Hi guys,
and the tweak goes on… Finally I have reached the coupling capacitors of the negative input of my DIY Aleph4 mono amp. Sorry, if this issue was handled here before, I did not find it.
The issue: the original pass design uses a 220uF electrolytic cap at the negative input branch. It decouples the output DC offset from the 10k input resistor and so the complete negative input. This cap has no defined polarity and voltage pre-load. This brings the poor polarized electrolytic to sound not optimal. As far as I understand the cap value shall be that big not to change the feedback factor 100k versus 10k at low frequencies. 220uF just have 72Ohms of impedance at 10Hz.
The old trick to make DC-unloaded electrolytics sound good: use two in line electrolytics with inverted polarity and pre load the center pin via a high resistance with a DC, e.g. 12V created by an easy Zener regulator.
BUT: here it does not work too well! The load up current through the 220uF cap flows via the 100k resistor to output ground and creates DC offset voltage of easily some hundreds mV for many ten minutes... I have tested it, the sound was much better, but with 4MOhm and 2 x 220uF the DC offset starts with 150mV with a RC-time constant Tau = ca 20min. No go!
So next thoughts: why this 220uF cap? If the cap is out, the 10k and the input branch gets full output DC offset. Is that any problem? The tricky cap pair or a smaller one could be put at very first income position, no problem with loading them. What do you think about this solution?
Of cause, the easiest way would be to use bipolare electrolytics like Black Gate NX in the original Pass design. I will test this in parallel…
Thanks for your comments on!
Regards
Klaus
and the tweak goes on… Finally I have reached the coupling capacitors of the negative input of my DIY Aleph4 mono amp. Sorry, if this issue was handled here before, I did not find it.
The issue: the original pass design uses a 220uF electrolytic cap at the negative input branch. It decouples the output DC offset from the 10k input resistor and so the complete negative input. This cap has no defined polarity and voltage pre-load. This brings the poor polarized electrolytic to sound not optimal. As far as I understand the cap value shall be that big not to change the feedback factor 100k versus 10k at low frequencies. 220uF just have 72Ohms of impedance at 10Hz.
The old trick to make DC-unloaded electrolytics sound good: use two in line electrolytics with inverted polarity and pre load the center pin via a high resistance with a DC, e.g. 12V created by an easy Zener regulator.
BUT: here it does not work too well! The load up current through the 220uF cap flows via the 100k resistor to output ground and creates DC offset voltage of easily some hundreds mV for many ten minutes... I have tested it, the sound was much better, but with 4MOhm and 2 x 220uF the DC offset starts with 150mV with a RC-time constant Tau = ca 20min. No go!
So next thoughts: why this 220uF cap? If the cap is out, the 10k and the input branch gets full output DC offset. Is that any problem? The tricky cap pair or a smaller one could be put at very first income position, no problem with loading them. What do you think about this solution?
Of cause, the easiest way would be to use bipolare electrolytics like Black Gate NX in the original Pass design. I will test this in parallel…
Thanks for your comments on!
Regards
Klaus
The spec is for DC voltage. Ordinarily you look to the
ripple current spec to assess the maximum AC voltage,
and this assumes that the cap is not being significantly
(if at all) operated with a reverse voltage.
The coupling caps you cite might see 1 mA or a few mV on
a bad day, and they can take that until they slowly dry out.
ripple current spec to assess the maximum AC voltage,
and this assumes that the cap is not being significantly
(if at all) operated with a reverse voltage.
The coupling caps you cite might see 1 mA or a few mV on
a bad day, and they can take that until they slowly dry out.
Hi,
Thanks, Henrik, for the link.
Well, no doubt, any Aleph sounds almost perfect with a normal electrolytic in. But guys who are interested in tweaking and so getting really 100% performance out of any device should just test it. The sound improvement is really remarkable.
I have also tested the version "without any cap", but without stronger adjustments it disturbs the hole well matched circuit. My Aleph block runs from less than 10mV DC up to over 200mV. To compensate this change the bias could be adjusted, but maybe the matched balance would be disturbed. I will not test it.
So I will buy 220uF 6.3V Black Gate NX plus 0.1uF NX bypass. They shall be such a kick. Compared to the total amount of an Aleph amp the 10€will be ok.
Regards
Klaus
Thanks, Henrik, for the link.
