| MaxS |
Hello :D
After readings about regulatots, I have Jung super regulator ( schematics attached ) but I'd get your advice about some substitions parts.
AD 825- JFET input
- High output current
I want to replace it with TL 071 but i'm afraid by a low output current of this one which makes the transistor didn't provide too much current in the case of an HFE around 40. Otherwise, i canuse a most expensive one as OPA 13x.
2N5087
I've read I can replace it with 2N2907, so I can replace it with a BC 327. Am I wrong ?
Yhanks you for your help |
|
|
| AndrewT |
Hi,
5087 is a CCS.
Since the circuit works well at high speed I would expect the transistor to also be highish in speed and similar (or better) gain. |
|
|
| mitwrong |
MaxS:
Analyze the schematic provided, the output tr. better replaced by a darling ton. and the ccs 2n5087 is redundancy, Just a single resistor could do the job fine. it only provides the bias current to drive the output stage, due to an IC is employed for error amp, also U use a LM317 for pre-reg. this psu is good for most application, more than enough as a bench supply. IC is better to choose a high slew rate, then any change of the output voltage can immediately corrected.
Your IC took the voltage from the output side is not advised, it better get the voltage from a regulated source.
:bigeyes: :bigeyes: :smash: :smash: :smash:
WE IMPROVE BETTER DESIGN |
|
|
| janneman |
| quote: | Originally posted by mitwrong
MaxS:
Analyze the schematic provided, the output tr. better replaced by a darling ton. and the ccs 2n5087 is redundancy, Just a single resistor could do the job fine. it only provides the bias current to drive the output stage, due to an IC is employed for error amp, also U use a LM317 for pre-reg. this psu is good for most application, more than enough as a bench supply. IC is better to choose a high slew rate, then any change of the output voltage can immediately corrected.
Your IC took the voltage from the output side is not advised, it better get the voltage from a regulated source.
:bigeyes: :bigeyes: :smash: :smash: :smash:
WE IMPROVE BETTER DESIGN |
As being involved in the development of the 'Jung' superregulator, you CANNOT replace the CCS by a resistor and expect the same performance. ALL parts of this design are carefully selected to give overall top performance. Before starting to modify it (which will most likely make things worse anyway) you MUST read the articles.
Taking the opamp supply from the REGULATED output is REQUIRED for this performance. Taking it from the source will make performance worse.
I am afraid that your 'advises' make a good design worse...
A TL071 will have no output current problems; the max output current depends on the ccs current and not on the opamp. However, the '071 has less gain at higher frequencies so the Zout will increase and the ripple and noise rejection will be a bit less at the top of the audio spectrum with respect to using one of the recommended high speed opamps. Whether you can hear that depends on a lot of other factors, you could just try it out.
Jan Didden |
|
|
| mitwrong |
Jan Didden:
I do learning English, but I will never speak to a Lawyer.
May be my suggestion makes your design worse? I pointed out the reason is base on theory, have you try and test them?
I was a psu designer, would you give a comparison on the item below and see the difference?
Regulation:1 at no load
2 at full load
3 from load to off load ( transient )
4 ripple
5 holdup time ( rise and fall time )
6 noise ( spikes )
7 input AC variation ( + - 20 % )
8 stability ( change in volts duration for 8 hrs )
IC works better on regulated voltage, If the supplied voltage varies, the working condition will be change, as you use the ic as voltage error amplifier, with the supplied voltage varies, would you get stable output?
With constant current supplied to the series output transistor base, also the output of U1, How can Ib of Q1 change?
Have you checked for circuit oscillates?
This is Forum! not a place for arguing, I just point out what we have seen, From our experience, your circuit seems has a little problem. ( not at good working condition )
I've been study the cct. from famous brand psu manufacturer like Kepco, Lambda, Soresen, TDK, Astec ------etc, nothing designed as such like???????
:bigeyes: :bigeyes: :smash: :smash: :smash:
WE CHECK CCT. LOGICAL AND SENSIBLE |
|
|
| Nordic |
| Mitwrong, thanks for the thought provoking post. Would you care to post a full schematic of your modified version? |
|
|
| AndrewT |
Hi Mit,
this Forum is a place for arguing.
It's the traditional and expected way we exchange views.
That way some may come to change their view for better or worse or to expand one's knowledge.
Long may the arguments continue! |
|
|
| AndrewT |
Hi Mit,
let's see how you have changed the Jung Super Reg and give us some comparative measurements for the parts of the spec you have managed to improve. |
|
|
| mitwrong |
Nordic, Andrew T:
Thanks for your supports, Argument is accept, but don't pin point on person! We just want to discuss technical matter, design, schematic, or material ( parts ). The way we argue is base on facts, theory, results, experience, supporting figures.
sometimes may use logical sense. That's the way what we like.
For the schematic Nordic requested, I will post till I have drawn it.
Sorry for delay.
:bigeyes: :bigeyes: :smash: :smash: :smash:
WE ARGUE WITH TECHNICAL SUPPORTS |
|
|
| janneman |
| quote: | Originally posted by mitwrong
Jan Didden:
I do learning English, but I will never speak to a Lawyer.
May be my suggestion makes your design worse? I pointed out the reason is base on theory, have you try and test them?
I was a psu designer, would you give a comparison on the item below and see the difference?
Regulation:1 at no load
2 at full load
3 from load to off load ( transient )
4 ripple
5 holdup time ( rise and fall time )
6 noise ( spikes )
7 input AC variation ( + - 20 % )
8 stability ( change in volts duration for 8 hrs )
IC works better on regulated voltage, If the supplied voltage varies, the working condition will be change, as you use the ic as voltage error amplifier, with the supplied voltage varies, would you get stable output?
With constant current supplied to the series output transistor base, also the output of U1, How can Ib of Q1 change?
Have you checked for circuit oscillates?
This is Forum! not a place for arguing, I just point out what we have seen, From our experience, your circuit seems has a little problem. ( not at good working condition )
I've been study the cct. from famous brand psu manufacturer like Kepco, Lambda, Soresen, TDK, Astec ------etc, nothing designed as such like???????
:bigeyes: :bigeyes: :smash: :smash: :smash:
WE CHECK CCT. LOGICAL AND SENSIBLE |
I worked on this circuit with Mr Jung. I invite you to replace the ccs with a resistor and compare your measurements with ours.
If you want comparative data, please see the article series in Audio Amateur with full curves and specs.
Also, please look again at the circuit. The opamp IS supplied from the regulated output voltage.
You apparently do not really understand the workings of the regulator, if you say 'how can it work with a constant current?'. The base current changes because the opamp regulates the difference between the ccs and the required base current. Using a ccs instead of a resistor means that the loop gain is much greater; the opamp works with a load of the very high dynamic impedance of the ccs versus the much lower impedance in the resistor case. And, the higher the dynamic resistance, the higher the loop gain and the tighter the regulation.
If you would read the article series in Audio Amateur in I think 1994, there is a whole section where I explain stability issues and oscillation and make recommendations for the best opamp.
BTW, your suggestion to replace the series transistor by a darlington will virtually guarantee oscillations due to the increased gain with increased phase shift within the loop.
And please, don't compare a high performance circuit like the Jung reg with run-of-the-mill Lambda or Kepco. There is really no comparison!
Jan Didden |
|
|
| mitwrong |
| quote: | Originally posted by janneman
And please, don't compare a high performance circuit like the Jung reg with run-of-the-mill Lambda or Kepco. There is really no comparison!
Jan Didden |
Excuse me! I don't understand what do you mean by no comparison ? I do know US Army, Air Force, Navy, uses The above mentioned PSU manufacturer. But I never heard of Jung regulator they been employed.
Is there any Jung reg. sell on market? I really want to buy one for research, that's our interest!
We are dig into any technique that are claimed " high performance and high technique ". Please do let us know and guide us to get one.
