What's wrong with the kiss, boy?

[rickmcinnis]

Scheming to build the POOR MAN'S shunt reg - I noticed the JUNG article on cascoding the LM317 with a DN2540 which seems to offer improvement compared to being used by itself.

I also wonder - since I have already built separate raw supplies for plus and minus why not use the positive reg for both polarities? The LM337 is not very highly regarded. I know this is routinely done with series regs but need to ask if it works with shunt regulators also?

I know I am gilding but I cannot help it.

But the big question is what to use in place of the obsolete BF245B?

 

[noelectrix]

which then drives the 1:5 SLAGLE autoformer

wasn´t it recommended not to exceed a 1:2 ratio? maybe that´s a point.

 

[rickmcinnis]

That is not the point because that part of the spectrum is just fine.

The problem was when using the buffer between the DAC and the autoformer attenuator.

I do not remember reading that was a potential problem. I think you are confusing the fact that ZM said any 1:1 transformer can be used as a 1:2 autoformer.

I can assure you for frequencies below 500 hz (which is all I can attest to) there is no problem with the arrangement whatsoever.

 

[Zen Mod]

Rick , now I remember that I owe you reply or two here .... will do that later tonight

for now , just few pictures ...... so called preliminary Iron Pumpkin prototype is done (preliminary being pot-buffer-6db autoformer)

what was important :
-to confirm flawless function of Good Gemini shunt reg ,
-to develop and finalize my own Diamond Buffer ........ which is nothing more but nothing less than drekload of parts replacing simple and great complimentary JFet buffer , but with similar THD spectra and substantially lower Rout
-to solve full basket of problems with balanced iteration : 1. to keep same gain in both Bal/Bal and SE/Bal working modes , 2.to maintain flawless THD spectra in both modus operandi , 3.to avoid/cure/prevent clicking and crackling noises when switching between various modus operandi

now , few days of some thinking what's conceptually best , to send to 6L6 for store , and to make that mishmash

after that , full forces at SE and Bal Metglas AVC development

 

[6L6]

Looking very good!!!

🙂 🙂 🙂

 

[rickmcinnis]

Dear ZM,

You owe me nothing! You are the one owed for ALL of your good advice and even better ideas! Plus I figured you were busy with this project and any question of mine has a lower priority than the IRON PRE.

Balanced sounds like the proverbial balancing act. Balanced autoformer attenuator sounds like asking for trouble to me but if anyone can pull it off ...

Curious to hear how you like the sound of the diamond compared to the simple complementary. Always hoping the diamond just might be even better ...

Your implementation looks really good. You are working towards a very different aesthetic than I. Is that case lined with copper sheet?

Is the GOOD GEMINI shunt reg an evolution of the POOR SERBIAN one or the same basic thing?

Can't wait to hear more.

Nobody keeps the excitement level up like ZM.

Take care,

 

[Dennis Hui]

To use a ZM-ism: Me Likey... 🙂

 

[Zen Mod]

first kitten

lousy screenshot

I'm sleepy ......

SE , stereo pcb - common shunt reg , cap multis per each buffer

buffers complementary JFets , Jensens flpchhhhsomething

nightynighty .......

 

[noelectrix]

😴



wow, looks great. I thought we have single pcb´s without input select, but it´s all there already 🙂

maybe two holes on the right (input) side wouldn´t hurt (drilling them won´t either)

tnx,
st.

 

[Zen Mod]

as I said - first kitten - not finished

there are 2 holes more , just moved outa screen

probably several more hours in optimizing , then few more for sorting documentation (names , values optimizing etc. )

 

[Zen Mod]

I have been working on the IRON PRE concept the last month or so.

I see the pcbs are not too far into the future but since they were not available I went ahead and made mine point to point.

My buffers end up being almost ball shaped to keep leads to a minimum. I have the two FETs facing each other and taped together to keep them at similar temperatures. I have no idea if this is good or bad so would appreciate ZM's guidance.

well done

This is how I have it set up - DAC to SLAGLE AUTOFORMER ATTENUATOR two outputs one to the SIT 1 and the other to a buffer which then drives the 1:5 SLAGLE autoformer within the J2 to equalize levels between drivers and a YAMAHA crossover (old one using PCM63s).

I initially had the buffer between the DAC and the attenuator and thought it sounded good. My initial impression was that the bass was improved by a good margin and the rest was left untouched. I then returned to no buffer and noticed that the buffer was not as transparent as I had thought. I then went with the scenario first described and I am getting the best of both worlds.

as I told you in our private correspondence - less things in chain , the better ;
in this case - you tried and determined that your DAC is perfectly capable of driving Slagle AVC directly .... which logically can't be the case with 1:5 gain autoformer ...... that thingie is needing closely connected cojones to sing

.....

I had planned to use buffers before the attenuator and after. The cable to the SIT 1 was never to be after the second buffer, as an aside. But now I guess I only need one before the 1:5 autoformer ...

UNLESS my buffer construction is responsible for the unwanted smoothing in the range above 500 hz.