Well, no doubt, any Aleph sounds almost perfect with a normal electrolytic in. But guys who are interested in tweaking and so getting really 100% performance out of any device should just test it. The sound improvement is really remarkable.
I have also tested the version "without any cap", but without stronger adjustments it disturbs the hole well matched circuit. My Aleph block runs from less than 10mV DC up to over 200mV. To compensate this change the bias could be adjusted, but maybe the matched balance would be disturbed. I will not test it.
So I will buy 220uF 6.3V Black Gate NX plus 0.1uF NX bypass. They shall be such a kick. Compared to the total amount of an Aleph amp the 10€will be ok.
Regards
Klaus
tests finished, BG wonderland...
Hi,
I have finished my tests on the Aleph amp coupling caps. Here are the experiences for interested guys…
1. 220uF 25V Panasonic FC electrolytic plus 10nF ERO1830 bypass: I have frequently wondered why my fully tweaked Aleph Amp still sounds somewhat soft and a bit coated with a layer of distortions, specially at the heights – this mentioned restrictions in spite of the fact that I love the sound of the Aleph and prefer it to all other solid state amps I know! Well, after this test I understood…
2. test with the "tricky polarized" coupling cap: two times two parallel 220uF FCs antiparallel in line create a over all 220uF cap with middle pin to apply polarization voltage. 10nF bypass, of course. The polarization was done by a well noise free 12V source. BUT: this circuit causes massive DC offset in the aleph circuit for many ten minutes, no go! Just for test issues. But the test results were impressive: the sound is more tight and solid, more transparency and more neutral. Leaner Bass with more punch and crystal clear hights, better focus. And all positive properties of the Aleph remain, they are just enhanced. This would have been the ultimate solution – if the circuit would allow it…
3. 220uF Black gate NX + 0.1uF NX bypass + 10nF bypass: unbelievable! All enhancements from test #2, but even clearer, deeper room, more dynamics. The sound is as clear as with good solid state amps with tons of feedback. But still the lovely Aleph sound remains. I burned the BGs in for 6 day at the FM-tuner noise. BUT, there must be a "but". Like at other tests with BGs at PSU I got this special BG sound character: I would call it "to clear to be true". The sound is hard to the limit of being to cool, artificial, the musicians play at fresh autumn wind in a concrete tunnel without enough warmth. Hopefully the BG will burn in further; otherwise I have to learn to love it…
What are your BG experiences, similar effects? Thanks for feedback.
Regards
Klaus
Hi,
I have finished my tests on the Aleph amp coupling caps. Here are the experiences for interested guys…
1. 220uF 25V Panasonic FC electrolytic plus 10nF ERO1830 bypass: I have frequently wondered why my fully tweaked Aleph Amp still sounds somewhat soft and a bit coated with a layer of distortions, specially at the heights – this mentioned restrictions in spite of the fact that I love the sound of the Aleph and prefer it to all other solid state amps I know! Well, after this test I understood…
2. test with the "tricky polarized" coupling cap: two times two parallel 220uF FCs antiparallel in line create a over all 220uF cap with middle pin to apply polarization voltage. 10nF bypass, of course. The polarization was done by a well noise free 12V source. BUT: this circuit causes massive DC offset in the aleph circuit for many ten minutes, no go! Just for test issues. But the test results were impressive: the sound is more tight and solid, more transparency and more neutral. Leaner Bass with more punch and crystal clear hights, better focus. And all positive properties of the Aleph remain, they are just enhanced. This would have been the ultimate solution – if the circuit would allow it…
3. 220uF Black gate NX + 0.1uF NX bypass + 10nF bypass: unbelievable! All enhancements from test #2, but even clearer, deeper room, more dynamics. The sound is as clear as with good solid state amps with tons of feedback. But still the lovely Aleph sound remains. I burned the BGs in for 6 day at the FM-tuner noise. BUT, there must be a "but". Like at other tests with BGs at PSU I got this special BG sound character: I would call it "to clear to be true". The sound is hard to the limit of being to cool, artificial, the musicians play at fresh autumn wind in a concrete tunnel without enough warmth. Hopefully the BG will burn in further; otherwise I have to learn to love it…
What are your BG experiences, similar effects? Thanks for feedback.
