individual self bias, yes. But all the above advantages always apply.
"All Generalizations Have Exceptions" - Me
Last edited:
I gathered new recent experience with real toroidal output transformers and also EI cores, standard transformers.
There is one major problem, in push-pull , any small magnetization reduces the INDUCTANCE of the push-pull, there is a small within -/+ 0.5ma bias where it has a very good bass response and BEAUTIFUL sound, and you can only set this up with a high definition 192khz analyser....
So, cathode bias doesn't guarantee a 0 net magnetization, some adverse current could still flow or the transformer sections would not be equal by 0.5 ohm which the cathode or AUTOMATIC bias would not detect!!!!!!!!
The EI cores are surprisingly nearly as bad as toroids or as quad C cores etc.
This is VERY important with the 6c33c with drifting bias and a long warm up time.
BTW I designed a driver and input stage for this tube which doesn't require any feedback..
The major reason I didn't build it is that there was no suitable SET transformer... and the power supply passive with at least 1mf (high currents0 was very costly. If you run this tube with anything below 1mf there is a compromise, maybe I am exaggerating, but I calculated it at the time... it needs a substantial chokes and capacitor bank, the pre-sections need a dedicated transformer to separate from the high current demand of the output tubes.
There is one major problem, in push-pull , any small magnetization reduces the INDUCTANCE of the push-pull, there is a small within -/+ 0.5ma bias where it has a very good bass response and BEAUTIFUL sound, and you can only set this up with a high definition 192khz analyser....
So, cathode bias doesn't guarantee a 0 net magnetization, some adverse current could still flow or the transformer sections would not be equal by 0.5 ohm which the cathode or AUTOMATIC bias would not detect!!!!!!!!
The EI cores are surprisingly nearly as bad as toroids or as quad C cores etc.
This is VERY important with the 6c33c with drifting bias and a long warm up time.
BTW I designed a driver and input stage for this tube which doesn't require any feedback..
The major reason I didn't build it is that there was no suitable SET transformer... and the power supply passive with at least 1mf (high currents0 was very costly. If you run this tube with anything below 1mf there is a compromise, maybe I am exaggerating, but I calculated it at the time... it needs a substantial chokes and capacitor bank, the pre-sections need a dedicated transformer to separate from the high current demand of the output tubes.
Would you be willing to consider an output transformerless design? There are several schematics out there using a pair of 6C33C tubes to get about 25W into 8 ohms. I've been running these OTLs for a few years now, with no problems at all.
It is 'easy' to build a transformer with low inductance, low input impedance, ok ok...
but the current is massive there,
for a push pull, just by memory it is around 220ma per side, , 440 ma
for a stereo, 880ma !!!
so there the core needs to be cut and put back together... ok easy
then the windings are NEVER even , so with perfectly equal bias in a typical PP with 100ma bias total...
follow this, 1% tolerance bias resistor + 0.05% voltmeter accuracy so the bias is around +/- 1ma and with a 0.5 % winding inequality the bias could fluctuate around 1.5ma
nothing too bad,
but if the current is 400ma, the tolerance will reach 6ma, this is enough to ruin the bass response and generate a lot of harmonics, reduce the inductance too much.
but the current is massive there,
for a push pull, just by memory it is around 220ma per side, , 440 ma
for a stereo, 880ma !!!
so there the core needs to be cut and put back together... ok easy
then the windings are NEVER even , so with perfectly equal bias in a typical PP with 100ma bias total...
follow this, 1% tolerance bias resistor + 0.05% voltmeter accuracy so the bias is around +/- 1ma and with a 0.5 % winding inequality the bias could fluctuate around 1.5ma
nothing too bad,
but if the current is 400ma, the tolerance will reach 6ma, this is enough to ruin the bass response and generate a lot of harmonics, reduce the inductance too much.
Well,
This somewhat unbalanced tube current problem is the perfect opportunity to use 2 Single Ended Transformers per channel to do Pseudo Push Pull (to solve the unbalance problem).
Start with 2 SE transformers; each should have very good coupling from primary to secondary. And each should have low DCR primaries and low DCR secondaries.
Connect Triode 1 plate to SE transformer #1 primary plate lead (Blue).
Connect SE transformer #1 primary B+ lead to B+ (Red).
Connect SE transformer #1 Secondary Common lead to ground, and 'common speaker output connector'.
Connect SE transformer #1 Secondary 8 Ohm tap lead to '8 Ohm speaker output connector'.
Connect Triode 2 plate to SE transformer #2 primary B+ lead (Red).
Connect SE transformer #2 primary B+ lead (Blue) to B+.
Connect SE transformer #2 Secondary Common lead to ground, and 'common speaker output connector'.
Connect SE transformer #2 Secondary 8 Ohm tap lead to '8 Ohm speaker output connector'.
The push pull out of phase triode # 1 versus triode #2 drives the reversed primary lead connections.
The output secondaries are in-phase, and aiding.
Since these are SE transformers, any unbalanced DC of the tubes does not affect the core laminations (but it would have on a true push pull transformer).
If you can follow all this, you ought to try it.
I have done this before, and it works very well.
I thought I was the first to do this, but then I was told the French had done it already.
