As this is the listening impression thread, I want to ask if anyone ever experimented with different amounts of capacitance on the power supply side of the F5? Any difference in sound?
I know for sure that when I had only 30000uF per rail (CRC), I found that bass was lacking and the sound was really lean and even the highs were grainy.
Then I added another 30000uF per rail (60000uF per rail (CCRCC) and bass filled in, no more leanness, and the grain went away.
Then I decided to get even larger caps and now I have 94000uF per rail (still in CRC) and even more bass filled in and the sound is more liquid.
Is there a point where it becomes diminishing returns or even affecting the sound in a negative way? Would it be advisable to build a separate power supply with lots of caps, and have several just before the amp in the case?
Any thoughts / impressions?
I know for sure that when I had only 30000uF per rail (CRC), I found that bass was lacking and the sound was really lean and even the highs were grainy.
Then I added another 30000uF per rail (60000uF per rail (CCRCC) and bass filled in, no more leanness, and the grain went away.
Then I decided to get even larger caps and now I have 94000uF per rail (still in CRC) and even more bass filled in and the sound is more liquid.
Is there a point where it becomes diminishing returns or even affecting the sound in a negative way? Would it be advisable to build a separate power supply with lots of caps, and have several just before the amp in the case?
Any thoughts / impressions?
Afair, the F5 has a noise figure of -90dBW with 30,000 uF per rail.
Such a number implies that the F5 is not in the same lousy noise reduction class as e.g. the Hiraga works.
If you can still hear a difference with 94,000 uF per rail, it means you either have exceptional hearing or the loudspeaker you use is a suck-up artist.
The point of diminishing returns is where your transformer and rectifier diodes are starting to moan.
Such a number implies that the F5 is not in the same lousy noise reduction class as e.g. the Hiraga works.
If you can still hear a difference with 94,000 uF per rail, it means you either have exceptional hearing or the loudspeaker you use is a suck-up artist.
The point of diminishing returns is where your transformer and rectifier diodes are starting to moan.
Last edited:
The experiments that Rick did on his F5 (see the OP) showed that what was done in the PS has a significant effect on the resulting sound... feel free to try different schemes in the PS area... that includes filtering schemes, rectifiers, caps, bypassed caps, wiring, layout, grounding, etc... it all has "an effect".
_-_-bear
PS. jacco, if the PS starts to hum does that mean that it just doesn't know the words? 😉
_-_-bear
PS. jacco, if the PS starts to hum does that mean that it just doesn't know the words? 😉
Hi dazed2
I have tried hi bias F5 3A on 32 V rails with toshibas and was realy impressed
I have 10mF Caps 50m F each bank in CRCRC the last C split between the 2 Chanels total off 200mF
As you found large capacitance works realy well for dinamics and bass slam.
I have 200mF of Rifa caps due for delivery soon (beter quality than the ones I am using at present)
Maybe you could try CRCRC and see?
I am also going for separate box for psu.
Jacco my Transformer hums a bit (maybe just me being to fussy)
Would to much capacitance cause this?
Is there any way to reduce hum but keep the capcitance?
Genuine questions I am new at this and would realy apreciate your help.
I think that it may be a badly built transformer (1000VA 24 V split secondary) but before I get a new one I would try to see if humis caused by sometink else.
I have tried different bridges normal metal blocks ones and faster Vishay PB3508)and DC stopper but still hum.
Maybe less with slower rectifier.
Would more resistance after the first C bank help?
I have tried hi bias F5 3A on 32 V rails with toshibas and was realy impressed
I have 10mF Caps 50m F each bank in CRCRC the last C split between the 2 Chanels total off 200mF
As you found large capacitance works realy well for dinamics and bass slam.
I have 200mF of Rifa caps due for delivery soon (beter quality than the ones I am using at present)
Maybe you could try CRCRC and see?
I am also going for separate box for psu.
Jacco my Transformer hums a bit (maybe just me being to fussy)
Would to much capacitance cause this?
Is there any way to reduce hum but keep the capcitance?
Genuine questions I am new at this and would realy apreciate your help.
I think that it may be a badly built transformer (1000VA 24 V split secondary) but before I get a new one I would try to see if humis caused by sometink else.
I have tried different bridges normal metal blocks ones and faster Vishay PB3508)and DC stopper but still hum.
