Unfortunately, the BC series 114 / 115 has been discontinued, they had 3mOhm ESR / 20 nH ESL.
The current series 101/102 and 106 don't come close to this. Perhaps you can find some 114.
Here is the link to the data sheet: BC series 114
Jochen
The current series 101/102 and 106 don't come close to this. Perhaps you can find some 114.
Here is the link to the data sheet: BC series 114
Jochen
BC capacitors
I am planning to use the BC 058 series.
10 *10000uf/50V in parallel per rail which should give me a rated ripple current of 50A and an ESR of 3.7mOhm.The inductances would also be parallel.
www.vishay.com/docs/28342/058059pl.pdf
I am planning to use the BC 058 series.
10 *10000uf/50V in parallel per rail which should give me a rated ripple current of 50A and an ESR of 3.7mOhm.The inductances would also be parallel.
www.vishay.com/docs/28342/058059pl.pdf
Capacitor size for the GP
Why focus only on the larger values like eg. big 47,000uF caps for the capacitor banks?
It may be a gain in occupied space, but is the ripple supression of a well designed PSU with multiple rc-sections (= more and smaller C's and smaller R's eg. (R)CRCRCRC) or just more paralelled C's) not better than or at least equal to the same total capacitance achieved with say just 2 large (expensive?) capacitors?
Advantages could be:
In this view I would suggest a max. capacitance of 22000uf per device.
Any thoughts?
Why focus only on the larger values like eg. big 47,000uF caps for the capacitor banks?
It may be a gain in occupied space, but is the ripple supression of a well designed PSU with multiple rc-sections (= more and smaller C's and smaller R's eg. (R)CRCRCRC) or just more paralelled C's) not better than or at least equal to the same total capacitance achieved with say just 2 large (expensive?) capacitors?
Advantages could be:
- Lower inrush current phenomena?
- A lower price for the larger total quantities ordered in the group purchase?
- More flexibility for people participating in the GP to choose the total capacity in the PSU, eg. for people who wish to build a "standard" AlephX (20 - 40 Watts) requiring less capacitance in the PSU.
In this view I would suggest a max. capacitance of 22000uf per device.
Any thoughts?
The price of a 47000uF cap is usually much lower than 4 times the price of a 10000uF cap!
To parallel for example 2*22000uF shurely gives a better result but it is also more expensive. In the case, that a lower price for the larger total quantities ordered in the group purchase equals the higher costs
for the smaller caps, this would shurely be the better solution.
illusionxx.
To parallel for example 2*22000uF shurely gives a better result but it is also more expensive. In the case, that a lower price for the larger total quantities ordered in the group purchase equals the higher costs
for the smaller caps, this would shurely be the better solution.
illusionxx.
Maybe I stay with my first idea, CRCRC with the Panasonic 150,000uF 5mOhm caps (Eur28/pc). Ain't that expensive compared to the other solutions.
Hello all,
I didn't go through the complete thread, but did you guys read the article on Class A power amp supplies by Rod Elliott?
In the article, he prefers a greater number of smaller caps as opposed to a couple of big ones, based on the lower cost and larger ripple current rating.
I agree that the difference in costs between 10mF and, for instance, 22mF is not that big but the difference between 10x4700uF and 1x 47mF is about €3,60 (according to the excellent online catalog of Reichelt).
Regards,
Jarno.
I didn't go through the complete thread, but did you guys read the article on Class A power amp supplies by Rod Elliott?
In the article, he prefers a greater number of smaller caps as opposed to a couple of big ones, based on the lower cost and larger ripple current rating.
I agree that the difference in costs between 10mF and, for instance, 22mF is not that big but the difference between 10x4700uF and 1x 47mF is about €3,60 (according to the excellent online catalog of Reichelt).
Regards,
Jarno.
Jarno said:
Read some more :
Gunther mania
Grundlagen der audio technik
Stereoplay januari till august 1988
Reinhard Wachowiak
Audiolabor Fein + brillant
www.Audiolabor.de
Michiel de Goeij en edwin weetikveel
Sphinx project 12
Scylla Mosfet amplifiers
ESR ratio on average for capacitors of the same construction is Sqrt(C1/C2).
