It's hard to find such small values in polystyrene. The smallest Xicons I can find are 100pF.
I did find some lower prices on Xicon Silver Micas, which could reduce kit price a bit (a buck or two), thoough it's another supplier, so more shipping charges need to be figured. I'll look at it tonight.
I did find some lower prices on Xicon Silver Micas, which could reduce kit price a bit (a buck or two), thoough it's another supplier, so more shipping charges need to be figured. I'll look at it tonight.
Panasonic Film caps in the kit (DigiKey part numbers)
P4656-ND CAP .1UF 100V POLY B SERIES C4, C5, C7, C17-C20
P4570-ND CAP .0001UF 50V POLY B SERIES C10
P4574-ND CAP .00022UF 50V POLY B SERIES C12
E1105-ND 1.0UF/100VDC METAL POLY CAP C13
P4551-ND CAP .001UF 50V POLYESTER C30
P4572-ND CAP .00015UF 50V POLY B SERIES C32
Panasonic EL caps in the kit
P6690-ND CAPACITOR 10000UF 50V ELECT TSHA C3, C8
P10325-ND CAP 220UF 50V ELECT FC RADIAL C1, C2, C9, C14
P11213-ND CAP 22UF 25V ELECT FC RADIAL C15
P10269-ND CAP 100UF 25V ELECT FC RADIAL C6, C11, C16
The FCs are "low-ESR" which is a relative term, I suppose.
P4656-ND CAP .1UF 100V POLY B SERIES C4, C5, C7, C17-C20
P4570-ND CAP .0001UF 50V POLY B SERIES C10
P4574-ND CAP .00022UF 50V POLY B SERIES C12
E1105-ND 1.0UF/100VDC METAL POLY CAP C13
P4551-ND CAP .001UF 50V POLYESTER C30
P4572-ND CAP .00015UF 50V POLY B SERIES C32
Panasonic EL caps in the kit
P6690-ND CAPACITOR 10000UF 50V ELECT TSHA C3, C8
P10325-ND CAP 220UF 50V ELECT FC RADIAL C1, C2, C9, C14
P11213-ND CAP 22UF 25V ELECT FC RADIAL C15
P10269-ND CAP 100UF 25V ELECT FC RADIAL C6, C11, C16
The FCs are "low-ESR" which is a relative term, I suppose.
I am now running a Rev C an i am very impressed. the balance is very good and it is certainly the version to go for.
I have one request , if somebody could help me. i need to roll of the tops a bit more. i have very efficient ribbons doing duty from 900hz to 44khz and i find that the amp is a touch on the bright side, compared to my JLH. any ideas?
i would like to get a bit of the richness of the mids into play
Maybe if i can get the Microcap file from Mauro then i can play around a bit with simulation
thanks in advance.
Rudi
I have one request , if somebody could help me. i need to roll of the tops a bit more. i have very efficient ribbons doing duty from 900hz to 44khz and i find that the amp is a touch on the bright side, compared to my JLH. any ideas?
i would like to get a bit of the richness of the mids into play
Maybe if i can get the Microcap file from Mauro then i can play around a bit with simulation
thanks in advance.
Rudi
> How could someone just wire in a 12V wall wart to mains to power a 60mm computer cpu fan?
> A computer HSF is alot more affordable than a $20.00 heat sink from some guy in Michigan.
That seems like a lot of effort to me. Why not just drop a 7812 across one of the bridges in your existing PSU, and shove the output at your fan? (Something like the quickie below. Usual caveats, especially about the untested and quickly-drawn nature apply).
Don't forget to heat sink it. A wussy heat sink will be plenty. The tab is connected to the middle pin, so either float the heatsink or use a mica pad (same as the MUR860s) to isolate it.
Wes
> A computer HSF is alot more affordable than a $20.00 heat sink from some guy in Michigan.
That seems like a lot of effort to me. Why not just drop a 7812 across one of the bridges in your existing PSU, and shove the output at your fan? (Something like the quickie below. Usual caveats, especially about the untested and quickly-drawn nature apply).
Don't forget to heat sink it. A wussy heat sink will be plenty. The tab is connected to the middle pin, so either float the heatsink or use a mica pad (same as the MUR860s) to isolate it.
Wes
Attachments
Hi Guys,BrianDonegan said:
Just came in from a day on the White river catching more than my fair share of brown and rainbow trout.
Yes, I will provide boards to Brian for kits, so those of you who want the complete kit will only need pay one shipping charge.