:bigeyes: :bigeyes: :smash: :smash: :smash:
BIGEYES AND SMASH LOOKING FOR A JUNG REG. |
|
|
| janneman |
| quote: | Originally posted by mitwrong
Excuse me! I don't understand what do you mean by no comparison ? I do know US Army, Air Force, Navy, uses The above mentioned PSU manufacturer. But I never heard of Jung regulator they been employed.
Is there any Jung reg. sell on market? I really want to buy one for research, that's our interest!
We are dig into any technique that are claimed " high performance and high technique ". Please do let us know and guide us to get one.
:bigeyes: :bigeyes: :smash: :smash: :smash:
BIGEYES AND SMASH LOOKING FOR A JUNG REG. |
The supplies you mentioned are orders of magnitude worse in specs that are important to audio, that is what I mean by no comparison. I am familiar with military hardware, and I am sure a Lambda will continue to work at 105 degrees Celcius as well as survive a plane crash. Surely that is not the first priority you look for in low level audio supplies?
Many people on this forum offer Jung-type supplies, just do a search. The whole design has been extensively discussed on this forum and published in Audio Amateur, see www.audioxpress.com for more info.
Mr Jung used to be a contributor at this forum. One of the reasons he left was becaus people were thougthlessly (sp?) throwing around 'improvements' to his design without even understanding how it works.
Jan Didden |
|
|
| jackinnj |
some of the things which were learned about the super reg were published in TAA over 10 years ago -- that the AD797 might not have been the best opamp for the design, a reader with start-up problems, curing oscillation -- the one thing stressed again and again by the authors was the need for careful layout.
i had the happy occasion to build a copy of the noise measurement apparatus which was described in TAA (the TAA design used SSM2017s which are now "unobtanium"). This lead to other experiments in noise quantification, and my realization that noise, EMI, RFI were not phantoms but real gremlins in the audio chain, and this lead to a modded Adcom GFP-565 which is in constant use.
As I recall, Jan was most painstaking in making the impedance measurements of the design.
Old Colony sound sells the boards for the original design -- and TAA tells how to update them for the most recent design. the design is also discussed by WJ in EDN and is in their archives. Regrettably, TAA has not yet put out a compendium CD for their various publications of the 1980's and 1990's. |
|
|
| mitwrong |
Jan:
Sorry for my misunderstanding on the Jung reg. In Asia, we only get in touch with famous brands, and dig into it. we, the first time heard about the Jung reg. that's why argument raise up.
On the other hand, we really want to research on this reg. Please tell us more about it. Thanks, anyway.
Kindly, guide us to the web, or net where these information could be obtain. In our side, may be our info. left behind.
Anyway, thank you very much.
WE ARE HUMBLE TO LEARN NEW TECHNIQUE |
|
|
| AndrewT |
| sounds like another circuit is about to be pirated and sold on for a profit. |
|
|
| jcx |
ah yes, the ongoing "controversy"
can a ~20 year old published circuit which was never patented be "stolen"
which circuit itself could be considered as derivative of the vacuum tube "super regulator" printed in Horn&Horn nearly 50 yrs ago
If the authors wish to keep their names off of someone’s product they do have some rights, but not to the circuit itself at this point
What part of “public domain” do you not understand?
are Hugh Dean, Rod Elliot "pirates" for profiting from kits - their circuits certainly look like the "prior art" to me, just how many products out there truly have unique, enforceable IP, and does society benefit any less because some can profit filling needs with known technology?
no need to repeat this thread:
http://www.diyaudio.com/forums/show...3428#post243428
|
|
|
| peranders |
Someone (forgot who) gave me this info, from a book 1977, Mr. Kaneda. Anyone seen any older info?
Compare the schematic, it's almost identical (in general) to the Jung Regulator with the exception of newer and MUCH better parts. Mr. Jung has also changed some critical design elements. The sharp eyed person notices also that the transistor at the opamp output was later added as an improvement in the late articles. The trick with this regulator is to choose the right parts and make the pcb pattern otpimal and that is more important if you want a fast regulator. |
|
|
| janneman |
| quote: | Originally posted by mitwrong
WE ARE HUMBLE TO LEARN NEW TECHNIQUE |
Nobody knows everything, but that means we need to do our homework before venturing opinions on merits of a design.
I told you in an earlier post where the information can be found. Use the forum's search options.
Jan Didden |
|
|
| mitwrong |
| quote: | Originally posted by AndrewT
sounds like another circuit is about to be pirated and sold on for a profit. |
[QUOTE]Originally posted by jcx
[B]ah yes, the ongoing "controversy"
can a ~20 year old published circuit which was never patented be "stolen"
Hey, Gents:
In this world, not be everybody loves money, or like to steal design.
We are people just like to research on new things, a group of people like audio, psu, auto, mobile,------ etc.
We are not on business, NOTHING TO SELL, we do research on our interest, not to gain any profit out of our research. we have all types of circuitry and design, and we kept them just for our references, and we will never disclose these info to other or public, when we finished one research, the related info will be destroyed. So don't worry about we will steal other design for profit. Thats the way what we do.
If one doesn't want his or her design or schematic open to public, or let other people to know or make use of, please kept them in a safe, don't post on public!
:bigeyes: :bigeyes: :smash: :smash: :smash:
MAY GOD BLESSED THOSE DESIGN AND SCHEMATIC |
|
|
| AndrewT |
Hi,
for "Personal use" is completely different from for "commercial gain".
We have to take your word for we have no way to police your final intention.
Was the original series published in Audio Amateur? |
|
|
| jcx |
misremembered the reference
http://www.diyaudio.com/forums/show...7658#post627658
a differential input 3/4 tube "op amp" controlled paralleled pass tubes, the amplifier and the voltage reference "bootstrap" powered from the regulated output V
all of the elements of the "super regulator", including use of the name "super regulator"
Nelson's permitting "personal use" of his patented designs is the only meaningful "grant" that can be offered, because he has the right to control the use of the IP of his patents during their term
published but unpatented circuits are not controllable IP, they become “prior art” - if the authors really wanted to control it their only options were patenting or attempting to practice as a trade secret - and reverse engineering has long been recognized as legal, a sale of a device employing a patentable technology is for the patent office the same as publication
this is how technology advances; circuits, techniques, patterns of problem solving are published and become available for all to use
|
|
|
| WaltJ |
| quote: |
misremembered the reference
http://www.diyaudio.com/forums/show...7658#post627658
a differential input 3/4 tube "op amp" controlled paralleled pass tubes, the amplifier and the voltage reference "bootstrap" powered from the regulated output V
all of the elements of the"super regulator", including use of the name "super regulator" |
There's one earlier yet than Korn and Korn 2d ed., (Fig. #?) as I had noted some time back, here:
US claiming IP on feedback regulators? Why not? For example, see the 1946 MIT Rad Lab series #21, "Electronic Instruments", section 16-6, 'Practical Regulator Design / Precision DC Voltage Supplies', in Fig. 16-27 for example. A tube based feedback regulator with the VR105 reference tube fed from the regulated output. A 6SL7 differential input amplifier is used, with a 6Y6 pass tube. This section was authored by A. Jacobsen and J. V. Holdam, Jr.
This clearly predates any solid state/op amp based feedback regulators, which as far as I know came about in the 1960s and 1970s. It is very interesting to see just how much the early technology had been optimized. The circuit in question had an output impedance of 0.8 ohms, and a regulation of 0.05%. But these were serious times indeed, there was a war ongoing.
Don't look for this work online (except perhaps a purchase via www.abebooks.com , where I got my copy). Not everything of value exists on the internet, and never will. Some very good stuff still takes some hard work and digging. When one goes through this process, the knowledge gained and subsequently absorbed has greater staying power than that which is simply dropped into one's lap.
