Does anyone recommend this Schematic ? try googling but could not find that high output.
anyone ?
cal
anyone ?
cal
As said in the other thread, you really need two pairs of 6550, in parallel PP, to get this much power in triode-mode. (If you were prepared to use ultralinear connection, then one pair of 6550 would be enough.)
6550 40W se
Maybe you should have a look at this project. The designer and builder (jane) is also a diyaudio member. He has done many amps, and I`ve read on a norwegian forum that this one sounds very good.
Maybe you should have a look at this project. The designer and builder (jane) is also a diyaudio member. He has done many amps, and I`ve read on a norwegian forum that this one sounds very good.
the 3x6550 only 24 W in triode mode.
where is the triode and Ultra linear switch ?
I need more POWER on triode mode.
but this schematic is so simple....
I don;t think I can get the sowter OT anywhere locally except from ebay.
that's costly.
I wonder how much it cost to create this project ?
cal
where is the triode and Ultra linear switch ?
I need more POWER on triode mode.
but this schematic is so simple....
I don;t think I can get the sowter OT anywhere locally except from ebay.
that's costly.
I wonder how much it cost to create this project ?
cal
If you REALLY need more power, forget triode and single ended.
Two 6550 will give you more than 60W in penthode mode at much lower cost.
Yves.
Two 6550 will give you more than 60W in penthode mode at much lower cost.
Yves.
yes but I really need the triode mode, because of the sound quality reason.
why doesn't anyone design that triode high power ?
must be very very difficult ?
cal
why doesn't anyone design that triode high power ?
must be very very difficult ?
cal
calico88,
Going SE whatever tubes you use by paralleling more than 2 is costly
and need perfectly matched tube pairs or triplets or quadruplets.
{And in those cases tubes must be replaced more often}
From a DIYaudio point of vue more than 2 will become instable
and will need potentiometers in each cathode to adjust bias current
for each tube ; otherwise distortion will rise quickly.
One other solution would be to have 1 primary for each tube
and there the transformer will be much more costly...
So going after more than 2 tubes in parallel in SE mode is not douable
for you! with 6550's. It's why so few designs exists!
One Italian designer did a 3 * 6550 SE but uses custom made automatic
bias adjustment, and special opt's ...
A design has been published by Ari Polisois in AudioXpress last year using
3 parallel 6C33C-B tubes in SE configuration to get 45 watts ;
but it's not a simple project! {custom SE opt's ; big heat dissipation
and a bigger PS }
Further to get 40 watts pure class A in SE you need a tube {'s} with a plate
dissipation of more than 140 watts ; and with low internal impedance.
And a costly SE opt to get the best of this mode. Very few tubes with those
type of parameters exist!
Regards.
Alain.
Going SE whatever tubes you use by paralleling more than 2 is costly
and need perfectly matched tube pairs or triplets or quadruplets.
{And in those cases tubes must be replaced more often}
From a DIYaudio point of vue more than 2 will become instable
and will need potentiometers in each cathode to adjust bias current
for each tube ; otherwise distortion will rise quickly.
One other solution would be to have 1 primary for each tube
and there the transformer will be much more costly...
So going after more than 2 tubes in parallel in SE mode is not douable
for you! with 6550's. It's why so few designs exists!
One Italian designer did a 3 * 6550 SE but uses custom made automatic
bias adjustment, and special opt's ...
A design has been published by Ari Polisois in AudioXpress last year using
3 parallel 6C33C-B tubes in SE configuration to get 45 watts ;
but it's not a simple project! {custom SE opt's ; big heat dissipation
and a bigger PS }
Further to get 40 watts pure class A in SE you need a tube {'s} with a plate
dissipation of more than 140 watts ; and with low internal impedance.
And a costly SE opt to get the best of this mode. Very few tubes with those
type of parameters exist!
Regards.
Alain.
Alain,
thanks again.
I would love to see that italian guys Schematic.... some how.
I wonder how much it cost me to build that Italian design.
I don;t think I can get the audio express magazine, anyone ?
cal
thanks again.
I would love to see that italian guys Schematic.... some how.
I wonder how much it cost me to build that Italian design.
I don;t think I can get the audio express magazine, anyone ?
cal
Calico88.
Here they are:
The first design SE with 6C33C-B from Ari Polisois:
http://www.valvediy.com/simplexpg1.html
The last article from AudioXpress was in:
July 2004 revue: High power SE 6C33C Amp
3 * 6C33C-B by channel. I still have the revue...
The only difference are 3 * 50 Ohms pots paralled with a 100 Ohms resistors in each
cathode, and a 1 Ohm for current measure ; the first stage and PS are similar.
Take a look at my web site for a "cute" integrated single tube version!
The Unison Research "Performance"
http://www.unisonresearch.com/valve/prodotto.aspx?idp=17
But it's a SE Ultra linear design...
Regards.
Alain.