I have not yet settled on the price, as I hav not got the estimated numbers and costs completely worked out yet, plus there may be a few last minute modifications to the PCB prior to the production run.
The cost will likely be in the $23-26 range for a pair as there are more square inches of PCB with these monoblocs(well for 2). Also note that these are double sided boards and not single as before, and they will be 2oz copper. all of the above results in about 30% increase in cost over the single sided REV A boards. The big benefit is that there are no jumpers, and the ground scheme is better optimized.
Cheers!
Russ
rudi said:
Hi Rudi,
I have run my Rev C amps driving my Cryolites which have a BG ribbon tweeter, but they are not terribly sensitive. I would not say it sounded bright at all, in fact probably somewhat less bright than my Rev As, though not by much. I would not say either is "harsh" in the high end.
I would be interested in the specifics of your setup, as I wonder if you maybe getting some secondary sybilance.
Cheers!
Russ
Russ White said:
Hi Russ. I do not think it is to bright for the average guy.
i think it is more a matter of taste. what i am trying to get is a bit more of the richness in the mids. for all other applications it is perfect. and you are right the Rev C is much less intrusive than the Rev A. one can listen for a very long time to the Rev Cmy current setup is :
speakers designed by myself. using Morel 5" Mw144 (inseparate inclosures and differant cut offs. Pinsh Ribbon - 900hz - 44lhz bipolar (not well know but is one of the best i have ever heard)
Modified Rotel frontend and an DAC using PCM 63 with a Siemens CCA valve output. no pre just a passive pot (alps)
interconnect - Audioquest
i include a picture of the speakers
Attachments
Very nice Rudi! I will have to see if I can find that tweeter! As I have a few RS180s which I would like to try them with to test again the BGs.rudi said:
Hi Russ. I do not think it is to bright for the average guy.
my current setup is :
speakers designed by myself. using Morel 5" Mw144 (inseparate inclosures and differant cut offs. Pinsh Ribbon - 900hz - 44lhz bipolar (not well know but is one of the best i have ever heard)
Modified Rotel frontend and an DAC using PCM 63 with a Siemens CCA valve output. no pre just a passive pot (alps)
interconnect - Audioquest
i include a picture of the speakers
I like the BGs, but I wonder if you Pinsh are better.
I aggree with you on taste, sometimes you want a little more warm and less bright, but for that I have nothing against a good parametric EQ.
Cheers!
Russ
Hi all:
Various:
- The low cap( 10 to 220pF ) are able be COG multilayer ceramic ( or NP0 ) 200Vdc, high quality and low price.
- The 100nF 100V if is multilayer ceramic is more efficient (PS noise filtering).
- The ver3 (revC) is more sensible to the components on the signal path. They use EL cap and film cap of high quality, that have not given you problems in other circuits. A thing that has struck me is that all they have thought about improves my layout, that have done in 2 hours but it is already good, or the PS, but any has recommended a bypass of ELcap 220uF on NFB with a film cap, does that it may be a lot of more useful ( and demonstrable ) of the PS bypass.
- Rsense( 0.47ohm 7W ): it has to be absolutely not-inductive. All the films & oxide res are inductive, to part of the special models on pure ceramic or glass. They improve it are the wire cement ( Brian and Russ used type ), that they have the low inductance of all the other ( and they cost less ). If they use a different type are able have problems, in particular in the verC.
- Diode Bridge: My choice is a "low cost" compromise, and ( to differance of the other solutions ) I dispose of data of the performances of this configuration ( FFT in all the conditions ).
- Main layout & main ELcap: 10000uF is a "suitable" value for common use of this 40W/8ohm. It is not "forbidden" to use cap more great, but often don't improve the performances (this circuit dont' have very high current capable).
- General PSU: I study Power Amp from about 20 years, and I have make a will nearly all the techniques of the "guru", ( is opamp that SS power amp ). The "low cost" compromises that I propose am not "casual" but "serious". If I have not put LM78xx or 317 are because has not given concrete ameliorations, not because I have not thought !
- I have read of the " critical" on the layout PS.
I have worked a lot of years with precision hi-tech circuits, and not don't need "lessons" on the PS bypass from any.
One's own my professional experience has taught me that are a lot of more critical the super-stabilized circuits of that "floating". To the critics, suggestion to study the rules of base of the PSRR and of the " virtual GND". My circuit uses as " virtual GND" the Main GND, that it has the 50 or 60 Hz ripples, but it is "common mode" to the PS, and all the circuit "floating " together ( without residues on the signal ).
