I'm planning to give the 2SK246Y a try, but it's going to be a while before I can.
Thanks for sharing your findings.
I've one reservation about the 2sk170/2sk246y
They don't have the same dynamic range and definition as the diodes and lm334 where as the 2sk's seem to have the lowest noise floor
It's just id prefer both qualities
Sent from my iPad using Tapatalk
They don't have the same dynamic range and definition as the diodes and lm334 where as the 2sk's seem to have the lowest noise floor
It's just id prefer both qualities
Sent from my iPad using Tapatalk
Last edited:
How do you deduct they have less dynamic range/definition, is this your personal listening observation or measured? I don't understand how this would be the case.
You don't have to believe me if you don't want to
Personally I can't understand how you have opinion without even trying the different options
I can keep my thoughts to myself if it makes you uncomfortable
Personally I can't understand how you have opinion without even trying the different options
I can keep my thoughts to myself if it makes you uncomfortable
Nige, Sorry, you are misreading my question. I didn't express an opinion, and I'm not questioning your opinion or judgment either, how could I? 🙂
Reading in between the lines you have answered, never the less. Thanks.
Last edited:
Carlsor,
You have removed the resistors in the I2S circuit.
Being of similar mind - why have a component where one is not needed - what was the purpose of the 100R resistors? I have no idea.
Does your set-up have something to do with your decision or would you have done it you were using WAVE IO?
Please let me know.
Thanks,
You have removed the resistors in the I2S circuit.
Being of similar mind - why have a component where one is not needed - what was the purpose of the 100R resistors? I have no idea.
Does your set-up have something to do with your decision or would you have done it you were using WAVE IO?
Please let me know.
Thanks,
The resistors are good practice in digital design, to decouple the individual circuits.
It reduces high frequency currents between logic chips, hence reducing RMI and jitter. It is recommended to not get them out.
It reduces high frequency currents between logic chips, hence reducing RMI and jitter. It is recommended to not get them out.
ok jumped the gun a bit there, i had thought it was a cheap shot at my decision making capabilities
although anyone can make a mistake, surely it cant be necessary to measure everything especially if you think its a clear cut decision
and if we cant mostly trust our ears to assess i think we've missed the point
was listening to lm334/2sk170 bl/ 2sk246y over the last couple of days
dont think any of them is optimal, but they show substantial sq differences
which gives scope for something much better
rax posted walt jungs articles a while back
the jfet was only one method should we not try the others?
youll either need them or you wont
best without if you can
but its whatever sounds best
Hi Nige, I totally agree with you, ears have the final say. In general, before I try a new audio tweak I’d like to at least have some sort of idea/understanding, why and in which way something would bring an improvement, and then later let my ears decided if it actually is. That was the intent behind the initial message.
you can correlate that only to measurements (Doede having AP) , due to fact that we (well , at least me) don't know enough what's in DAC chip , so sorta playing with Black Box
crucial is - is it soft CCS better to ears , or hard one (disregarding tempco for a moment) .......
easy to reply with listening sessions , but nothing what I can transfer in full electronic analysis
what I really tried to say , I agree with you ...... 😉
crucial is - is it soft CCS better to ears , or hard one (disregarding tempco for a moment) .......
easy to reply with listening sessions , but nothing what I can transfer in full electronic analysis
what I really tried to say , I agree with you ...... 😉
I don't wanna squabble it's unproductive
Zen
The three 2sk jfet I've tried sound quite similar in comparison to the diodes or lm334 they seem to have a lower noise floor maybe that's better rejection but they also seem to smooth or soften the sound to the extent that bass control and dynamics or impact gets affected when comparing lm334 to 2sk 246y
Whether it's the voltage swing in the dac chip having an effect on the ccs or not I'm not sure
Any thoughts?
