I think we can summerize this regulator so far:
The design is more than 30 years old, maybe it's going back to the tube age.
Improvements have been possible through the years by better components
The articles in TAA served a purpose => use the circuits, anyone, no restrictions what so ever
Everyone who want credit gets it.
The design is universal so you can use the parts you like or can get as long as they are at least technically correct.
69 boards are ordered to 40 SEK a piece which is a good price (no comments on that from the gentlemen here)
The interest for more is zero. => Commercial value zero
I have got many requests for hole mounted regulators and I have redirected those to ALW.
Some people really want to discuss the regulator when they are in a good mood and when they aren't.....
I have noticed also that AudioXpress still is selling the old original pcb (with a quality standard not year 2004) and have not updated it to the lasted version. I wonder why? I suspect this project is dead or they are selling only a few pcb's now.
Jan, why not update your pcb, since you made it, to version 2004, with enhancements? Why don't you do it in private along with Fred, Andy, Walt and Jocko? I'm sure you can make something great out of it.
What do you say, all considerations taken, professional pcb made by a professional CAD software, holemounted, SMD miniature and SMD ultra miniature? The SMD pcb's will be available ready made and also tailored for different applications. A special version for digital circuits and another one for analog circuits. Also a complete version with rectifer bridge, smoothing caps, +,- regs etc. I am convinced that you can do something great.
The design is more than 30 years old, maybe it's going back to the tube age.
Improvements have been possible through the years by better components
The articles in TAA served a purpose => use the circuits, anyone, no restrictions what so ever
Everyone who want credit gets it.
The design is universal so you can use the parts you like or can get as long as they are at least technically correct.
69 boards are ordered to 40 SEK a piece which is a good price (no comments on that from the gentlemen here)
The interest for more is zero. => Commercial value zero
I have got many requests for hole mounted regulators and I have redirected those to ALW.
Some people really want to discuss the regulator when they are in a good mood and when they aren't.....
I have noticed also that AudioXpress still is selling the old original pcb (with a quality standard not year 2004) and have not updated it to the lasted version. I wonder why? I suspect this project is dead or they are selling only a few pcb's now.
Jan, why not update your pcb, since you made it, to version 2004, with enhancements? Why don't you do it in private along with Fred, Andy, Walt and Jocko? I'm sure you can make something great out of it.
What do you say, all considerations taken, professional pcb made by a professional CAD software, holemounted, SMD miniature and SMD ultra miniature? The SMD pcb's will be available ready made and also tailored for different applications. A special version for digital circuits and another one for analog circuits. Also a complete version with rectifer bridge, smoothing caps, +,- regs etc. I am convinced that you can do something great.
AD797
The AD797 is a remarkable chip, but not in this reg.
I've always recommended and supplied the AD825 as per Walt's later recommendations and that of others who used their ears. The AD825 is cheaper, wonderfully stable and sounds best in my variation on Walt's theme.
I've tried lots of alternatives and nothing yet beats the AD825 in analogue applications. I have one AD797 based reg for a very specific application.
How many regs have you built Per? Have you put them on real-world loads? I'm willing to bet you two of my PCB's that if you build ten that not all will be unconditionally stable.
Not one of my AD825-based units has ever oscillated ,
The AD797 is a remarkable chip, but not in this reg.
I've always recommended and supplied the AD825 as per Walt's later recommendations and that of others who used their ears. The AD825 is cheaper, wonderfully stable and sounds best in my variation on Walt's theme.
I've tried lots of alternatives and nothing yet beats the AD825 in analogue applications. I have one AD797 based reg for a very specific application.
How many regs have you built Per? Have you put them on real-world loads? I'm willing to bet you two of my PCB's that if you build ten that not all will be unconditionally stable.