Walt Jung
I cannot help but note the irony of some comments on the developments of these regulators. The "nothing new" I find extremely odd, as if to imply that anything beyond the very first instance in a developmental chain is unworthy. By that line of thought, a 747 isn't an improvement over the Wright brother's flyer, right?
There's also some very fuzzy attributions along the way. FYI, it is Mike Sulzer, not Dave. Reading his articles does reveal this.
WaltJ |
|
|
| jcx |
Certainly the concept has deep roots and has been modified, adapted and improved in multiple incarnations, that is my point, the topology has been improved by people likely selling for profit their implementations while allowing them to become part of the published ”art” for future engineers to build on.
With no innovations the articles on your website would be a valuable contribution to teaching the “art” to new engineers, the real improvements are great to have too.
If authors in a hobby magazine were themselves selling the circuit in question I can see the courtesy of buying their version in the face of competing copies, but I don’t see how anyone could expect a published circuit to be restricted to “personal use only” forever forward. |
|
|
| janneman |
Guys,
I am not sure why this thread now seems to gyrate to the legality or not of building each other's circuits. Let me clarify my point. That was that a poster started to throw around offhand 'improvements' to the 'Jung regulator' without (as became clear quickly) even understanding how it worked. That's a great disservice to unsuspecting members that ask questions, and who are completely misled. That's not to say I or anybody else have a patent to The Trvth (tm), but at least we should engage our brain before operating the keyboard and do our homework.
I don't want to speak for Walt, and I am sure he will correct me if necessary, but at least for me it was not a point of building that design. After all, copying is the sincerest form of flattery ;)
Jan Didden |
|
|
| mitwrong |
Hi,
" TOO MUCH COOK SPOILED THE SOUP "
Am I the cook? from my point of view, I was not offhand to give " suggestion " on Jung reg. I was based on theoretical point, and no intention to disservice one's design. This is forum, we utilize it to discuss circuitry, design, as well as parts. Mainly to exchange the idea from each other, learn from each other.
No suggestions, there will be no improvement. As an engineer, he targets on technical, invention, new products as their goal to research, to find out the truth.
If any suggestion that came from us and cause your inconvenience, we say sorry about that.
PS: what is the purpose of this forum for? technical argument?
:bigeyes: :bigeyes: :smash: :smash: :smash:
LEARN FROM EACH OTHER IS ANOTHER WAY OF LEARNING |
|
|
| peranders |
| quote: | Originally posted by WaltJ
The "nothing new" I find extremely odd, as if to imply that anything beyond the very first instance in a developmental chain is unworthy. |
I think you missundestand me at least. There is a basic idea and the time has made this idea better and better by small changes in design and _big_ improvements in parts. Since I don't have the whole story from the tube regulator over to Mr. Kaneda in 1977 over to the Jung/Didden incarnation I can't really judge who made what. What I can see is that a resistor is changed to a constant current generator which certianly is smart and a real improvement. Changing from a 709 to an AD825 has to do with time only.
Walt, when you talk about improvements I think noone really has ment electrical improvement, more like practical improvement such as easier mounting, smaller pcb, more options for different parts. |
|
|
| AndrewT |
Hi Mit,
you offered improvements, whether in performance or parts simplification.
We asked you to post them with the test results of your research.
I am still waiting.
Maybe the other readers are too. |
|
|
| peranders |
| quote: | Originally posted by MaxS
Hello :D
After readings about regulatots, I have Jung super regulator ( schematics attached ) but I'd get your advice about some substitions parts.
AD 825- JFET input
- High output current
I want to replace it with TL 071 but i'm afraid by a low output current of this one which makes the transistor didn't provide too much current in the case of an HFE around 40. Otherwise, i canuse a most expensive one as OPA 13x.
2N5087
I've read I can replace it with 2N2907, so I can replace it with a BC 327. Am I wrong ?
|
Back to topic...Depending of desired output current you can use almost any small signal transistor with sufficient ratings. Hfe should be >100, better with 300 or more. fT should be > 100 MHz or more.
BC327/337/54x/55x/560 are OK examples. I have used BC550C and BC5560C with good results. The power transistor should be fast, D44H11/D45H11 are excellent but BD139/140 are also OK for smaller currents. |
|
|
| mitwrong |
Andrew T:
Thanks your support on our suggestion on the reg. improvement, Those points we suggested are base on theoretical background, and experiences.
As another member just pointed out his schematic which the ccs is replaced by a resistor, and the op-amp is replaced by 709. of cause, we haven't done this experiment, we can't say any thing on it. We suggested this item several threads before, due to we have seen this type of design several times. but this is the first time I saw design using ccs. And also the first time heard about Jung regulator.
On our market, we do know Lambda, Kepco, KDK, that's what the argue started with.
We do things not by venturing! We must have or get technical background to support what we have said. By the meantime, we are working on another project, and have to idle the regulator research, it may take some time to do it, we don't want to reveal the result without being done the test. By the time when we have done the reg. analysis and evaluation, we will announce it.
:bigeyes: :bigeyes: :smash: :smash: :smash:
WE ANNOUNCE BY FACTS |
|
|
| peranders |
| quote: | Originally posted by mitwrong
As another member just pointed out his schematic which the ccs is replaced by a resistor, and the op-amp is replaced by 709. of cause, we haven't done this experiment, we can't say any thing on it. |
I'm afraid you misunderstand a bit. My example was anno 1977 and Mr. Jung/Didden was an example of an improved verison. Not the other way around. I think examine a 709 is rather pointless. |
|
|
| AndrewT |
Hi Mit,| quote: | Originally posted by mitwrong
....Analyze the schematic provided, the output tr. better replaced by a darling ton. and the ccs 2n5087 is redundancy, Just a single resistor could do the job fine. it only provides the bias current to drive the output stage, due to an IC is employed for error amp, also U use a LM317 for pre-reg. this psu is good for most application, more than enough as a bench supply. IC is better to choose a high slew rate, then any change of the output voltage can immediately corrected.
Your IC took the voltage from the output side is not advised, it better get the voltage from a regulated source....... |
AND| quote: | Originally posted by mitwrong
........I was a psu designer..... |
Now completely the opposite.| quote: | Originally posted by mitwrong
Those points we suggested are base on theoretical background, and experiences...............his schematic which the ccs is replaced by a resistor, and the op-amp is replaced by 709. of cause, we haven't done this experiment, we can't say any thing on it. We suggested this item several threads before, due to we have seen this type of design several times. but this is the first time I saw design using ccs. |
Are you TELLING Janneman how to design it better or ASKING how it works and IF there MAY be room for improvement?
You came barging in telling everyone this and that and how it will be better and why parts were redundant.
Keep in mind readers of all levels of expertise read your posts and some may infer from your language that you know what you claim to understand.
Then when we probe you crawl out and admit you know nothing. In the meantime someone may have picked up the advice in your first post and gone and---well how bad can it get?
Make up your mind!
Are you an expert or student? |
|
|
| peranders |
| No need to being rude, Andrew. Why don't you wait and see if Mr. mitwrong will have something to show us later on. Every design can be improved and the super regulator is no exception. |
|
|
| AndrewT |
I am being rude.
Mit made out HE was the EXPERT and told us all how to do it and what bits to miss out.
My last post compared his first with his last and suddenly we see he has built nothing like it and even admits that it is beyond his experience.
That annoys me.
I hope he got the message.
I and many others go to some lengths to explain how something works, we expect newcomers to join us and we provide help and advice willingly.
I too am a relative newcomer & I get help, repeatedly. I appreciate all those efforts.
Mit could do with modifying his attitude, or I'll be rude again. |
|
|
| SY |
:cop:
Rudeness is not acceptable. It is OK to criticize ideas, not other members. |
|
|
| mitwrong |
Andrew T, and other members:
No need to care who I am, I never said I am an expert, as I replied in post 26, I gave out was " SUGGESTION ", not order! if you trust the suggestion, try it, this is FORUM, is not in court, this place is for discussion, may have arguments, anybody can give suggestion on here, right?
