Hi Ian,
>Love your descriptions.
--- sometimes the vocabulary of hifi freaks must sound like voodoo…
>I plan to try 220uf Panasonic FC as coupling capacitor with 0.047uf 1837 capacitors in my High Low Pass.
>I know your have been referring to 0.01uf, just 0.047 seems to cover more than just the highs and according to my sums 220uf looks more like short circuit above about >100 hertz and the large value just preserves phase shift and group delay.
>I am sure I saw 0.047 as by Pass in a shematic with 220uf FC's somewhere.
>Any thoughts?
My only thought: your statements are 100% correct, I do not know. To be honest: I have never tested it in detail. My master of bypassing was Mr. Hartwig from Thel in Germany. Here I saw these small MKP-bypasses first. I mean highest quality bypass, not the common bypass used in any consumer device and responsible for much of the shrillness in sound… Thel uses 22nF, not more. Why ever. By telephone he told me that 10-22nF shall be sufficient. I found those 10nF KP1830 in a box here at my employer, so I took them. I would recommend you just to test it.
I think in the D1 the first time bypasses were used. But if it was really tested by listening sessions? I am not sure…
Regards from old Europe
Klaus
>Love your descriptions.
--- sometimes the vocabulary of hifi freaks must sound like voodoo…
>I plan to try 220uf Panasonic FC as coupling capacitor with 0.047uf 1837 capacitors in my High Low Pass.
>I know your have been referring to 0.01uf, just 0.047 seems to cover more than just the highs and according to my sums 220uf looks more like short circuit above about >100 hertz and the large value just preserves phase shift and group delay.
>I am sure I saw 0.047 as by Pass in a shematic with 220uf FC's somewhere.
>Any thoughts?
My only thought: your statements are 100% correct, I do not know. To be honest: I have never tested it in detail. My master of bypassing was Mr. Hartwig from Thel in Germany. Here I saw these small MKP-bypasses first. I mean highest quality bypass, not the common bypass used in any consumer device and responsible for much of the shrillness in sound… Thel uses 22nF, not more. Why ever. By telephone he told me that 10-22nF shall be sufficient. I found those 10nF KP1830 in a box here at my employer, so I took them. I would recommend you just to test it.
I think in the D1 the first time bypasses were used. But if it was really tested by listening sessions? I am not sure…
Regards from old Europe
Klaus
Hi Klaus,
like I said. All the ingredients are there but something ?? just isn´t really right at the moment. The investment wasn´t that big and I´m shure I can use those caps somewhere in case I´ll change back.
I thought about possible causes last night but the only thing I came up with is the fact that I´m possibly listening at higher levels as before (changed the position of my volume knob at the same time
)
I´ll just wait a bit longer
William
P.S. concrete failures cause bridges to collapse
William
like I said. All the ingredients are there but something ?? just isn´t really right at the moment. The investment wasn´t that big and I´m shure I can use those caps somewhere in case I´ll change back.
I thought about possible causes last night but the only thing I came up with is the fact that I´m possibly listening at higher levels as before (changed the position of my volume knob at the same time
)I´ll just wait a bit longer
William
P.S. concrete failures cause bridges to collapse
William
Hi William,
Could it be that you just miss these brute (pseudo) attack of the NX? They IMHO seem to "spit" into ones face, fun, especially with Rock music. The N sounds fuller, more colored, laid back heights – it is competitive unspectacularly. But how to explain this contradiction: more details and "inner dynamic" can be heard. Maybe listen to acoustical music, here the N should show what "neutral" mean. But however, there is no wrong and right. If your setup benefits from the NX sound, there we are. Maybe my set hates NX effects, because already my speakers sound "frontal" and it became too much. It is a pity that these tests are so expensive…
Regards
Klaus
Could it be that you just miss these brute (pseudo) attack of the NX? They IMHO seem to "spit" into ones face, fun, especially with Rock music. The N sounds fuller, more colored, laid back heights – it is competitive unspectacularly. But how to explain this contradiction: more details and "inner dynamic" can be heard. Maybe listen to acoustical music, here the N should show what "neutral" mean. But however, there is no wrong and right. If your setup benefits from the NX sound, there we are. Maybe my set hates NX effects, because already my speakers sound "frontal" and it became too much. It is a pity that these tests are so expensive…
Regards
Klaus
Hi Klaus,
I believe I got the cause of my little unhappiness. Because of the somewhat laidback character and the repositioning of my volume knob I´ve been listening at somewhat higher levels than before wich probably causes my A5 to distort a bit (only 30-35watts into 2.3 Ohms).
