100W Caddock is overkill IMO. And you are paying $8 for additional 1.5W.
2W resistor in that location is still OK in most applications.
2W resistor in that location is still OK in most applications.
Peter Daniel said:
Here is a link to the catalog page that has all of the prices listed for the Caddock products carried by Mouser.com:
http://www.mouser.com/index.cfm?handler=fra_pdfset&pdffile=232
--
Brian
So, the MP930, rated at 2.25W without heatsinks would be alright for this?
Prices for the 0.2 ohm 930:
1 $ 4.150
10 $ 3.320
25 $ 3.070
50 $ 2.910
100 $ 2.740
200 $ 2.570
500 $ 2.490
--
Brian
Prices for the 0.2 ohm 930:
1 $ 4.150
10 $ 3.320
25 $ 3.070
50 $ 2.910
100 $ 2.740
200 $ 2.570
500 $ 2.490
--
Brian
Strange... I had to change my timezone to -3:30 time zone (Newfoundland) to get my time to display properly. It was an hour and a half early for me before I made this change.
Is the board having problems with times?
--
Brian
Is the board having problems with times?
--
Brian
Brian,
If that was because my comment I think it was outside of time zone issues. I read your post as 100 W being overkill, replied and when my post was saved I saw you had added the free air dissipation ratings. That was my point about the 100 W.
As for the 2.25 W being enough I think that 5 W probably would be desired as the bias, if I am remembering correctly from rtirions schematic, is in the order of 2.2 A (and I can't remember the exact value) as he said each output FET dissipates 27 W each at 12 V. With the source resistor at 0R22 that comes to some 2.2*2.2*0.22 = 1.06 W which is not unreasonable but will definitely heat it up. On top of that I suppose the comes the AC current? MP9100 is not unreasonable but for the money I would probably attach a small aluminium strip or sink to the resistors and use MP930. I myself is using the Panasonics for everything with a power prefix.
/UrSv
If that was because my comment I think it was outside of time zone issues. I read your post as 100 W being overkill, replied and when my post was saved I saw you had added the free air dissipation ratings. That was my point about the 100 W.
As for the 2.25 W being enough I think that 5 W probably would be desired as the bias, if I am remembering correctly from rtirions schematic, is in the order of 2.2 A (and I can't remember the exact value) as he said each output FET dissipates 27 W each at 12 V. With the source resistor at 0R22 that comes to some 2.2*2.2*0.22 = 1.06 W which is not unreasonable but will definitely heat it up. On top of that I suppose the comes the AC current? MP9100 is not unreasonable but for the money I would probably attach a small aluminium strip or sink to the resistors and use MP930. I myself is using the Panasonics for everything with a power prefix.
/UrSv
Ralf,
in my opinion the 30watt output is about the upper bound for the concept, it's not very practical, too many amps to go around, too many watts to dissipate and you still need capactor banks before and after to really store some energy.
I think that in this case you are better off having some wirewound enamel R's or some thermistors take the heat. Granted you have less control over the voltage but who cares you have the X config. at work for you.
in my opinion the 30watt output is about the upper bound for the concept, it's not very practical, too many amps to go around, too many watts to dissipate and you still need capactor banks before and after to really store some energy.
I think that in this case you are better off having some wirewound enamel R's or some thermistors take the heat. Granted you have less control over the voltage but who cares you have the X config. at work for you.
UrSv said:
Me too.😉
Probably the best alternative to Panasonics would be Mills 5W wirewounds, which spacing is compatible with Panasonic resistors anyway. I mentioned Caddocks just in case someone had them.
Hi
Has anybody tried Isabellenhütten thick-film resistors, I think the PBH type looks very promising. But the price might be a little high.
Spec and prices
Claus
Has anybody tried Isabellenhütten thick-film resistors, I think the PBH type looks very promising. But the price might be a little high.
