Hello Steen - and others....
Its Ok that you don’t have anything to complain about - neither do I, BUT - I have built 3 Pearls in the last couple of months, and the all sounded different from each other, so there are definitely some parameters that you can change, and it will affect the sound, maybe to your liking, maybe not.
Supply noise - I have a inefficient set of Proac clones, just like you - on them, there are almost no noise - switching to my Snell E II (+5dB in sensitivity)- there are definitely noise to be heard, and more noise than ex. with the Ono. Hence, I build a version with an even more overkill PSU, with 80000 uF in the PSU (plus the 43000 uF on board) - and added two extra RC filters - the result was even less noise, but more pronounced was the reproduction of space - the room 'grew' in some way. It is quiet, but ex. the Ono is quieter......but its ok for me.
I have also builded one in a wooden box, and one in a metal box - there weren’t more noise in the wooden one, but the "metal" had more bass, and more pronounced bass - why I don’t know.
Have built it with and without battery, and there were no obvious difference in sound - except its easy to hear when the battery runs down, and that happens once a month......
I don’t have any trouble with it - builded it just like Wayne in his article - have tried a lot of different groundings - about 3 hours with the soldering iron, and the RIAA turned on, to find the grounding that made the least noise - and it was the "wayne way".
I don’t think that you can say what is best for you, if you have not heard them all - i have not heard them all, therefore I experiment.
I wont just drop the source resistors, don’t worry ;- ) - but I recall that nelson suggested that one could lower the value to 18 ohms, given that the devices was matched well enough.
Cheers !
Buhl
Its Ok that you don’t have anything to complain about - neither do I, BUT - I have built 3 Pearls in the last couple of months, and the all sounded different from each other, so there are definitely some parameters that you can change, and it will affect the sound, maybe to your liking, maybe not.
Supply noise - I have a inefficient set of Proac clones, just like you - on them, there are almost no noise - switching to my Snell E II (+5dB in sensitivity)- there are definitely noise to be heard, and more noise than ex. with the Ono. Hence, I build a version with an even more overkill PSU, with 80000 uF in the PSU (plus the 43000 uF on board) - and added two extra RC filters - the result was even less noise, but more pronounced was the reproduction of space - the room 'grew' in some way. It is quiet, but ex. the Ono is quieter......but its ok for me.
I have also builded one in a wooden box, and one in a metal box - there weren’t more noise in the wooden one, but the "metal" had more bass, and more pronounced bass - why I don’t know.
Have built it with and without battery, and there were no obvious difference in sound - except its easy to hear when the battery runs down, and that happens once a month......
I don’t have any trouble with it - builded it just like Wayne in his article - have tried a lot of different groundings - about 3 hours with the soldering iron, and the RIAA turned on, to find the grounding that made the least noise - and it was the "wayne way".
I don’t think that you can say what is best for you, if you have not heard them all - i have not heard them all, therefore I experiment.
I wont just drop the source resistors, don’t worry ;- ) - but I recall that nelson suggested that one could lower the value to 18 ohms, given that the devices was matched well enough.
Cheers !
Buhl
Allthough dependent or your specific set of 2Sk170's , but you won't run in to problem if you drop the source resistors. The bias point will change offcourse but there is room enough for that. Transductance will be somewhat higher (good for MC) en noise will be somewhat lower (again good for MC) .. THD won't change much , since this is primairily set but the output stage.
GR maybe for no source resistors
The 2SK170 came in three grades, GR,BL, and V. The differences are in "on resistance" or current flow. If you replace the BL with V, the current maybe too high. The current through GR would be low in this circuit. Tied straight to ground they would run along fine.
The 170's are low current fets. No resistors will also make matching them up more critical.
George
The 2SK170 came in three grades, GR,BL, and V. The differences are in "on resistance" or current flow. If you replace the BL with V, the current maybe too high. The current through GR would be low in this circuit. Tied straight to ground they would run along fine.
The 170's are low current fets. No resistors will also make matching them up more critical.
George
Hans, after reading your last post, I realise that I might have been to fast, out of the gates
I never heard a better RIAA than the Pearl (on my system), thats why I will think twice about modding on it. Your findings tells us, that maybe something could be gained anyway
If you didn't read them yet, here is a couple of nice articles about J-fets. They say a lot about biaspoint and Idss limiting via the source resistor and so on. "J-fet's, the new frontiers" are the ones.
Before anyone starts to fiddle with the resistor values, remember that the SK170's in Pass's Q-pack's were not matched. (Or so I was told at least)
Steen
http://www.borbelyaudio.com/special_articles.asp
I never heard a better RIAA than the Pearl (on my system), thats why I will think twice about modding on it. Your findings tells us, that maybe something could be gained anywayIf you didn't read them yet, here is a couple of nice articles about J-fets. They say a lot about biaspoint and Idss limiting via the source resistor and so on. "J-fet's, the new frontiers" are the ones.