Is not a case that little hi-tech amp are so "silent".
Humbly, advice to learn, some time, instead of pretend to teach it he has many things that know.
- In connection with the "presumed" problems of gnd spKs layout and relay contact, seems me that the experiences of anyone has built the circuit it's "good sound", and I have analysed all the forms of THD-IMD possibles of the complete circuit, without finding differences. Other thing is that the PS up & down is able to create HF oscillations (no good for the TWs), non only "thump". The "Thump" are able even is produced from the preamp or from the "main black out ", and I don't accept risks for my "beloved" spks, over all in the name of "paranoias" it doesn't show. Other technical element, the res. of the relays contact does part of the original circuit, and if don't use the relay, it is has to think about lead in series to the output a res of about 0.01-0.1 ohms, to improve the stability on capacitive load.
- Maintain always isolated the circuit of GND of the amp from the main ground (safety), and the signal GND from the Power GND...
Ciao
Mauro
Various:
- The low cap( 10 to 220pF ) are able be COG multilayer ceramic ( or NP0 ) 200Vdc, high quality and low price.
- The 100nF 100V if is multilayer ceramic is more efficient (PS noise filtering).
- The ver3 (revC) is more sensible to the components on the signal path. They use EL cap and film cap of high quality, that have not given you problems in other circuits. A thing that has struck me is that all they have thought about improves my layout, that have done in 2 hours but it is already good, or the PS, but any has recommended a bypass of ELcap 220uF on NFB with a film cap, does that it may be a lot of more useful ( and demonstrable ) of the PS bypass.
- Rsense( 0.47ohm 7W ): it has to be absolutely not-inductive. All the films & oxide res are inductive, to part of the special models on pure ceramic or glass. They improve it are the wire cement ( Brian and Russ used type ), that they have the low inductance of all the other ( and they cost less ). If they use a different type are able have problems, in particular in the verC.
- Diode Bridge: My choice is a "low cost" compromise, and ( to differance of the other solutions ) I dispose of data of the performances of this configuration ( FFT in all the conditions ).
- Main layout & main ELcap: 10000uF is a "suitable" value for common use of this 40W/8ohm. It is not "forbidden" to use cap more great, but often don't improve the performances (this circuit dont' have very high current capable).
- General PSU: I study Power Amp from about 20 years, and I have make a will nearly all the techniques of the "guru", ( is opamp that SS power amp ). The "low cost" compromises that I propose am not "casual" but "serious". If I have not put LM78xx or 317 are because has not given concrete ameliorations, not because I have not thought !
- I have read of the " critical" on the layout PS.
I have worked a lot of years with precision hi-tech circuits, and not don't need "lessons" on the PS bypass from any.
One's own my professional experience has taught me that are a lot of more critical the super-stabilized circuits of that "floating". To the critics, suggestion to study the rules of base of the PSRR and of the " virtual GND". My circuit uses as " virtual GND" the Main GND, that it has the 50 or 60 Hz ripples, but it is "common mode" to the PS, and all the circuit "floating " together ( without residues on the signal ).
Is not a case that little hi-tech amp are so "silent".
Humbly, advice to learn, some time, instead of pretend to teach it he has many things that know.
- In connection with the "presumed" problems of gnd spKs layout and relay contact, seems me that the experiences of anyone has built the circuit it's "good sound", and I have analysed all the forms of THD-IMD possibles of the complete circuit, without finding differences. Other thing is that the PS up & down is able to create HF oscillations (no good for the TWs), non only "thump". The "Thump" are able even is produced from the preamp or from the "main black out ", and I don't accept risks for my "beloved" spks, over all in the name of "paranoias" it doesn't show. Other technical element, the res. of the relays contact does part of the original circuit, and if don't use the relay, it is has to think about lead in series to the output a res of about 0.01-0.1 ohms, to improve the stability on capacitive load.
- Maintain always isolated the circuit of GND of the amp from the main ground (safety), and the signal GND from the Power GND...
Ciao
Mauro
mauropenasa said:
Mauro, thanks for your input!
It is very much appreciated.
I will try bypassing the 220uf cap in the NFB loop with 100nf and report my results. I think it would be good to add a spot for this in the production monobloc PCBs, and I will do so.
I was also wondering if the 220uf cap would benefit being a Low ESR part like a pnansonic FC type cap.As far as PS and the shunt regulator I would never argue with a master. I am very much a beginner compared to you Mauro, but I wanted to experiment with different configurations. Just to satisfy my incurable curiosity. I have now experimented with a 317/337 prototype and I can verify that there is no noticible change.