Not one of my AD825-based units has ever oscillated ,
Re: No contest
Somewhere in Walt's text I read that he hs had good exerience with AD797 but there has been some problems. In the end Walt says that you should deside for yourself which one to choose. I'm quite convinced that there are dozens of opamps which are suitable for the task. Note also that I have different transistors. The power transistor is 130 MHz for instance.
They sounds alright so far but I will definitely try AD825 and others later on.ALW said:
Somewhere in Walt's text I read that he hs had good exerience with AD797 but there has been some problems. In the end Walt says that you should deside for yourself which one to choose. I'm quite convinced that there are dozens of opamps which are suitable for the task. Note also that I have different transistors. The power transistor is 130 MHz for instance.
peranders said:
Actually, P-A, I am doing something great at the moment.
But it isn't regulators....Seriously, the PCB could be updated, but I don't see a '1995' or ' 2004' standard. The AA PCB was the best I could do at the moment for that particular circuit. The changes and enhancements that were published later, mainly by Walt can be accomodated on the PCB, but I agree that a revamp would make it easier for the builder. Note I said ' easier' not that it would be 'better', whatever that is. But I am against making 'universal' PCB's, the performance will undoubtedly suffer because of the inevitable compromises. Andy did a great job with his PCB, you have made a version, I don't see the need to do yet another 'me too' version.
And I am not a man to make PCBs for other circuits. I like to develop systems integrally, because you can only get optimum results if you do everything, from concept to circuit to PCB and several iterations. I can remember when Walt approached me to work together on the article series in AA. I immediately send of a list of changes I wanted to make to 'his' circuits because I had different ideas... In the end we worked very well together and I think we both learned from it, I certainly did.
But surely you noticed that Walt turned away in disgust (that's how I read it at least) because everybody was building a 'me too' Jung regulator that was only remotely recongisable as his original work. He explicitly asked this forum's members to stop calling it "Jung" regulators, but, hey, he's only the original author, so to hell with that. (Not you Andy, you are carefull in this).
No, I am not motivated to update the PCB. I like to point issues out to make people think about what they are doing. Some pick it up, come back with more questions, and embark on a wonderfull discovery in audio electronics. I enjoy that greatly, reminds me of my own journey. Some are annoyed that I don't give them ready-made answers. I give those up. Others clearly are more interested in scoring points than in finding out how things are really working. Waste of time.
That, P-A, in a nutshell is how I look at these things. That also should explain why I sometimes seem abrasive to forum members (including you). Not out of a lack of respect, but impatience and all the waste of bandwidth and effort and lack of clear thinking, especially that. It's about the *only* thing that makes us unique in this universe (so far), and we underutilize it big time. It continues to amaze me that people accept uncritically advice without explanation. If someone tells them "buy these capacitors for your preamp, they are great", they happily spend their money. I would ask, really, why? What's the difference with the other brand? Why should that difference make it a better choice? And if there is no answer or just handwaving, to hell with that 'advice'. Etc, etc. If you don't, what's the point of this wonderfull hobby of ours? You might as well collect stamps.
Sorry, got carried away.
janneman
I don't quite catch your drift.......
"Fred you know one big problem with the 797 in these regulators design is the power supply that feeds them. Using only one rail for power plays hell with the opamps PSRR and this will cause it to break in to oscillation."
The negative rail is ground. Errors between and the ground reference appear as noise and are reflected in the Negative PSRR, which is most often worse than the positive rail PSRR. While remote sensing can introduce both power supply referenced error and common voltage error signals, A good low impedance connection between the ground reference for the op amp and the negative supply eliminates the most sensitive input for noise, the negative power supply rail. I think at least couple the op amp ground reference to the negative rail (which functions as the input ground for the single supply op amp regulator) for good noise rejection at high frequencies where the PSRR ratio falls off. I would RC decouple the positive rail to op amp for the same rational of falling positive supply PSRR. Remote sensing, while great for reducing the effect of wire/trace resistance between the regulator and audio circuit, can make for some real stability headaches when using fast op amps in the regulator.