Spec and prices
Claus
The Isaplans are metal foil types and indeed very good. Better prices here: http://www.buerklin.com/gruppen/KapE/E073910.asp
I am going to use them on my aleph 5, see: http://www.diyaudio.com/forums/showthread.php?threadid=7233
Original manufacturer's data: http://www.isabellenhuette.de/inhalt/produkte/prodci2.htm
I am going to use them on my aleph 5, see: http://www.diyaudio.com/forums/showthread.php?threadid=7233
Original manufacturer's data: http://www.isabellenhuette.de/inhalt/produkte/prodci2.htm
Try to buy ISAPLAN directly by Isabellenhütte!!!
They made min. order 100EURO.
Is cheaper then bürklin.🙂
Andreas
They made min. order 100EURO.
Is cheaper then bürklin.🙂
Andreas
Thanks Holger and Andreas.
I ones bought a few Isaplan PBH at Conrad, but they don´t sell them any more, or at least they are no longer in there catalog.
I have newer used them, they where for another project, that has been looking at me for years.
How are they compared to other high power resistors? Film resistors are often told to be analytic and bright, are that true for the Isaplan? I have no doubt that they are very high quality resistors but do they sound as good as they look.
Claus
I ones bought a few Isaplan PBH at Conrad, but they don´t sell them any more, or at least they are no longer in there catalog.
I have newer used them, they where for another project, that has been looking at me for years.
How are they compared to other high power resistors? Film resistors are often told to be analytic and bright, are that true for the Isaplan? I have no doubt that they are very high quality resistors but do they sound as good as they look.
Claus
revised layout
Does this look better for component placement? I checked it over and I believe that it follows grey's original schematic. What significant changes should I make to this, besides the revised CCS. I still need to add more ground traces, mounting holes and finish the other side.
After I finish with this board, I can revise it with the new CCS and other features and get another set made.
For the other side, I am planning on just using terminal blocks for the output devices, so I am planning on keeping the order reversed, so that I can make it perfectly symmetical.
Any comments appreciated.
--
Brian
Does this look better for component placement? I checked it over and I believe that it follows grey's original schematic. What significant changes should I make to this, besides the revised CCS. I still need to add more ground traces, mounting holes and finish the other side.
After I finish with this board, I can revise it with the new CCS and other features and get another set made.
For the other side, I am planning on just using terminal blocks for the output devices, so I am planning on keeping the order reversed, so that I can make it perfectly symmetical.
Any comments appreciated.
--
Brian
Attachments
Brian
I don't see why r5s an 6s can't be moved up by 1/4in. Also try to have provisions for vertical multiturn trimpots (see my layout).
I don't see why r5s an 6s can't be moved up by 1/4in. Also try to have provisions for vertical multiturn trimpots (see my layout).
grataku said:
Thanks for the suggestions. Should I just change over to multiturn pots, or should I leave spacings for both?
R5s and R6s can't be moved up, since they straddle some traces. What would be the benefit of moving these components?
Thanks,
--
Brian
Single-turn pots are more reliable and robust, and have lesser tendancy to move from the point of adjustment when subjected to vibration or shock, but are more difficult to adjust precisely than muti-turn pots. It's a trade-off. I recommend multi-hole footprints to allow for all of the three common pin spacings. this also gives a little more flexibility when purchasing components.
Re: revised layout
So if someone want's to mount the FETS directly to the board, they would be outside the board on one side and under the board on the other side, right?
At least keep the same spacing for the terminal blocks as for FETs. I wouldn't like to use wires, and if I would be forced to use them I'd rather solder the wires directly to the PCB. This is not something you'll be dissasembling once a week for clean up😉 and having terminal blocks do nothing but degrades the performance.
One other remark in respect to symmetry. Although your board may have perfect symmetry, when you use wires to attach output devices, symetry might be lost.😉
BrianGT said:
So if someone want's to mount the FETS directly to the board, they would be outside the board on one side and under the board on the other side, right?
