I played with reducing R1 to about 550 ohms. Using the 2SK170, Q5 was just starting to come out of saturation.
Using R1 at 550 ohms may move Q5 enough from the edge to a more flat area of the curve for the 2SK209.
With that initial testing, I am planning to run R1 at 750 ohms and use the 209s. Staying more with the original an see how it sounds. After things settle in I may look at reducing R3 if it seems Q5 would benefit.
Using R1 at 550 ohms may move Q5 enough from the edge to a more flat area of the curve for the 2SK209.
With that initial testing, I am planning to run R1 at 750 ohms and use the 209s. Staying more with the original an see how it sounds. After things settle in I may look at reducing R3 if it seems Q5 would benefit.
Hello,
Taking a closer look.
I plugged in the the original set of 2SK209 and noted that one had a little more current than the other 3. I removed the odd Jfet and 3 Jfets remained plugged in.
Input 5mV
Gain 50.25dB
Output 1.627Vrms
The noise floor is lower and you can see some power supply harmonics above the grass.
THD+N 0.008%
SINAD 82.2dB
The 2nd HD dropped more than a little.
With only 3 Jfets installed the current through the input stage dropped by 25% plus a little. I do not know if the cascode likes less current?
See the attached FFT.
Thanks DT
There is now less current through R1 and so less voltage drop across R1; now greater voltage across Q5, moving it further away from saturation toward a more favorable operating point.
Hello,
Quite some time ago I recall a discussion here about how the volts are divided up across R1, Q5 and the Jfets.
With one fewer Jfet the voltage drop across R1 is reduced. R1, 5.21V at 6.94ma.
The delta V across Q5 is 2.57V. I do not have curves for the PNP's. I want to see Mr. Early's effect. Later.
The Drain Source delta is 6.18V. I do have curves for the 2SK209's
Thanks DT
Going for a little road trip for a few days, I will be back.
At first I thought it was an unfavorable operating point on the Ids(-Vgs) characteristic curve, but a quick calculation and a look at the data sheet showed no problems.
But, if I briefly check your two cases statically, it follows that
1. Vce(Q5) = 1.06V with Vcb(Q5) = 0.36V
2. vce(Q5) = 2.76V with Vcb(Q5) = 2.06V
Auricle has already shown what happens in the first case in his simulations - but so far no one has run an FFT over it and looked at the resulting spectrum. If you force this point by simply injecting a current that is too high (for Q5), everything is back to normal because Q5 now represents a short circuit.
Obviously, the collector base voltage must also be statically higher than in your documented 1st case.
It is almost impossible to provide specific assistance or to identify the cause of the problem remotely - because you are not on site and cannot turn the measuring devices and power supplies.
Just leave out the cascode and reduce R1 to 590 ohms.
When analyzing a circuit, you have to proceed systematically and comparatively. It may well turn out that you have to increase the supply voltage of the BLs to up to 30Vdc so that you can be shure ...
But, if I briefly check your two cases statically, it follows that
1. Vce(Q5) = 1.06V with Vcb(Q5) = 0.36V
2. vce(Q5) = 2.76V with Vcb(Q5) = 2.06V
Auricle has already shown what happens in the first case in his simulations - but so far no one has run an FFT over it and looked at the resulting spectrum. If you force this point by simply injecting a current that is too high (for Q5), everything is back to normal because Q5 now represents a short circuit.
Obviously, the collector base voltage must also be statically higher than in your documented 1st case.
It is almost impossible to provide specific assistance or to identify the cause of the problem remotely - because you are not on site and cannot turn the measuring devices and power supplies.
Just leave out the cascode and reduce R1 to 590 ohms.
When analyzing a circuit, you have to proceed systematically and comparatively. It may well turn out that you have to increase the supply voltage of the BLs to up to 30Vdc so that you can be shure ...
Bon Voyage !
I'll certainly be keeping your solution in mind.
Higher supply V might help as well. Wayne mentioned 7818/7918 during his Burning Amp presentation, but background noise drowned him out.
