A while ago ago i had an idea - an amplifier that uses solid state devices for current gain and transformers for voltage gain.
there would be ZERO voltage gain from any active device. all the voltage gain would come only from transformers.
theoretically such an amp could be fairly linear without any feedback - no ?
what i was told then ( at partsexpress.com ) is that there is no need for this because you can simply use an op-amp to get all the clean gain you want. i am hoping on this forum i will get a more intelligent answer 🙂
there would be no input or output transformers by the way. all the transformers would be internally used for gain.
there would be ZERO voltage gain from any active device. all the voltage gain would come only from transformers.
theoretically such an amp could be fairly linear without any feedback - no ?
what i was told then ( at partsexpress.com ) is that there is no need for this because you can simply use an op-amp to get all the clean gain you want. i am hoping on this forum i will get a more intelligent answer 🙂
there would be no input or output transformers by the way. all the transformers would be internally used for gain.
This has allready been done. Google for Susan Parker + Zeus amp. There is also a long thread in this subject at Solid State.
i guess this is it:
http://www.diyaudio.com/forums/showthread.php?t=42259
her post also precedes my join date on partsexpress.
good work.
http://www.diyaudio.com/forums/showthread.php?t=42259
her post also precedes my join date on partsexpress.
good work.
actually i take it back - her design is different.
she uses transformers at input and output and transistors in the middle.
i was proposing the use of transistors at input and output and transformer in the middle.
she uses transformers at input and output and transistors in the middle.
i was proposing the use of transistors at input and output and transformer in the middle.
Good idea; let's see your proposed circuit.
(You might want to transfer this thread to the Solid State part of the forum).
i don't know how to build or design amplifiers 🙂
as far as moving the thread i don't think i can do it ? mods feel free though.
If you can control what the driving source is in terms of output impedance, maximum output voltage and DC offset, you can simplify it to a transformer at the input and transistors at the output.
se
Why not try placing say a 10:1 transformer in front of a First Watt F4 (PCBs are available). If it doesn't work, well then at least you'll have an F4.
-j
-j
The only issues I have with this idea is that transformers generally work better when stepping down. Stepping up and capacitance comes into play and kills the high frequency response. Also transformers will tend to ring which you need a good signal generator and a scope to fault find. These problems may not be as great with solid state devices rather than valves due to their much lower output impedance, but miller capacitances on Solid state tend to be greater than for valves.
Shoog
Shoog
How much high frequency response do you need?
Is 100kHz not sufficient?
The better manufacturers don't force you to go finding it and provide information on the proper loading of their transformers.
se
I think what you are talking about is coupling two amp stage with a transformer, a "transformer coupled amp". These have been build but the problem is that there is a cheaper way to get better quality.
Always you have to think about budget. It doesn't matter if you have $10,000 or $10 to spend, you always want the best that your budget can buy. I think if you were to cost out the parts yo'd find the transformer used to couple the stages costs 10X or more then capastor coupling
Also you have to remember that you would have to use a "real" transformer and not the theoretical transformer. The real transformer will have resistance and capacitance and some kind of frequency response. You could make all of this work very well but the question is still if you couldn't get better sound for the same cost using a different design
IMO the best use for transformer coupling is in radios because (1) you can take advantage of the limited frequency response and use the transformer as a filter and (2) the transformers designed for those frequencies are small and much lower cost. In radios you see transformer use for voltage gain all the time. It's very common there but not cost effective for audio.
I guess it depends on how one defines "quality."
Also depends on how one defines "best."
And whether one is mostly interested in looking at numbers, or enjoying listening to reproduced music.
Not aware of any capacitors that provide voltage gain.
As will any other circuit.
Your point?
And how exactly can this question ever be answered without first having tried the transformer approach?
Just as you do in professional microphone preamps.
se
Borat,
First, let me point you to someone who is experimenting with SS devices and transformers. I only do this because I have heard the amp, through high resolution speakers and it is superb.
http://www.pimmlabs.com/
You should look at the Tabor SS. Though this is not his latest topology, it will show you where his thoughts have traveled. The transformers involved in his latest iteration are not cheap. For the interstage, nickle core and 1:1 PP, $280 ea. For the OPT level 3 and $560ea. Quite obviously these prices are for the very best parts available.
Sonically, the equivalent interstage capacitor would be a reliacap / multicap tin & poly PPFX type at $31.50 for 1uf. The only difference between the sound of the cap and transformer has to do with perception of the sound field and is true of these two types of devices at all price points.
With a capacitor as interstage, when a single instrument that is embedded in a sound field illusion plays much louder than surrounding instruments, it blooms out from that illusion into a larger and seemingly much closer instrument.
With a transformer this bloom does not occur, the illusion remains stable and the instrument just plays louder.
These characteristics will hold true for groups of instruments also.
In a general sense the audio quality will be very similar until you get into large orchestral groups, played through high resolution speakers that are very efficient (98 db or more).
