Simple Super Shunt and Simple Super Shunt w/CCS vs. JSR03/05 (and other series regs)

In the Feb 2008 issue of AudioXpress (US print magazine) pp. 30-37, the article "Shunt or Not" by author Are Waagbo presents two "sophisticated shunt regulator version":

AudioXpress-02-2008_p34_fig-5_simple_shunt_reg.jpg


AudioXpress-02-2008_p34_fig-5_simple_shunt_reg_with_CCS.jpg


The author claims that in comparison to his shunts, the JSR03/05 was “a waste of money” (noting a simple zener shunt was better). He also indicates other series-regulator designs (e.g., Borbely and Sulzer) are not much better.

Along with his own, Waagbo does credit Borbely’s shunt regs.

What do you folks think? Can a simple Zener-resistor-capacitor shunt reg. outperform JSR or Sulzer. If so, then the more-sophisticated shunts (above) must be really spectacular – or are they?

Thx for any feedback you can provide.
 

Onvinyl

Member
2002-08-02 10:41 am
Germany
hollowman said:
Along with his own, Waagbo does credit Borbely’s shunt regs.

What do you folks think? Can a simple Zener-resistor-capacitor shunt reg. outperform JSR or Sulzer. If so, then the more-sophisticated shunts (above) must be really spectacular – or are they?

Thx for any feedback you can provide. [/B]

No wonder he credits borbely's designs -- the shown circuits are almost entirely a downgrade of those!

The claim 'better than jung' is, in this form, simply ridiculous. Specwise, jung beats anything so simple. The sound is a matter of taste, of course. My main concern with the proposed circuit would be to replace the zener with a really low noise assembly.

I *do* like shunt regs, even the fancier ones -- there are quite some examples here on diyaudio during the last few months, it's just simply both silly and unfair to bash the jung-didden circuit in comparison.

And yes, I think the shown ones will work pretty good!
Rüdiger
 
Re: Re: Simple Super Shunt and Simple Super Shunt w/CCS vs. JSR03/05 (and other series re

Onvinyl said:
The sound is a matter of taste, of course. My main concern with the proposed circuit would be to replace the zener with a really low noise assembly.
Can you suggest a "really low noise assembly"? In any case, the author concurs about the noise in the article. But I think he ultimately feels the noise is a fair trade-off for whatever sonic gains/benefits the simple zener-based ckt provides.
I *do* like shunt regs, even the fancier ones -- there are quite some examples here on diyaudio during the last few months, it's just simply both silly and unfair to bash the jung-didden circuit in comparison.
Can you point to some of those "examples here on diyaudio during the last few months"? Links to the best of those would be helpful. Thx!
And yes, I think the shown ones will work pretty good!
If you build them, let us know. I'm having problems finding some of the parts.

BTW (and IIRC) ... about a year ago there were a series of articles in this same mag by Jung...praising shunts. I think the full versions of these 2007 articles may be found here:
http://waltjung.org/Late_Articles.html
Maybe the guy is re-evaluating his beliefs?

FWIW ...what I can says is that quite a lot of the really high-end CD/digital gear (Spectral, Wadia, Muse, etc.) opt for carefully-implemented shunt regulation, not series.
 

Onvinyl

Member
2002-08-02 10:41 am
Germany
Hollowman,
if you do a search with the criteria 'shunt' and 'search titles only' you may find e.g.:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=110424&highlight=

http://www.diyaudio.com/forums/showthread.php?s=&threadid=114070&highlight=

http://www.diyaudio.com/forums/attachment.php?s=&postid=840668&stamp=1139437025

http://www.diyaudio.com/forums/showthread.php?s=&postid=1399081#post1399081

With really low noise I mean to avoid noisy zeners. You might try LED's , or use a low noise diode like LM329DZ, or might even get by with a CCS+resistor.

Don't get the impression I don't like shunt regs! I do love them. I wouldn't place my bets that a superreg couldn't sound darn good, that's all.
I'm trying quite some at the very moment, and will make a shootout between different regs with a mc-phono head amp (that is still on the to build list, anyway...)

Rüdiger
 
shunt regs+last cap.

Hi Gents,

The last time i'm playing with different regulators for my UGS-line amp(see UGS-thread on Pass Labs). I was very supprised last week when i reed the review in audioxpress about those shunt regs.
In fact i use the same simple shunt reg which was also a publication in AX a few years ago. Norman Taggard(i hope the name is correct) was the author.
I''m very satisfied about that circuit but replacing the shunt resistor(56R in my case) for the Borbely current source made no difference for me.
I.M.H.O. the power supply rejection of the used preamp makes the difference but the author of the shunt article gave not much information about the used setup and It's all written with a huge pro borbely cab.
As i said, i'm fooling around with different(shunt) regulator configurations but everytime that last cap is the bottleneck.
Yes, the panasonics FC-series are also very good for me. When i swap it for an elna audio grade i become depressive. But in fact we are talking about colouration.
How to solve that?