Maybe less with slower rectifier.
Would more resistance after the first C bank help?
Jacco,
The NP specs are for 60000uF per rail. At the time I was lazy thinking since my caps were 15000uF each and it was only 30W amp, 30000uF per rail would be enough. Well I was wrong. If the spec says 60000uF per rail, that should be the minimuim to use. I don't know I and a few others heard the difference at only 30000uF per rail on different systems.
bksabath,
have you tried different configurations of CRC, CRCRC etc etc ?
Could you say what sound characteristics of each? I'd love to know in order to go towards a direction.
Also wouldn't caps that are very high capacitance ie 90000uF an above increase ESR thus making it less dynamic? Would you have to add say an small poly pro cap to counter that?
Bear, do you know where in the thread Rick posted his results? Even this thread has like 80 pages. 🙁
The NP specs are for 60000uF per rail. At the time I was lazy thinking since my caps were 15000uF each and it was only 30W amp, 30000uF per rail would be enough. Well I was wrong. If the spec says 60000uF per rail, that should be the minimuim to use. I don't know I and a few others heard the difference at only 30000uF per rail on different systems.
bksabath,
have you tried different configurations of CRC, CRCRC etc etc ?
Could you say what sound characteristics of each? I'd love to know in order to go towards a direction.
Also wouldn't caps that are very high capacitance ie 90000uF an above increase ESR thus making it less dynamic? Would you have to add say an small poly pro cap to counter that?
Bear, do you know where in the thread Rick posted his results? Even this thread has like 80 pages. 🙁
For your convenience : www.firstwatt.com/pdf/prod_f5_man.pdf
Says 29,000 uF per rail, Raw Supply as in no CRC.
The finished product has a CRC with two 10,000 uF per rail, drops the noise by another 40 micro-Volts, iirc.
(Papa is cheapscate-r with caps)
Last edited:
Hi Dazed
I have 10mF caps cheaper to replace and faster than larger ones + price breack for 40 was good
CRCRC gives more riple reduction I think.
As Papa sugested I have put few uF of small panasonic caps on the last bank.
As I said I am waiting for beter ones especialy as you mention low ESR is a good thing to have. (Hi ESR may be cause for trafo to Hum?)
My F5 is wery much work in progress I was so happy with the Mule that IMO is well worth triing different things (loads of fun as well)
Bear Yes please could you post that link to Rick post?
I have 10mF caps cheaper to replace and faster than larger ones + price breack for 40 was good
CRCRC gives more riple reduction I think.
As Papa sugested I have put few uF of small panasonic caps on the last bank.
As I said I am waiting for beter ones especialy as you mention low ESR is a good thing to have. (Hi ESR may be cause for trafo to Hum?)
My F5 is wery much work in progress I was so happy with the Mule that IMO is well worth triing different things (loads of fun as well)
Bear Yes please could you post that link to Rick post?
Jacco,
Hmm I didn't see that part. I just looked on pg 18 of the manual that you posted the link there and used that power supply diagram which shows a CCRCC config with 15000uF per C.
BKsabath, the way mine is built I have just a CRC config in the power supply section of the amp. I kinda want to keep the same case, so I'm even mroe limited.
I wonder if using small say 1uF polypro caps would eliminate the ESR issue with large caps... thereby using 100-200mF caps 😀
Hmm I didn't see that part. I just looked on pg 18 of the manual that you posted the link there and used that power supply diagram which shows a CCRCC config with 15000uF per C.
BKsabath, the way mine is built I have just a CRC config in the power supply section of the amp. I kinda want to keep the same case, so I'm even mroe limited.
I wonder if using small say 1uF polypro caps would eliminate the ESR issue with large caps... thereby using 100-200mF caps 😀
I use CCRCCRC config per rail with each C being 22000 uF. My transformer used to hum earlier when I had 15000 uF caps in CCRCC, but disappeared completely with the caps change. I used to suspect my transformer as well. Don't know what causes this trafo hum and why it disappears with increased filter caps.
That PS schematic refers to the commercial version, 1 PS for 2 channels.
8 times 15,000 is 120,000 uF, still makes 30,000 uF per rail of each channel.
(the piccies i recall from the 6moons article seemed to appear as 10,000uF instead of 15,000, could be that one had 18,000 uF versions)
Not that it is of any interest if you prefer it Big.