Listening levels only become clear when trying to hear the phone (didn´t) or doorbell (didn´t).
I turned the volume down two clicks wich helped
William
I believe I got the cause of my little unhappiness. Because of the somewhat laidback character and the repositioning of my volume knob I´ve been listening at somewhat higher levels than before wich probably causes my A5 to distort a bit (only 30-35watts into 2.3 Ohms).
Listening levels only become clear when trying to hear the phone (didn´t) or doorbell (didn´t).
I turned the volume down two clicks wich helped
William
wuffwaff said:
Being an aspirering Aleph-X-builder, I too have wondered whether it would be worthwhile to do a little tweaking on the inputcaps. Would it be overkill to try Klaus’ composite cap here?
Wuffwaff have you build and Aleph-X, and if yes, what caps have you used for inputcaps?
For the other 220uf-caps I have decided on Panasonic FC’s with the MKP1837 cap, any objections to that, or perhaps a better idea?
I know that in the end it all depend on my own ears, and I plan to do some testing with different caps, but since I haven’t build the darn thing yet, I still have to decide what I want to try first.
lpm76 said:
To remember: in principal just caps with less than 2V and not defined polarity of polarization voltage need to be BG to achieve best results.
Another effect: Pana FCs shall sound much better than any MKP. By the time I got to know two guys who changed from 10-20uF MKP to 33 or 47uF Pana FC plus 10nF MKP bypass and were much more satisfied. And the price differences...
Ipm 76,
still not ready yet I´ve got most of the parts by now and will start the metalwork soon.
220uF Panasonic FC plus ERO 10nF sound good. You can use this everywhere where the cap is biased with a small (or bigger) voltage. If this is not the case (coupling caps) it depends on how big they need to be.
I wouldn´t use a 470uF cap on the X-input. You could go for a smaller (10 or 20uF BG-N) or a film cap bypassed with the famous 10nF. The advantage of the BG is size (compared to a 10uF film cap) and price (against the more expensive film caps).
William
still not ready yet
I´ve got most of the parts by now and will start the metalwork soon.220uF Panasonic FC plus ERO 10nF sound good. You can use this everywhere where the cap is biased with a small (or bigger) voltage. If this is not the case (coupling caps) it depends on how big they need to be.
I wouldn´t use a 470uF cap on the X-input. You could go for a smaller (10 or 20uF BG-N) or a film cap bypassed with the famous 10nF. The advantage of the BG is size (compared to a 10uF film cap) and price (against the more expensive film caps).
William
lpm76 said:Thank you Klaus and Wuffwaff for you crystal clear answers. That is much appreciated. I have been reading this thread with such and interest, that I have begun to look up what you 2 guys have posted in other threads. Take that as a compliment.
I am just a stupid and curious tweaky, Mr. Pass is the guy to get complements! W/o his circuits for free my love to repoduced music would have stopped one day. Now, with tweaked Pass DIY equipment I can not beleave what I hear.
regards
Klaus
Ofcourse you are right Klaus, but my thanks to Mr Pass will be plenty, when the Amp is done.
By the way, is see alot of people mention Elna Cerafine-caps and few people thinks it is worth the money it costs - even though that aint much. What puzzles me is that nobody seem to talk about the Elna Silmic even though it is supposed to be a better cap. Do you have any experience with it? The Elna Silmic that is...
By the way, is see alot of people mention Elna Cerafine-caps and few people thinks it is worth the money it costs - even though that aint much. What puzzles me is that nobody seem to talk about the Elna Silmic even though it is supposed to be a better cap. Do you have any experience with it? The Elna Silmic that is...