A couple of things that have been overlooked are the response of the op amp to lower rail voltages when using a single supply than two supplies. GBW usually drops a lot at lower voltages and will have an effect on stability. I believe PSSR and common mode numbers are effected as well. Great care should be taken that the reference voltage is within acceptable common mode range for the op amp at the lower supply voltage. Using the regulated output voltage for the error amp to increase effective PSRR with respect to the unregulated input voltage to the regulator is a good idea .....
if you are aware of the issues above. Ignore them and might be scratching your head for quite awhile as to why the circuit
doesn't behave as expected. While some of the new op amps are
designed to operate even with 5 single supplies, the AD797 and 825 are not but are designed for minimum supply voltage of + and - 5 volts or +10 volts and ground, with the inputs in an exeptable range near 5 volts.
"Fred you know one big problem with the 797 in these regulators design is the power supply that feeds them. Using only one rail for power plays hell with the opamps PSRR and this will cause it to break in to oscillation."
The negative rail is ground. Errors between and the ground reference appear as noise and are reflected in the Negative PSRR, which is most often worse than the positive rail PSRR. While remote sensing can introduce both power supply referenced error and common voltage error signals, A good low impedance connection between the ground reference for the op amp and the negative supply eliminates the most sensitive input for noise, the negative power supply rail. I think at least couple the op amp ground reference to the negative rail (which functions as the input ground for the single supply op amp regulator) for good noise rejection at high frequencies where the PSRR ratio falls off. I would RC decouple the positive rail to op amp for the same rational of falling positive supply PSRR. Remote sensing, while great for reducing the effect of wire/trace resistance between the regulator and audio circuit, can make for some real stability headaches when using fast op amps in the regulator.
A couple of things that have been overlooked are the response of the op amp to lower rail voltages when using a single supply than two supplies. GBW usually drops a lot at lower voltages and will have an effect on stability. I believe PSSR and common mode numbers are effected as well. Great care should be taken that the reference voltage is within acceptable common mode range for the op amp at the lower supply voltage. Using the regulated output voltage for the error amp to increase effective PSRR with respect to the unregulated input voltage to the regulator is a good idea .....
if you are aware of the issues above. Ignore them and might be scratching your head for quite awhile as to why the circuit
doesn't behave as expected. While some of the new op amps are
designed to operate even with 5 single supplies, the AD797 and 825 are not but are designed for minimum supply voltage of + and - 5 volts or +10 volts and ground, with the inputs in an exeptable range near 5 volts.
jackinnj said:
GBW is important because you want the error amp to drive the pass transistor to keep the output equal to the DC reference (x ratio), so the error amp should be able to work with whatever is impressed on the output, especially from digital loads. Also, the input noise etc should be attenuated, ideally to zero, by the (ideally zero) opamp output impedance, also for hf noise and emi that enters from the mains. This will go better if the opamp works closer to an ideal opamp up to high freqs. You need high GBW for that.
Low distortion I would think is not important for this, but low noise is, because the opamp noise will be reproduced at the output of the regulator. You could try to filter the noise from the opamp output before going into the pass xsistor, but that will also decrease the loop GBW, so you change one bad thing for another. Opamp DC offset clearly is unimportant, I mean who cares that you have 15.1234VDC instead of 15.1245VDC at the output.
A low noise high GBW opamp definitely is the best.
Jan Didden
jackinnj
It happens that many of the high-speed opamps which are low noise have a high GBW. Of course the problems is in getting a part that is fast, with low noise and has the BW to react to fast changes in response to the load. Many of these part types are not stable t low gains or even unity meaning one. So it all a trade off.
I have used a OP27 which is a very low noise opamp in the circuit however it just does not have the speed to react to load.
I had the inclination I would do some impulse testing to find out the differences between the performances of the opamps. However, I believe the work that Jung has done is clear that the HF performance is an improvement with the fast opamp.