Regards
Klaus
Re: tests finished, BG wonderland...
Klaus ,
Check out this link
http://www.blackgate.jp/english.htm
Try running the BG's back to back as per this article .
Failing that, upgrade the Aleph to X Aleph...all the fuzz will disappear.
macka
Klaus said:
Klaus ,
Check out this link
http://www.blackgate.jp/english.htm
Try running the BG's back to back as per this article .
Failing that, upgrade the Aleph to X Aleph...all the fuzz will disappear.
macka
Re: tests finished, BG wonderland...
Remove the 10nF ERO cap, the black gates are good enough in this position and a bypass might do more worse (read constricted, soft, artificial sound). Try only the 220uf BG for a couple of days then add the 0.1uf NX to see if it improves over the single cap.
Klaus said:
Remove the 10nF ERO cap, the black gates are good enough in this position and a bypass might do more worse (read constricted, soft, artificial sound). Try only the 220uf BG for a couple of days then add the 0.1uf NX to see if it improves over the single cap.
Hi again,
@ macka: thanks for back to back hint, but it means two BGs per block… I will check the Aleph x soon!
" Remove the 10nF ERO cap, the black gates are good enough in this position and a bypass might do more worse (read constricted, soft, artificial sound). Try only the 220uf BG for a couple of days then add the 0.1uf NX to see if it improves over the single cap."
I have not described it in detail, not to bore you. I checked in detail how the sound improvements of the BGs are created, step by step with hours of listening not to get lost. First I tested the 220uF BG without anything. Good… Then the NX0.1 bypass came in – much better, but to cool… The small 10nF bypass improves the scenario much without any negative influence. They clean up last effects and give stronger and more neutral coloration. This special type of small low voltage (63V AC/100V DC) polypropylene caps with ultra low inductance seem to be the only thing which can really do so. I can not imagine that the 0.1NX can "clean" that perfect HF effects. I use this 10nF caps everywhere, at PSU, at speaker cross over and at coupling caps. The improvements are everywhere the same, the character as well.
Regards
Klaus
@ macka: thanks for back to back hint, but it means two BGs per block… I will check the Aleph x soon!
" Remove the 10nF ERO cap, the black gates are good enough in this position and a bypass might do more worse (read constricted, soft, artificial sound). Try only the 220uf BG for a couple of days then add the 0.1uf NX to see if it improves over the single cap."
I have not described it in detail, not to bore you. I checked in detail how the sound improvements of the BGs are created, step by step with hours of listening not to get lost. First I tested the 220uF BG without anything. Good… Then the NX0.1 bypass came in – much better, but to cool… The small 10nF bypass improves the scenario much without any negative influence. They clean up last effects and give stronger and more neutral coloration. This special type of small low voltage (63V AC/100V DC) polypropylene caps with ultra low inductance seem to be the only thing which can really do so. I can not imagine that the 0.1NX can "clean" that perfect HF effects. I use this 10nF caps everywhere, at PSU, at speaker cross over and at coupling caps. The improvements are everywhere the same, the character as well.
Regards
Klaus
irritation...
Hi Niclas,
it is a bit of irritating... I got the history of this caps like this:
There was a company EVOX-Rifa producing this shining blue caps first http://www.evox-rifa.de/cap_catalog/pulsecap/pfr.pdf , Polypropylene film/foil capacitors.
Later the name was changed to ERO, RO für Roederstein, why ever. The name of this cap became KP1830. I guess KP means MKP = metallized foil polyprop cap. KP must be an incomplete name, without a metal foil = FKP or a metallization = MKP no cap is given. Then Vishay bought Roederstein and now a cap MKP1837 can be found. I have tested both, "old" KP1830 I got and new MKP1837. I can find no difference. So for the future I should just talk about the available (e.g. FARNELL) type MKP1837. Farnell offers two other types of this low voltage high precision MKP caps from other suppliers. Hifi-Fellows told me they work also fine. An other type is the orange EMZ-type like THEL is using http://www.thel-audioworld.de/pics/3KP.JPG . EMZ was SIEMENS in former times, another irritating story…
BTW, to quote guys talking about this wonder MKP-stuff I would like to give feedback from Tony Gee to read. Tony is a nice guy from Dutchland with a famous cap test homepage. It seems to be offline currently, to much traffic…
This small MKPs work the same good everywhere! So transform the crossover-application report from Tony to PSU and to coupling caps!
