Technics Class A Biasing

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi there,

Can some one explain how this bias circuit works?


Legend:

PRE Drive Q3 & Q4
Output stage Q5 & Q6
Bias Diode QD1
Bias Transistor Q7
Stabistor pack DP1
Overload det Transistor Q8
 

Attachments

  • technics explanation schematic.jpg
    technics explanation schematic.jpg
    25.4 KB · Views: 973
Hi,
does it matter if the non operation device passes near constant current or is actually switched off?

A single ended ClassA amplifier never switches off it's output device. The device remains in control of the output current.
If the output current nears or exceeds the bias current then the stage overloads/clips.
Similarly if the output device passes zero current the amp is no longer in ClassA.

A push-pull ClassA stage has an upper and lower pair of devices.
Both these devices remain in control of the output current.
If one or other of the devices drops out of control then the stage reverts to ClassAB operation to supply the extra current above the ~ 2times bias current.

If one device passes a constant (low) current and the other device is passing more than double the bias current to the load then the stage is operating in ClassAB.
It matters not that neither device has switched off. It is limited to ~twice bias current for ClassA operation.
 
there are a lot of designers and builders that disagree with me, notably Prof. Leach where he describes the operation of his drivers in LO Tim as staying in ClassA simply because he has arranged that they never switch off. I contend that he is wrong.

You have to make up your own mind.

My "clarification" is not fact. It is my opinion.
 
I appreciate it Andrew. But you must have some reasoning before taking any view.

I was looking for class-A performance but still was not ready to accept the inefficiency of it. However, I thought if the cross-over distortion, a drawback of proper class-B, can somehow be eliminated then it would be a good trade-off in between class-A and non-switching class-B.

I thought technics chip was doing that.

I really didn't want to start a philosophical debate about class conflicts and its subtleties.

cheers
 
Disabled Account
Joined 2006
Think of it this way, these schemes help because the outputs are kept at a bias level and never go through a point where the devices switches completly off, theres more to the theory but Im really bad at explaining these things, andrewt or andyc would be much better at this. Asked them nicely, Im sure they will help you out. There are some explanations in the krill amp thread and the thread connected with it, also read the theory behind the cambridge azur840 amp, theres a white paper on the azur range.
 
D.Self's Cambridge work is very similar to the conversion of opamp output stages to single ended ClassA using a current source to one supply.

All this does is increase the level of low current ClassA available before the amp reverts to ClassAB for the higher current outputs.
I wonder if the asymmetrical nature of this ClassA to ClassAB handover is audible?

The opamps ClassA current is determined in conjunction with the load to never handover to ClassAB. Cambridge cannot achieve this trick with their power amp.
 
Disabled Account
Joined 2006
I have modified a cambridge 840 or 2, it sounds ok, I couldnt pick up when the exchange happens, feel the frontend needs attention, I wasnt really deep listening for this technique they use, but it does have its flaws. The measurements they did seem to show some improvement.
 
AndrewT said:
there are a lot of designers and builders that disagree with me, notably Prof. Leach where he describes the operation of his drivers in LO Tim as staying in ClassA simply because he has arranged that they never switch off. I contend that he is wrong.

You have to make up your own mind.

My "clarification" is not fact. It is my opinion.

Andrew,

I agree with you. Just because one of the transistors is conducting some current does NOT make the operation class A and does NOT eliminate static crossover distortion. The spirit of class A (and the good performance that derives therefrom) is that both transistors contribute to the signal over the full cycle. If a transistor goes into essentially a constant current mode by its being prevented from cutting off, it is not contributing to the signal.

Cheers,
Bob
 
atiq19, Although I cannot answer your questions about Technic's biasing scheme I would suggest that the inspiration for it came from Peter Blomley's February/March 1971 articles in Wireless World. The basic thinking behind Blomley's idea is that if neither class B output device enters cutoff the linearity in the crossover region becomes more controllable. An added complexity is that we need a "signal splitter" stage to direct the audio signal half cycles to the appropriate output devices.
On this forum I did a search using the terms "Blomley Amplifier" and found several threads. One is called "Question about signal disection" that includes a link to the circuit diagram. At the time of publication the amplifier had a distortion performance that exceded the capabilities of the measurement instrumentation available. It was not a DIY construction project as only the circuit schematic was given.
Keith
 
If this biasing scheme or Krill amp biasing scheme prevents crossover distortion & only leaves static distortion can anyone quantify these?

What I am asking is lets not deal in absolutes - how big is crossover distortion relative to static distortion - is it 90% compared to 10%? If so we are removing 90% of the class AB distortion which seems like a good idea to me - no?

Sometimes we dismiss good ideas because they do not give us the 100% solution. So lets quantify what we're talking about
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.