Hi,
slightly OT on this thread but I can't find my answer and this seemed like a good place to start.
In the Aleph 30 there are 2 small caps. A 10pf that seems to be in the feedback loop and a .001uf that seems to be in the ccs feedback loop. How important is it to stick to these values? How far off can I be...what are the repercussions of maybe using a .1 instead of .001 etc? I have access to a ton of high quality caps (board farming.. poly/silver/tant. etc.) None as small as I need though.
Thanks guys
slightly OT on this thread but I can't find my answer and this seemed like a good place to start.
In the Aleph 30 there are 2 small caps. A 10pf that seems to be in the feedback loop and a .001uf that seems to be in the ccs feedback loop. How important is it to stick to these values? How far off can I be...what are the repercussions of maybe using a .1 instead of .001 etc? I have access to a ton of high quality caps (board farming.. poly/silver/tant. etc.) None as small as I need though.
Thanks guys
Hi Mpmarino,
these are compensation caps. The value depends a bit on layout, used FEts, wiring etc.
I did the following in my Aleph-X to determine the right value:
1. disconnect active current source (leave out resistor or 220uF cap) and look at a 10kHz square wave with an 8 ohm load. If you´ve got overshout try a 10pF cap in the feedback loop. If there´s still overshoot try a 22pF one etc. etc. etc.
2. connect active current source again. Now you will probably have some nice oscillation........Put in a 1nF and look at the square wave again. If overshoot --> bigger value, if too rounded use a smaller value.
You can´t use much bigger values in the feedback loop cause this will reduce gain at higher frequencies. I´ve used silver mica for the 10pF ones and WIMA FKP2 for all the other small values.
William
these are compensation caps. The value depends a bit on layout, used FEts, wiring etc.
I did the following in my Aleph-X to determine the right value:
1. disconnect active current source (leave out resistor or 220uF cap) and look at a 10kHz square wave with an 8 ohm load. If you´ve got overshout try a 10pF cap in the feedback loop. If there´s still overshoot try a 22pF one etc. etc. etc.
2. connect active current source again. Now you will probably have some nice oscillation........Put in a 1nF and look at the square wave again. If overshoot --> bigger value, if too rounded use a smaller value.
You can´t use much bigger values in the feedback loop cause this will reduce gain at higher frequencies. I´ve used silver mica for the 10pF ones and WIMA FKP2 for all the other small values.
William
Panasonic FC for this agonizing little capacitor
Evening All,
I am just putting the final bits onto my digikey order for an Aleph 5(or 5'ish - actually an Aleph 3 with a balanced 5 front end) amplifier ...
I have studied the spec sheet for the Panasonic FC caps.
It would appear that it would be OK with a 6.3V rating, but it has a VERY short life (1000 hrs).
Looking at the higher voltage capacitors, I note they have MUCH longer lives (either 3000 or 5000 hrs), but interestingly a lot LOWER impedance at 100kHz. The low
Intuitively I would have thought:
a) that lower impedance would be a "good thing", and that;
b) it would "sound better".
What do you wise ones recommend?
Regards,
Yours urgently,
George.
Specifically: (Voltage; Impedance; Endurance)
6.3V 0.350R 1000 hr
50V 0.110R 5000hr
50V 0.090R 3000hr
63V 0.085R 5000hr
63V 0.147R 3000hr
Evening All,
I am just putting the final bits onto my digikey order for an Aleph 5(or 5'ish - actually an Aleph 3 with a balanced 5 front end) amplifier ...
I have studied the spec sheet for the Panasonic FC caps.
It would appear that it would be OK with a 6.3V rating, but it has a VERY short life (1000 hrs).
Looking at the higher voltage capacitors, I note they have MUCH longer lives (either 3000 or 5000 hrs), but interestingly a lot LOWER impedance at 100kHz. The low
Intuitively I would have thought:
a) that lower impedance would be a "good thing", and that;
b) it would "sound better".
What do you wise ones recommend?
Regards,
Yours urgently,
George.
Specifically: (Voltage; Impedance; Endurance)
6.3V 0.350R 1000 hr
50V 0.110R 5000hr
50V 0.090R 3000hr
63V 0.085R 5000hr
63V 0.147R 3000hr
I have just been reviewing this thread.
I think its a case by case thing and depends on whether there is a Dc potential and if a large value is required as large values have poor Hf properties and inductance .
A bit off topic but I recently I did some research for coupling capacitors in my High Low Pass active crossover. Very little offset about 10 mv but too much for a dc coupled power amp
I tried:
Panasonic 220uf Fc bypassed with MFP1837 0.047.
Warm and engaging bloom, but coloured.
Nichison Muse FG 10 uf not bypassed.
Tighter bass and high were exagerated.
BG NX 4.7 anti parellel
Dull mid range and veiled highs, loss of Hf details
BG N 4.7 ant parellel
Cleaner and more open than BG NX bu lacked transient speed compared not direct coupling.
Auricaps
Very slight bloom but otherwise invisible and superior to all other capacitors I have tried to date.
But I found that I needed 10 uf for a 22kohn differential balanced output on the low pass outputs , 4.7 uf value caused some audible phase shift.