Is that any suggestion or ideas given in MUST BE CORRECT OR ACCURATE? I think everybody may have mistake, I am not an exception. Every body has his own knowledge, what happen if he got his knowledge out dated, or with mistake. You guys will blame him for it? is it fair? I think Mr. Andrew is a teacher, in school, is that teacher always right? no mistake at all? If a teacher has done something wrong or mistake in teaching, will the education board take immediately action?
On post 26, I already say " sorry " due to my suggestion, You guys don't know how to " forgive ". If I still see some message pin point on me, I will never show up at this forum again.
:bigeyes: :bigeyes: :smash: :smash: :smash:
FORGIVE BETTER THAN BLAME ON SOMEONE |
|
|
| mitwrong |
Mr. SY ( moderator }
One thing I want to make clarify, is anybody give in " SUGGESTIONS ' must be accurate and correct, must have evidence, or have done the test or experiment ahead on it? How about by knowledge, or passed experience?
Is it possible to have done the test, or certified the circuitry before provide " SUGGESTION ".
In other forum, that is a place where members exchange their ideas and discuss on everything. How about here? ( SAY WITH RESPONSIBILITY OR LIABILITY)
Another thing I wanted to clarify. I am an Asian, not that good in English, may be I used incorrect wordings, every time when I reply a post, there must be a dictionary I consulted with. That's why I join in this forum to learn Technical and English. I am a new member, not that clear about the rules although I have read thru.
:bigeyes: :bigeyes: :smash: :smash: :smash:
LEARNING IS ENDLESS |
|
|
| mitwrong |
| quote: | Originally posted by AndrewT
Hi Mit,
this Forum is a place for arguing.
It's the traditional and expected way we exchange views.
That way some may come to change their view for better or worse or to expand one's knowledge.
Long may the arguments continue! |
Hi, Lads and Gents:
" DOES IT MAKE ME CONFUSE "
:bigeyes: :bigeyes: :smash: :smash: :smash:
EXPERIENCE COMES FROM MISTAKE |
|
|
| radianceaudio |
Hi everyone, Walt, Jan, et al;
I think this is my first post here, I hope you will be tolerant of my really dumb questions, I am a student with no pretenses whatsoever of more than the most rudimentary knowledge. Watching me trace a circuit with my fingers is kinda like Math teachers watch a kid count on his fingers. I will try hard, however, to keep my questions as few as possible and comments scarce.
The Jung Super Regulator has me fascinated and impressed, I've spent two days now going through the circuit and I think I understand all of it except the CCS part, always a thorn in my paw, and I am about to build it up for a 6GM8 25 Volt guitar preamp aikido-incorporated project I'm working on.
I have three questions:
My first, big $64,000 question is this: what about using a MOSFET for the pass device?
I realize the driver circuitry would undoubtedly have to change....which I would not know how to do.
My desire to use a MOSFET rather than a transistor is admittedly irrational and emotional; I have been told that MOSFET'S are closer to tubes, and since the pass device is unavoidably well within the audio path of the amp, I would much rather the pass device be a tube - but needing at least 50 mA at 25 V, a tube as pass device doesn't seem, even to me, practical.
So, would anyone humor me with info on how to use a MOSFET instead, OR if using a MOSFET instead is a really bone-head idea, then I would appreciate knowing it and finding out why.
2nd big $64,000 question: if I stick with the design as-is, would someone mind guiding me to the "best" version, using opamps that are readily available from digikey?
3rd and final: why are the two opamps paralleled? Is it simply to source enough current for the driver stage?
best regards to all, charles
:D |
|
|
| janneman |
| quote: | Originally posted by radianceaudio
Hi everyone, Walt, Jan, et al;
I think this is my first post here, I hope you will be tolerant of my really dumb questions, I am a student with no pretenses whatsoever of more than the most rudimentary knowledge. Watching me trace a circuit with my fingers is kinda like Math teachers watch a kid count on his fingers. I will try hard, however, to keep my questions as few as possible and comments scarce.
The Jung Super Regulator has me fascinated and impressed, I've spent two days now going through the circuit and I think I understand all of it except the CCS part, always a thorn in my paw, and I am about to build it up for a 6GM8 25 Volt guitar preamp aikido-incorporated project I'm working on.
I have three questions:
My first, big $64,000 question is this: what about using a MOSFET for the pass device?
I realize the driver circuitry would undoubtedly have to change....which I would not know how to do.
My desire to use a MOSFET rather than a transistor is admittedly irrational and emotional; I have been told that MOSFET'S are closer to tubes, and since the pass device is unavoidably well within the audio path of the amp, I would much rather the pass device be a tube - but needing at least 50 mA at 25 V, a tube as pass device doesn't seem, even to me, practical.
So, would anyone humor me with info on how to use a MOSFET instead, OR if using a MOSFET instead is a really bone-head idea, then I would appreciate knowing it and finding out why.
2nd big $64,000 question: if I stick with the design as-is, would someone mind guiding me to the "best" version, using opamps that are readily available from digikey?
3rd and final: why are the two opamps paralleled? Is it simply to source enough current for the driver stage?
best regards to all, charles
:D |
Hi Charles,
Welcome to the forum! Hope you like it here.
That ccs is not a mystery. Do you understand that it sets up a constant current that feeds the base of the pass transistor? That the opamp siphons off any part of this constant current that is too much for the load current and that would drive up the output voltage?
So, about the mosfet. Yes, you can do it. I am not an expert in mosfets, but off the bat I see two issues:
- you will more voltage between the rectified 'raw' DC and the output voltage of the reg (the in/out difference) because the mosfet needs more drive voltage than the bjt which needs basically 0.65 V Vbe or so. On the other hand, there are mosfets that work with 3 or 4 V between gate and source so the delta is not that much and it can anyway be taken care off with enough rectified voltage. I guess you want 25VDC regulated output? So you would want at least 30VDC input or a transformer that gives off some 20VAC (or 20 x 20VAC centertapped).
- a mosfet has less gain/transconductance than a bjt, so the total loop gain available for the control loop is less with a mosfet so the measured performance will be not as good. Things like Zout, ripple rejection, rejection of ripple impressed by the load, will be up to 10 times worse. So, whatever 'magic' the mosfet brings to the circuit has to offset this. I don't know how this all would be audible.
The matter of opamps is whatever you like. I also read the reports that this or that opamp sounds more this or that, honestly I take that with a big bag of salt, but precisiely because this is all so subjective my or anybody's opinion can be opposite to your opinion. My only advice would be to stay away from the AD797 as it is prone to instability here.
Opamps paralleled: There are no opamps paralleled in this circuit. What you may have seen is peranders' circuit board for this reg where you see two different patterns for two types of opamps; you can use either one but of course NOT both at the same time.
Cheers,
Jan Didden |
|
|
| radianceaudio |
Ah, thank you so much, Jan, your information is a huge help.
:D
I'm very glad indeed to find out the schematic I looked at was just showing hookups for two different opamps; I was scratching my head indeed over why there would need to be two of them in parallel!
I'll just stick with the transistors; your comments about the disadvantages of MOSFETs sound convincing.
I found the Walter Jung circuit posted earlier in this thread, so I think I'm in good shape; I have several opamps in the solid state drawer, so I'll pull 'em out and try one out!
Cool.
I also wish to build a very good rather high voltage bench supply, something like 0 to 700 V, and am going to track down that circuit mentioned earlier in this thread that is the tube precursor for the solid state "super regulator" for lower voltages. When I find it I'll post it and ask for help re-dimensioning it.
best, charles |
|
|
| radianceaudio |
OK, I don't know what I'm doing wrong, but I can't find my way through the posted links back to the earlier tube-based "super regulator" circuit.