Here they are:
The first design SE with 6C33C-B from Ari Polisois:
http://www.valvediy.com/simplexpg1.html
The last article from AudioXpress was in:
July 2004 revue: High power SE 6C33C Amp
3 * 6C33C-B by channel. I still have the revue...
The only difference are 3 * 50 Ohms pots paralled with a 100 Ohms resistors in each
cathode, and a 1 Ohm for current measure ; the first stage and PS are similar.
Take a look at my web site for a "cute" integrated single tube version!
The Unison Research "Performance"
http://www.unisonresearch.com/valve/prodotto.aspx?idp=17
But it's a SE Ultra linear design...
Regards.
Alain.
Calico, I hesitated, at first, to suggest you look at Koren's design because it is anything but simple.
Koren uses parallel push-pull 6550s, in triode-mode class AB2 PP, to get a c laimed 80 watts RMS. Most of the complexity is in the power supply, caused by his extensive use of SS control and regulation circuits. However, you could probably still get 40w or more in class AB1, which would avoid the need for cf drivers. His automatic bias is a luxury; manually adjusted bias could do the job just as well.
Koren uses parallel push-pull 6550s, in triode-mode class AB2 PP, to get a c laimed 80 watts RMS. Most of the complexity is in the power supply, caused by his extensive use of SS control and regulation circuits. However, you could probably still get 40w or more in class AB1, which would avoid the need for cf drivers. His automatic bias is a luxury; manually adjusted bias could do the job just as well.
alain,
where is the last article using 3x 6550 ?
I built 300B SE, and I love SE, but it's time for me to put more juice.
can I see the 3x6550 Schematic ? I wonder how much that cost
Ray,
that korean guy is the best article DIY I ever seen.
very detail of how to do it. how you found it ?
have you built it ?
too bad for me it's too complex for my knowledge, but that's what I need ... in term of power.
cal
where is the last article using 3x 6550 ?
I built 300B SE, and I love SE, but it's time for me to put more juice.
can I see the 3x6550 Schematic ? I wonder how much that cost
Ray,
that korean guy is the best article DIY I ever seen.
very detail of how to do it. how you found it ?
have you built it ?
too bad for me it's too complex for my knowledge, but that's what I need ... in term of power.
cal
The guy's name is Alan Koren. He isn't Korean! I came across it ages ago, when browsing. Sure, it's complicated but it needn't be. There's no need to have such a complex power supply and, as mentioned before, if you don't try to drive it into class AB2 there is no need for cathode follower drivers, eliminating 2 6SN7 tubes.
Note: the driver tubes are marked 12AU7 on his schematice but he explains that he actually used 6SN7 tubes, because of the high plate-cathode voltage needed.
Note: the driver tubes are marked 12AU7 on his schematice but he explains that he actually used 6SN7 tubes, because of the high plate-cathode voltage needed.
ray_moth said:
can you explain more detail on this ?
which 2 6SN7 to be eliminate ?
how much power I will get if I do this ? same ?
cal
Without diverging into a long discussion about different classes of operation, what is at issue in this particular case is whether to use class AB1 or class AB2 for the 6550s in triode mode. Koren explains something about this in his article, including the fact that quite a bit of extra power (almost double) is available from triodes in AB2. He reckons he can get 80w output. In class AB1. he could probably get a bit over 40w.
The most important difference between AB1 and AB2 is that AB2 requires the grids to be driven positive for part of the cycle. This means that their input impedance will drop from several hundred kilohms to about 1 kilohm, every time the grid voltage goes positive. The driver needs to be able to supply this current without distortion and cannot be capacitor-coupled to the 6550. Koren chose a direct-coupled cathode follower using 6SN7 to do the job. He could have used a transformer or, nowadays, a suitable MOFET source-follower instead.
To get 40 watts from 6550s in triode-mode, you could either:
(a) use just one pair of PP tubes in AB2, which would mean that you would still need a current-supplying driver of some sort; or
(b) use two pairs of 6550 tubes in parallel PP in AB1, which would considerably simplify the driver requirements. In AB1, you could do away with the existing direct-coupled 6SN7 cathode followers and couple the preceding 6SN7 differential amplifiers to the 6550 grids using coupling capacitors if the 6SN7 could drive the significant Miller capacitance of parallel triode-strapped 6550s. It might be advisable to retain one of the 6SN7 cathode-followers to drive a pair of 6550s but capacitor-coupled to simplify the bias arrangements.
Either option would require considerable redesign, especially option (b), but each option offers the chance to simplify things. The output transformer would need to be changed with option (a), because with only two tubes in PP the plate-to-plate impedance is doubled. Plitron might have a suitable transformer, I don't know, but Pitron is an expensive choice anyway and other quality brands are available at lower prices.