I also found the 220 uf value on the Pass Fet regulator had an effect and I found Elna Cerefine 200uf the best. I also found that bypassing each discrete opamp with 10 uf blurred the sound so i removed them and use only 0.1 mkt bypass to block any HF breakthrough.
I figure these investigations might be useful for the BOSOZ also.
macka
I think its a case by case thing and depends on whether there is a Dc potential and if a large value is required as large values have poor Hf properties and inductance .
A bit off topic but I recently I did some research for coupling capacitors in my High Low Pass active crossover. Very little offset about 10 mv but too much for a dc coupled power amp
I tried:
Panasonic 220uf Fc bypassed with MFP1837 0.047.
Warm and engaging bloom, but coloured.
Nichison Muse FG 10 uf not bypassed.
Tighter bass and high were exagerated.
BG NX 4.7 anti parellel
Dull mid range and veiled highs, loss of Hf details
BG N 4.7 ant parellel
Cleaner and more open than BG NX bu lacked transient speed compared not direct coupling.
Auricaps
Very slight bloom but otherwise invisible and superior to all other capacitors I have tried to date.
But I found that I needed 10 uf for a 22kohn differential balanced output on the low pass outputs , 4.7 uf value caused some audible phase shift.
I also found the 220 uf value on the Pass Fet regulator had an effect and I found Elna Cerefine 200uf the best. I also found that bypassing each discrete opamp with 10 uf blurred the sound so i removed them and use only 0.1 mkt bypass to block any HF breakthrough.
I figure these investigations might be useful for the BOSOZ also.
macka
Mmm George,
the 1000 Hrs you mention is likely to be the life expectancy at 100 or 105C, the max operating temperature of the Panasonic capacitors. The datasheet of the FCs should have it mentioned somewhere at which temperature the 1000 Hrs is specified.
The lifespan is probably something as 4000 Hrs at 80C, and 100.000 Hrs at 40 C.
the 1000 Hrs you mention is likely to be the life expectancy at 100 or 105C, the max operating temperature of the Panasonic capacitors. The datasheet of the FCs should have it mentioned somewhere at which temperature the 1000 Hrs is specified.
The lifespan is probably something as 4000 Hrs at 80C, and 100.000 Hrs at 40 C.
Thank-you everybody for your responses.
I understand your point about the lifepan of the capacitors. I thought that this would be under "extreme" circumstances.
Does anyone have any thoughts about the sound or the logic of the different capacitors, namely:
Is using the "little" 6.3 volt with the "high" impedance at 100kHz better or worse than using the "big" 63 Volt capacitor with the "low" impedance as the coupling capacitor?
With thanks (again
),
George
I understand your point about the lifepan of the capacitors. I thought that this would be under "extreme" circumstances.
Does anyone have any thoughts about the sound or the logic of the different capacitors, namely:
Is using the "little" 6.3 volt with the "high" impedance at 100kHz better or worse than using the "big" 63 Volt capacitor with the "low" impedance as the coupling capacitor?
With thanks (again
),George
I installed the BG N 470uF as decoupling caps in my Aleph 1.2s. The first two days, about 15 hours of listening they made my aleph sing like nothing I ever have heard before. Just plain wonderfull. Nice and warm midtones with a lot of detail and very thight bass and powerful bass. However, afterwards the sound has become cold and very forward.
Is this due to the burning in of the black gate capasitors? What should I expect? Will the fantastic sound every return?
Tnx!
F
Is this due to the burning in of the black gate capasitors? What should I expect? Will the fantastic sound every return?
Tnx!
F
Hi Mr. Nelson!
Thank you so much for answering me. I am quite the novice and did not understand which resistor you recommended me to change. Which resistor(s?) do you refer to (in the service manual schematc -R9, R10 and R11)?
The Aleph 1.2 made my speakers sing as they have never done before. It was really amasing. I spent all last weekend listening to as many cds as possible!
Thank you for all the help, and thank you for making this project possible!!
Best regards,
Folke
Thank you so much for answering me. I am quite the novice and did not understand which resistor you recommended me to change. Which resistor(s?) do you refer to (in the service manual schematc -R9, R10 and R11)?
The Aleph 1.2 made my speakers sing as they have never done before. It was really amasing. I spent all last weekend listening to as many cds as possible!
Thank you for all the help, and thank you for making this project possible!!
Best regards,
Folke
Just my experience:
I have tried the BG N and NX several times in anti parallel mode (the theorists have wonderful arguments on paper) but I was never pleased with it. Just one plain BG-N or NX always sound ways better! Much more open and fast than two in anti parallel.
BTW, Claus did extensive testing with these 220 uF caps in the Alephs. See earlier in this thread.
My own favourite is just one BG-N without any other cap.
Loudspeaker optimized MKP's are much slower and constrainted at this position.
Lucas
I have tried the BG N and NX several times in anti parallel mode (the theorists have wonderful arguments on paper) but I was never pleased with it. Just one plain BG-N or NX always sound ways better! Much more open and fast than two in anti parallel.
BTW, Claus did extensive testing with these 220 uF caps in the Alephs. See earlier in this thread.
My own favourite is just one BG-N without any other cap.
Loudspeaker optimized MKP's are much slower and constrainted at this position.
Lucas
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.