This is a quote of Nelson Pass taken from another thread:
"The cap gives the amplifier unity gain at DC and
is used to give low DC offset to the output and
protect the speaker against DC at the input of the
amp. You can safely remove it, but you will find
yourself making adjustments to the offset as the
temperature. The best place for adjustment would
be either R3 or R24 on the Aleph 4 (ref Left channel).
The Aleph 0 had a pot at this location, and no capacitors at all, but there was simply too much field adjustment
required to keep the offset low, and too many users had
DC coming out of preamps."
I took the cap out of my Aleph 4 and I thought it was a nice improvement. The amp was a bit 'spongy' or 'wooly' sounding prior to to removing the (Panasonic FC) cap. I also tried tantalum caps and much prefered these to the FC caps, the tantalums seem to posess a strong individual set of (psuedo) characteristics. I do like the added bass slam of tantalums and the way they highlight the treble when listening to modern rock and roll but over all I prefer the clarity and added subtlety that comes with omitting the cap all together.
Refering to the Aleph 4 schematic after removing the cap I let the amp come up to operating temp and then tweeked R24 (loading resistor on the drain of the diff pair) by adding some resistance in parrallel untill the DC offset was sufficiently low.
Removing the cap is certainly the cheapest way to go and represents the least compromise in sound quality but of course one has to live with a little DC offset during warm up but after that the offset has been low and fairly steady. If one doesn't have DC coming for the preamp removing the cap seems like a pretty darn good option IMHO.
"The cap gives the amplifier unity gain at DC and
is used to give low DC offset to the output and
protect the speaker against DC at the input of the
amp. You can safely remove it, but you will find
yourself making adjustments to the offset as the
temperature. The best place for adjustment would
be either R3 or R24 on the Aleph 4 (ref Left channel).
The Aleph 0 had a pot at this location, and no capacitors at all, but there was simply too much field adjustment
required to keep the offset low, and too many users had
DC coming out of preamps."
I took the cap out of my Aleph 4 and I thought it was a nice improvement. The amp was a bit 'spongy' or 'wooly' sounding prior to to removing the (Panasonic FC) cap. I also tried tantalum caps and much prefered these to the FC caps, the tantalums seem to posess a strong individual set of (psuedo) characteristics. I do like the added bass slam of tantalums and the way they highlight the treble when listening to modern rock and roll but over all I prefer the clarity and added subtlety that comes with omitting the cap all together.
Refering to the Aleph 4 schematic after removing the cap I let the amp come up to operating temp and then tweeked R24 (loading resistor on the drain of the diff pair) by adding some resistance in parrallel untill the DC offset was sufficiently low.
Removing the cap is certainly the cheapest way to go and represents the least compromise in sound quality but of course one has to live with a little DC offset during warm up but after that the offset has been low and fairly steady. If one doesn't have DC coming for the preamp removing the cap seems like a pretty darn good option IMHO.
Hi William,
I have been trying different bias currents for the diff pair and the last time I rebiased the pair I didn't bother to get to fussy in respect to the DC-offset because my KEF 105 II's use coupling capacitors...at least for now
I've been meaning to try a different value of current for the diff pair. In the next couple of days I'll rebias the diff pair and I'll try to get the DC-offset as low as possible after warm up then I will post the DC-offset values from cold to warm up.
I for one would love to try a DC-servo, please tell us what you have in mind!
I have been trying different bias currents for the diff pair and the last time I rebiased the pair I didn't bother to get to fussy in respect to the DC-offset because my KEF 105 II's use coupling capacitors...at least for now
I've been meaning to try a different value of current for the diff pair. In the next couple of days I'll rebias the diff pair and I'll try to get the DC-offset as low as possible after warm up then I will post the DC-offset values from cold to warm up.
I for one would love to try a DC-servo, please tell us what you have in mind!
Hi Gaucho,
just a simple dc-servo build around an LF411 or OP77GP. It consist of an input filter (low pass) of 220nF and 1MOhm (<2Hz)and an output resistor and is connected to the -IN of the diff pair. Power supply is taken from the main rail with resistors and decoupled localy. I´ve got the schematic at home and I can post it if you´re interested.