I'm partial to anything biggy.
Tanks Anilva
Hum and caps back there again just got to wait for new caps before I blame trafos
interesting tanks.
Some how different point of view from Jaco
Quote>The point of diminishing returns is where your transformer and rectifier diodes are starting to moan.<end quote
Hum and caps back there again just got to wait for new caps before I blame trafos
interesting tanks.
Some how different point of view from Jaco
Quote>The point of diminishing returns is where your transformer and rectifier diodes are starting to moan.<end quote
The quality of the caps IS important to the sound, not just the size - 30,000uF/ rail of Panasonic isn't going to break the bank but with Siemens Sikorel, Nichicon Super Thru, etc it's a rather different story and fast diminishing dollars would suggest a good look at simple C-multipliers - better results generally for the loss of a few volts .....
Jacco,
I think there must have been a miscomunication on my part.
I meant 4 x 15000uF on the -ve side and 4x15000uF on the +ve side.
Therefore 60000uF +ve and 60000uF-ve . I thought that was what rail meant. Of course if you then divide that by the number of transistors (4 in a stereo amp) then yes it would be 30000uF per transistor.
Anyways, yes quality of the caps would matter, also physical size.
I'm using 50V Kendeils so they are kinda big, and cost is not too bad... about $25 @ from this site.
I think there must have been a miscomunication on my part.
I meant 4 x 15000uF on the -ve side and 4x15000uF on the +ve side.
Therefore 60000uF +ve and 60000uF-ve . I thought that was what rail meant. Of course if you then divide that by the number of transistors (4 in a stereo amp) then yes it would be 30000uF per transistor.
Anyways, yes quality of the caps would matter, also physical size.
I'm using 50V Kendeils so they are kinda big, and cost is not too bad... about $25 @ from this site.
Rick did not post his results, I posted my impressions of what he did which was based in part on discussions we had...
_-_-bear
Maybe old news, but did you check for DC with two butt-to butt caps at the primaries ?
(badly wound transformers can be cured with a vacuum job and injecting varnish. Back in the old days of limited transformer pickings, i even did it with impregnating resin, before an epoxy cast)
Tanks Jacco
Got a DC filter with bridge diode and 8 back to back 12mU 6.8 V caps
supose drawing will be handy it is the standard back to back caps published several times.
And got a traffo back today posh tape and vacuum impregnated
Strange thing it has got a Avell Lindberg return addres on the box.
which realy make me wonder about a couple of things.
1)somethink else wrong
or 2) I am to fussy.
Any way cant wait to set it up and see/Ear
Got a DC filter with bridge diode and 8 back to back 12mU 6.8 V caps
supose drawing will be handy it is the standard back to back caps published several times.
And got a traffo back today posh tape and vacuum impregnated
Strange thing it has got a Avell Lindberg return addres on the box.
which realy make me wonder about a couple of things.
1)somethink else wrong
or 2) I am to fussy.
Any way cant wait to set it up and see/Ear
Well probaly out of topic but maybe worth a mention.
While working on the F5 my Girlfriend is plaiinng Cat Stevens with the once perfectley good old amplifier.
Yours trully walked past the speakers and had a bit of a shock.
At first I tought that Girlfriend was getting her own revengenge after me hiding the toilet paper and decided to put a big thick blanket maybe a 20Tog sleeping bag over the speakers.
Well no blanket there just the difference betwenn F5 and the rest and this after a few days of no music.
While working on the F5 my Girlfriend is plaiinng Cat Stevens with the once perfectley good old amplifier.
Yours trully walked past the speakers and had a bit of a shock.
At first I tought that Girlfriend was getting her own revengenge after me hiding the toilet paper and decided to put a big thick blanket maybe a 20Tog sleeping bag over the speakers.
Well no blanket there just the difference betwenn F5 and the rest and this after a few days of no music.
Well said and true in my case as well.
Earlier I could listen to all kinds of music recordings. But after switching completely to Pass designs, I am always on the lookout for prime recordings as these amps show up poor quality so easily. And I realize that I am listening to lot of new type of music (for me) just because they are recorded well.
well my friends
I kind of hate to say this 😱
but maybe not solely 'blame' the recordings
you amps more likely reveal your speakers as the weak link now
you now have top quality amps
but do you also have good speakers ?