Andrzej Sochon said:
I'm afraid you can't think of them as dc stable sub-rails - it is a trick by NP to make life a little easier on the jfets. If you don't end up with a ground loop, I think you get 6db attenuation with this configuration, the output decoupling however should be perfectly fine!
cviller said:I'm afraid you can't think of them as dc stable sub-rails - it is a trick by NP to make life a little easier on the jfets. If you don't end up with a ground loop, I think you get 6db attenuation with this configuration, the output decoupling however should be perfectly fine!
I am not sure if I understand you well, Cviler, but I do not observe any attenuation with my configuration. More: I have a feeling that all details of the stage are more clear. Maybe another NP trick. Or something wrong with my ears or with my "single grey matt between ears", as my Wife says...
Time to sleep...
a.
Babowana said:How big difference from the original? Tks.
Hi Babowana,
Difficult to explain, I have language problems, sorry.
SOZ was my f4 predecessor, great sound, but catching details of an acoustic stage was difficult for me. Before listening I always precisely checked super symmetry positions of speakers in the room. Then I sat at the triangle corner. Many times during listening I moved myself slightly to front or back to find the best stage symmetry and details. Tiring kind of listening.
Now, with f4, listening is much easier and gives much more pleasure. Super symmetrical positioning of speakers is not the most important thing, moving ears forward and back brings no significant improvement.
Some mentioned capacitors more and – a new discovery: I can seat not exactly on symmetry line. Stage is not dramatically shifted to the right side if I am one meter to the right from symmetry line - Acoustic stage remains natural. In other words the listening space transformed from triangle to an isosceles trapezium. Listener (listeners) are on the shorter segment of trapezium.
What more: now I have better balance between medium and hi frequency. Before adding capacitors I had to much medium tones. But maybe this is a result of warmed up elements.
I do not have pre-amp. My f4 are driven directly from DAC balanced outputs.
a.
Yeah Maousse,
"snubbers" - sort of.
I think it's technically a "resonance control" device - lots of theory about these things on the "Capacitor Bipass" thread.
Rather pleased at improvement in overall clarity and base detail particularly with the B.Gates - rather surprised, too.
Rather useful little thing, this.
Andrzej,
I just added some 220uF (and the bipass R-C, as usual) from point R5/R7 to Gnd (and P2/R6, too) for some "rail isolation" from power Fet modulation - pretty standard procedure and seems to work okay - can't see why it would present any problems - not sure about your Signal Gnd point, tho.....
There's something about the D3,4 diodes that look "naked" - and those TL431s are sometimes noisy, no?
The simple R-C network has been added to caps C1,2 (B.Gate N),3,4 (1000uF Nichi KZs)and 5,6 (Rubi ZL) by simply adding them directly across the electro cap pins underneath the pcb - easier with SMD resistors, tho.
What speakers are you using, and room Q?
....... Hope this is of assistance.
"snubbers" - sort of.
I think it's technically a "resonance control" device - lots of theory about these things on the "Capacitor Bipass" thread.
Rather pleased at improvement in overall clarity and base detail particularly with the B.Gates - rather surprised, too.
Rather useful little thing, this.
Andrzej,
I just added some 220uF (and the bipass R-C, as usual) from point R5/R7 to Gnd (and P2/R6, too) for some "rail isolation" from power Fet modulation - pretty standard procedure and seems to work okay - can't see why it would present any problems - not sure about your Signal Gnd point, tho.....
There's something about the D3,4 diodes that look "naked" - and those TL431s are sometimes noisy, no?
The simple R-C network has been added to caps C1,2 (B.Gate N),3,4 (1000uF Nichi KZs)and 5,6 (Rubi ZL) by simply adding them directly across the electro cap pins underneath the pcb - easier with SMD resistors, tho.
What speakers are you using, and room Q?
....... Hope this is of assistance.
Andrzej Sochon said:
Thanks for the very interesting info.
When I saw your additional caps first, I thought them just as
a simple reinforcement for cleaner dc-supply. But, they seem to
be doing more work than expected.
I also try to find the sweet listening spot, moving my seat around
as you do, for serious listening. Your variation is quite simple and
seems to be worth trying . . .
No pre? If you like tube sound, I would like to inform you that
the good parts of tube pres are well powered by F4 . . .
jameshillj said:
Hi JH,
You are right about my Signal Gnd point, maybe better is to use PSU Gnd point instead. But at the moment it is difficult to modify.
I do not have any ideas about D3,4 diodes, and TL431’s. I do not observe noise, maybe future preamp will unveil noisy points of my f4. Or maybe my ears are to old.
I also had an idea to add small parallel caps to C1,2, but I did not, just because of lack of patience. I have to buy BG capacitors and compare with present accidental Philips.
What about C3,4 – why do you add RC network here – does it affect sound? I think R24C4 and R25C3 are important at power ON/OFF temporary states.