Before anyone starts to fiddle with the resistor values, remember that the SK170's in Pass's Q-pack's were not matched. (Or so I was told at least)
Steen
http://www.borbelyaudio.com/special_articles.asp
Hi !
Now this thread has come out of the sauerkraut and back to the technical. So I´ll try again.
Here is an idea I got from an article by Jean Hiraga:
Try lowering the voltage over the FETs ( and increase it over the bipolar ). You can go as low as around 3 volts. This reduces Igx , excesive gate current. It worked well for me in a Monster. Gave a more natural sound with more wood in wood instruments , more brass in the brass etc.
Any experiences or comments ?
Thorsten ( another one )
Now this thread has come out of the sauerkraut and back to the technical. So I´ll try again.
Here is an idea I got from an article by Jean Hiraga:
Try lowering the voltage over the FETs ( and increase it over the bipolar ). You can go as low as around 3 volts. This reduces Igx , excesive gate current. It worked well for me in a Monster. Gave a more natural sound with more wood in wood instruments , more brass in the brass etc.
Any experiences or comments ?
Thorsten ( another one )
thorstenlarsen said:Hi !
Now this thread has come out of the sauerkraut and back to the technical. So I´ll try again.
Here is an idea I got from an article by Jean Hiraga:
Try lowering the voltage over the FETs ( and increase it over the bipolar ). You can go as low as around 3 volts. This reduces Igx , excesive gate current. It worked well for me in a Monster. Gave a more natural sound with more wood in wood instruments , more brass in the brass etc.
Any experiences or comments ?
Thorsten ( another one )
Hello Thorsten
Sorry if it was me that brought the sauerkraut into this thread - bu, having chosen Computer Science, instead of EE - I lack very fundamental knowledge about electronics.....
How - do you lower the voltage over the Jfets, and how do you do that ? By higher value of the source resistors ?
And the bipolars - I expect those to be the N-channel (eg. ZTX450) that are used as cascoding and regulating devices - how do you raise the voltage here - by lovering their collector value ?
I sorry, but you have to "cut this one out in cardboard" to anglophile a well known danish expression ;-)
Cheers !
Buhl
Now... I have 4 PCBs for Ono an 2 for a Pearl.....for me and my brothers....( we probably will have to draw tickets when I'm all done..)
- and I have a bucket of SK170 GR.. judging from the above, I wil do fine by shorting the source resistors... but wouldn't lowering them so -say 1-5 ohms instead help the balancing between them..??
- and I have a bucket of SK170 GR.. judging from the above, I wil do fine by shorting the source resistors... but wouldn't lowering them so -say 1-5 ohms instead help the balancing between them..??
Thorsten,
I found te following information in an article from Klaus Boehning detailing his phono pre design, at http://www.klaus-boening.de/html/lab_page.html - check the chapter Determining the lower limit of UDS . Some other references tend to also prefer a higher VDS for the common source transistor but I cannot really say what is better.
Btw., where can I find the Hiraga article, please?
The article was about a hybrid amplifier called " le Passific ". Two stages with j-fet/bjt cascode input differential driving two el84 push-pull outputs. I read a photo-copy of the article around 12 years ago. I think it was from "l`Audiophile".
The decision to lower the fet-voltage was made on subjective grounds. The lower gate leakage current was mentioned as a possible explanation for the better sound quality.
I tried it out on a "Monster amp" together with a friend, and we came to the same conclusion as Jean Hiraga.
To Buhl: On the Pearl schematics at passdiy.com, you can see two voltage-dividers each with two 10 kOhm resistors. The middle points connected to bases of bjt´s of the cascodes. To decrease the voltage over the fet, you just have to decrease the value of the lower resistor. For listening tests, just solder another over the existing one. Remember that a new solder-joint takes around 15 minutes of playing to burn in.
Just try it out. But let us know what you think.
Thorsten
( another one )
The decision to lower the fet-voltage was made on subjective grounds. The lower gate leakage current was mentioned as a possible explanation for the better sound quality.
I tried it out on a "Monster amp" together with a friend, and we came to the same conclusion as Jean Hiraga.
To Buhl: On the Pearl schematics at passdiy.com, you can see two voltage-dividers each with two 10 kOhm resistors. The middle points connected to bases of bjt´s of the cascodes. To decrease the voltage over the fet, you just have to decrease the value of the lower resistor. For listening tests, just solder another over the existing one. Remember that a new solder-joint takes around 15 minutes of playing to burn in.
Just try it out. But let us know what you think.
Thorsten
( another one )
Hans, I am sorry to have caused disruption on your thread! I see now that my post didnt have the impact, it was supposed to, at all! I did read my first post again, and didnt find it expressed, what I meant it to! The tone was too harsh, no matter how I try to turn it
I am sorry for that The sauerkraut is all mine, Hans I will try to contribute in a more polite and hopefully enlightening manner for the future Steen
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.