Regards,
Tony Z
Now let's just look for the best V_base and the correct, but lowest Vcc - for our application. Minimum distortion and minimum noise with maximum possible dynamic signal modulation /range, in the open and closed loop ... etc. pp.
Has the value of R8 been optimally selected for our application?
These are rather the interesting questions, related to the synthesis and analysis ..!?
Regards,
HBt.
Has the value of R8 been optimally selected for our application?
These are rather the interesting questions, related to the synthesis and analysis ..!?
Regards,
HBt.
Attachments
Last edited:
Rolling the opamps in the blind....
I received 7 pair of opamps to test in my Pearl3. They are coded
A318, A170, A785 A848 , A957, A093 and A539. I added to those anonomous opamps the JRC opamps I received with the kit and a pair of Muses02 I allready had.
Mainly tested at my place with a heavily damped Lenco L75 with an Ortofon MC30 cartridge. Preamp is the balanced Iron pre from Zenmod, the 6-24 crossover from this Nelson guy, two modded ACA’s for mid and midbass, and a Reckhorn amp for the Slotted OB bass. Speakers are mahogony horns with JBL drivers from 480hz, Volt 12 inch on the SLOB baffle (60 to 480 hz) and Augie 15 inch for the slot loaded part. These tests I did together with my wife, no electronicly interested molecule in her body but she loves listening to music.
Secondly I did the comparison with a vintage Stax electrostatic headphone, very nice sound but compromised in volume. Tests done on my own.
And to verify that I was not completely imagining things I brought my Pearl3 to a diy friend with a very nice diy system ( I think) with a Lenco using a Denon DL-103 and all good Pass stuff. Together with a third audio fanatic (a semi-professional musician) I asked them to judge the 4 best opamps from my evaluation. While I did my best not to show my order of preference they agreed with me. Relieve!
So far so good, agreement between the four people involved was almost 100%.
Now to the disappointing part, of course it is all very system and record dependent, still very subjective but worse is that I find it very hard to describe the differences.
Records used were Friday Night in San Francisco (Al DiMeola Paco de Lucia John McLaughlin) Famous Blue Rain Coat (Jennifer Warnes) Ben Webster meets Oscar Peterson, Sonny Rollins Colossus, Miles Davis Ascenseur pour lÉchafaud, Dire Straits Love over Gold and J.J. Cale Troubadour.
To my ears the positives and negatives of the opamps did not change a lot with the different music. After the first round of listening the top two opamps were found and it required quite a bit of discipline to go back and listen to all of them, again and again….
I tried to bring some system in my judgment by focusing on low (clear bass notes, bass drum impact) mid (voices, guitar, sax) and high frequency (cymbal, high notes on instruments), speed and life feel. And tried to give an overall score, from 1 to 10.
I will tell my verdict of them in my/our order of preference with special aspects they did well or not so well in my opinion
JRC overall score 7. Nice sound, could easily live with those if I had not heard the others. Lowest score of all on life feel and involvement. Bit bland sound
A957 overall 7.2. no special remarks to make
A848 7.4 Did pretty well in overall sound and different frequency areas but no ‘speed’ or life feel.
A318 7,5 Weaker on bass to our ears, did well on speed and attack
A170 7,7 Nice, did well on almost all aspects but lacked life feel and involvement
A785 7,8 Really nice, Of the top 4 it lacked however most in speed and life feel
A093 7,8 Very hard to tell apart for me from A785. Just a little bit better in dynamics I feel
Muses02 8 Good, could be a permanent choice in my chain. Did everything really well in comparison to the rest of the field
A539 8, 2 In this shootout the best, more involving and more life feel, really excellent in speed and attack.
I am really looking forward to see the overall results of this (mostly) blind test and to see to what extent I agree and disagree with my fellow testers. And of course to find out where I completely was fooled by my ears...