Speakers: Suspense’93 (Manfred Zoller project, kits of Focal speakers, year 1993), 8Ohms,approx. 91 dB. I made housings according to the Mr. Zoller design. In spite of no preamps sound level is quite big. Each speaker consists of two bass speakers, one mid and one tweeter, bi-wired.
Room Q is approx 50 cubic meters.
Best Regards,
a.
Babowana said:Thanks for the very interesting info.
When I saw your additional caps first, I thought them just as
a simple reinforcement for cleaner dc-supply. But, they seem to
be doing more work than expected.
I also try to find the sweet listening spot, moving my seat around
as you do, for serious listening. Your variation is quite simple and
seems to be worth trying . . .
No pre? If you like tube sound, I would like to inform you that
the good parts of tube pres are well powered by F4 . . .
Hello Babowana,
Paraphrasing Carpenter: “Isn't it amazing how a few of capacitors can organize your world“. In the past, applying additional capacitors as close as possible to supply legs of digital and analog IC’s, I made sound improvements of my CD player.
Regarding pre and tube sound – to be frank wit you I do not know how tube sounds. Some friends tells me: Andrzej, forget transistors, focus on tubes. But I do not like high voltages...
Best Regards,
a.
Andrzej Sochon said:
My shoulder is improving--thank-you for asking.
I've been putting the finishing touches on my DIY hobby room. It's giving me plenty of exercise. Cabinets installed, now I'm loading them up with all my diy parts, books, and equipment... I'm very happy.
Maybe this month I can go back to work.
carpenter said:
My shoulder is improving--thank-you for asking.
I've been putting the finishing touches on my DIY hobby room. It's giving me plenty of exercise. Cabinets installed, now I'm loading them up with all my diy parts, books, and equipment... I'm very happy.
Maybe this month I can go back to work.
give us piccies of new hobby room ;
if ya think that here is inappropriate - just post them in any - ya know which one(s) - thread
cviller said:I'm afraid you can't think of them as dc stable sub-rails - it is a trick by NP to make life a little easier on the jfets. If you don't end up with a ground loop, I think you get 6db attenuation with this configuration, the output decoupling however should be perfectly fine!
Hello Cviller,
You are right, my input capacitors provide an AC output to the input GND. I focused on various versions of f4 schematics. I have difficulties in understanding all aspects of f4 schematics, especially at higher frequencies, having no laboratory equipment.
I also found Mr. Nelson Pass very interesting comments regarding bootstraping and improving frequency response (posts #397, 416, 426). Let me ask a risky question: what is an influence of PC board design? Maybe bootstraping and 1K resistor in series with the C3/C4 caps – maybe this is panacea for frequency response in this particular case of NP PC boards design. I see there are few very long paths here (f4 User and Service Manual). Long parallel near (close) paths may create problems at higher frequencies, I don’t know if 1 MHz is enough or not. I am sure Mr. Pass could tell us many interesting stories about how his projects behave at high frequency. (Long time ago somebody told me that it is much easier to make generator than an amplifier).
In comparison Peter Daniel boards (amp) are very compact, they have rather short paths. Did somebody try to analyze or measure frequency response of f4 build on Peter board? My risky question comes again: does f4 depend on board design? If yes, maybe bootstrapping and hi frequency peaks compensation are not always necessary. I will appreciate very much Mr. Nelson Pass opinion.
I remember somebody asked a question regarding oscillations observed on oscilloscope. What are the final results of his tests? Was it NP PC board design or Peter Daniel, how the sound results are? Could you remind me?
Apart from anything else I like NP boards and housing design very much, all is very clear, all parts are easy available, not like at my deep cramped tunnel, see my avatar.
Best Regards,
a.
Andrzej,
Amazed you actually use those sinks this way around - try turning them "inside out" (fin in)and cool the assembly with a slow running, quiet computer fan and add a temp control cct for faster, more accurate pwr fet temp setting - plenty of examples.
There are any number of cheap fans like Cool-Master, Nexus, etc and mount with soft rubbers, add a 40mm fan to h.sink space, ensure no sharp edges, etc.
I've thought about using the larger, high flow, low turbulance fans which are "totally" silent but quite expensive and require a better system of temp/power control, unfortunately, so still using the C.Masters - silent in use, with baffle plate.
Amazed you actually use those sinks this way around - try turning them "inside out" (fin in)and cool the assembly with a slow running, quiet computer fan and add a temp control cct for faster, more accurate pwr fet temp setting - plenty of examples.
There are any number of cheap fans like Cool-Master, Nexus, etc and mount with soft rubbers, add a 40mm fan to h.sink space, ensure no sharp edges, etc.
I've thought about using the larger, high flow, low turbulance fans which are "totally" silent but quite expensive and require a better system of temp/power control, unfortunately, so still using the C.Masters - silent in use, with baffle plate.
jameshillj said:
Hello JH,
Thank you very much for fantastic idea with turning sinks - my cables are long enough.
In the past SOZ was inside, including typical computer fan at the top. I even applied simple electronics with thermistor to slow down propeller after a cold start. I will use it again.
Best Regards,
a.