"Hi Klaus,
Well, I've tested them. I tried them as a bypass for the tweeter series caps in my Progress speaker and I must say I am very impressed! To get straight to the point they don't change a Supreme into a Supreme Silver-Oil but they realy do clear things up. I must admit I was scepticle at first as the value is only 8,2nF (0,0082uF) - measured 8,48nF - and the caps in the Progress are 12,6uF.
The difference is most noticable with classical music but also good quality recordings of jazz and fusion benifit:
- no change in soundstage width or depth but there is more "concert hall acoustics" that let you get into the recording more.
- not as liquid as silver/oil but they take away the "grainy" edge from the Supreme's.
- a gain in clarity and transparancy making instruments better seperable from each other, the violins in an orchestra are a group of individual violins instead of one mass.
- jazz drum brushes sound more like a brush than a "shush".
So all in all a very nice tweak! Thanks for the tip! How much do these things cost? Do you want your cap's back?
---
regards,
Tony Gee
tweaky regards
Klaus
stappvargen said:
Hi Niclas,
it is a bit of irritating... I got the history of this caps like this:
There was a company EVOX-Rifa producing this shining blue caps first http://www.evox-rifa.de/cap_catalog/pulsecap/pfr.pdf , Polypropylene film/foil capacitors.
Later the name was changed to ERO, RO für Roederstein, why ever. The name of this cap became KP1830. I guess KP means MKP = metallized foil polyprop cap. KP must be an incomplete name, without a metal foil = FKP or a metallization = MKP no cap is given. Then Vishay bought Roederstein and now a cap MKP1837 can be found. I have tested both, "old" KP1830 I got and new MKP1837. I can find no difference. So for the future I should just talk about the available (e.g. FARNELL) type MKP1837. Farnell offers two other types of this low voltage high precision MKP caps from other suppliers. Hifi-Fellows told me they work also fine. An other type is the orange EMZ-type like THEL is using http://www.thel-audioworld.de/pics/3KP.JPG . EMZ was SIEMENS in former times, another irritating story…
BTW, to quote guys talking about this wonder MKP-stuff I would like to give feedback from Tony Gee to read. Tony is a nice guy from Dutchland with a famous cap test homepage. It seems to be offline currently, to much traffic…
This small MKPs work the same good everywhere! So transform the crossover-application report from Tony to PSU and to coupling caps!
"Hi Klaus,
Well, I've tested them. I tried them as a bypass for the tweeter series caps in my Progress speaker and I must say I am very impressed! To get straight to the point they don't change a Supreme into a Supreme Silver-Oil but they realy do clear things up. I must admit I was scepticle at first as the value is only 8,2nF (0,0082uF) - measured 8,48nF - and the caps in the Progress are 12,6uF.
The difference is most noticable with classical music but also good quality recordings of jazz and fusion benifit:
- no change in soundstage width or depth but there is more "concert hall acoustics" that let you get into the recording more.
- not as liquid as silver/oil but they take away the "grainy" edge from the Supreme's.
- a gain in clarity and transparancy making instruments better seperable from each other, the violins in an orchestra are a group of individual violins instead of one mass.
- jazz drum brushes sound more like a brush than a "shush".
So all in all a very nice tweak! Thanks for the tip! How much do these things cost? Do you want your cap's back?
---
regards,
Tony Gee
tweaky regards
Klaus
Thanks for clearing that up Klaus.
I was going through Vishays datasheets and they have two caps:
http://www.vishay.com/docs/26017/mkp1837.pdf
http://www.vishay.com/docs/26016/kp1830.pdf
The mkp1837 seems only to be found in a 160VDC version and since you often quoted 63V 10nF (which kp1830 offers), I was starting to think mkp1837 wasn't the cap you were using. Just wanting to order the _right_ cap to do some serious tweaking. It's a disease of the mind!
/Niclas
I was going through Vishays datasheets and they have two caps:
http://www.vishay.com/docs/26017/mkp1837.pdf
http://www.vishay.com/docs/26016/kp1830.pdf
The mkp1837 seems only to be found in a 160VDC version and since you often quoted 63V 10nF (which kp1830 offers), I was starting to think mkp1837 wasn't the cap you were using. Just wanting to order the _right_ cap to do some serious tweaking. It's a disease of the mind!