Thinking more about it, I suspect I would rather have option (a), one pair of 6550s in AB2. With only one pair to worry about, there is no real need to go for automatic bias, unless you sepcially want to. Koren's motivation for this was mostly due to the fact that he was using a toroidal OPT and couldn't afford to let the currents of the four 6550s get out of balance, for fear of having a significant net DC in the OPT primary and saturating its core. Toroids are very sensitive to this.
Frankly, I wouldn't start with Koren's design to build a 40w amp, because I think that would be doing things the hard way. I'd rather start from scratch or find another, simpler model on which to base my amp. That's why I hesitated to mention it to you before.
Of course, you could decide to build exactly what Koren built, in which case you'll need a lot of money, time and patience. The end result would be worth it, I'm sure, but only if you are up to the challenge. BTW, you need to be especially careful handling MOSFETs because they can be destroyed by an electrostatic charge from your body.
The most important difference between AB1 and AB2 is that AB2 requires the grids to be driven positive for part of the cycle. This means that their input impedance will drop from several hundred kilohms to about 1 kilohm, every time the grid voltage goes positive. The driver needs to be able to supply this current without distortion and cannot be capacitor-coupled to the 6550. Koren chose a direct-coupled cathode follower using 6SN7 to do the job. He could have used a transformer or, nowadays, a suitable MOFET source-follower instead.
To get 40 watts from 6550s in triode-mode, you could either:
(a) use just one pair of PP tubes in AB2, which would mean that you would still need a current-supplying driver of some sort; or
(b) use two pairs of 6550 tubes in parallel PP in AB1, which would considerably simplify the driver requirements. In AB1, you could do away with the existing direct-coupled 6SN7 cathode followers and couple the preceding 6SN7 differential amplifiers to the 6550 grids using coupling capacitors if the 6SN7 could drive the significant Miller capacitance of parallel triode-strapped 6550s. It might be advisable to retain one of the 6SN7 cathode-followers to drive a pair of 6550s but capacitor-coupled to simplify the bias arrangements.
Either option would require considerable redesign, especially option (b), but each option offers the chance to simplify things. The output transformer would need to be changed with option (a), because with only two tubes in PP the plate-to-plate impedance is doubled. Plitron might have a suitable transformer, I don't know, but Pitron is an expensive choice anyway and other quality brands are available at lower prices.
Thinking more about it, I suspect I would rather have option (a), one pair of 6550s in AB2. With only one pair to worry about, there is no real need to go for automatic bias, unless you sepcially want to. Koren's motivation for this was mostly due to the fact that he was using a toroidal OPT and couldn't afford to let the currents of the four 6550s get out of balance, for fear of having a significant net DC in the OPT primary and saturating its core. Toroids are very sensitive to this.
Frankly, I wouldn't start with Koren's design to build a 40w amp, because I think that would be doing things the hard way. I'd rather start from scratch or find another, simpler model on which to base my amp. That's why I hesitated to mention it to you before.
Of course, you could decide to build exactly what Koren built, in which case you'll need a lot of money, time and patience. The end result would be worth it, I'm sure, but only if you are up to the challenge. BTW, you need to be especially careful handling MOSFETs because they can be destroyed by an electrostatic charge from your body.
andRookie said:
To me the distortion is way too high and only measured at 10W ..this is very misleading and why isn't it done at industry standard i.e, a dB below full power ? Everyone knows the figure will be way higher but :-
Can anyone give me a viable explanation why this isn't done ?
richj
That's SE for you! Most of that will be 2nd harmonic, giving it a 'sweet' tone, which is what SE is noted for. Such THD figures are misleading and it would be much more informative to give distortion per harmonic.
As to why the distortion is expressed as a %, it was traditional to do so. A typical Fourier analysis would more likely show it in dB, I agree.
In the 3X6550 project there are some differences in the distortion pattern between triode and UL mode. In triode mode HD is mainly 2nd harmonic up to the clipping, at 10W HD is 0.6% and at 20W+ HD is ~1.5% so HD is more or less proportional to power output and the dominant harmonic is second, all the way (with my equipment/analyzer it is difficult to measure distortion less than 65-70dB).
In UL the distortion pattern is more complex. At 10W HD is less than in triode mode, about 0.25% and 2nd harmonic is dominant. At 20W there are still more 2nd harmonic than 3rd harmonic, but 3rd harmonic starts to increase. Total HD at 20W is ~1%. At 40W+ 2nd and 3rd harmonic are about equal and total HD=2-3%.
Misleading or not, I’m not manufacturing amps for other than my self. 😉
Jan E Veiset
In UL the distortion pattern is more complex. At 10W HD is less than in triode mode, about 0.25% and 2nd harmonic is dominant. At 20W there are still more 2nd harmonic than 3rd harmonic, but 3rd harmonic starts to increase. Total HD at 20W is ~1%. At 40W+ 2nd and 3rd harmonic are about equal and total HD=2-3%.
Misleading or not, I’m not manufacturing amps for other than my self. 😉
Jan E Veiset
- Status
- Not open for further replies.