This somewhat unbalanced tube current problem is the perfect opportunity to use 2 Single Ended Transformers per channel to do Pseudo Push Pull (to solve the unbalance problem).
Start with 2 SE transformers; each should have very good coupling from primary to secondary. And each should have low DCR primaries and low DCR secondaries.
Connect Triode 1 plate to SE transformer #1 primary plate lead (Blue).
Connect SE transformer #1 primary B+ lead to B+ (Red).
Connect SE transformer #1 Secondary Common lead to ground, and 'common speaker output connector'.
Connect SE transformer #1 Secondary 8 Ohm tap lead to '8 Ohm speaker output connector'.
Connect Triode 2 plate to SE transformer #2 primary B+ lead (Red).
Connect SE transformer #2 primary B+ lead (Blue) to B+.
Connect SE transformer #2 Secondary Common lead to ground, and 'common speaker output connector'.
Connect SE transformer #2 Secondary 8 Ohm tap lead to '8 Ohm speaker output connector'.
The push pull out of phase triode # 1 versus triode #2 drives the reversed primary lead connections.
The output secondaries are in-phase, and aiding.
Since these are SE transformers, any unbalanced DC of the tubes does not affect the core laminations (but it would have on a true push pull transformer).
If you can follow all this, you ought to try it.
I have done this before, and it works very well.
I thought I was the first to do this, but then I was told the French had done it already.
Last edited:
Tony,
I started with a stereo 300B SE amp.
I had a medium level phase splitter module.
I connected the module 2 phase output to the triode input/driver of the stereo amp's L and R stage.
I connected the SE output transformers as described in my Post #12
That made the stereo SE amp into a mono-block pseudo push pull amp.
One key to doing it this way is to parallel the output secondaries.
For an 8 Ohm speaker, use the ground and 16 Ohm taps; get 2x the power and the same damping factor that the Stereo SE amp had.
If you series the two output secondaries, you degrade the damping factor, and also get more 3rd harmonic distortion.
I started with a stereo 300B SE amp.
I had a medium level phase splitter module.
I connected the module 2 phase output to the triode input/driver of the stereo amp's L and R stage.
I connected the SE output transformers as described in my Post #12
That made the stereo SE amp into a mono-block pseudo push pull amp.
One key to doing it this way is to parallel the output secondaries.
For an 8 Ohm speaker, use the ground and 16 Ohm taps; get 2x the power and the same damping factor that the Stereo SE amp had.
If you series the two output secondaries, you degrade the damping factor, and also get more 3rd harmonic distortion.
Start with a recording of a bass note that has a rich harmonic structure.
If that is a 30 Hz bass note, and your loudspeakers cut off sharply at 60 Hz, you will not
hear the 30Hz, and the 2nd harmonic at 60 Hz will be somewhat low in amplitude.
The third harmonic at 90Hz will be heard, as well as the 4th harmonic at 120 Hz, and 5th harmonic at 150Hz will be heard.
It is a reasonably well established fact that when our ears hear the upper harmonics (in this case spaced by 30Hz each), you will "hear" the "missing" 30 Hz fundamental.
Your ear interprets, and "tells" you that the unheard fundamental is 30 Hz.
Also, if your amplifier and/or speaker creates its own 60 Hz and 90 Hz harmonics, the (missing) bass note of 30 Hz will sound louder than it would if there was no 2nd or 3rd harmonic generated by your amp and/or speaker.
Also, our ears are limited by the Fletcher - Munson Effect.
The lower the frequency, the more amplitude you need in order to be able to hear it.
Is this perhaps the effect you have heard about, and are mentioning?
The science of sound, and the science of hearing.
P.S. I was in the PI decades ago. Nice country.
If that is a 30 Hz bass note, and your loudspeakers cut off sharply at 60 Hz, you will not
hear the 30Hz, and the 2nd harmonic at 60 Hz will be somewhat low in amplitude.
The third harmonic at 90Hz will be heard, as well as the 4th harmonic at 120 Hz, and 5th harmonic at 150Hz will be heard.
It is a reasonably well established fact that when our ears hear the upper harmonics (in this case spaced by 30Hz each), you will "hear" the "missing" 30 Hz fundamental.
Your ear interprets, and "tells" you that the unheard fundamental is 30 Hz.
Also, if your amplifier and/or speaker creates its own 60 Hz and 90 Hz harmonics, the (missing) bass note of 30 Hz will sound louder than it would if there was no 2nd or 3rd harmonic generated by your amp and/or speaker.
Also, our ears are limited by the Fletcher - Munson Effect.
The lower the frequency, the more amplitude you need in order to be able to hear it.
Is this perhaps the effect you have heard about, and are mentioning?
The science of sound, and the science of hearing.
P.S. I was in the PI decades ago. Nice country.
Last edited:
Yes, my original intention was to build Tim Mellow's OTL amp, however the crowd is split regarding OTL operation, some say it's good or can be good, others say nooo way, again others say 6S33S tubes are not that reliable in general, so at the moment I'm thinking about some kind of classic triode design.
But if you say Tim's OTL is great and reliable for years, then I might reconsider it. So many opinions out there...
Picking up this project and starting just now, with some delay...
- Do you mean 7/24 ON full-time for 4 days ?
- Heaters only ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.