Surely, LT1636 is not a low noise opamp when compare with others.
49710, OPA227, OPA211 and even 5534 may give better noise figures.
But LT1636 sounds most neutral to my ears. Ad797 is also may favourite.
Most of the OPA type will color the sound.
Don't believe me.
Try them.
PS. My testing circuit differs from the 2000 version slightly. See attached schem.
Attachments
Yes, I am using several Golden Mini Regulators.
It performs very well in my gears.
When used with OPA134 or 49710, it gives an excellent C/P ratio.
I think that it is designed basing on the Super reg.
The main difference between the Golden Mini and the Super Reg is the use of compensation caps.
The possible drawback of the Golden regulators is that the pcb layout isn't 100% perfect. It is impossible to get it working with AD825. The reason for this is the grounding. The paths for power and signal ground aren't separated.
Last edited:
Good point. The originals are very thoughtfully laid out to take this into account. The Gold Reference doesn't provide for remote output sensing which is very important to maintaining low impedance.:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
HI,
Does anyone have the dropout volatge for this regulator to work?
Also for the dual +/- board, is the transformer required to have just 1 secondary?
Thanks
Does anyone have the dropout volatge for this regulator to work?
Also for the dual +/- board, is the transformer required to have just 1 secondary?
Thanks
2.5 V approx. It depends of the used LED. Blue and white LED will create higher dropout voltage.
It's best if the transformer has two secondary windings but you can also get negative voltage using one winding with half wave rectification.
It's best if the transformer has two secondary windings but you can also get negative voltage using one winding with half wave rectification.
Thanks for the info 🙂
Btw, is the opa 627ap interchangable with the lme49710 (with adjustment of voltages for the holden regulator, as i have got some no longer used 627's to play around with.
Thanks again
Btw, is the opa 627ap interchangable with the lme49710 (with adjustment of voltages for the holden regulator, as i have got some no longer used 627's to play around with.
Thanks again
John Atkinson discussed the Ayre Acoustics KX-R Twenty, a $27,500 pre -- looks like they used a shunt regulator -- perhaps picking up on some wisdom dispensed in Audio Xpress a few years ago...
Hi.
I want to make the SuperJung regulator for JLH class A amplifier. I want to run it with 2 Amperes quiescent current.
error amp will be AD797, and the pass transistor has a beta of min 120.
2000mA/120 = 17mA , and JLH with 15V +- rails, swings from 0,8A to 2,8A
2800mA/120 = 24mA
would it be somehow beneficial to omit the LT1084 preregulator, input CCS, and go straight for direct PNP pass transistor ? AD797 should easily sink those 17-24mA ?
error amp would have to equalize against refernce both ripple and load, but that his job anyway.
Or is it better to stabilize the input of SuperJung with the preregulator and CCS feeding both the AD797 and NPN pass transistor?
I want to build a pre-pre-pre reg i.e. a "normal" SuperJung, with LM317 prereg, and change the ccs to fig. 13C from article here:
https://audioxpress.com/assets/upload/files/Sources_101_P2.pdf
and build additional jfet CCS to feed the voltage reference built discrete instead of lm329, from article here:
https://refsnregs.waltjung.org/2New_Vrefs_082715_Draft.pdf. I want 3 red leds and ztx951 ztx851, according to other thread in this forum (most quied diodes), leds like it most at 5mA.
this first Superreg is supposed to supply only about 100mA of current, only for the first two CCS on the JLH from here: The Class-A Amplifier Site - JLH Class-A Update. ,
and its own voltage reference. Discrete voltage reference fed from jfet ccs and relaxed Superjung shall service both error amps in the "small" and "big" Superregs.
the power transistors however, i wanted to feed from a "direct" superreg, PNP to AD797 without preregs and ccs, 2 amperes, 1 volt ripple (22mF after bridge, SiC schottky diodes, snubbed).
cost does not matter. Does it make any sense from the musical point of view ?
I want to make the SuperJung regulator for JLH class A amplifier. I want to run it with 2 Amperes quiescent current.
error amp will be AD797, and the pass transistor has a beta of min 120.
2000mA/120 = 17mA , and JLH with 15V +- rails, swings from 0,8A to 2,8A
2800mA/120 = 24mA
would it be somehow beneficial to omit the LT1084 preregulator, input CCS, and go straight for direct PNP pass transistor ? AD797 should easily sink those 17-24mA ?
error amp would have to equalize against refernce both ripple and load, but that his job anyway.
Or is it better to stabilize the input of SuperJung with the preregulator and CCS feeding both the AD797 and NPN pass transistor?