With my buffer all grounds are brought to a single point. I am not using the ZM shunt supply. I am using a JUNG super reg for each buffer. I am wondering: is that reg the problem or is it my construction method? I know I am bordering on apostasy with my reg choice. Have to say below 500 hz the thing sounds incredible. Bass viol and guitar have a harmonic wholeness - the upper range tones and lowest range tones are in balance - there is real impact that is not limited to just bloat in the lowest registers and the real tonality only coming from the mid bass. If that tortured sentence can be decrypted.

well - whatever active stage you have , it will do some things ; only your eras (I repeat) are able to determine how much steps back and how much steps front ....

regarding Jung - man knows his drek , no doubt ; however , I'm proponent of shunt regs , even if I can't give you any sort of scientific proof (besides so called scientific bable why something may work better ) that they're really better sounding than series regs
to my ears -yes , and that's it - I passed my opinion 😉

I am asking for hunches not answers since I know there is not enough here to give an answer.

I do plan on trying the shunt reg on another set of buffers since it seems like a good idea to buffer the output of the SOEKRIS DAC raw output and the attenuator.

I had read many times that this buffer is not particular about how it is powered so I thought the JUNG would be fine. I know I will only know when I try shunt regulation.

ZM - are you still recommending the regulator in the POOR SERBIAN MAN ...COOKBOOK?

Take care,

shunt reg from PSM LS is handy one , not best world in regulation , made originaly with idea to be cheap and to have awesome ratio of money/time/result

last one (Good Gemini ) is , say, simplified Shunty (which is damn overkill and demonstration of Force ) but also damn effective

 

[Zen Mod]

........

Curious to hear how you like the sound of the diamond compared to the simple complementary. Always hoping the diamond just might be even better ...

Your implementation looks really good. You are working towards a very different aesthetic than I. Is that case lined with copper sheet?

Is the GOOD GEMINI shunt reg an evolution of the POOR SERBIAN one or the same basic thing?

......

preamp is made for a friend (having no less than 3 pairs of speakers based on Mighty ZM's brainfarts) , as interlink to final Iron Pumpkin (buffer to AVC with gain) so it'll be easy operation - ALPS out , old Iskra repeaters out , new AVCs in ....
that - looking as copper , is nothing else than sides mades of some wakoo wood

sound - he's thrilled , to put it mildly ;
even early SE prototype , shown in dreky casset-deck case sounded better than his BA3 preamp ;
well - I didn't made that one , and - as I said to him - where BA3 FE is really shining is in original role - as full banging gain stage , driving hefty output stage ..... so optimizing it for mild gain is not an easy task and really not fair to expect from it nature of Swiss knife 🙂

so - to cut it short - I believe Diamond buffer is sounding damn good , in this iteration

 

[Nelson Pass]

Dear ZM, You owe me nothing! You are the one owed for ALL of your good
advice and even better ideas! Nobody keeps the excitement level up like ZM.

I will second that.

 

[aspringv]

...I'm looking forward to your diamond buffer! I've been on the lookout for what you had in mind there for quite a few months!

 

[Zen Mod]

regular routine

.

 

[rickmcinnis]

I will second that.

Well, I was speaking of earthlings when I said that. Not, you know, audio gods!

 

[Zen Mod]

to chill out from SE (stereo pcb) iteration , working on Bal/SE in/Bal out (mono pcb) iteration

- problem with both pcbs is that I ooked proper Eagle routine - instead of importing properly (complete pcbs+sch) subcircuits , so having them on different sheets in project as a whole ........ I simply copy/pasted brd subcircuits, leaving sch parts laying somewhere in electronic mist (non-annotated to pcb files)

so - will need later to produce complete schmtcs backwards , just because being too lazy to abandon several hours of work on big pcb files

anyway - had some fun tonight :

 

[noelectrix]

that´s like routinely drama for photographers, "save" instead "save as..", always destroying several hours of work if you need to go little step back. 🙄

But question to general concept: the mighty TVC upgrade (once the donut exists) could be implemented "instead" of the jensen tranny? Or will that be extraextra PCB?
Hope the Jensen will survive vacuum pump desoldering...

 

[Zen Mod]

again - these will be (pot-JFet buffer-autoformer ) packages for DiyA Store , 6L6 asked for ;

one is gonna be SE , stereo pcb , with shunt reg common to both channels , but with local cap multis , so no worry
second one is gonna be BAL/SE in , mono pcb , so shunt reg per channel (still with cap multis , why not)

in both cases 5 inputs , relay switched , plain rotary selector for these
Bal/SE - number of SE or Bal inputs configurable by builder - simply put one 1N4148 diode for each SE input you want

I'm working on other solution , which is selector-Diamond Buffer -AVC (so no ordinary pot) , both in SE and Bal (as above) , fed with same Good Gemini shunt

when done , I'll publish all schematics , but will preserve fabrication of both AVC and pcbs for my self
then you'll be able to get them from me , either just AVCs (SE and Bal) , or as complete packages
anyway , as life taught me , that Bunny Wabbit's back legs are still in da woods
so , don't take me too seriously , as always

almost forgot - yes , you will be perfectly able to shun pot from these , and substitute AVC block(s) instead of Jensen(s)

 

[noelectrix]

ahh, finally even i seem to get it 😀, "poti" is not just let out "somehow" in AVC version, somehow made with "override" jumpers or so in this PCB, it is real different schematic 🙂

tnx!

mmhmm, haha, you edited and leave me still numb : ) (don´t waste your time too much, i´ll keep reading!)