Gr.
Johan
 

Attachments

  • pre_shunt_reg.png
    pre_shunt_reg.png
    17.6 KB · Views: 4,814
Re: shunt regs+last cap.

joho said:
I.M.H.O. the power supply rejection of the used preamp makes the difference but the author of the shunt article gave not much information about the used setup and It's all written with a huge pro borbely cab.
Nice work on your ckt! I found it somewhat similar to Peufeu's CCS/shunt PS. The AudioXpress author found the use of pre-reg 317 limiting (music more compressed), but maybe your design does not behave that way.

I, too, found that AudioXpress article confusing in certain regards (maybe it's the author, but the chief editor should've done a better job, too). Some main points of confusion regard component suggestions and labeling, notably transistors:

In the schematic below, what is:
- the FET transistor above the 221R (and what's its negative version)? (See below)
- the optimal negative version of BC549C and BC559C in the positions they are placed in the schematic? Are they indeed BC559C and BC459C, respectively?

AudioXpress-02-2008_p34_fig-5_simple_shunt_reg_with_CCS.jpg


The negative (-) version of a regulator -- sometimes, for better performance -- may not use the complimentary opposite transistor and/or topology. And, in any case, model numbers (nomenclature) for negative/positive pairs may not be so apparent; E.g. for output transistors … Positive N-channel Toshiba 2SK1530 compl. to Negative p-channel Toshiba 2SJ201. For shunts, author notes use either same as output OR Positive P-channel IRFP9140 and Negative N-channel IRFP150.
 
Hi Hollowman,

The use of a pre-reg before the normal LM317/337 circuit isn't my own idee. Walt Jung and ALW did this before. I haven't the feeling that there is any compression. I can very easy make an A/B check and i didn't hear any difference but i found out that the last reg(the shunt reg) is very dominant. When i start to test the circuit i had some oscillation trouble but C2/C7 solved the problem.

The schematics in AudioXpress are correct but i used some screab.
IRF610/9610 and bc550/bc560.
So above the 221R a p channel fet (9610) and for the negative version mirror the circuit or look at the super shunt regulator circuit.
Bc459 huh ??? 549 i suppose.

In summary;positive supply, P-channel mosfet+pnp in CCS and P-channel mosfet in shunt. Negative supply N-channel mosfet+npn in CCS and N-channel mosfet in shunt.

Gr.
Johan

B.t.w. it can also different. Please take a look at the "shunty design" of ZENMOD on the pass forum.
 
The schematics in AudioXpress are correct but i used some screab.
"screab"??
IRF610/9610 and bc550/bc560.

So, bc550/bc560 are your preferred substitutes for the orig authors choice of BC549/BC559, respectively?

So above the 221R a p channel fet (9610) and for the negative version mirror the circuit or look at the super shunt regulator circuit.
Bc459 huh ??? 549 i suppose..
Ooops: I meant BC549 (npn)! And, yes, BC559 (pnp) is the compliment.
In summary: Positive supply, P-channel mosfet+pnp in CCS and P-channel mosfet in shunt. Negative supply N-channel mosfet+npn in CCS and N-channel mosfet in shunt.
Right. But that leaves out the npn BC549C, in the SSR w/CCS schematic (see image posted earlier), whose base is between Vz and 100R. I assume this is changed to BC559C (PNP) for the neg supply.
c560 are your preferred substitutes for the orig authors choice of BC549/BC559, respectively?
 
hollowman said:


So, bc550/bc560 are your preferred substitutes for the orig authors choice of BC549/BC559, respectively?

Right. But that leaves out the npn BC549C, in the SSR w/CCS schematic (see image posted earlier), whose base is between Vz and 100R. I assume this is changed to BC559C (PNP) for the neg supply.
c560 are your preferred substitutes for the orig authors choice of BC549/BC559, respectively?


Hi Hollowman,

BC550/560 are "lower"noise than the BC549/559 but i think it doesn't matter which one you use in this application.

For the shuntpart of the schematics you need a driver for the mosfet. Please look careful at the schematic in post 6!!!
If you like you can change R4/R8(56R) for the Borbely or Zenmod current source. For me it didn't made any difference but for you???
What preamp do you use? What speakers? What source?

I.M.H.O. the positioning of the loudspeakers and the roomaccoustics do much more than exclusive high audio grade components. So use standard industry components and everything will be o.k.