I want to build a pre-pre-pre reg i.e. a "normal" SuperJung, with LM317 prereg, and change the ccs to fig. 13C from article here:
https://audioxpress.com/assets/upload/files/Sources_101_P2.pdf
and build additional jfet CCS to feed the voltage reference built discrete instead of lm329, from article here:
https://refsnregs.waltjung.org/2New_Vrefs_082715_Draft.pdf. I want 3 red leds and ztx951 ztx851, according to other thread in this forum (most quied diodes), leds like it most at 5mA.
this first Superreg is supposed to supply only about 100mA of current, only for the first two CCS on the JLH from here: The Class-A Amplifier Site - JLH Class-A Update. ,
and its own voltage reference. Discrete voltage reference fed from jfet ccs and relaxed Superjung shall service both error amps in the "small" and "big" Superregs.
the power transistors however, i wanted to feed from a "direct" superreg, PNP to AD797 without preregs and ccs, 2 amperes, 1 volt ripple (22mF after bridge, SiC schottky diodes, snubbed).
cost does not matter. Does it make any sense from the musical point of view ?
Hi Simulacrum
I installed four regulators in my modified JLH69 a year ago.https://www.diyaudio.com/community/threads/jlh-10-watt-class-a-amplifier.3075/post-6050539 They are also based on a WJ regulator(Improved_PN_Regs) with small modifications so that they can give over 4A. I drastically increased the current through the current source so that I could give enough base current for the pass transistor and so that the regulator could start at all. That's why I changed all transistors and added Q2 to unload the operational amplifier. The regulator has 2SC1383 / 2SA683 as drivers and KSC5242 / KSA1962 as output transistors. I did not use any pre-regulator because the input voltage is already sufficiently filtered by a CRCRC filter.
The change in sound was immediately heard and drastically so it raised the JLH69 by at least one class above. Meanwhile, Hiraga Le Monstre arrived in the same box instead of JLH69, it’s hard for me to say which is better, both are similar in sound and equally addictive.That is why Hiraga amps can be seen in the pictures now.
A couple of days ago I replaced the LM329 with the WJ LED reference you mentioned in your post, so I made four new references on a 15x15mm PCB. In these few days as I listened to the amplifier I realized that I will never use LM329 for a low noise reference again, I am very pleased with the result.
In my DAC I have already replaced the ADR525 (better than the LM4040) with a 2.5V LED reference, the regulators are again a WJ but shunt version because I use them to power the digital section. The result was the same, it was immediately heard that the new reference was better.
So in some last month I have made three 2.5V and four 6.9V Walt Jung LED references. I have to mention that I am now supplying all these references with FET CCS instead of resistor.
last picture is WJ shunt reg. (+/- 5V) with built-in 2.5V LED references and FET CCS
I installed four regulators in my modified JLH69 a year ago.https://www.diyaudio.com/community/threads/jlh-10-watt-class-a-amplifier.3075/post-6050539 They are also based on a WJ regulator(Improved_PN_Regs) with small modifications so that they can give over 4A. I drastically increased the current through the current source so that I could give enough base current for the pass transistor and so that the regulator could start at all. That's why I changed all transistors and added Q2 to unload the operational amplifier. The regulator has 2SC1383 / 2SA683 as drivers and KSC5242 / KSA1962 as output transistors. I did not use any pre-regulator because the input voltage is already sufficiently filtered by a CRCRC filter.
The change in sound was immediately heard and drastically so it raised the JLH69 by at least one class above. Meanwhile, Hiraga Le Monstre arrived in the same box instead of JLH69, it’s hard for me to say which is better, both are similar in sound and equally addictive.That is why Hiraga amps can be seen in the pictures now.
A couple of days ago I replaced the LM329 with the WJ LED reference you mentioned in your post, so I made four new references on a 15x15mm PCB. In these few days as I listened to the amplifier I realized that I will never use LM329 for a low noise reference again, I am very pleased with the result.
In my DAC I have already replaced the ADR525 (better than the LM4040) with a 2.5V LED reference, the regulators are again a WJ but shunt version because I use them to power the digital section. The result was the same, it was immediately heard that the new reference was better.
So in some last month I have made three 2.5V and four 6.9V Walt Jung LED references. I have to mention that I am now supplying all these references with FET CCS instead of resistor.
last picture is WJ shunt reg. (+/- 5V) with built-in 2.5V LED references and FET CCS
Attachments
Last edited by a moderator:
I don't think there is a need for a special regulator for the two CCS in the amplifier. I also modified them from bipolar to jfet CCS. The new CCSs also helped in the performance of the amplifier itself. CCS were done according to Walt’s Source101 articles.
I tested the regulator at full power amplifier, the voltage of +/- 19V is stable and I did not notice any signal influences in the supply voltage.
I am very pleased with the sound and I think that only with these modifications JLH69 can match Hiraga Le Monstre.
amplifier
first regulator test