/Niclas
Re: tests finished, BG wonderland...
Why don't you try BG type N. I was never too impressed with NX caps, I use them only when price is consideration.
N type sounds less colored to me. And don't use ANY additional bypasses, cap by itself sound most natural (I mean BG here). Also , orientation of BG cap influences sound. Start of the foil is marked with short lead (or NON POLAR print).
I like those blue 1837 caps, and I used them as bypasses in my preamp. How do they compare to those green ROE caps (in your opinion)?
Klaus said:
Why don't you try BG type N. I was never too impressed with NX caps, I use them only when price is consideration.
N type sounds less colored to me. And don't use ANY additional bypasses, cap by itself sound most natural (I mean BG here). Also , orientation of BG cap influences sound. Start of the foil is marked with short lead (or NON POLAR print).
I like those blue 1837 caps, and I used them as bypasses in my preamp. How do they compare to those green ROE caps (in your opinion)?
Re: Re: tests finished, BG wonderland...
Hi Peter,
what do you mean with "sound less colored"? Following your fist sentence I understand it as positive comment, but what do you mean in detail? Or better asked, does the N type sound also "cool&clean" like the NX, or better regarding this issue? Thanks for you comment on!
regards
Klaus
Peter Daniel said:
Hi Peter,
what do you mean with "sound less colored"? Following your fist sentence I understand it as positive comment, but what do you mean in detail? Or better asked, does the N type sound also "cool&clean" like the NX, or better regarding this issue? Thanks for you comment on!
regards
Klaus
Actually it seems to sound clener. By comparison NX sound a bit softer and veiled to me. I'm using N type as coupling caps and also as PS bypasses and I never noticed they were cold or excessively clean, just sounded right (at least to me).
But again, the sound of a given cap will depend on the rest of the circuit and it's location in such circuit. It's hard to give one answer.
For instance, I was modifying ML 38 preamp and I couldn't stand the sound of AD8610 chip when I placed it in the input gain location. It's sounded too bright, OPA627 sounded better.
Later I replaced all original electrolytics (Chemicom) to BG type N, and now the OPA sounded rather dull and noninvolving, with excess of coloration. I replaced it with AD8610 and this chip wasn't bright anymore. I decided to keep it.
I also had those 1837 Vishay caps as bypasses at the chip, but Panasonics sounded better in there (more appropriate tonal balance), but in other places in the circuit, 1837 caps were preferred to Panasonics (best grade from DigiKey).
But again, the sound of a given cap will depend on the rest of the circuit and it's location in such circuit. It's hard to give one answer.
For instance, I was modifying ML 38 preamp and I couldn't stand the sound of AD8610 chip when I placed it in the input gain location. It's sounded too bright, OPA627 sounded better.
Later I replaced all original electrolytics (Chemicom) to BG type N, and now the OPA sounded rather dull and noninvolving, with excess of coloration. I replaced it with AD8610 and this chip wasn't bright anymore. I decided to keep it.
I also had those 1837 Vishay caps as bypasses at the chip, but Panasonics sounded better in there (more appropriate tonal balance), but in other places in the circuit, 1837 caps were preferred to Panasonics (best grade from DigiKey).
Thanks, Peter, for detailing. It is funny, I can 100% confirm your try&error story effects - the complete circuit forms the sound. So the benefit from OPAs, bypasses, cap types and topology of caps close to the consumer or far away at the PSU side can vary from one extreme to the other. Tweaking means testing without to much thoughts as EE… No model of EE can describe these effects!
I will check the N type at next occasion, or go for Aleph X…
I will check the N type at next occasion, or go for Aleph X…
Klaus,
You are right about the effects on the '"whole".
Some have written about the sound of NX and the effect of E cap.
I have used the C BlackGate with the BOSOZ and it performs superbly in that application un bypassed.
Another option is to try the North Creek Cresendo, Kent and Nelson have used them with favourable results.
Ian
You are right about the effects on the '"whole".
Some have written about the sound of NX and the effect of E cap.
I have used the C BlackGate with the BOSOZ and it performs superbly in that application un bypassed.
Another option is to try the North Creek Cresendo, Kent and Nelson have used them with favourable results.
Ian
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.