But there's large sq differences between ccs sources
I don't think we should stop searching for something better
I also found the lsk170 to be a poor imitation of the 2sk170 it's got much higher noise
Zen
The three 2sk jfet I've tried sound quite similar in comparison to the diodes or lm334 they seem to have a lower noise floor maybe that's better rejection but they also seem to smooth or soften the sound to the extent that bass control and dynamics or impact gets affected when comparing lm334 to 2sk 246y
Whether it's the voltage swing in the dac chip having an effect on the ccs or not I'm not sure
Any thoughts?
But there's large sq differences between ccs sources
I don't think we should stop searching for something better
I also found the lsk170 to be a poor imitation of the 2sk170 it's got much higher noise
yup
what we need certainly must have two things : low noise and good tempco (even if later one I'm not finding so important , always thinking about temp. equilibrium state , not state before that)
now we only need to deduce what's preferable - finite or infinite dynamic impedance of CCS
with finite , current will vary more but we'll have smaller voltage spikes at pin 20
with infinite , current will vary less , but spikes will be greater
sorry for not contributing in flesh , I'm having too much other things to do these days
what we need certainly must have two things : low noise and good tempco (even if later one I'm not finding so important , always thinking about temp. equilibrium state , not state before that)
now we only need to deduce what's preferable - finite or infinite dynamic impedance of CCS
with finite , current will vary more but we'll have smaller voltage spikes at pin 20
with infinite , current will vary less , but spikes will be greater
sorry for not contributing in flesh , I'm having too much other things to do these days
Would as you say infinite dynamic ccs be more reliant on the impedance of the power available to the analog section of the dac chip?
irrelevant , in this context - CCS being situated between pin 20 and gnd
however , as I wrote before , in case of placing CCS between pin 20 and (hypothetic) negative rail , that negative rail must not be perfect ...
however , as I wrote before , in case of placing CCS between pin 20 and (hypothetic) negative rail , that negative rail must not be perfect ...
Thanks Doede for offering your help. I obviously am not trying to go around you with five DAC boards and two mainboards in my possession! (well, the blue board is in transit) Trying to find the best packaging for the battery scheme and I appreciate your understanding, both personal and electronic!
nige2000, you seem frustrated with the CCSs. I do not think anyone searches for something better harder that you. Just to keep things in perspective, give us your current pecking order, or more precisely, where each of these approaches exist in your pecking order of goodness.
You have made it clear that either CCS is better than the resistor but I am hoping you can give an assessment of how much better than the resistor the CCS. Is it greatly better or just better.
You had mentioned you had used either the TX2575 or something equivalent in quality - how far from this have you got with the CCS? If such a thing can be put into words - not looking for an essay.
nige2000, you seem frustrated with the CCSs. I do not think anyone searches for something better harder that you. Just to keep things in perspective, give us your current pecking order, or more precisely, where each of these approaches exist in your pecking order of goodness.
You have made it clear that either CCS is better than the resistor but I am hoping you can give an assessment of how much better than the resistor the CCS. Is it greatly better or just better.
You had mentioned you had used either the TX2575 or something equivalent in quality - how far from this have you got with the CCS? If such a thing can be put into words - not looking for an essay.
think a real good (undiscovered) ccs could hit into the top 3 or 4
for lowest noise the 2sk170 bl is greatly better
for control, dynamic, and some aspects of detail lm334 is best (but is noisy by comparison) either is much better than a resistor
id imagine the sq of the 2sks will suit most peoples taste
marginal improvement use a current diode
its the only quick fix
Thanks, nige2000 for the answers.
It would seem that ZEN MOD's suggestions from a month back of using a good old transistor might be worth exploring.
I readily admit to not having the ability to pioneer so defer to those who do.
Maybe the FET is not the answer - or worse - maybe the simple solution is too simple? A discreet version of the LM334 or something similar?
It would seem that ZEN MOD's suggestions from a month back of using a good old transistor might be worth exploring.
I readily admit to not having the ability to pioneer so defer to those who do.
Maybe the FET is not the answer - or worse - maybe the simple solution is too simple? A discreet version of the LM334 or something similar?