I would really like to build the tube-based super regulator!!
With help from you guys on redimensioning it for higher voltage and to use a couple of these gu50's that I got a whole bunch of...........
So, anyway, could someone kindly supply the circuit or the direct link to the circuit for the tube-based super regulator??
very best, charles |
|
|
| radianceaudio |
By the way, I DID find the book Mr. Jung referred to at ABE Books, but it's 75 bucks plus shipping, :bigeyes:
soooooooooo, if somebody has the circuit, it sure would be nice to save the money (that I don't have)......
but, sigh, if nobody has it, I will somehow spring for the book. :sigh:
best, charles :D |
|
|
| peranders |
| I don't know which book you mean exactly but the "Opamp applications handbook" can be bought but also be downloaded for free from analog.com Those in the business get the book for free. |
|
|
| peranders |
| quote: | Originally posted by radianceaudio
I also wish to build a very good rather high voltage bench supply, something like 0 to 700 V, and am going to track down that circuit mentioned earlier in this thread that is the tube precursor for the solid state "super regulator" for lower voltages. When I find it I'll post it and ask for help re-dimensioning it. |
Making a lab PS out of a Super Regulator is not easy and probably not recommendable. It's probably better to have a separate PS for the control section of the PS. |
|
|
| peranders |
| quote: | Originally posted by janneman
Opamps paralleled: There are no opamps paralleled in this circuit. What you may have seen is peranders' circuit board for this reg where you see two different patterns for two types of opamps; you can use either one but of course NOT both at the same time. |
The future is dark for DIL08 parts and bright for any surface mounted one so it's very convenient to be able to choose SO08 opamps which are rather easy to solder. Many new exciting opamps can't be found in DIL08. |
|
|
| peranders |
As Mr. Didden says, MOSFET's will work but there are pros and cons. A possible con may be higher capacitances which may be a problem in order to block incoming noise but MOSFET's can be seen in the Audiocom's Invisus Super Regulator, see picture below.
http://www.diyaudio.com/forums/show...9863#post639863 |
|
|
| radianceaudio |
Ah....yes, good point, Peranders, thanks,
I see it: as the regulated output varies/drops, especially under 100 V, it's going to be hard to keep the difference amp working.....and to maintain the ratios for the difference amp.....
so, the solution is to have two sections to the regulated output; a well regulated output that provides the voltage for the difference amp, and the variable output that the difference amp actually controls. And, if I'm going to have a 0 to 700 V positive rail, the difference amp probably will need to be referenced to a negative rail negative to neutral....
I'll get back to you on this as soon as I can get my hands on that :cuss: :cuss: :cuss: :cuss: tube op amp super regulator circuit......
:D
best, charles |
|
|
| radianceaudio |
P.S. The book mentioned by, I think it was, Mr. Jung, is here:
1. ELECTRONIC ANALOG COMPUTERS. (D-C Analog Computers) 2nd ed.
KORN, Granino A. & KORN, Theresa M.
Bookseller: Elgen Books
(Rockville Centre, NY, U.S.A.) Price: US$ 75.00
[Convert Currency]
Quantity: 1 Shipping within U.S.A.:
US$ 4.50
[Rates & Speeds] Add Book to Shopping Basket
Book Description: NY,McGraw-Hill,1956. 452pp., illus., ex-lib(call no.removed from spine) VG. Bookseller Inventory # 25.1720
best, charles |
|
|
| radianceaudio |
Hi Jannemann,
(it struck me that if you sign with your full name I guess I should not assume I can call you by your first....so I'll go with your username until told otherwise) :cool:
Thanks, yeah, I'm a religious reader of Mr. Broskie, this is on my list of things to work out, :)
but I'm stubbornly determined to get to the heart of this tube version of the super regulator, so I ordered the book; :crossfingers:
stubborn persistence, I have found, does make up somewhat for talent; :headbash:
after I thought about it, it seemed to me that a rather hefty reference book on tube analog computing circuits, including the super regulator, could be a treasure store of very nifty goodies indeed....
so, as soon as it gets here, and I located the schematic in question, I'll post it, probably post my own proposed modification - if I'm not baffled - and ask for feedback.
By the way, your explanation of the constant current source part of the Jung super regulator clarified that for me, thank you very much. :cheers:
best, charles |
|
|
| janneman |
| quote: | Originally posted by radianceaudio
[snip]stubborn persistence, I have found, does make up somewhat for talent; :headbash:[snip] |
In fact, I believe that what people call 'talent' is mostly stubborn persistence ;)
| quote: | Originally posted by radianceaudio
[snip][snip]after I thought about it, it seemed to me that a rather hefty reference book on tube analog computing circuits, including the super regulator, could be a treasure store of very nifty goodies indeed....
so, as soon as it gets here, and I located the schematic in question, I'll post it, probably post my own proposed modification - if I'm not baffled - and ask for feedback.
[snip]best, charles |
Sounds like a good plan.
Jan |
|
|
| jackinnj |
| quote: | Originally posted by janneman
In fact, I believe that what people call 'talent' is mostly stubborn persistence |
Among others, that quote can be attributed to homespun Thomas Edison.
http://www.thomasedison.com/edquote.htm |
|
|
| radianceaudio |
Well, I'll be.
And here I thought that was kinda original.
There's not much that hasn't already been said or thought, is there?
My thought about stubbornness substituting for talent is probably at least 100,000 years old, now that I think about it........probably not a lot of Einstein's on the savannah's of Africa back then...except relatively speaking, of course.....
:D
best, charles |
|
|
| alexan |
Hi
I've built the the following ALWSR regulator which is a based on a Jung regulator with a few changes, you can find the schematic and the circuit description in the following link
ALWSR regulator manual
Can you please tell me if you think that this circuit is better or worse from the original regulator?
Alex |
|
|
| HiFiNutNut |
Hi,
I have been very interested in building the super regulator for my line level XO network, EQ, CD player, etc.
I need +/-15V to drive up to 7 opamps.
Where can I find the schematic? Are there any kits available? I have briefly read through this site: http://waltjung.org/Regs.html however I have not found the -15V version and am not sure which schematic is the latest.
Any help would be much appreciated.
Regards,
Bill |
|
|
| radianceaudio |
I hope you guys aren't asking me, because I sure don't know the answer to either question, except that there are plenty of posted schematics for the jung super regulator in SS, but I have no idea which is best. Question: if the noise is so low your scope won't measure it, which is probably the case with any of the posted schematics, unless you've got one helluva Tek scope, why would it matter which version you use?
My own interest is in the tube version of the super regulator, and when that book finally arrives, I'll post it here and maybe we can have some tube fun.
best, charles |
|
|
| peranders |
| quote: | Originally posted by HiFiNutNut
however I have not found the -15V version and am not sure which schematic is the latest. |
It's pretty easy.
PNP->NPN
swap pin 4 and 7 of the opamp
flip the LM329
flip the zener
LM317 -> LM337 |
|
|
| HiFiNutNut |
Peranders,
Thanks. I knew it would be like that but am too lazy to work it out and wanted to be sure about it.
Somebody mailed me that he sells PCBs. I guess when in due time I will go for that.
Regards,
Bill |
|
|
| HiFiNutNut |
Actually the reason I want to try the Jung regulator is because I am a bit worried about my LM317/LM337 regulator. I followed the data sheet closely to design it and it has been working mostly very well for over 2 years. I am not too worried about the noise or the impedance because obviously the +-15V regulator gives pretty high-end sound.
The only issue is with possible conciliations. I found having low capacitance after the LM317/LM337 (less than 500uF) does not sound good. Adding a Rubycon ZL 2,200uF sounds very good. Of course, for each opamp, I use 125uF per rail, bypassed by 0.01uF film, and a 0.1uF film cross the rails. This sounds the best for me. Noise or impedance is not an issue.
