> Can replacement the 220uF capacitors to 2uF?
Good question.
The 220uF caps are there to provide a low impedance path for the signal current.
If you replace 220uF by 2uF, you will increase the resistance of that path by a factor of 110.
That in turn will cause the floating supply to swing a factor of 110 more relative to signal ground.
Until it runs out of headroom.
Let's assume your R_iv to be 680ohm. And signal current is 4mA. And you use a 24V supply.
So a signal of 1Hz 4mA will cause the top rail to swing between 15V and 9V rel to signal Gnd.
i.e. about the same amplitude as the output signal, as expected.
If you change the cap to 47uF, the top rail will now swing between 19V and 5V.
i.e. 8V pk-pk more than the signal itself.
Anything less than 47uF will cause the JFETs to run out of Vds headroom, and signal will clip.
The situation will improve if the signal current is less.
e.g. with 1.2mA and R_iv 2.4k, and 47uF caps, the signal will cause the top rail to swing between 15V and 9V rel to signal Gnd.
Changing the cap to 10uF will cause the rail to swing between 22V and 2V.
So to be on the save side, I would not use anything less than 220uF for signal current up to 2.8mA rms (or +/-4mA).
If you want to use higher signal current, you should increase the cap value proportionately.
Patrick
Good question.
The 220uF caps are there to provide a low impedance path for the signal current.
If you replace 220uF by 2uF, you will increase the resistance of that path by a factor of 110.
That in turn will cause the floating supply to swing a factor of 110 more relative to signal ground.
Until it runs out of headroom.
Let's assume your R_iv to be 680ohm. And signal current is 4mA. And you use a 24V supply.
So a signal of 1Hz 4mA will cause the top rail to swing between 15V and 9V rel to signal Gnd.
i.e. about the same amplitude as the output signal, as expected.
If you change the cap to 47uF, the top rail will now swing between 19V and 5V.
i.e. 8V pk-pk more than the signal itself.
Anything less than 47uF will cause the JFETs to run out of Vds headroom, and signal will clip.
The situation will improve if the signal current is less.
e.g. with 1.2mA and R_iv 2.4k, and 47uF caps, the signal will cause the top rail to swing between 15V and 9V rel to signal Gnd.
Changing the cap to 10uF will cause the rail to swing between 22V and 2V.
So to be on the save side, I would not use anything less than 220uF for signal current up to 2.8mA rms (or +/-4mA).
If you want to use higher signal current, you should increase the cap value proportionately.
Patrick
Last edited:
Hi All
I tried to breaboardthe CEN, for my CDplayer PCM63 based
But, the first board have 120mV Input offset
Fet are from Spencer, so well matched
The termic coupled it's "rudimental"
I have not a scope
There is something wrong in my P2P?
Or i have to try gate resistors?
Thank in advance
Marco
I tried to breaboardthe CEN, for my CDplayer PCM63 based
But, the first board have 120mV Input offset
Fet are from Spencer, so well matched
The termic coupled it's "rudimental"
I have not a scope
There is something wrong in my P2P?
Or i have to try gate resistors?
Thank in advance
An externally hosted image should be here but it was not working when we last tested it.
[/IMG]Marco
An externally hosted image should be here but it was not working when we last tested it.
[/IMG]![IMGDEAD]](/community/proxy.php?image=http%3A%2F%2F%5BIMGDEAD%5Dhttp%3A%2F%2Fi43.tinypic.com%2F2hs0j00.jpg%5B%2FIMGDEAD%5D&hash=a3cd5a3d194d5ff7e45e68364ee62736)
![IMGDEAD]](/community/proxy.php?image=http%3A%2F%2F%5BIMGDEAD%5Dhttp%3A%2F%2Fi44.tinypic.com%2F25tgw1s.jpg%5B%2FIMGDEAD%5D&hash=4ba0a3cba1a91141e7adec4a7f5eb08b)
Input offset is mainly influenced by matching between 2SK170 & 2SJ74.
