You can either pair up speakers or tubes to get a certain acoustical power out. At low frequencies speakers pay off much more. In that respect, the idea of a array of speakers IS a good one. If impedance and damping are the only reasons to use so many speakers and tubes may I suggest to go even further than that. Use instead of the choke a "tapped" choke (a autotransformer) of suitable design. For example, with centertap it transforms your impedance fourfold and transformed power gets cut in half. You could use a quarter of the tubes you intended. A reduced amount of paralelled tubes helps also with the dc-current load in the choke. If you need really 60-80W, go "circlotron" with autotransformer. With two 6C33 you will have plenty of power and dc is taken almost completely out of the equation.
Last edited:
OTL is after the circlotron or McIntosh the next best thing .
All of them can/or can not (depends) be improved by a suitable autotransformer
All of them can/or can not (depends) be improved by a suitable autotransformer
cap and transformer when fed from low impedance, is that a good idea?
I have a plitron toroid, even the sligthest amount of dc kills the otherwise excellent qualities. I thougth about Cap coupling to get rid of dc, but....
Anatoly, SRPP or white CF woud be nice if only the load would not wander all over the place as it usually does
I have a plitron toroid, even the sligthest amount of dc kills the otherwise excellent qualities. I thougth about Cap coupling to get rid of dc, but....
Anatoly, SRPP or white CF woud be nice if only the load would not wander all over the place as it usually does
it appears that an OTL amplifier will output more power into higher Z loads.
This is like an SE amp. Bass response will improve as well as treble response
This is like an SE amp. Bass response will improve as well as treble response
Anatoly, a SRPP has max output and minimal distortion only with one specific load impedance.
Feedback will not change that fact, nor will it do anything to the "all over the place" load impedance, it can only correct the loads influence on outputvoltage.
Feedback or not, the SRPP has to be optimized for the lowest load impedance.
When this impedance increases distortion will increase.
This is quite contrary to what happens in PP class A amplifiers wich do drive the grids indepedendly and in wich case the distortion goes DOWN, NOT UP as in the SRPP.
Anyway, to get low distortion and decent damping (in SRPP even more so because optimized SRPP dictates heavy loading and optimized load impedance will approach Rplate) you propable will always want to apply nfb.
Do you view this differently?
Last edited:
most OTL work on the brink of theyr current capabilities.
Imagine a car driving uphill in 5th gear.
A transformer is like your gearbox.
^that is why in my OTL amp, i used a power traffo with a core capacity of over 1000 volt amperes in an amp that only delivered 25 watts per channel..
Yes, I view it differently. Nothing is perfect, the best results are optimal, so one more stage before SRPP, and a feedback around them would minimize distortions.
SE amp is a different pair of shoes, I would not compare them.
What will improve with for the tubes more suitable higher Zload is
efficiancy, tube lifespan, reliability, output power.
BW, distortion and damping depend on the OTL in question and the way Zload is made higher.
You gain some, loose some, no free lunch
Thats the way to go (unless you need really big power that can only be obtained by using an insane amount of tubes)
Anatoly, you rather take 1 more stage and fb to straigthen crooked things out than by using 1 more stage to drive both grids, get more power, higher damping, lower distortion?
I prefere to do it differently, drive both grids and use fb only to improve things that cannot be improved otherwise.
Transcendent 6C33C
Agree with leadfinger.
I built the amps laid out in his patent. 4x6C33C in each mono block make 110W per channel. Easily drives 4, 8 or 16 ohms speakers. 16 is a great match for these amps. I use vintage alnico Coral full range 16 ohm drivers. Your speakers may be interesting, but not required.
Excellent sound. Totally reliable. I have enjoyed them now for years with no issues.
There is quite a following for Transcendent amps because they are very good and very reliable. Most of the concerns expressed in this thread regarding OTL does not apply due to his patented design.
Agree with leadfinger.
I built the amps laid out in his patent. 4x6C33C in each mono block make 110W per channel. Easily drives 4, 8 or 16 ohms speakers. 16 is a great match for these amps. I use vintage alnico Coral full range 16 ohm drivers. Your speakers may be interesting, but not required.