Meaning : 10 parallel 4700uF caps have an ESR value 0.0316 of that of a single 47000uF cap.
jacco vermeulen said:
Are you sure it's not 1 / (1/C1) + (1/C2) + etc.?
PSU Designer can't have caps in parallel, so I have to calculate them myself.
TheJudge said:
Nope !
Sqrt( 4700/47000) = Sqrt (0.10) =0.316
0.316 / 10 = 0.0316
1/ (10 * 1/1) = 1/10
This would mean that a single 4700 µF would have a lower ESR than a single 47000µF, which is not the case. Perhaps you swapped C1 and C2, so sqrt(47000/4700) = sqrt(10) = 3.16 (the 47000 is 3.16 times better than a single 4700). So 10x4700 are 10/3.16 times better regarding ESR than a single 47000.
Mmhhh. looks more realistic.
Mmhhh. looks more realistic.
Hi illusionxx,
Some time ago I did some calculations in Excel for myself.
I was aiming at using the F&I capacitors from Conrad, 22000uF/63V or 47000/63V and using their quantity discount rates.
At a total quantity of 104 (13 sets, being 2 banks of 4 caps of 22000uf per set) to get the 100+ discount, versus 52 (13*2*2 of 47000uF) to get about the same capacity it came out just under €5 difference in price per set!?
So pricewise there was no big difference.
Taken into account the other mentioned advantages and the fact that quantum discounts for 100+ pcs. are not that good at Conrad, I think the comparrison could be even more in favor of the smaller caps and that there is no real need to go for the bigger caps?
TheJudge
I also tried putting in caps in paralel in PSU-designer.
I (kind of) suceeded only by using CR sections with the lowest possible R of 0.0011 Ohm (1.1 mili-Ohms), the resistance you might expect from wiring- and contact resistance etc.
I was suprised though by the results; the total voltage drop over these resistances was rather big and I don't know if I can trust these results.
Some time ago I did some calculations in Excel for myself.
I was aiming at using the F&I capacitors from Conrad, 22000uF/63V or 47000/63V and using their quantity discount rates.
At a total quantity of 104 (13 sets, being 2 banks of 4 caps of 22000uf per set) to get the 100+ discount, versus 52 (13*2*2 of 47000uF) to get about the same capacity it came out just under €5 difference in price per set!?
So pricewise there was no big difference.
Taken into account the other mentioned advantages and the fact that quantum discounts for 100+ pcs. are not that good at Conrad, I think the comparrison could be even more in favor of the smaller caps and that there is no real need to go for the bigger caps?
TheJudge
I also tried putting in caps in paralel in PSU-designer.
I (kind of) suceeded only by using CR sections with the lowest possible R of 0.0011 Ohm (1.1 mili-Ohms), the resistance you might expect from wiring- and contact resistance etc.
I was suprised though by the results; the total voltage drop over these resistances was rather big and I don't know if I can trust these results.
Did anyone of you guys read the posting of Mr Lgreen ?
He made the suggestion of a group buy and matching for Mosfets.
I am building 4 more Aleph's and 2 Zen XA's for which i need some 100 Mosfets.
I am supposing some of you may be in need for Mosfets too.
He made the suggestion of a group buy and matching for Mosfets.
I am building 4 more Aleph's and 2 Zen XA's for which i need some 100 Mosfets.
I am supposing some of you may be in need for Mosfets too.
Hello Jacco,
Maybe the guy(s) at Specmail can give you a quantity discount on matched sets? You might need less Fets if they are already matched, and maybe those guys are interested in delivering larger quantities?
BTW, that's a reallly nice stash of caps! You really have great dumpsters in which to dive!
Regards,
Jarno.
Maybe the guy(s) at Specmail can give you a quantity discount on matched sets? You might need less Fets if they are already matched, and maybe those guys are interested in delivering larger quantities?
BTW, that's a reallly nice stash of caps! You really have great dumpsters in which to dive!
Regards,
Jarno.
Hi Jacco,
It was my reasoning to set up a WIKI for the Dutch group purchase, ie. to get an idea of the real demands for the different parts as a group purchase.
I was too busy however with other things and did not have the time to look into it yet how to do this and do some basic page set up.
I will look into it this evening (night? 😀) and try to set it up.
