skyraider said:
I'm not an EE either.
Class *2 operation is where the valve is operated with positive grid bias for some or all of the cycle. This means that the control grid will draw some current, which ideally does not happen when the grid is biased in class *1 (negative grid bias). Class *2 operation allows a greater operating range, hence greater efficiency for most valves. For high-µ lazy transmitter types (say, 811s) it's the only way to go.
If you want to read more about this, there is quite a detailed explanation of this on Steve Bench's website, along with full schematics for class A2 841 and 572-160 SE amps (albeit with valve driver stages)
I thought most audiophiles hated opamps because they're very complicated internally and utilise oodles of feedback 😕skyraider said:
Most traditional power triodes are relatively low µ, so they'll need a quite a lot of grid swing to work nicely. There may be some exceptions, but I can't think of any off the top of my head right now...skyraider said:
Consulting with my local trans winder, he proposed 2k:50k interstage. The only thing I dont understand is how come he can come up with the Av=5.......Isnt the number of wire turns are porpotioned to the impedance?
Anyway, based on his design, this is what I have in mind. To be frank, I have good experience with opa627 and several opamps before this, and I feel they should be given a chance to drive tubes directly.
But this design has limitation to the quality of the interstage trans.
Anyway, based on his design, this is what I have in mind. To be frank, I have good experience with opa627 and several opamps before this, and I feel they should be given a chance to drive tubes directly.
But this design has limitation to the quality of the interstage trans.
Attachments
ah.. the miseries of having golden ears! Yes, many people believe opamp are not for serious listener... but the last thing I want is to turn this thread into silly 'opamp vs tube' debate.
Lets just say the thread is only open for people who can accept opamps. ok... ok.. I know Im in the wrong place😀 😀 😀
Unless there is 'proper' justification why this design should not be done, I am open for feedback.
thanks. More reading to do..
I dont understand how 2k:50k interstage transform into Av=5... This is the figure given by my trans winder guy.
Anyone please explain?
Anyone please explain?
skyraider said:
The impedance ratio is the square of the turns ratio. In this case, the impedance ratio is 50K/2K=25, hence the turns ratio is the square root of that, 5. This is conveniently the same as the voltage ratio, so the transformer steps up the voltage by a factor of 5 (note that it will step down the current by a factor of 5 as well)
Skyraider,
What you have is the impedance ratio. Impedance ratio is the square of the turns ratio.
So 2K/50 K impedance ratio is 1:25 and the turns ratio is the square root of that 1:5 .
Voltage ratio is the same as "turns ratio" 1:5.
So from the smaller winding to the larger winding the voltage gain is 5.
Cheers,
Ashok.
What you have is the impedance ratio. Impedance ratio is the square of the turns ratio.
So 2K/50 K impedance ratio is 1:25 and the turns ratio is the square root of that 1:5 .
Voltage ratio is the same as "turns ratio" 1:5.
So from the smaller winding to the larger winding the voltage gain is 5.
Cheers,
Ashok.
Colin
You can omit all the items in red. However 2a3 grid stopper may be nessecary. But i never found a need for them in my 2a3 n 300b amps
Regarding the question about the capacitance. I am not sure until i wind the trans and work it out. What you should be more worried is the capacitance on the secondary reflected. That could cause some distress to the opamp if iit is not loaded with enough current. However i recommend to use 15ma for the opamp and you are home free. Considering you are using buffer stage, your current drive is high with lower output impedance
Therefore driving the 2k trans should be not a problem
Someone beat me to the explanation of the gain ratio
Hope to hear from your requirement of the trans then
CHeers
Nicholas
You can omit all the items in red. However 2a3 grid stopper may be nessecary. But i never found a need for them in my 2a3 n 300b amps
Regarding the question about the capacitance. I am not sure until i wind the trans and work it out. What you should be more worried is the capacitance on the secondary reflected. That could cause some distress to the opamp if iit is not loaded with enough current. However i recommend to use 15ma for the opamp and you are home free. Considering you are using buffer stage, your current drive is high with lower output impedance
Therefore driving the 2k trans should be not a problem
Someone beat me to the explanation of the gain ratio
Hope to hear from your requirement of the trans then
CHeers
Nicholas
Leave the grid stopper there ( close to the grid ) and the input resistor to ground will only be useful if you plan to remove the trafo for any reason and leave the grid open.
The resistor from plus input terminal of the input stage to ground is useful . While operating the volume pot , if any small discontinuity occurs, the output dc will change causing scratch noises at the least. With high quality pots this problem will be minimal. You could avoid that resistor - one component less that might color the sound !
I'm not sure about the resistors at the output of the BUF634. The resistor in series and shunt with the transformer will surely have some affect on frequency response but that depends on the circuit following that and the behaviour of the transformer.
If you have some recommended values with a specified transformer , go ahead and use it.
Cheers.
The resistor from plus input terminal of the input stage to ground is useful . While operating the volume pot , if any small discontinuity occurs, the output dc will change causing scratch noises at the least. With high quality pots this problem will be minimal. You could avoid that resistor - one component less that might color the sound !
I'm not sure about the resistors at the output of the BUF634. The resistor in series and shunt with the transformer will surely have some affect on frequency response but that depends on the circuit following that and the behaviour of the transformer.
If you have some recommended values with a specified transformer , go ahead and use it.
Cheers.
Sorry I've been out of this thread a while...
Well anyway, let me get this straight. Help me out here in case I get something wrong.
You DON'T like coupling capacitors.
You DO approve of op-amps. (Nothing wrong with that, but highly impractical here.)
BUT! As op-amps are slightly hard to find in high-voltage models at the local RadioShank, you find a need to step up the output voltage.