Excellent sound. Totally reliable. I have enjoyed them now for years with no issues.
There is quite a following for Transcendent amps because they are very good and very reliable. Most of the concerns expressed in this thread regarding OTL does not apply due to his patented design.
Do you thinck a patent can overcome boundries layed out by physics? 110W into 4 ohm is 7,4A peak current! That would be in your case 3,7A per tube! If that is of no concern to you then I can only conclude that you lack understanding of what is going on in the tube.
16 ohm is much more doable but a different story and the fact still remains that you need huge NFB to straigthen out the mess created by loading the tubes in a most unfavourable way.
The only reason that this amplifier would be able to work for years into 4 ohms can only be fact that it never even got close to 110W
16 ohm is much more doable but a different story and the fact still remains that you need huge NFB to straigthen out the mess created by loading the tubes in a most unfavourable way.
The only reason that this amplifier would be able to work for years into 4 ohms can only be fact that it never even got close to 110W
Last edited:
Douglas Self has an article from Wireless World, February 1957, scanned on his site: The Wireless World Archive
Wow, thanck you very much Anatoly, a real treasure anyone should dig in.
I grew up reading WW and Funkschau but saved only WW past 73.
This link would deserve a special place so every tubehead can easely find it
I grew up reading WW and Funkschau but saved only WW past 73.
This link would deserve a special place so every tubehead can easely find it
No, I was trying to suggest it's a magical audiophile device that runs on water and cures hearing disability. 🙂
Actually, I did not mean to suggest it makes 110W into 4ohms. I believe its 110 at 8. The point is it is not flea powered and makes great music while driving a wide range of speakers likely to be found in reader's homes. It will not drive 2ohms but it does not need exotic speakers to produce bass and it will not self destruct or fry your speakers. Heat output is 350W per channel which is 1/2 typical OTL of similar power. Transcendent customers are a pretty happy bunch. The fact that the design is available to the DIY community at a nominal fee is potentially useful to some readers.
Gorgon53, you bring up a great point. How can the tube ( 40w plate ) dissapate the 100's of watts of heat generated when driving speakers to even 2 -3 amps peak with 150v across each tube.
A SS amplifier would melt down and everyone would say it's a terrible design guaranteed to fail.
A SS amplifier would melt down and everyone would say it's a terrible design guaranteed to fail.
Stocktrader, other than a SS a tube can take severe beating but within limits offcourse.
The anodes thermal capacity may be ok with short overloading, but things are different at the cathode, it will get stripped once the electrodecloud is depleted and 3,7A are certainly out of the question. One migth get away with maybe 2A or so for some short peaks. But, as fare as I know, Rozenblit drives the grids RC -coupled and this migth safe the tubes because anything even close to 3,7A wont happen.
The anodes thermal capacity may be ok with short overloading, but things are different at the cathode, it will get stripped once the electrodecloud is depleted and 3,7A are certainly out of the question. One migth get away with maybe 2A or so for some short peaks. But, as fare as I know, Rozenblit drives the grids RC -coupled and this migth safe the tubes because anything even close to 3,7A wont happen.
Last edited:
It's not quite 100's of watts. An OTL with two 6C33C output tubes will reach about 25 watts into 8 ohms, which means a peak current of 2.5 amps. A given output tube only passes current for one half of the audio cycle. If I have my sums right, it works out as about 107 watts average dissipation per tube for a sine wave played at 25 watts.
And in practice, one doesn't listen to a sine wave at full volume for an extended period. Typically, with music as opposed to a full-volume sine wave, the average power dissipations will be a lot less than the 107 watts, and probably comfortably within the 60 watt dissipation limit for the tube even if the music is played loud.
Using four 6C33C tubes instead of two should roughly quadruple the output power capability, so about 100 watts into 8 ohms. The calculation of the average power dissipated per tube would be about the same as above, namely about 107 watts per tube when delivering the full 100 watts into 8 ohms (actually somewhat less in this case, more like 94 watts, because more of the 150 volts is dropped across the 8 ohm load during the half cycle of conduction). Again, much less than that with typical music, even when at very high levels.
Last edited:
- Status
- Not open for further replies.