In my own system I deliberately use a PPFX capacitor in the interstage portion of my preamp and a nickle core OPT, for a SE cathode coupled balanced output, to the grids of the amplifier input. The amplifier is all transformer coupled. I find this to be the best balance, with the capacitor providing just a small amount of "extra" 3 dimensional illusion.
Chris A's comments about step up are generally correct and though you can drop the internal capacitive losses in the transformer to offset the miller effect for the following transistors, you would spend a bit of money getting the balance correct for the transformer. When you begin to tighten up the ratio between coupling capacitance and distributed capacitance, thereby cutting the distributed capacitance in the transformer, you must match the load capacitance from the driven circuit to that lost distributed capacitance, in the transformer, or you will have a hi Q resonant peak from the transformer, at around 35k and from 3 to 24 dB.
Bud
First, let me point you to someone who is experimenting with SS devices and transformers. I only do this because I have heard the amp, through high resolution speakers and it is superb.
http://www.pimmlabs.com/
You should look at the Tabor SS. Though this is not his latest topology, it will show you where his thoughts have traveled. The transformers involved in his latest iteration are not cheap. For the interstage, nickle core and 1:1 PP, $280 ea. For the OPT level 3 and $560ea. Quite obviously these prices are for the very best parts available.
Sonically, the equivalent interstage capacitor would be a reliacap / multicap tin & poly PPFX type at $31.50 for 1uf. The only difference between the sound of the cap and transformer has to do with perception of the sound field and is true of these two types of devices at all price points.
With a capacitor as interstage, when a single instrument that is embedded in a sound field illusion plays much louder than surrounding instruments, it blooms out from that illusion into a larger and seemingly much closer instrument.
With a transformer this bloom does not occur, the illusion remains stable and the instrument just plays louder.
These characteristics will hold true for groups of instruments also.
In a general sense the audio quality will be very similar until you get into large orchestral groups, played through high resolution speakers that are very efficient (98 db or more).
In my own system I deliberately use a PPFX capacitor in the interstage portion of my preamp and a nickle core OPT, for a SE cathode coupled balanced output, to the grids of the amplifier input. The amplifier is all transformer coupled. I find this to be the best balance, with the capacitor providing just a small amount of "extra" 3 dimensional illusion.
Chris A's comments about step up are generally correct and though you can drop the internal capacitive losses in the transformer to offset the miller effect for the following transistors, you would spend a bit of money getting the balance correct for the transformer. When you begin to tighten up the ratio between coupling capacitance and distributed capacitance, thereby cutting the distributed capacitance in the transformer, you must match the load capacitance from the driven circuit to that lost distributed capacitance, in the transformer, or you will have a hi Q resonant peak from the transformer, at around 35k and from 3 to 24 dB.
Bud
You're pretty much talking apples and oranges here.
How you can compare big *** PP interstage coupling transformers in a tube circuit where the active devices are providing the voltage gain to what the original poster was talking about?
There would be no miller effect as the "following transistors" would not be used as voltage amplifiers.
Again, you're comparing apples and oranges here.
se
Gary used the same transformers for the output of both the tube and solid state Tabor amps. The tube units used relatively large Lundahl interstage units. The Solid State amp has used smaller inductance units for the IT and I am pretty sure he has also tried a direct coupled scheme.
I agree this is basically an apples and oranges expression in hardware. However the usefulness of transformers with solid state devices is usually overlooked and I wanted to point to a successful venture, to encourage the out of box thinking. I must admit to surprise when confronted with how much more musical and transparent the use of transformers with SS is, over the direct drive from SS device to speaker.
Gary's test data speaks for the quantitative side of this combination.
Bud
About 4 or 5 years ago, i had this idea because the voltage gain from a transfomer 1:10 turn ratio (i.e.) is most linear than you can give from best linear triode. So i tought about an amp with voltage gain obtained from step up transformer and current gain from active devices.
Then, when i have discussed about this project with Mr. Imai, hi told me the following:
-for a good drive, the active devices must have 1/8 times output impedance than a transfomer at least
-because the active device must be empolyed as current amplifier, the output signal must be take from source (or cathode) as in follower schematic, so we have 100% feedback and this is not good for the sound.
Despite this answer, i have always though to build ones, but i not have made yet (Hope in the next).
Best regards, Francesco.
Then, when i have discussed about this project with Mr. Imai, hi told me the following:
-for a good drive, the active devices must have 1/8 times output impedance than a transfomer at least
-because the active device must be empolyed as current amplifier, the output signal must be take from source (or cathode) as in follower schematic, so we have 100% feedback and this is not good for the sound.
Despite this answer, i have always though to build ones, but i not have made yet (Hope in the next).
Best regards, Francesco.
Last edited:
Fair 'nuff.
And I agree regarding transformers and solid state. Though personally I prefer doing it inside out though compared to what Gary's doing.
I don't have a problem directly driving the speakers with solid state. It's keeping with the solid state > coil of wire > solid state > coil of wire scheme of things that I prefer upstream.
se
- Status
- Not open for further replies.