At least keep the same spacing for the terminal blocks as for FETs. I wouldn't like to use wires, and if I would be forced to use them I'd rather solder the wires directly to the PCB. This is not something you'll be dissasembling once a week for clean up😉 and having terminal blocks do nothing but degrades the performance.
One other remark in respect to symmetry. Although your board may have perfect symmetry, when you use wires to attach output devices, symetry might be lost.😉
Re: Re: revised layout
Good points. I am considering just keeping the devices mounted on the board, then mounting them directly to a heatsink. I just need to find two large heatsinks for each channel. I am aiming for 100W into 4 ohms, but would settle for less. What is the rationale for calculating heat dissipation for the AX?
The boards are 6" x 6" now. I could make them smaller if I wanted to put components closer together. I like having them a little farther apart and having large ground planes.
For mounting to heatsinks, could I just mount the boards perpendicular to the heatsink, then have one side of devices point down and the other up? This would set the internal chassis with of my amp to about 6", but I can deal with that.
I received some samples in the mail today: some AD-797s, some microcontrollers from microchip.com and some Hyperfast-Soft Recovery(Stealth™ Family) diodes from fairchild semiconductor.
Peter, I can send you some of those diodes if you want some.
Datasheet: http://www.fairchildsemi.com/ds/IS/ISL9R3060G2.pdf
--
Brian
Peter Daniel said:
Good points. I am considering just keeping the devices mounted on the board, then mounting them directly to a heatsink. I just need to find two large heatsinks for each channel. I am aiming for 100W into 4 ohms, but would settle for less. What is the rationale for calculating heat dissipation for the AX?
The boards are 6" x 6" now. I could make them smaller if I wanted to put components closer together. I like having them a little farther apart and having large ground planes.
For mounting to heatsinks, could I just mount the boards perpendicular to the heatsink, then have one side of devices point down and the other up? This would set the internal chassis with of my amp to about 6", but I can deal with that.
I received some samples in the mail today: some AD-797s, some microcontrollers from microchip.com and some Hyperfast-Soft Recovery(Stealth™ Family) diodes from fairchild semiconductor.
Peter, I can send you some of those diodes if you want some.
Datasheet: http://www.fairchildsemi.com/ds/IS/ISL9R3060G2.pdf
--
Brian
Re: Re: Re: revised layout
This could do it, but remember that the width of a chassis is mostly defind by the diameter of a transformer. In case of my 750W Plitron for A75 it was just 3/8" more than 6". Also, proposed by you method, while doable, still presents some difficulties in mounting the FETs. The other alternative is either using some 1/4" angles to mount devices to heat sinks or slab of aluminum at least 1/2" thick. You mount the board with devices to it and then with the screws from the outside attach this module to the sinks. But having width of the inside 6" wide asks for a big chunk of metal. For my amp I'm planning the inside to be 3" wide and preferably the slab (which connects both heatsinks and is the mount for FETS and board) would be made out of copper. I check the pricing, 2 pieces 11 x 0.75 x 3" of copper would be $140.
As to the diodes I would be glad to receive some, but you know, it's either 4 or 8. If it's too many, just e-mail me the contact details and I could ask for samples myself.😉
BrianGT said:
This could do it, but remember that the width of a chassis is mostly defind by the diameter of a transformer. In case of my 750W Plitron for A75 it was just 3/8" more than 6". Also, proposed by you method, while doable, still presents some difficulties in mounting the FETs. The other alternative is either using some 1/4" angles to mount devices to heat sinks or slab of aluminum at least 1/2" thick. You mount the board with devices to it and then with the screws from the outside attach this module to the sinks. But having width of the inside 6" wide asks for a big chunk of metal. For my amp I'm planning the inside to be 3" wide and preferably the slab (which connects both heatsinks and is the mount for FETS and board) would be made out of copper. I check the pricing, 2 pieces 11 x 0.75 x 3" of copper would be $140.
As to the diodes I would be glad to receive some, but you know, it's either 4 or 8. If it's too many, just e-mail me the contact details and I could ask for samples myself.😉