And of course many thanks for this fun project to Mark
Leo
I received 7 pair of opamps to test in my Pearl3. They are coded
A318, A170, A785 A848 , A957, A093 and A539. I added to those anonomous opamps the JRC opamps I received with the kit and a pair of Muses02 I allready had.
Mainly tested at my place with a heavily damped Lenco L75 with an Ortofon MC30 cartridge. Preamp is the balanced Iron pre from Zenmod, the 6-24 crossover from this Nelson guy, two modded ACA’s for mid and midbass, and a Reckhorn amp for the Slotted OB bass. Speakers are mahogony horns with JBL drivers from 480hz, Volt 12 inch on the SLOB baffle (60 to 480 hz) and Augie 15 inch for the slot loaded part. These tests I did together with my wife, no electronicly interested molecule in her body but she loves listening to music.
Secondly I did the comparison with a vintage Stax electrostatic headphone, very nice sound but compromised in volume. Tests done on my own.
And to verify that I was not completely imagining things I brought my Pearl3 to a diy friend with a very nice diy system ( I think) with a Lenco using a Denon DL-103 and all good Pass stuff. Together with a third audio fanatic (a semi-professional musician) I asked them to judge the 4 best opamps from my evaluation. While I did my best not to show my order of preference they agreed with me. Relieve!
So far so good, agreement between the four people involved was almost 100%.
Now to the disappointing part, of course it is all very system and record dependent, still very subjective but worse is that I find it very hard to describe the differences.
Records used were Friday Night in San Francisco (Al DiMeola Paco de Lucia John McLaughlin) Famous Blue Rain Coat (Jennifer Warnes) Ben Webster meets Oscar Peterson, Sonny Rollins Colossus, Miles Davis Ascenseur pour lÉchafaud, Dire Straits Love over Gold and J.J. Cale Troubadour.
To my ears the positives and negatives of the opamps did not change a lot with the different music. After the first round of listening the top two opamps were found and it required quite a bit of discipline to go back and listen to all of them, again and again….
I tried to bring some system in my judgment by focusing on low (clear bass notes, bass drum impact) mid (voices, guitar, sax) and high frequency (cymbal, high notes on instruments), speed and life feel. And tried to give an overall score, from 1 to 10.
I will tell my verdict of them in my/our order of preference with special aspects they did well or not so well in my opinion
JRC overall score 7. Nice sound, could easily live with those if I had not heard the others. Lowest score of all on life feel and involvement. Bit bland sound
A957 overall 7.2. no special remarks to make
A848 7.4 Did pretty well in overall sound and different frequency areas but no ‘speed’ or life feel.
A318 7,5 Weaker on bass to our ears, did well on speed and attack
A170 7,7 Nice, did well on almost all aspects but lacked life feel and involvement
A785 7,8 Really nice, Of the top 4 it lacked however most in speed and life feel
A093 7,8 Very hard to tell apart for me from A785. Just a little bit better in dynamics I feel
Muses02 8 Good, could be a permanent choice in my chain. Did everything really well in comparison to the rest of the field
A539 8, 2 In this shootout the best, more involving and more life feel, really excellent in speed and attack.
I am really looking forward to see the overall results of this (mostly) blind test and to see to what extent I agree and disagree with my fellow testers. And of course to find out where I completely was fooled by my ears...
And of course many thanks for this fun project to Mark
Leo
I have split off a brand new thread to discuss the Pearl 3 blind opamp rolling experiments which member @Lvandoorn mentions above. This will make it easier for interested members to find and study the posts about Pearl 3 opamp rolling ... ... and it will also make it easier for UNinterested members to never see posts about Pearl 3 opamp rolling. 
The new thread is
Blind test: opamp rolling in the Pearl 3 phonostage
If you want to discuss opamp rolling in Pearl 3, please use the new thread. And please enjoy or ignore, however you prefer.
The new thread is
Blind test: opamp rolling in the Pearl 3 phonostage
If you want to discuss opamp rolling in Pearl 3, please use the new thread. And please enjoy or ignore, however you prefer.
Do it with fun.