One has to see how the startup behaviour is but there really isn´t much to it. I´ve used these for years in my Crescendo amp.
I don´t know why they weren´t used in the Alephs. Complexity? Cost? Sound? The first two reasons would be no problem, the last one would.
William
just a simple dc-servo build around an LF411 or OP77GP. It consist of an input filter (low pass) of 220nF and 1MOhm (<2Hz)and an output resistor and is connected to the -IN of the diff pair. Power supply is taken from the main rail with resistors and decoupled localy. I´ve got the schematic at home and I can post it if you´re interested.
One has to see how the startup behaviour is but there really isn´t much to it. I´ve used these for years in my Crescendo amp.
I don´t know why they weren´t used in the Alephs. Complexity? Cost? Sound? The first two reasons would be no problem, the last one would.
William
Sound (again)
Hi,
I think I found the cause of my not being quite happy with the BG setup. I´ve put the 470uF cap underneath the circuit board (not shielded from the power supply) and the rest of the caps on the other side. Maybe this way some power supply noise (the cap was directly above the chokes) got into the signal
Well I changed this so now the caps are all on the right side of the ground plane. Now I´m happy again. Sound is as described before but without the slightly irritating in it.
William
Hi,
I think I found the cause of my not being quite happy with the BG setup. I´ve put the 470uF cap underneath the circuit board (not shielded from the power supply) and the rest of the caps on the other side. Maybe this way some power supply noise (the cap was directly above the chokes) got into the signal
Well I changed this so now the caps are all on the right side of the ground plane. Now I´m happy
again. Sound is as described before but without the slightly irritating in it.William
HI,
here are 3 dc-servos published in elektuur over the last 10 years. A drawing of a possible power supply is in the lower right corner.
The resistor at the output of the opamp should be chosen so that the servo puts out around 3-6volts. Depending on the offset and input impedance this can vary quite a lot (I think I´ve got around 22k in my Crescendo).
William
P.S. I used the third one
here are 3 dc-servos published in elektuur over the last 10 years. A drawing of a possible power supply is in the lower right corner.
The resistor at the output of the opamp should be chosen so that the servo puts out around 3-6volts. Depending on the offset and input impedance this can vary quite a lot (I think I´ve got around 22k in my Crescendo).
William
P.S. I used the third one
Attachments
DC-offset
I rebiased the differentail pairs (to 80mA!) in my Aleph but I don't like the effect, it sounds to edgy, I was tempted to pull some feedback out of the amp to address this problem but I have a little hum as it is so I'll put the diff bias back to the 40mA I was running it at before.
Before I mucked around with the bias I did check the DC-offset over a 1 hour warm-up period. Here are the numbers:
1 sec: -325mV
1 min: -180mV
2 min: -53mV
3 min: +2.5mV
4 min: -9.5mV
5 min: +17.2mV
10 min: +25mV
15 min: -11mV
35 min: -35mV
65 min: +108mV
I should point out I don't use a CCS for the diff pair just a resistor and that my diff pair mosfets were matched (at 10mA each) and are not thermally linked. The heatsinks I use are fairly low mass for their given dissipation. The offset was fairly stable after about an hour, so it should be possible to get rid of most of the final DC. I'll see if I can improve upon these numbers.
I rebiased the differentail pairs (to 80mA!) in my Aleph but I don't like the effect, it sounds to edgy, I was tempted to pull some feedback out of the amp to address this problem but I have a little hum as it is so I'll put the diff bias back to the 40mA I was running it at before.
Before I mucked around with the bias I did check the DC-offset over a 1 hour warm-up period. Here are the numbers:
1 sec: -325mV
1 min: -180mV
2 min: -53mV
3 min: +2.5mV
4 min: -9.5mV
5 min: +17.2mV
10 min: +25mV
15 min: -11mV
35 min: -35mV
65 min: +108mV
I should point out I don't use a CCS for the diff pair just a resistor and that my diff pair mosfets were matched (at 10mA each) and are not thermally linked. The heatsinks I use are fairly low mass for their given dissipation. The offset was fairly stable after about an hour, so it should be possible to get rid of most of the final DC. I'll see if I can improve upon these numbers.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.