But in a few occasions, which I can not replicate because it is intermittent, I saw my 12" woofers moving in and out at maximum excursion slower with no signals! I reckoned it was some resonance at only a few Hz. Where did they come from, I guess they may come from the low impedance large caps after the LM317/LM337. Because of the low impedance and low ESR, the resonance can not be damped sufficiently.
But would I get the same problem with the Jung regulator? I mean if I add large capacitance after the Jung regulator, would it conciliate at some points? You may say no capacitance is needed after the Jung regulator. Would it sound good in that case? Would it be able to deliver the transient at demands without the capacitor?
Has anyone done any critical A/B test between the sound of a Jung regulator and a very well implemented LM317/LM337 regulator with large caps?
Regards,
Bill |
|
|
| Damon Hill |
Concilate? I think you mean oscillate?
Your low-frequency instability suggests "motor-boating", which is a type of instability probably within whatever amplifier you're using, or one of your regulators is severely misbehaving.
http://cache.national.com/ds/LM/LM117.pdf
Follow the instructions on the data sheet on page 8 and 9. Excessive capacitance on the output is probably not very helpful; 10 uF - 100 uF
on the adjust pins of the regulators are helpful and larger amounts will not have much useful effect.
Large amounts of capacitance on the output of the Jung regulators are unnecessary; the design already has a very low output impedance at all audio frequencies--when constructed properly! Read the application notes very carefully and follow them to the letter. This is a very sophisticated and optimized design; I had to study the notes for quite a while to comprehend them and properly implement them in my applications. |
|
|
| peranders |
| The purpose of adding capacitance at the regulator output is to lower the output impedance but the super regulator has already extremely low output impedance, like microohms. Therefore adding something with milliohms seems rather pointless. The super regulator creates in an artifical way an extremly "large" capacitor. My advice is to have 100-470 uF max at the output, preferrable down towards 100 uF. |
|
|
| janneman |
| quote: | Originally posted by HiFiNutNut
Actually the reason I want to try the Jung regulator is because I am a bit worried about my LM317/LM337 regulator. I followed the data sheet closely to design it and it has been working mostly very well for over 2 years. I am not too worried about the noise or the impedance because obviously the +-15V regulator gives pretty high-end sound.
The only issue is with possible conciliations. I found having low capacitance after the LM317/LM337 (less than 500uF) does not sound good. Adding a Rubycon ZL 2,200uF sounds very good. Of course, for each opamp, I use 125uF per rail, bypassed by 0.01uF film, and a 0.1uF film cross the rails. This sounds the best for me. Noise or impedance is not an issue.
But in a few occasions, which I can not replicate because it is intermittent, I saw my 12" woofers moving in and out at maximum excursion slower with no signals! I reckoned it was some resonance at only a few Hz. Where did they come from, I guess they may come from the low impedance large caps after the LM317/LM337. Because of the low impedance and low ESR, the resonance can not be damped sufficiently.
But would I get the same problem with the Jung regulator? I mean if I add large capacitance after the Jung regulator, would it conciliate at some points? You may say no capacitance is needed after the Jung regulator. Would it sound good in that case? Would it be able to deliver the transient at demands without the capacitor?
Has anyone done any critical A/B test between the sound of a Jung regulator and a very well implemented LM317/LM337 regulator with large caps?
Regards,
Bill |
Hi Bill,
Having worked with Walt Jung on his super reg project, let me give you my take on this.
First of all, your slow moving woofers are not caused by any power supply issues. There is no 'direct' connection from the supply to the woofers. Even if the supply would slowly vary, still the amp controls the speaker. What you might have witnessed is a very low freq amp oscillation which in the old tube days we called 'motorboating'because of the sound it made. I have no idea how that happened in your amp, possibly you have ground loop issues.
To the regs: The 317/337 regs are among the best of the integrated regs, but are no match to a well designed discrete reg like Walts. Now I base this on measurements and how well the circuit emulates an 'indeal' voltage source. Whether there is an audible difference in practise depends on a lot of other things, of course. Many amps simply don't have the quality to show up these differences. For instance, the Jung reg can only shine if the layout and grounding (and if used the remote sense) is meticulously executed as described in the original article. You mess that up, and you'r no better than any run-of-the-mill reg. Same care for grounding etc should go into a 317/337.
It really doesn't pay to have large caps at the output of either reg, because the regs are best in exactly the freq range where the cap would work: on the lower side. The very best is to have enough cap at the reg to keep it stable (and these should NOT be low ESR caps) and then put your best caps directly at the circuit you are powering, NOT at the reg side.
In the end, what's best? From a theoretical view, the superregs are best, but unless you are very meticulous in the layout and grounding and fully understand the issues, you probably would have the same final quality with a well executed 317/337.
Jan Didden |
|
|
| consort_ee_um |
| With a large capacitor on the output of a voltage regulator you can get this happening- At switch on the capacitor has a large inrush current so the voltage regulator current limits, then relaxes again but the capacitor demands more current so the regulator limits again. You get a slow oscillation seen as your cones moving. As Janneman said you do not need a lot of capacity on the output of a regulator. I would go lower than peranders say 1uF to 10uF |
|
|
| janneman |
| quote: | Originally posted by consort_ee_um
With a large capacitor on the output of a voltage regulator you can get this happening- At switch on the capacitor has a large inrush current so the voltage regulator current limits, then relaxes again but the capacitor demands more current so the regulator limits again. You get a slow oscillation seen as your cones moving. As Janneman said you do not need a lot of capacity on the output of a regulator. I would go lower than peranders say 1uF to 10uF |
The cone movement has NOTHING to do with any slow moving supply. And the supply oscilation you mention is extremely rare and only happens with specific low frequency loop gains. The reason for not having a large cap at the output is simply because it is a waste of money and pcb space.
Jan Didden |
|
|
| darkfenriz |
Sorry for a bit off topic, but has anyone examined transient response of 317/337 regs?
They seem to have about 10kHz ringing with purely capacitive load, don't they?
regards |
|
|
| janneman |
| quote: | Originally posted by darkfenriz
Sorry for a bit off topic, but has anyone examined transient response of 317/337 regs?
They seem to have about 10kHz ringing with purely capacitive load, don't they?
regards |
I'm not sure about the exact frequency, but this is the reason why you should NOT use low ESR caps directly after the reg. Farther away, at the circuit to be powered, is usually OK since the reg and low ESR cap are isolated by the wiring inductance.
Jan Didden |
|
|
| consort_ee_um |
Janneman Nothing to do with loop gains merely the characteristics of the current limit. Turn on transient with large cap on output of regulator could cause the speaker movement mentioned.
With the integrated voltage regulators you need some output capacitance for stability typically 0.1 to 1uF |
|
|
| radianceaudio |
uhh....this may sound dumb, but sometimes dumb is right because it got overlooked.
HiFiNutNut: are you using a turntable/vinyl? are you only seeing this woofer excursion while a needle is tracking in between musical sections? in a system with extreme low end response, a little bit of warp in the record, or even a center hole slightly offset, will cause woofer excursion.
That, or motorboating in the amp itself (mostly likely the output section of the amp), would be the most likely causes, not the power supply.
best, charles |
|
|
| HiFiNutNut |
Oh, did I say concilate instead of oscillate? I obviously spelled it wrong when I first typed it then I used a spell checker that picked it up but in a hurry I must have picked a wrong word then did a global replacement ending up with concilate!
Thank you for all your enlightenments.
That motorboating happened a few times, not at switch-on time. Lately I have not seen it again. Thank you for pointing it out that it is not an issue with the regulator.