Gr.
Johan
 
joho said:
What preamp do you use? What speakers? What source?
Not sure why/how these components relate to the discussion but...
Pre-amp/integrated (headphone amps): heavily-modded Van Alstine FET-Valve; heavily-modded Meijer Corda headphone amp; PPA2 DIY headphone amp; modded GoVibe headphone amp (using as "pre-amp" of sorts!)
Speakers: modded Paradigm Export Monitor Mk2; modded Vifa Studio; various DIYs based on projects/plans avail. on the Web. But I mostly use headphones or earphones, like Senn 650, Shure 530, Beyer 880 and Grado 325i
Source: Rarely use analog VPI 19 Jr/Rega 300/Sumiko BP Spec.; mostly use heavily-modded Philips 650, modded Pioneer DV59AVi and heavily-modded MF A324 DAC
I.M.H.O. the positioning of the loudspeakers and the room acoustics do much more than exclusive high audio grade components. So use standard industry components and everything will be o.k.
While I agree that those other (and unrelated) high-fidelity standards do matter a lot (esp. re: high-grade/boutique audiophool parts), we're getting off-topic. The topic at hand is comparing/contrasting specific types of shunt regs. Speaking of which, here's another one (mostly based on Ed Simons' design in AudioXpress June 2007):
+/-15VDC, 400mA, MOSFET Shunt Regulator -- schematic
 
Hi Hollowman,

If you mostly use headphones, positioning of speakers and roomaccoustics don't count,that's clear.
What i tried to say is; you can discuss about component/part
choices p.e. bc550 contra some japanese counter part. Or even a bc550 brand philips contra fairchild. If your listening conditions/your golden ears are not very,very good, you hear no difference.(english isn't my native language so i hope you understand me well. If not there is always Jacco).
You're audio set looks serious. You collect headphones?
A friend of mine has the topmodel of grado and i must say it sounds very impressive.

O.k to the point. Different shunt regs... the only thing i can say...
build one and listen.

At the moment i'm listening to a series regulator. It's a two stage design. In fact a zener stage around a current source+emittor follower and a local mosfetfollower with a expensive 0.1uF polyp.cap as last capacitor.
And i like it. It's better than the shuntreg but i'm waiting till the weekend for the end conclusion.

Gr.
Johan
 
joho said:
At the moment i'm listening to a series regulator. It's a two stage design. In fact a zener stage around a current source+emittor follower and a local mosfetfollower with a expensive 0.1uF polyp.cap as last capacitor.
And i like it. It's better than the shuntreg but i'm waiting till the weekend for the end conclusion.
Can you post a schematic of this series regulator? Also, do you recommend any particular brand/type of Zener -- both for your shunt reg and this series regulator?
 
Re: shunt regs+last cap.

joho said:
Hi Gents,

<snip>
As i said, i'm fooling around with different(shunt) regulator configurations but everytime that last cap is the bottleneck.
Yes, the panasonics FC-series are also very good for me. When i swap it for an elna audio grade i become depressive. But in fact we are talking about colouration.
How to solve that?

Gr.
Johan
It is possible to find films at that size for not insane prices (they are big though), but at that current is 100uF really necessary?
http://www.northcreekmusic.com/Zen.htm

How about 200uF film for $20.00
http://www.northcreekmusic.com/Specials.htm
 
serie reg.

Hello again,

Attached the schematic of the regulator i'm testing/listening to.
You must try different zener/diode combinations to get the output voltage you need. Please also test it with the current you need because there is no feedback and the voltage get's a bit lower when the current raise.
The mostfetfolower is a la Charles Hansen's Ayre V3 if i'm right.
Personally i like Mundorf Suprime cap's but they are not cheap and small. The northcreek cap's are new for me. The price is better than the Mundorfs. For sure when the dollar is low and the euro high. But shipping cost...
I do not prefer any particular brand/type of Zener. When i buy some in the shop at the corner it's always a supprise what you get.

Gr.
Johan
 

Attachments

  • ugs_supply_rev3.png
    ugs_supply_rev3.png
    14.2 KB · Views: 3,279

Onvinyl

Member
2002-08-02 10:41 am
Germany
Hi,
zeners as references are pretty noisy. Their Vn can be as high as 2µV/1/2Hz, where even the 1M-Resistor shown has 128 nV!
It is however not so easy to use low noise references since they are normally low voltage. I use bypassed resistor stabliized with a FET CCS, but it is not easy to fix the voltage.
Any idea to circumvent this is welcome, and, at least for high gain circuits, pretty much a must.
I don't see how the proposed local regulator could work. Does it depent on leakage current?
Rüdiger