Hi Patrick.
Any further advice as to where you are up to with your testing and tweaking for this design.
Have you had any interest at all when you make the Gerber files available from anyone willing to organise and run a GB on these pcb's?, as I think you said earlier that you would not be running a GB.
Any further advice as to where you are up to with your testing and tweaking for this design.
Have you had any interest at all when you make the Gerber files available from anyone willing to organise and run a GB on these pcb's?, as I think you said earlier that you would not be running a GB.
Another brief update for those interested in understanding circuit development.
As already mentioned, the proto according to Post #1 has the right distortion spectrum but some 12dB too high in level.
Measurements and analyses were done to identify where the lack of NFB came from.
And a series of experiments were carried out to identify recovery means.
We have now a solution for the front end that allows us to recover 6dB H2 without changing the ratio of H2:H3 (in Spice).
Verifications by measurements are currently limited by the Alpha Tester's equipment (0.012% H3).
So he needs to improve that by another 10~20dB first before we can go further.
Will take quite some time. But we know pretty sure how to get to 0.035% and 0.035%, as in the J2 measured by Stereophile.
Patrick
.
As already mentioned, the proto according to Post #1 has the right distortion spectrum but some 12dB too high in level.
Measurements and analyses were done to identify where the lack of NFB came from.
And a series of experiments were carried out to identify recovery means.
We have now a solution for the front end that allows us to recover 6dB H2 without changing the ratio of H2:H3 (in Spice).
Verifications by measurements are currently limited by the Alpha Tester's equipment (0.012% H3).
So he needs to improve that by another 10~20dB first before we can go further.
Will take quite some time. But we know pretty sure how to get to 0.035% and 0.035%, as in the J2 measured by Stereophile.
Patrick
.
Attachments
Patrick,
Would you consider trying a version using dual LSK170B? Those seem to be available, even though in limited quantities.
LSK170B-TO-92 | LINEAR SYSTEMS | NAC Semi
Would you consider trying a version using dual LSK170B? Those seem to be available, even though in limited quantities.
LSK170B-TO-92 | LINEAR SYSTEMS | NAC Semi
The PCB is compatible with 2SK170.
But it is a low cost project with easily available parts.
Hence 2SK209s.
Patrick
But it is a low cost project with easily available parts.
Hence 2SK209s.
Patrick
And in case you ask why not 2SK2145.
I have measured thousands.
I know how well (or not) they are matched in one package.
So unless I run out of space, I still prefer 2SK209s.
Patrick
I have measured thousands.
I know how well (or not) they are matched in one package.
So unless I run out of space, I still prefer 2SK209s.
Patrick
After playing with lots of circuit variations, we managed to get H2 under 0.03%, but not yet H3 to 0.003% while retaining good stability.
Still some more variations to investigate, including alternative power devices.
Once the picture is complete, we can then recommendations, depending on your objectives.
Cheers,
Patrick
.
Still some more variations to investigate, including alternative power devices.
Once the picture is complete, we can then recommendations, depending on your objectives.
Cheers,
Patrick
.
Attachments
In case you wonder why there is no news, it is because the post is on strike (until December) !!!!!!!!!!!!
Postal workers extend strike until December | Yle Uutiset | yle.fi
As a result, the 2x FQA9P25 that the Alpha tester ordered for test is not arriving any time soon.
We test what we can without those.
For example, raising rail voltage from 21V to 24V brings another 2.7dB on H2 and 4.5dB on H3.
The best we have till now is 0.022% H2 and 0.005% H3.
In the meantime, patience is called for.
Patrick
Postal workers extend strike until December | Yle Uutiset | yle.fi
As a result, the 2x FQA9P25 that the Alpha tester ordered for test is not arriving any time soon.
We test what we can without those.
For example, raising rail voltage from 21V to 24V brings another 2.7dB on H2 and 4.5dB on H3.
The best we have till now is 0.022% H2 and 0.005% H3.
In the meantime, patience is called for.
Patrick
The result is more important than the speed of the project. I think most of us are patient enough to allow you guy's to work on it with peace in mind 🙂
I like the THD spectrum - nicely done 🙂 - I might need to add the designated amp design to my to do list.
Again, the mission here is to develope a circuit that can built using readily and widely available FETs,
while at the same time resembles a similar distortion spectrum as a single ended Class A amplifier,
such as the FW J2.
This mission is, apart from a few more tests, essentially accomplished.
Cheers,
Patrick
while at the same time resembles a similar distortion spectrum as a single ended Class A amplifier,
such as the FW J2.
This mission is, apart from a few more tests, essentially accomplished.
Cheers,
Patrick