My speakers are the John K's NaO. My power amps are my DIY dual 400W monoblocks and dual 135W monoblocks. They are Randy Slone's Optimos. The amps have been working fine but lately I found when turning off the 400W monoblocks it generated a tiny bit of noise at the woofers. I guess I will check them out and see if there are something wrong. I built my own line level XO circuit using my own LM317/337 regulators. The preamps are the 12B4 or Aikido. The CD players are my modified NAD542 or Marantz SA11.
Now back to the topic of the regulator.
Yes I have been having this puzzle. Given that the impedance of the LM317/337 regulators is much lower than even a low ESR cap at low frequencies, the capacitor should have no impact on the sound because power should be drawn from the lower impedance source. However, in practice, you can EASILY find a large low ESR cap after the regulator having a MAJOR influence on the sound. I have tried a number of times on my NAD542 CD player placing the caps after the regulators for the analogue stage. The analogue stage only uses one dual opamp per channel.
With about 10uF Nichicon Fine Gold after the regulator (as the way it was manufactured), the sound is thin and dry.
With a 4,700uF Panasonic FC, the sound is messed up completely. The cap was placed about 50mm from the regulator.
With a 2,200uF Rubycon ZL, the sound is full, balanced and musical.
So why? it does not make sense to me.
The same is applied to my LM317/337 regulator with my line level XO. I used a high ESR 100uF cap after the regulator. 60mm away, I used the 2,200uF Rubycon ZL. The sound is good. I tried twice taking out the 2,200uF, and found the sound thin and dry. So the 2,200 was back on the board.
Also a new Question: all opamp books recommend bypassing the opamps with good caps - shunting with film, ceramic or tant, etc. These are low ESR caps. I have been using LOW ESR types (Panasonic FC, Rubycon ZL) 125uF to bypass the +-15VDC rails just next to each opamp, also shunt with a 0.01uF polystyrene, as well as using a 0.1uF MKP directly cross the positive and negative rails. Would this arrangement work with the Jung regulator?
Many thanks.
Bill |
|
|
| Damon Hill |
If you'll read the documentation and study the schematic, you'll see that the Jung regulator is already thoroughly well by-passed--exactly where it needs it. You can experiment, but I can't promise that the results will be good.
What sort of test equipment do you have? Possibly an oscilloscope?
Go forth now, and read the documentation. You really need to do this first. Your questions imply that you haven't done so. |
|
|
| darkfenriz |
Thank you Jan
Now I use 1 ohm resistor directly after regulator and then capacitor and this seems to cure the situation, doesn't it?
Adam |
|
|
| mr.duck |
I've got quite a good idea of how the jung reg works but I have a few basic questions.
1) The current source is formed with the LED, the resistor in series with the LED, and the small signal transistor? (or is it just the LED and resistor?).
2) Is output impedance affected much by the output impedance of the tracking pre regulator (LM317)?
3) How does the error amp control the output voltage? Does it just output a voltage (little or no current) to control how much current flows into the base of the pass transistor from the current source? Or does it control the output voltage by driving more current into the base of the pass transistor?
Thanks for your help. I have a couple more questions depending on what the answer is to that last one. |
|
|
| peranders |
1 The LED is creating a constant voltage for the current generator formed by a teansistor and a resistor (22-100 ohms). The light is a nice side effect :nod:
2 No
3 The opamp tries to lower the output voltage. Inactive opamp = full output voltage. Base current through the pass transistor comes from the current generator. |
|
|
| janneman |
| quote: | Originally posted by darkfenriz
Thank you Jan
Now I use 1 ohm resistor directly after regulator and then capacitor and this seems to cure the situation, doesn't it?
Adam |
I hope you didn't put the resistor in series with the output (between output and load)?
Jan Didden |
|
|
| radianceaudio |
oh, yeah, that's what he did, I was sure of that when I read it. He essentially inserted a long cable before the bypass cap at the powered device.
I don't think he's understanding that the pass transistor in the output of the Jung S R is providing both regulation and very low impedance, that there's no need for a capacitor at all at the output terminal......that, as you well know, adding a cap just messes things up..... :laugh:
best, charles
still waiting for that book from ABE....aaarrggg!! |
|
|
| janneman |
| quote: | Originally posted by radianceaudio
oh, yeah, that's what he did, I was sure of that when I read it. He essentially inserted a long cable before the bypass cap at the powered device.
I don't think he's understanding that the pass transistor in the output of the Jung S R is providing both regulation and very low impedance, that there's no need for a capacitor at all at the output terminal......that, as you well know, adding a cap just messes things up..... :laugh:
best, charles
still waiting for that book from ABE....aaarrggg!! |
I resent your post. Making fun of anybody who knows different things then you (if you know these things at all, which is not at all clear to me) is a cheap shot. I hope you enjoy it in my ignore file.
Jan Didden |
|
|
| darkfenriz |
| quote: | Originally posted by janneman
I hope you didn't put the resistor in series with the output (between output and load)?
Jan Didden |
hmmm... yes I did, why not? |
|
|
| janneman |
| quote: | Originally posted by darkfenriz
hmmm... yes I did, why not? |
Well, you neatly isolated the low Zout of the reg from the load! Your reg (when implemented well) will have a Zout of a few milli-ohms. (An ideal voltage source would be zero of course). Now with the 1 ohms, the Zout will be 1 ohms in parallel with the impedance of that cap, but that varies with frequency. You more or less throw away the reg with this!
What you could do is leave the R, C in but take the load, as normally done, from the reg, on the left side of the 1 ohms!
Your R,C will still damp the reg, and you still have low Zout. Eat your cake and have it too.
Anyway, I think that 1 ohms and the cap is really not necessary anyway. Any old crappy electrolytic will do. The crappier the better ;)
Jan Didden |
|
|
| mr.duck |
| quote: | 1 The LED is creating a constant voltage for the current generator formed by a teansistor and a resistor (22-100 ohms). The light is a nice side effect
2 No
3 The opamp tries to lower the output voltage. Inactive opamp = full output voltage. Base current through the pass transistor comes from the current generator. |
ah so when picking an opamp, output current capabilities are not important. AD9631 should be a good alternative to AD825? It has 1.3kV / uS slew rate :) maybe that opamp would be too fast and would not be stable. |
|
|
| radianceaudio |
Good heavens, Janneman,
I had no intention of making fun or taking a cheap shot.
I certainly apologise for being sloppy in my language and giving that impression.
I made the mistake of posting quickly without taking the time to use super-polite language to make my intent clear. My bad.
Email, which is what this essentially is, is a very unforgiving communication medium, and it's equally easy to mis-post intentions and misinterpret what one reads.
Really, I was just stating an observation, an observation that was perfectly correct. If darkfenriz wants to take offense I'll happily apologize to him....when I put :laugh: there I meant it as friendly "we're all human and sometimes misunderstand", I've never knowingly taken a cheap shot at anyone in my life,
and sheez you seem a bit touchy, frankly.
If you want me to leave I'll do so. Luckily, this is not the only forum on the planet, and you are not the only expert on regulators.
best, probably about to say goodbye to this forum forever, charles |
|
|
| peranders |
| quote: | Originally posted by radianceaudio
Good heavens, Janneman,.....
and sheez you seem a bit touchy, frankly.
|
I couldn't either undertand the touchness :scratch:
| quote: | Originally posted by radianceaudio
Luckily, this is not the only forum on the planet, and you are not the only expert on regulators. |
True, but you have not so many around. I think Mr. Didden knows quite a lot :nod: |
|
|
| pwillard |
I think it was pointed out...
With a good power supply, (and I think it was made clear that an LM317 can make a good supply), the Low Frequency oscillations seen at the amp output are rarely due to power supply issues.
Randomly tossing in "two pinners" is also likely to have marginal success unless there is good reason to do so. The Data Sheet makes it clear that lower value, not higher value capacitance should be used right at the output pins of the regulator.