People can state their interest(s) there and we can use it to see if a group purchase is feasable.
I will post a link here when it's up.
I will also have a look at the thread of Mr Lgreen, maybe we can work together with them.
Personally I am interested only in high quality transformers and -capacitors, (matched-) IRFP9610 and maybe other board components.
It was my reasoning to set up a WIKI for the Dutch group purchase, ie. to get an idea of the real demands for the different parts as a group purchase.
I was too busy however with other things and did not have the time to look into it yet how to do this and do some basic page set up.
I will look into it this evening (night? 😀) and try to set it up.
People can state their interest(s) there and we can use it to see if a group purchase is feasable.
I will post a link here when it's up.
I will also have a look at the thread of Mr Lgreen, maybe we can work together with them.
Personally I am interested only in high quality transformers and -capacitors, (matched-) IRFP9610 and maybe other board components.
byteboy,
i`m not against small caps, but the calculation you did depends on the pricerange of the caps. More expensive caps will have a higher difference in price.
I was thinking of the Krummer caps also because of their long livespan of 560.000 h @ 40°C.
But taking into acount my financial situation i must agree that the best solution will be resonably priced small caps (let`s say 10000uF). This way it will be easy to hit the 100pc. pricebreak.
I think you`ll take some into the WIKI?!
In my opinion it`s the best to take the ESR values out of the datasheet and calculate with these numbers. So the ESR of 10 caps in parallel is 1/10 of the ESR of a single cap (simply parallel R`s). Same with ESL.
For example:
47000uF : ESR = 5mOhm
10000uF : ESR = 15mOhm
4 parallel 10000uF caps : ESR = 15mOhm/4 = 3.75mOhm ;
So we have 3.75mOhm against 5mOhm.
illusionxx
i`m not against small caps, but the calculation you did depends on the pricerange of the caps. More expensive caps will have a higher difference in price.
I was thinking of the Krummer caps also because of their long livespan of 560.000 h @ 40°C.
But taking into acount my financial situation i must agree that the best solution will be resonably priced small caps (let`s say 10000uF). This way it will be easy to hit the 100pc. pricebreak.
I think you`ll take some into the WIKI?!
In my opinion it`s the best to take the ESR values out of the datasheet and calculate with these numbers. So the ESR of 10 caps in parallel is 1/10 of the ESR of a single cap (simply parallel R`s). Same with ESL.
For example:
47000uF : ESR = 5mOhm
10000uF : ESR = 15mOhm
4 parallel 10000uF caps : ESR = 15mOhm/4 = 3.75mOhm ;
So we have 3.75mOhm against 5mOhm.
illusionxx
So, if you are hunting for low ESR, which is the better of the two ?
btw :
it is common practice to place small caps on pcb's just before the sources/collectors of the Q's.
The reason being that small capacitors are much faster for impuls power than large caps.
Maybe one should differentiate between current going in the cap and current going out.
For low ripple in class A amplifiers i switched to very large capacitors long ago, going to higher and higher capacity caps.
I built Jean Hiraga's Le Class A for which i employed extreme big SCR caps. It worked well for the froggies, it worked for me.
Nevertheless, i still use large banks of small capacitors placed on powersupply PCB's too.
In my view combining a large capacitor in combination with a number of small ones makes much more sense than investing in esoteric capacitors like the one from Thel.
Magura's link for the Rifa 169 and 200 is high quality for the buck,
the Rifa's are among the best.
btw :
it is common practice to place small caps on pcb's just before the sources/collectors of the Q's.
The reason being that small capacitors are much faster for impuls power than large caps.
Maybe one should differentiate between current going in the cap and current going out.
For low ripple in class A amplifiers i switched to very large capacitors long ago, going to higher and higher capacity caps.
I built Jean Hiraga's Le Class A for which i employed extreme big SCR caps. It worked well for the froggies, it worked for me.
Nevertheless, i still use large banks of small capacitors placed on powersupply PCB's too.
In my view combining a large capacitor in combination with a number of small ones makes much more sense than investing in esoteric capacitors like the one from Thel.
Magura's link for the Rifa 169 and 200 is high quality for the buck,
the Rifa's are among the best.
- Status
- Not open for further replies.