So, you use an INTERSTAGE TRANSFORMER.
So umm...are you even the slightest bit aware what an interstage transformer does to an electric signal?? Anyone??...
Tim
Well anyway, let me get this straight. Help me out here in case I get something wrong.
You DON'T like coupling capacitors.
You DO approve of op-amps. (Nothing wrong with that, but highly impractical here.)
BUT! As op-amps are slightly hard to find in high-voltage models at the local RadioShank, you find a need to step up the output voltage.
So, you use an INTERSTAGE TRANSFORMER.
So umm...are you even the slightest bit aware what an interstage transformer does to an electric signal?? Anyone??...
Tim
Quote:
So umm...are you even the slightest bit aware what an interstage transformer does to an electric signal?? Anyone??...
Tim
Tim, please tell us won't you?
So umm...are you even the slightest bit aware what an interstage transformer does to an electric signal?? Anyone??...
Tim
Tim, please tell us won't you?
burnedfingers said:
See attached.
It gets uglier the more precise you want it.
Who knows what the complex combination is on for example a multiwinding one (CT'd or multisection primary or secondary). I know for example Hept'AU7's has a 90 degree end-to-end phase shift at some frequency, I forget what (10 or 30kHz maybe?).
Tim
Attachments
I have to agree - this is the looniest tube thread ever. Someone who does not understand at all what transforemers do wants to create a new topology in which a transformer does most of the work. And to what end? To create a tube amp blessed with the sonic flavour of an opamp.
Besides the rant i am generally in favour of what interstage transformers do, if not to the electrical signal, then at least to reproduced music.
BUt only: if these transformers are of very high quality and if they are step-down or 1:1. The low driving impedance of an opamp probably makes designing such a transformer easier but certainly not trivial.
skyraider said:
I'm running a (PP version of something like this right now using an AD815/AD823 hybrid to drive it. The cheap EDCOR interstage is probably the biggest limitation.
I never did like the idea when colin proposed to me to get me to wind the 1:5 trans. But again sometimes its it good to challenge the norm. In the end game it is colin that will decide whether it is good or not.
I orginally proposed the 6c45 it to 2a3.
When he proposed the 1 to 5 idea, i dislike the ide but i knew he was serious with it. So instead of just critizing him i turn the energy into something more productive like how to make the trans. I suspect if he included a feedback loop around the interstage that would help the square waveform
cheers
nicholas
I orginally proposed the 6c45 it to 2a3.
When he proposed the 1 to 5 idea, i dislike the ide but i knew he was serious with it. So instead of just critizing him i turn the energy into something more productive like how to make the trans. I suspect if he included a feedback loop around the interstage that would help the square waveform
cheers
nicholas
scott wurcer said:
Yeah, that got a major chuckle from my input transformer guru.
Out of curiosity, have you measured the frequency response of the total amp?
SY said:
It seemed to follow what you would expect from the triode connected EL34's and the output tranny. I might get around to looking at the whole picture someday, but hey this is all for fun afterall.
BTW I just watched the Hokkaido oysters episode today Sakai-san won hands down. You do miss out, not partaking of oysters (a wink is as good as a nudge to a blind man, you know what I mean).
Plan B
If you really want to drive output tubes with an op-amp then try this instead:
1) Ground the Output Tube Grid
2) Put a darlington between cathode and 0V with a small (say 10R) resistor on the 0V side
3) wire the opamp to drive the darlington with the -ve input strapped back to the top of the 10R resistor.
There its done - no need for high voltage op-amps or any of that stuff. Its a variable current source which follows the +ve input. Just apply the audio signal to the +ve input with a bit of DC on it to give the required bias current in the tube
Now as to whether you'll get a valve sound from it - thats another kettle of fish - I'd doubt it.
I use something very similar with parallel power MOSFETS to drive 100A 100us wide controlled edge speed current pulses into Laser Diode arrays.
Cheers,
Ian
Tech Note: Why a Darlington and not a power FET? The Beta of a darlington increases with temperature and becomes easier to drive. The gate current of a FET increases with temperature and becomes more difficult to drive - above about 85 to 90 degrees C the darlington will be easier to drive than a FET.
If you really want to drive output tubes with an op-amp then try this instead:
1) Ground the Output Tube Grid
2) Put a darlington between cathode and 0V with a small (say 10R) resistor on the 0V side
3) wire the opamp to drive the darlington with the -ve input strapped back to the top of the 10R resistor.
There its done - no need for high voltage op-amps or any of that stuff. Its a variable current source which follows the +ve input. Just apply the audio signal to the +ve input with a bit of DC on it to give the required bias current in the tube
Now as to whether you'll get a valve sound from it - thats another kettle of fish - I'd doubt it.
I use something very similar with parallel power MOSFETS to drive 100A 100us wide controlled edge speed current pulses into Laser Diode arrays.
Cheers,
Ian
Tech Note: Why a Darlington and not a power FET? The Beta of a darlington increases with temperature and becomes easier to drive. The gate current of a FET increases with temperature and becomes more difficult to drive - above about 85 to 90 degrees C the darlington will be easier to drive than a FET.
Not all of them, only the expensive ones
yes (and.. yess)
Finally someone who understands me!!!😀 😀 😀
Back to the trans, I have a good trans winder here. so that solves a part of my problem.
I know I will have problem with leak inductance and parasitic capacitance.
But consider that transformer have been used in the input of many amps without problems???? More to say on interstage..
I know step-up trans driven by miniscule opamp is a different case, but lets see if any practical engineering skills can produce better results. The buffer can cope with 250mA output current, I believe this does provide room to work with.
- Status
- Not open for further replies.