I totally agree on your point about isolating the grounds, mains GND and power GND should never meet.
I also agree with your rightful disdain for the relay paranoia. One simply needs to try it to see there is no basis in fact for the superstitions and fear.
Also, I was quite skepticle(as noobs often are) at first of the PS arrangement as it is far different from what is recommended by a few very vocal others on the board. I was quite wrong, but I also quickly stated so in my first build of the REV A(actually the MY_INV prior). This is a very quiet amp it is the quietest I have ever measured including my Aleph30SE and my Krell, and none of my GCs previous to the REV A even come close. So there is something to this approach. People should build and test before the criticise. One reason this amp is so devoid of noise has to be the well thought out PS.
I also agree on cap selection. For the 100V 100uf I will change the pads on the monoblocs to better accomodate ceramics, but I can report that Mylar and MKT types worked fine here without any detriment.
For the other small values I did exactly as mauro recommends (mutilayer ceramics) which had great succss. I have also tried silver micas which worked the same, but cost much more.
I don't think PSSR could be tweaked much more.
It is already the best I have seen in an amp so simple, well actually the best I have seen so far. Also I can attest to the fact that omitting the relay(SP circuit) is not a good idea and has doubtful if not non-existant sonic benefit. It is also well known that the failure mode of these NS chips can in some cases result in rail voltage to the output. I ,as Mauro stated, value the speakers I worked hard to build too much to risk that. Also, keep in mind that the current in the circuit when the relay switches is typically single milliamps(mine about 6-7), so switching life in cycles for the relay is extremely high.
Thanks again Maruo, please do not take some of my amaturish questions as criticisms of your circuit, but as simple questions from one who wishes to learn as much as possible from you and others here.
Cheers!
Russ
The micas that Carlos mentioned are now in the BOM, and are only $.55USD each, so I feel comfortable with them. I could change to ceramics, but I don;t think the savings will be much.
For the 100nF, I could switch to ceramics easily enough. They are so cheap, I could actually include both film and ceramic. Perhaps I will include both for the beta kits, and get feedback (no pun intended), but I agree that ceramics would probably be fine.
Actually, I was looking through my Parts Express catalog the other day, in particular, the "electronic parts" section, where they have a bunch of really cheap Xicon caps and NTE resistors, etc. I was thinking of getting a bunch of parts from them and building a RevC channel to compare to the BETA kit, just to see if I could hear a difference. Reality checks are always good.
For the 100nF, I could switch to ceramics easily enough. They are so cheap, I could actually include both film and ceramic. Perhaps I will include both for the beta kits, and get feedback (no pun intended), but I agree that ceramics would probably be fine.
Actually, I was looking through my Parts Express catalog the other day, in particular, the "electronic parts" section, where they have a bunch of really cheap Xicon caps and NTE resistors, etc. I was thinking of getting a bunch of parts from them and building a RevC channel to compare to the BETA kit, just to see if I could hear a difference. Reality checks are always good.
Hi Mauro
I agree the speaker portection is required. i build mine without it just to find out how it sounds, my next version will certainly have it, and the bridge diode arrangement is purely due to what we can get hold of here in south africa.
we will all agree that you have very good reason for doing what you have done and all of us can just speculate about best practices.
I myself know nothing about electronics and are led by people like you. I am sure if you build a house you will come to a architect like myself.
at the end we all love the hobby, but we can't eliminate the Wisecracks around us.
Best regards and thanks for you informed advise
Rudi
I agree the speaker portection is required. i build mine without it just to find out how it sounds, my next version will certainly have it, and the bridge diode arrangement is purely due to what we can get hold of here in south africa.
we will all agree that you have very good reason for doing what you have done and all of us can just speculate about best practices.
I myself know nothing about electronics and are led by people like you. I am sure if you build a house you will come to a architect like myself.
at the end we all love the hobby, but we can't eliminate the Wisecracks around us.
Best regards and thanks for you informed advise
Rudi
Russ White said:
I like the BGs, but I wonder if you Pinsh are better.
Hi Russ
the tweeter is far better than the BG. first of all it acts a midrange/tweeter and super tweeter. = no cross over. being bipolar it gives a spacial sense that is hardly equaled. some designers put a tweeter at the back of the speaker to simulate this.
the not so nice side of it is that it costs $ 2300 for a pair.
But you would Love them once you have them.
Rudi