Genuine 2SK170 is a symmetrical part. That means, if you swap D and S, nothing happens. Neither Idss nor Vgs would differ. Of course, you have to keep Gate (middle pin).
So the currents are 783 and 843 mA, that is fine and verifies that the shunt is operating nicely. Also the CCS must be operating, otherwise there would be no 18V or the thing (heat sink) would be cooking.
First suspect should be the CVS (where the external power enters the PSU). If possible replace Q104/204 (BC337/327) with lower gain versions (say Hfe 300 or less) also replace R100/200 with 68 Ohm or parallel the 330 with 100 Ohm and see what happens if this does not help let me know.
(p.s. I will not be able to answer until tomorrow morning)
Changing q104/204 to hfe of 460 worked (original hfe of 460 and 560)! Seems that there are no more oscillations on that board!
When I tried the same trick on the other board, which suffered from similar frequency oscillations, it did not work, still oscillating. Note that I am still using j310 instead of 2sk170 for the njf. Will try tonight what happens if I also replace the njf....
The final word on this:
-Q104/204 is a bit hit and miss. I ended up with both around Hfe=460, although it is not a sure thing it will work with that value (I tried multiple with similar Hfe and some gave oscillation and some not)
-I had to remove the caps across Q104 and Q204 (see assembly guide), otherwise it wouldn't work.
-I had to KEEP the caps across Q102 and Q202 (2.7n, I used 2.2n), otherwise it didn't work.
-Q104/204 is a bit hit and miss. I ended up with both around Hfe=460, although it is not a sure thing it will work with that value (I tried multiple with similar Hfe and some gave oscillation and some not)
-I had to remove the caps across Q104 and Q204 (see assembly guide), otherwise it wouldn't work.
-I had to KEEP the caps across Q102 and Q202 (2.7n, I used 2.2n), otherwise it didn't work.
Many, many thanks for your effort and report
We gained lots of useful information, now I have to cookup (the spell checker objects to cookup but offers, as an alternative, the same with one 'o'. You can guess what I am objecting to) something for the builders guide.
Many, many thanks for your effort and report
Nope, thank you for your help in getting my boards to work!
What I wanted to ask:
I noticed the output offset drift is (after probably 24-48 hours running) quite high, up to several 100's mV's. After I placed a cardboard box over the pcb's the drift went down to probably usually +-25mV with peaks to 100mV.
1. I see people reporting output offset drift of 5mV. Was that measured using a DMM or a scope? When using the DMM, my pcb's show quite low drift (range towards 5mV, however using a scope gives the full picture as mentioned above.).
2. So the Paradise is very sensitive to thermal fluctuations (I have an aircon on the other side of the room and switching it off makes a huge difference already). This placed quite strict requirements on the enclosure for the Paradise. Air vents are probably a bad idea and heatsinks outside the enclosure a good one. Also thermal coupling of certain devices (or the whole thing) is a good idea. Anyone care to comment on how they achieved this?
3. I am obviously still waiting for the caps leakage to settle down. I am using the caps provided in the parts GB (panasonic apparently.) Would using different caps bring any improvement, either sonically or in the offset drift?
thanks
I noticed the output offset drift is (after probably 24-48 hours running) quite high, up to several 100's mV's. After I placed a cardboard box over the pcb's the drift went down to probably usually +-25mV with peaks to 100mV.
1. I see people reporting output offset drift of 5mV. Was that measured using a DMM or a scope? When using the DMM, my pcb's show quite low drift (range towards 5mV, however using a scope gives the full picture as mentioned above.).
2. So the Paradise is very sensitive to thermal fluctuations (I have an aircon on the other side of the room and switching it off makes a huge difference already). This placed quite strict requirements on the enclosure for the Paradise. Air vents are probably a bad idea and heatsinks outside the enclosure a good one. Also thermal coupling of certain devices (or the whole thing) is a good idea. Anyone care to comment on how they achieved this?
3. I am obviously still waiting for the caps leakage to settle down. I am using the caps provided in the parts GB (panasonic apparently.) Would using different caps bring any improvement, either sonically or in the offset drift?
thanks
2.1 Fully enclosed box with extra fins extentions to take the heat out See My Paradise.
2.2 Heat sink outside box it is realy a good idea
2.3 This came up but was not discuseed in depth IMO and it is worth same time and efort.
3 I TOLD YOU TO RUN IN THE CAPACITORS WHIT SIMPLE LM317
Nah only joking shouting just bit of fun do as you like but suggestion was there.
I wuld say try the Paradise with standard components, it takes a bit of listening to apreciate the honesty of this stage, then feel free to make changes to suit your ears and system.
For imput caps Pana FC are prety good same wuld argue that Pana FM sound better
AS I understand low leachagge current is the main parameter of interest in the input stage and if this is the case Hi voltagge rating for same Capacitance Value may present same improvment.
But I can not see any real issue on the R2 boards
PS You got PM
Here we are:
...almost,
Karfreitag will proof...
beautiful, great layout!
http://www.diyaudio.com/forums/analogue-source/218625-paradise-builders-27.html#post3346568
Did you change the resistor values as described in JJ's post? This effectively cures the offset problem and together with a power-on/off mute relay makes the Paradise safe to use in dc coupled systems. Together with the anti-oscillation caps it's probably the most important finding in this thread.