Not much you can do, other than try to get better matched devices.
Suggest you measure Idss of each device individually first.
You can also try to add gate stopper resistors to see if it alleviates the problem.
Output offset is near perfect, as it should be.
The rest looks OK.
Patrick
Not much you can do, other than try to get better matched devices.
Suggest you measure Idss of each device individually first.
You can also try to add gate stopper resistors to see if it alleviates the problem.
Output offset is near perfect, as it should be.
The rest looks OK.
Patrick
Hi Patrick
I misured the Fets and the matching it's near 0,5%
so i put gate resistance and problem solved!
Input offset below 1 mV for left and right channel
So, THANK YOU
Sound:
I think need a buffer, because i go into a 10K Shunt attenuator (my resistor I/V is 1,4K)
Sound it's clear, trasparent but suffer a bit into bass frequences and dinamics
what you think, Patrick? Some suggestion from you end others members will be welcome
Marco
I misured the Fets and the matching it's near 0,5%
so i put gate resistance and problem solved!
Input offset below 1 mV for left and right channel
So, THANK YOU
Sound:
I think need a buffer, because i go into a 10K Shunt attenuator (my resistor I/V is 1,4K)
Sound it's clear, trasparent but suffer a bit into bass frequences and dinamics
what you think, Patrick? Some suggestion from you end others members will be welcome
Marco
Thank you very much, Patrick
I tried my DCB1 but it have some problem
It's possible to buy PCB for Gerhard Buffer? Because now in my CDP have a 15V bipolar supply free (supplied previous output stage) so i can utilize it
for battery, may i use 24V? Cause i have 4 lead battery 12 V / 1200 mA and relative charger
Now i have 1nF parallel to resistor I/V, will try 680pF
Marco
I tried my DCB1 but it have some problem
It's possible to buy PCB for Gerhard Buffer? Because now in my CDP have a 15V bipolar supply free (supplied previous output stage) so i can utilize it
for battery, may i use 24V? Cause i have 4 lead battery 12 V / 1200 mA and relative charger
Now i have 1nF parallel to resistor I/V, will try 680pF
Marco
You can contact Gogowatch. We have spare ones (JG filter).
For the JG Filter, +/-15V is specified. I would not go lower.
But there is no need to go +/-24V. You just burn energy for no performance gain.
For the IV, you may use 18~24V floating.
And 1nF as C_iv is also fine. But try and adjust value according to taste.
Patrick
For the JG Filter, +/-15V is specified. I would not go lower.
But there is no need to go +/-24V. You just burn energy for no performance gain.
For the IV, you may use 18~24V floating.
And 1nF as C_iv is also fine. But try and adjust value according to taste.
Patrick
Now i don't follow...
Are you now talking about what to parallell to the electrolytic-cap, if yes - then i already have 2,2uF MKP there.
Why i ask is because is because i remember you talked earlier about not using to little cap with lower r_iv and higher current. But what is "to little" running ESS9018 = 8mA and 240R R_IV per SEN?
Are you now talking about what to parallell to the electrolytic-cap, if yes - then i already have 2,2uF MKP there.
Why i ask is because is because i remember you talked earlier about not using to little cap with lower r_iv and higher current. But what is "to little" running ESS9018 = 8mA and 240R R_IV per SEN?
Hi
Today i tried to put my DCB1 (source impedance trimmed to 24Kohm ) between CEN and Pot
So, now the sound was a way better, bass rock solid, mid and high crystal clear, no more grain.
This I/V it's a wonderful piece.
I sended a PM to gogowatch for Gerhard Buffer PCB .
Thank you very much, Patrick
Marco
Inviato dal mio GT-S5300 con Tapatalk 2
Yes, that is the one post! - i just thought i read it alot earlier in the thread.
Anyway, of choosing polarized caps (lite Silmic II) then 220uF would be fine, right? I have problem with larger size than that.