What are the pros and the cons of those 3 vas ?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hawksford cascoded VAS I show in my image are from the book of Douglas Self, and my "very little mod" are just adding the 10 R resistor (the magic resistor suggested by Peufeu in his web site).

Thank

Bye

Gaetan
D.Self's Second edition shows the basic cascode.
Sixth edition shows both the basic cascode and the Hawksford enhanced cascode. The Hawksford version shown includes both the AC bypassing capacitor and the VAS Re.
 
So can I summarise ... the 1970 cascode circuit I posted earlier used a low value resistor to limit the current in the so-called VAS transistor in the event of a current overload.
If a resistor is also used in the emitter of the lower device of the cascode pair, then connecting the lower end of the bias chain to the emitter, as Hawksford showed, merely keeps the voltage across the lower device more constant than it would otherwise and thus aims to reduce Early effect distortion further.
 
What edition is your book?
I have the 5th edition (Focal Press) and there is an emitter resistor for Q1.
That resistor is essential to make it a "Hawksford" cascode, as opposed to just a cascode.
It looked that you had removed this resistor for your "very little mod" but it seems D.Self or the book publisher messed this up.

Best wishes
David

Paper is here. AES E-Library Reduction of Transistor Slope Impedance Dependent Distortion in Large-Signal Amplifiers

Hello

Your right, in the aes paper there is an emitter resistor for Q1, I will use this paper.

Btw, can we use a cascoded VAS, without a buffer transistor, with a low power amp (up to 40 watt) ?

Thank

Bye

Gaetan
 
Apart from your modification, your picture looks exactly like my 5th edition except with the emitter resistor removed.
Be honest, did you remove that resistor?;)
Otherwise I still want to know what edition it was.

Best wishes
David

Hello

Some wrote that if Cascode vas is adopted, it must drive a high impedance load, but with a low power amp (30 to 40 w), does Cascode vas can drive a low power amp ?

I've simulated the Hawksford" cascode and with that 100 R resistor, and the thd was much much more higher, so I removed it, but after your post I replaced that 100 R resistor by a 10 R resistor and thd came back to normal (-85 db).

Thank

Bye

Gaetan
 
Hi guys,

I'd like to ask if there are added benefits when adding degenerators on this type of VAS. Other circuit had it. Just curious though if attaching small value resistors at the emitters will improve stability of operation.

Thanks!
 

Attachments

  • self_ch18_fig8at.jpg
    self_ch18_fig8at.jpg
    17 KB · Views: 308
Hi guys,

I'd like to ask if there are added benefits when adding degenerators on this type of VAS. Other circuit had it. Just curious though if attaching small value resistors at the emitters will improve stability of operation.

Thanks!

Hi Albert,

Degeneration, being a local NFB, decreases the local gain of the stage, increasing its DC operating point stability (including thermal stability) and widening the bandwidth of the stage at the same time. Decreased local gain results in decreased overall open-loop gain and - assuming the closed loop gain is constant, set by the global NFB network - decreased loop gain, leading to higher overall stability.

I would use some higher value degeneration resistors for the input LTP (say, 100R as a minimum), and lower value resistors for the balanced VAS LTP (some 22R...33R).

Cheers,
Valery
 
AndrewT,

I was referring to the ltp VAS, I took
the schematic at D. Self's book but it seems he never did an extensive study of the topology. I was aware that it was a Hitachi design. Correct me if I'm wrong.

Valery,

As always many thanks for the details.

The reason I asked was that I was using it in my other CFP output design amp. Upon further investigation while working on it, it seems oscillation problem comes at VAS stage. I find it out by installing and uninstalling miller comp caps, both at VAS stage and the outputs. Oscillation seemed to stop by adding compensation caps at base collector junction of VAS but I doubt it is enough as "a cure" perhaps it needs more fine tuning at VAS stage.

Sorry, O.T.
 
In the paper (A New Amplifier Topology) published by S.Groner in Linear Audio Vol 2, he showed the Cdom in the same location as your post87 pic. He also discussed the second Ccb and explained it needed to be moved from the collector to audio ground, i.e. it becomes Cgb connecting ground to base. He further stated that the values of Cdom and Cgb be accurately matched to maximise the benefit.

I see very many second stage VAS/TIS done as dual transistor in an LTP form, but all of them do not follow the S.Groner arrangement. As far as I have seen his implementation is unique.
Many place the Cdom on the wrong side of the LTP and omit the second. Some show both capacitors as you have, but connect as Ccb instead of Cgb.

Does any Member have any thoughts on the correct way to implement the compensation?

Is there a copy of this paper published elsewhere that is free to consult without having to pay Linear Audio?
 
Last edited:
In the paper (A New Amplifier Topology) published by S.Groner in Linear Audio Vol 2, he showed the Cdom in the same location as your post87 pic. He also discussed the second Ccb and explained it needed to be moved from the collector to audio ground, i.e. it becomes Cgb connecting ground to base. He further stated that the values of Cdom and Cgb be accurately matched to maximise the benefit.

I see very many second stage VAS/TIS done as dual transistor in an LTP form, but all of them do not follow the S.Groner arrangement. As far as I have seen his implementation is unique.
Many place the Cdom on the wrong side of the LTP and omit the second. Some show both capacitors as you have, but connect as Ccb instead of Cgb.

Does any Member have any thoughts on the correct way to implement the compensation?

Is there a copy of this paper published elsewhere that is free to consult without having to pay Linear Audio?

Here is the explanation - quoted from the article:
"Besides the Miller compensation capacitor C2 there is now an additional capacitor (C1) which has an (although only second order) influence on compensation. It is required to make the drive from the input stage single-ended at high frequencies, as otherwise the output current of Q2 would bypass the Miller compensation loop at high frequencies."
 

Attachments

  • new_amplifier_topology_fig3.jpg
    new_amplifier_topology_fig3.jpg
    26.4 KB · Views: 195
True, the cascode can source and sink the same current as any VAS but to achieve the maximum open loop gain it needs a very high load indeed. A simple Darlington type output would load the VAS far too much. Even with a triple output stage the use of an extra class A emitter follower is really worthwhile and the result clearly audible aspecially in the bass.

A cascode stage has a high output impedance. It does not HAVE to drive a high impedance load. A cascode stage can drive any load that a normal VAS can.
Your 100 ohm resistor probably reduced the open loop gain but not because of the Hawksford configuration necessarily.
John
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.