I am not sure if this VAS stage has been presented earlier but I have not seen it ever in any audio power amplifier and therefore presenting it for discussion.
Anyway the idea was to get feedback loop shorter in a bit similar way as in VSSA topology but still to have differential stage.
The idea is to move the VAS stage drive to another tail of differential stage.
Correct me if I am wrong but benefits are
1. emitter follower Q8 is very easy load for differential stage collector resistor (or it could be current mirror as well).
2. Q8 as emitter follower is very fast stage, significant delays nor BW limitation
3. Q6 is cascade stage (common base) which is also very fast
(neither of Q8 nor Q6 are limiting BW as traditional common emitter VAS stage)
4. final negative feedback is in a way shoter (as in VSSA or (other) current feedback topologies).
I have no idea if this is better than common emitter + cascade combination VAS stage but I see it interesting.
I did simple simulation with MicroCap demo version and it seems that R3 (differential stage collector resistor) size has very small effect to BW which is interesting. (I am no kind of expert running amp simulations. I have used MC mostly for X-over desings).
Comments?
Anyway the idea was to get feedback loop shorter in a bit similar way as in VSSA topology but still to have differential stage.
The idea is to move the VAS stage drive to another tail of differential stage.
Correct me if I am wrong but benefits are
1. emitter follower Q8 is very easy load for differential stage collector resistor (or it could be current mirror as well).
2. Q8 as emitter follower is very fast stage, significant delays nor BW limitation
3. Q6 is cascade stage (common base) which is also very fast
(neither of Q8 nor Q6 are limiting BW as traditional common emitter VAS stage)
4. final negative feedback is in a way shoter (as in VSSA or (other) current feedback topologies).
I have no idea if this is better than common emitter + cascade combination VAS stage but I see it interesting.
I did simple simulation with MicroCap demo version and it seems that R3 (differential stage collector resistor) size has very small effect to BW which is interesting. (I am no kind of expert running amp simulations. I have used MC mostly for X-over desings).
Comments?
Attachments
It's a Rush circuit (there is no such thing as a Rush cascode), also called series differential pair or emitters coupled pair.
It should be less linear than a single common emitter transistor but may have some other advantages.
I think it was used in an Akaï amplifier around 1980.
It should be less linear than a single common emitter transistor but may have some other advantages.
I think it was used in an Akaï amplifier around 1980.
based on this description it has been used in the input stage (like in the differential pair)https://mrevil.asvachin.eu/amp/topologies/rush/
In my drawing it is in the VAS stage to make it possible to use other half of input differential pair to drive VAS stage because the "Rush circuit / emitter couplled" does not swap phase 180 deg. as common emitter does. I think this might be something new.
Maybe you should have investigated the website you linked a bit further. Its very rare these days to find something new under the sun. https://mrevil.asvachin.eu/amp/amp4/sub/Amp_schematic.png
I did and the schema you linked has different idea. In that what you linked the VAS stage bases on the common emitter (Q2)which is current amplified by Q5 and then casccaded by X1. There is no emitter follower.
The VAS stage is driven by left side of the diff stage just in tradional amp topology. Right side (feedback side) of diff amp effect to VAS stage drive directly is only about 5%. Therefore also from this point of view this one is traditional desing.
- Status
- Not open for further replies.