It happens that many of the high-speed opamps which are low noise have a high GBW. Of course the problems is in getting a part that is fast, with low noise and has the BW to react to fast changes in response to the load. Many of these part types are not stable t low gains or even unity meaning one. So it all a trade off.
I have used a OP27 which is a very low noise opamp in the circuit however it just does not have the speed to react to load.
I had the inclination I would do some impulse testing to find out the differences between the performances of the opamps. However, I believe the work that Jung has done is clear that the HF performance is an improvement with the fast opamp.
Somewhere here in this thread someone claimed that it was impossible to make an universal pcb for both positive and negative voltages.
http://www.diyaudio.com/forums/showthread.php?postid=268689#post268689
Parts which are doubled: LM317/337, the reference (not entirely neccessary) and the zener. Other polarized parts ahve to be turned around, caps and LED. The only thing you may fear is the 431 which is displaced quite a bit in the negative version but you can always choose an another reference.
Isn't this task rather easy when you see the pictures?
This is the positive regulator.
http://www.diyaudio.com/forums/showthread.php?postid=268689#post268689
Parts which are doubled: LM317/337, the reference (not entirely neccessary) and the zener. Other polarized parts ahve to be turned around, caps and LED. The only thing you may fear is the 431 which is displaced quite a bit in the negative version but you can always choose an another reference.
Isn't this task rather easy when you see the pictures?
This is the positive regulator.
Attachments
SMT SuperRegulator
I finished up some layouts for an SMT version of the SuperRegulator -- this is what a 3D version looks like -- the pre-regulators are SOT-223 and Nat Semi has some specific recommendations for copper plane heatsinking:
I finished up some layouts for an SMT version of the SuperRegulator -- this is what a 3D version looks like -- the pre-regulators are SOT-223 and Nat Semi has some specific recommendations for copper plane heatsinking:
An externally hosted image should be here but it was not working when we last tested it.
I read that articles on your several permutations of the super-regulator -- I have nothing against the LM431 -- it just has to be compensated or it will oscillate.
Here's the setup I used to measure the LM329's -- the devices are housed in a cookie tin, balanced outputs are fed into my r.m.s microvoltmeter via Pomona RG58 cables. the tin itself is grounded to the voltmeter. with the inputs shorted inside the container the ambient noise is less than a microvolt.
with 1 diode the noise is ~31uV, with 2 it is 27uV, with 3 it is 10uV.
if you have a bunch of LM329's however, the noise will vary -- I don't think that this device was built like an REF-01 which has highly repeatable characteristics among a "bin" of devices.
I will repeat the tests with LM431's, REF-01's.
Here's the setup I used to measure the LM329's -- the devices are housed in a cookie tin, balanced outputs are fed into my r.m.s microvoltmeter via Pomona RG58 cables. the tin itself is grounded to the voltmeter. with the inputs shorted inside the container the ambient noise is less than a microvolt.
with 1 diode the noise is ~31uV, with 2 it is 27uV, with 3 it is 10uV.
if you have a bunch of LM329's however, the noise will vary -- I don't think that this device was built like an REF-01 which has highly repeatable characteristics among a "bin" of devices.
I will repeat the tests with LM431's, REF-01's.
Attachments
I have the 431 without any decoupling because I'll believe it isn't necessary. In the datasheet there is an area with unstableness with certain combinations of capacitance and current. I haven't ever experienced any oscillations nor have I provoked it to happen. I don't know how true the graphs are in the datasheet.
If you measure the noise you should do it after the LP-filter 1k+ 100 uF. I think the value at the opamp input is the important spot.
I say that LM431 LM329, LM385 and many more may be used. I don't prefer LM431 really. I think it is sufficiently good but there are better ones.
If you measure the noise you should do it after the LP-filter 1k+ 100 uF. I think the value at the opamp input is the important spot.
I say that LM431 LM329, LM385 and many more may be used. I don't prefer LM431 really. I think it is sufficiently good but there are better ones.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.