Hi @minek123
I am building my own board and want to use the Cascode implementation for the BJT Circlophone. I know you posted a schematic for the same in one of the previous posts.
1) How does it work, with reference to distortions compared to using 2 Zeners?
2) Also is there a difference in working for the CCS shown in the schematic vs the one below this line.
Could you or Elvee guide me to the correct one? The idea is to be independent of the supply changes if needed due to some unforeseen reason one being output power transistor availability.
Could Elvee or You explain how the Cascode transistor works? Does it work like a variable voltage-blocking Zener diode? I am pretty new in this discrete amplifier side.
I am building my own board and want to use the Cascode implementation for the BJT Circlophone. I know you posted a schematic for the same in one of the previous posts.
1) How does it work, with reference to distortions compared to using 2 Zeners?
2) Also is there a difference in working for the CCS shown in the schematic vs the one below this line.
Could you or Elvee guide me to the correct one? The idea is to be independent of the supply changes if needed due to some unforeseen reason one being output power transistor availability.
Could Elvee or You explain how the Cascode transistor works? Does it work like a variable voltage-blocking Zener diode? I am pretty new in this discrete amplifier side.
Last edited:
Hi @minek123
Thank you for the update on my query. 😊
Is the cascade modification voltage range dependent?
I will be running the circlophone via a 24-0-24 transformer
My configuration :
Output device is MJL21194 and VAS is 3503 also the current sense section is 2N5551.
I second Minek: all of these improvements should bring tangible benefits, but the measured differences are completely insignificant.
If you are patient enough, you can probably locate these results somewhere in thread, but it isn't worth the trouble.
The cascode and CCS have a practical advantage: they make the C completely supply-independent.
If you are patient enough, you can probably locate these results somewhere in thread, but it isn't worth the trouble.
The cascode and CCS have a practical advantage: they make the C completely supply-independent.
Hi @Elvee
I am facng some difficulty in using the attached ASC file with the input voltage set at 2VPP, the distortion are as displayed.
Also R9 and R12 ( R20 and R21 in ASC ) Rrespective values has been changed to 39 as per minek's asc changes, it seems that the simulation does not start with Cascode and 56 ohm.
along with R2 as 1k on the LTP common emittor resistor as per Minek's drawing
and some times it will not run only just keep re-iterating....
the below screenshot is when the input voltage is 2 volts and supply voltage is + / - 32volts,
something is making it break
there is something i am missing out ie way to operate and what should be value of
Below is with VIN as 1.5v set in the source module.
I am facng some difficulty in using the attached ASC file with the input voltage set at 2VPP, the distortion are as displayed.
Also R9 and R12 ( R20 and R21 in ASC ) Rrespective values has been changed to 39 as per minek's asc changes, it seems that the simulation does not start with Cascode and 56 ohm.
along with R2 as 1k on the LTP common emittor resistor as per Minek's drawing
and some times it will not run only just keep re-iterating....
the below screenshot is when the input voltage is 2 volts and supply voltage is + / - 32volts,
something is making it break
there is something i am missing out ie way to operate and what should be value of
Below is with VIN as 1.5v set in the source module.
Last edited:
Here is my sim (50V rails) with simpler CCS.
My values (especially gain) were adjusted to 50V rails; not 25V rails.
With lower rails, you need to lower the gain, or Vin to make sim work.
In your sim I see Vin=2V - that's way too much, especially for 25V rails, obviously the amp will be clipping.
I quickly run a sim of the almost-original circlophone (with zeners) and 50V rails, and it's clipping in much nicer way.
My values (especially gain) were adjusted to 50V rails; not 25V rails.
With lower rails, you need to lower the gain, or Vin to make sim work.
In your sim I see Vin=2V - that's way too much, especially for 25V rails, obviously the amp will be clipping.
I quickly run a sim of the almost-original circlophone (with zeners) and 50V rails, and it's clipping in much nicer way.
Last edited:
I intend to run from 30 to 35 volts (0-24 0-24 Transformer) and i observered that making 56ohms as 39 and reducing the FB to near half , starts the simulation instantaniously but if FB is at 10k and R9 and R10 at 56 ohms than the sim does not start.
Last edited:
From the Below Screenshot - I had reduced R23 ( FB ) to around 5K and R20 and R21 to 39ohm (which orginally are 56 Ohm) after which i was able to use input signal upto 2v ( atleast the simulation ran )
so as per your suggestion i have now reverted R23 to 10k and increased R36 to 1K from 470. I hope it is R36 as the below screen shot you were suggesting to increase.
The sim behaves oddly, but it is a LTspice issue, not a Circlophone one.
The cascode seems to degrade the clipping behaviour, in sim at least, but it might not be reflected in reality.
The clipping behaviour of the original C is very clean, even in sim: https://www.diyaudio.com/community/threads/my-little-cheap-circlophone-c.189599/post-2583998
The cascode seems to degrade the clipping behaviour, in sim at least, but it might not be reflected in reality.
The clipping behaviour of the original C is very clean, even in sim: https://www.diyaudio.com/community/threads/my-little-cheap-circlophone-c.189599/post-2583998
yes it did magic, thank you so much, also what is the role of those R20 and R21 ( as per my schematick ) 56ohm (original Value ), why does it not work with when input was 2v