My opinion is that the amps stability (and grounding) needs to be examined before the power source if the power source is of a standard design. |
|
|
| SY |
| quote: | Originally posted by darkfenriz
Sorry for a bit off topic, but has anyone examined transient response of 317/337 regs?
They seem to have about 10kHz ringing with purely capacitive load, don't they?
regards |
It depends on the cap. The output impedance of a 317/337 looks like an inductor of a few uH. With an undamped cap at the output, a high Q resonance can be formed. That's why it's sometimes useful to add a little ESR to the cap or use a big, sloppy, high ESR cap at the output. |
|
|
| janneman |
| quote: | Originally posted by radianceaudio
Good heavens, Janneman,
I had no intention of making fun or taking a cheap shot.
I certainly apologise for being sloppy in my language and giving that impression.
I made the mistake of posting quickly without taking the time to use super-polite language to make my intent clear. My bad.
Email, which is what this essentially is, is a very unforgiving communication medium, and it's equally easy to mis-post intentions and misinterpret what one reads.
Really, I was just stating an observation, an observation that was perfectly correct. If darkfenriz wants to take offense I'll happily apologize to him....when I put :laugh: there I meant it as friendly "we're all human and sometimes misunderstand", I've never knowingly taken a cheap shot at anyone in my life,
and sheez you seem a bit touchy, frankly.
If you want me to leave I'll do so. Luckily, this is not the only forum on the planet, and you are not the only expert on regulators.
best, probably about to say goodbye to this forum forever, charles |
OK, so I was a bit touchy. But as you yourself state, quick posting for essentially the whole world make the message very easy to misunderstand. I have seen too many personal ridicules at this place, so maybe I was a bit too primed to see it in your post. If anything it makes it clear that we all must engage brain before operating keyboard. Apologies for misjudging you.
And hey, reading your last sentence, who's the touchy one now ;) ??
Jan Didden |
|
|
| pwillard |
More from the datasheet:
Adding a capacitor to the ADJ pin will improve ripple rejection.
Adding a capacitor to the OUT pin improves transient response. 10uf to 25uF are ideal.
Quote from data sheet:| quote: | | Any increase of the load capacitance larger than 10 μF will merely improve the loop stability and output impedance. |
BUT... any value higher than 25uF will allow excessive current flow in the case of a short to INP pin and would likely damage the regulator unless protection diodes are used. I noticed from the design, sweet as it is... that they are not used. |
|
|
| mitwrong |
Janneman:
From the post argument between you and the other members, I think you are too protective and old minded, especially on the design of the Jung reg. When anybody said or gave idea on the Jung reg. design, you always come out and give some sort of argue or defence. As one member said: you are not the only reg. designer in this world. Technology is day by day going up. make your eyes wider and look things further. and also this is forum, everybody has his or her right to criticize or praise any design or ideas, but you always stay on the defence side, you should point out the reason what mistake the member has been made. in this manner, other people will agree your point more than to against. I think you are an engineer grade person, must accept challenge concept from others. Even yourself, should criticize your old design and work out a new improved version. Today there are lots of new improved parts that may replace the old designed parts.
I know, the Jung reg. is an excellent design, but I still think there must be room for improvement. only time can show us proof and evidence. I think even Walter J will agree my point.
If any impression caused by the above message, Have beg for your forgiving manner.
NO CHALLENGE, PEOPLE TODAY MAY NOT LANDED ON THE MOON |
|
|
| janneman |
| quote: | Originally posted by mitwrong
Janneman:
From the post argument between you and the other members, I think you are too protective and old minded, especially on the design of the Jung reg. When anybody said or gave idea on the Jung reg. design, you always come out and give some sort of argue or defence. As one member said: you are not the only reg. designer in this world. Technology is day by day going up. make your eyes wider and look things further. and also this is forum, everybody has his or her right to criticize or praise any design or ideas, but you always stay on the defence side, you should point out the reason what mistake the member has been made. in this manner, other people will agree your point more than to against. I think you are an engineer grade person, must accept challenge concept from others. Even yourself, should criticize your old design and work out a new improved version. Today there are lots of new improved parts that may replace the old designed parts.
I know, the Jung reg. is an excellent design, but I still think there must be room for improvement. only time can show us proof and evidence. I think even Walter J will agree my point.
If any impression caused by the above message, Have beg for your forgiving manner.
NO CHALLENGE, PEOPLE TODAY MAY NOT LANDED ON THE MOON |
Hi Mitwrong,
Thanks for your frank opinion. I don't want to be too defensive for these kind of things, and I have no obligation to WJ or anybody else. But you see, if someone proposes some change to the super reg, and I KNOW it will totally destroy the fantastic performance (because I tried it and learned the hard way), what do I do? Just shut up and let the other guy waste his time and money and fall on his face? Some of these guys don't even know how the thing works. Should I just shut up, LOL, and forget it?
Maybe I should, maybe that is the only way to learn, as we all did.
I hope (at least I try) to be open to genuine improvements to super regs, whatever. But if someone proposes to use an expensive supercap at the reg output, should I not point out that that is a) making it probably an oscillator instead of a reg, and b) money much better spend by putting the cap at the circuit to be powered? We all know that a person putting a 20 $ cap at the reg most probably will report an improvement, even when it oscillates, so maybe there's no point in trying to get something across.
I know, I'm too serious sometimes. It's only audio. ;)
Happy Holidays,
Jan Didden |
|
|
| mitwrong |
| quote: | Originally posted by janneman
Hi Mitwrong,
Thanks for your frank opinion. I don't want to be too defensive for these kind of things, and I have no obligation to WJ or anybody else. But you see, if someone proposes some change to the super reg, and I KNOW it will totally destroy the fantastic performance (because I tried it and learned the hard way), what do I do? Just shut up and let the other guy waste his time and money and fall on his face? Some of these guys don't even know how the thing works. Should I just shut up, LOL, and forget it?
Maybe I should, maybe that is the only way to learn, as we all did.
I hope (at least I try) to be open to genuine improvements to super regs, whatever. But if someone proposes to use an expensive supercap at the reg output, should I not point out that that is a) making it probably an oscillator instead of a reg, and b) money much better spend by putting the cap at the circuit to be powered? We all know that a person putting a 20 $ cap at the reg most probably will report an improvement, even when it oscillates, so maybe there's no point in trying to get something across.
I know, I'm too serious sometimes. It's only audio. ;)
Happy Holidays,
Jan Didden |
Hi, Mx Janneman:
First, can you tell us whether you are a gentleman or a lady, so that we can put the heading Mr. or Ms. before your name. And also what do you want the name we people call you, Janneman or Jan Diden, we don't want to cause any embarrassing.
As you are an expert in regs. we must respect whatever you said on the designed issue. But another important factor to destroy the performance is the GROUNDING, expensive caps, I know is not an essential factor. But somebody may like to use expensive parts. ( Like ladies want to buy famous brand products. Or audiophile want to use Mullard or Telefunken tubes.)
I have tried many experiment, Grounding loop is an important factor which can turned out good or bad results. Another component is the power transformer, it also takes an important place. The whole circuit wiring also plays an important route. A stand alone psu is differ from a psu that installed inside an equipment. Therefore you may instruct us how to precaution about the ground loop. ( is our request, not an obligation. )
Last, we want to thank you for guiding us on the straight road to the excellent design, and we don't want you to keep your mouth shut. We still need your help.
Merry Xmas, and happy new year
Mitwrong |
|
|
| jackinnj |
| quote: | Originally posted by mitwrong
Hi, Mx Janneman:
First, can you tell us whether you are a gentleman or a lady, |
Enough of this ill mannered troll. |
|
|
| SY |
| Mitwrong, Jan is a very well-known expert in regulator design and has done extensive collaboration with Walt Jung in the past. He is most definitely male. I'd probably listen carefully to what he has to say- you might disagree, and it's OK to do so within the conte | | | |
