Simple Symetrical Amplifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Thanks Jay, that explains why the conversation seemed to be going on ad infinitum.

It's funny though because I'm just about coming around to the idea that valves might be able to give more clarity & life than solid state - at least a valve front end.

Perhaps it's the caps & iron that can end up sounding mellow ? ?

mike
 
Thanks Jay, that explains why the conversation seemed to be going on ad infinitum.

It's funny though because I'm just about coming around to the idea that valves might be able to give more clarity & life than solid state - at least a valve front end.

Perhaps it's the caps & iron that can end up sounding mellow ? ?

mike

tube amps really excel in mini dynamics reproduction i.e. resolution.
That is why these give more clarity . While there are some tube MOSFET hybrids no one has tried a tube BJT hybrid ..that could be interesting
 
no one has tried a tube BJT hybrid

Wouldn't be so sure about that...... I'm sure they are around, they just don't seem to meet the commercial standard for some reason.

BTW, I agree with you about the non-switching Class AB. I have tried this one too, over some years, but never found the benefits I expected sonically. I really don't think that the switching of the output stage is such a big problem as long as it's clean and well controlled.

A good compromise is to ensure that the drivers never switch off.

Hugh
 
With K30ATM and BF245A had tested only. But mine's LAT and low driver bias suffices. Check if your input BJTs still run about the same Ie, and if the amp's output tracks sine wave well symmetrically when cranking it up on the scope. Say at 30V pk-pk. About HF noise, see if its acting up so to revisit comp caps or input filtering 'cause the CCS boosts its speed. Will some NTC find enough temp to react when bundled with two small BJTs running 30mW each? We must get to know which exact type works for LC first. Still the CCS is fixing the drivers bias stability not the input drift. But in a bridge everything interacts non the less.

Upon further review (tried a DN2540), the CCS prevents bias of the outputs from drifting too high. That part is fine. But the problem is that the voltage drop across the CCS is small until it begins limiting current, but then it continues to rise. This eats up the rail voltage on one side. I don't think this will work. I await the NTC recipe.

Sheldon
 
I used low pinch off Jfets for that reason. Up to what level did the DN2540 hold before starting to chew up?

The DN2540 holds up fine, it's a 400 volt device. It keeps the current through the bias diodes from drifting higher. The problem is that the driver transistor keeps trying to turn on harder, increasing the voltage at the top of the CCS. That effectively reduces the rail voltage on the rail driving the CCS. Clipping is asymmetrical, of course.

Sheldon
 
VAS cascodes has 10 mA collector bias current, I don't think they ever go to off state ... :rolleyes:

P.S. Vbe multiplier voltage is always present and quite constant, so no sign of FET crossover distortion. FET bias is independant from output stage and can be several hundred mA if you like. Voltage part of the circuit also has +/-10 V more PSU potential than the output stage. ;)

Ah
how would you bias the FETs for class AB?
 
Wouldn't be so sure about that...... I'm sure they are around, they just don't seem to meet the commercial standard for some reason.

BTW, I agree with you about the non-switching Class AB. I have tried this one too, over some years, but never found the benefits I expected sonically. I really don't think that the switching of the output stage is such a big problem as long as it's clean and well controlled.

A good compromise is to ensure that the drivers never switch off.

Hugh

It could well be that the idea of non switching class AB does not yield any sonic superiority, agreed. But I think the idea of only current amplification all over should have audible benefits . ( And hence a tube BJT hybrid !)
if it were not about those tubes made in China and Russia having unknown quality I had possibly thought deeper on a kind of "current dumping" tube BJT hybrid where the tube supplies the voltage and the BJT supplies the current.
BJTs are unbeatable if they do what they can do best, current amplification.
Same stuff with speakers, speakers are current controlled not voltage controlled.
 
But I think the idea of only current amplification all over should have audible benefits .

An amp without a VAS ? you'll need very low impedance speakers !

really do not understand ur logic here.

I thought that it was well understood, in the final analysis, that BJT's are in fact voltage controlled with finite resistance whereas FETs are voltage controlled with almost infinite resistance.

You seem to be attempting to wrap up your preference for BJTs in some kind of logic - but all I'm hearing is that you have a preference for BJTs.
 
An amp without a VAS ? you'll need very low impedance speakers !

really do not understand ur logic here.

I thought that it was well understood, in the final analysis, that BJT's are in fact voltage controlled with finite resistance whereas FETs are voltage controlled with almost infinite resistance.

You seem to be attempting to wrap up your preference for BJTs in some kind of logic - but all I'm hearing is that you have a preference for BJTs.

No BJTs are not voltage controlled. This is not the place to explain the physics of pn junction, though.
 
An amp without a VAS ? you'll need very low impedance speakers !

really do not understand ur logic here.

I thought that it was well understood, in the final analysis, that BJT's are in fact voltage controlled with finite resistance whereas FETs are voltage controlled with almost infinite resistance.

You seem to be attempting to wrap up your preference for BJTs in some kind of logic - but all I'm hearing is that you have a preference for BJTs.

A current amp can be equated to a current controlled current source CCS.
The CCS forces a current which is a linear multiple of control current. No need for Vas.
 
Hi hahfran ;)

That's completely wrong approach. All power amplifiers have to follow voltage gain linearly according to frequency. Why? Because all the speakers's sound pressure is linearized to input voltage at their terminals. Impedance curve is much more unlinear and has no correlation to sound pressure curve even at driver's resonance frequency.
Current amplifier would make a speaker sound completely wrong because SPL would follow impedance curve and that would be a real mess. We have to make amplifier's gain linear to input voltage and to have output impedance as low as possible (high damping factor). Power supply and numerous BJT output pairs enables high current capability. Basics of audio ... :)
 
Hi hahfran ;)

That's completely wrong approach. All power amplifiers have to follow voltage gain linearly according to frequency. Why? Because all the speakers's sound pressure is linearized to input voltage at their terminals. Impedance curve is much more unlinear and has no correlation to sound pressure curve even at driver's resonance frequency.
Current amplifier would make a speaker sound completely wrong because SPL would follow impedance curve and that would be a real mess. We have to make amplifier's gain linear to input voltage and to have output impedance as low as possible (high damping factor). Power supply and numerous BJT output pairs enables high current capability. Basics of audio ... :)

No it is the correct approach because B*l*I= d2/dt2+ d/dt+....
that is the mechanical motion equation.
All one has to do is to feed a current amp with the output of an "analog computer" ( 3 integrators and one adder ) to cancel out all of the mechanical
constants such as mass compliance...and electrical constants such as resistance and inductivity of the voice coil. So far only swedish co ACE employs this approach. It gets a bit simpler with positive feedback from a resistors in series with the speaker.
of course the real speaker does not follow a simple second order de but it is a simple and easy to implement approach.
 
I am not telling anything new here. Of course one has to map the mechanical stuff onto an electrical schematic and the whole trick is to give the amp a complex output impedance which is exactly the reverse ( the mirror image in complex plane) of the speaker's electrical equivalent. It took a very long time until makers of active speakers adopted this most simple idea but now all of them do it.
 
Back to the roots
provided the model provided by Bob Cordell is correct and the schematic is correct
I get an op of 1.8 amps drain current .
Is that correct?
 

Attachments

  • SSA op.jpg
    SSA op.jpg
    71.3 KB · Views: 455
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.