Yes. (although fairly minor) But a very simple fix - place the CRC filter in the same chassis as the transformer. Then a last set of filter caps in the chassis with the audio circuit.
Regulators should be close to the load. Having the reg in the PSU box isn't the best situation.
Screwed again!!! No way could I fit a (Salas) reg board and a BA-3 board in a 2U 288mm chassis, along with the rest of the parts.
So I guess it's back to the CRC psu in the supply chassis and at least 470uF of E-caps per BA-3 board.
BUT check this out http://diy.beyarn.com/bigdata/BIBguideRev2.pdf “There are advantages of very low and relatively even and extended output impedancewhen using such a PSU…….This PCB works on remote sensing 4 wire 'Kelvin'' outputs. Remote sensing is good for disregarding the wiring resistance issue to correctly measuring the load demands by the regulator's error amplifier and effectively brings the load in touch with the regulator no matter the distance. Keeps the impedance low and gives you freedom of proximity to the circuits and arrangement. The wiring can be manageable gauge that way. Should be twisted pairs which you can twist yourself.”
So he claims that impedance of this regulator output is low enough to where umbilical cables might be as long as 4 feet or so?
Last edited:
Remote sensing is a well-known technique that can work very well. The only thing to watch out for is instability due to the inductance/capacitance of the (long) sensing lines.
The diyaudio superreg has specific parts on the PCB to avoid instability with remote sensing.
Jan
The diyaudio superreg has specific parts on the PCB to avoid instability with remote sensing.
Jan
So using this Super Regulator V2.2 - Power Supplies and Accessories - Circuit Boards for each BA-3 will allow umbilicals as long as 4.5 feet without any sonic penalty? If yes, any other recommended precautions?
Why use 4.5 ft umbilicals? Remote sensing is meant to compensate for the fact that the reg is not immediately next to the load. But I have never tested it by putting the reg on the other side of the room.
Jan
Jan
Sorry, I thought that you may have been following along since my post #1914. I suffer from severe space limitations and will have to put at least the transformer and bridge rectifiers in a separate chassis, apart from the two BA-3 boards. But your reg boards might work for me: the BA-3 board is 140 mm L x 70mm W , one + & - pair of your reg boards are 75mm W. And the interior width of this chassis for each BA-3 board is 210mmW. Galaxy 288 (3mm Front) - Compact with Quasi Heatsinks - Chassis
So, given my unhappy space limitations, will your boards have a problem being 4 feet or so from the supply's 4700uF E-Caps?
For where the BA-3 boards have to go, max cubic space is Depth: 13.5” (342.9mm ), Height: 6.25” (158.75mm) { = Total height for TWO “stacked” enclosures }, Length: 9.89” (251.21mm). Stacking the two 288mm chassis (each containing a BA-3 board, attenuators, RCA jacks, et al) is exactly my plan. But no way is there available footprint for a chassis housing an entire psu and BA-3 board (even if two of those chassis were stacked vertically)-much less for all reg and BA-3 boards, transformer and all other parts to be in one big chassis.
So if the reg boards have to be close to the CRC filters then there's no advantage with putting a + & - pair of them in with a BA-3 board. I might as well just forget the regs and simply add at least 470uF across each of the rail pads on each BA-3 board.
PSU + one BA3 in one chassis = left channel mono block
PSU + one BA3 in one chassis = right channel mono block.
PSU + one BA3 in one chassis = right channel mono block.
No that would be no problem. But it would seem to me that placing a box 4.5 ft away will only aggravate the problem with umbilicals all over the place.
Why not the power box underneath the amp box, for a 1 ft cord, or is that also a problem?
Jan
No, I need to use each BA-3 FE (single-ended) with my two F4 amps to bi-amp my speakers. But see my last post; I just don't have enough room for a chassis to house one (BA-3 with a complete psu and all pots and jacks) X 2. Even if they were stacked vertically. That's why part of the supply needs to be in a separate chassis.
I meant stack the two BA-3 BOARDS (not enclosures) on top of each other in the same enclosure. Starting to sound like Andrews Idea is most plausible, but store is out of 288.....I was thinking I used that box, but I had the 388.
Russellc
Last edited:
As I explained to Russellc, I don't have enough vertical rack space to stack a BA-3 box and its psu box under it. I can only stack two BA-3 boxes-and that has to be inside of a 15” h x 17” d TV cabinet (LOL!), with the psu box (one toroid transformer for two bridges and two + & - CRC filters to run both BA-3 boards) sitting outside the TV cabinet on the carpet. So to reach the rear panels of those BA-3 boxes, the cables really can't be much shorter than 3 feet.
Also, to get to there, those cables would have to pass in front a floor standing multi driver speaker with a 15" woofer, where the flux field measures on my (pinned) 2 gauss scale magnetometer http://www.caltechindia.com/images/gauss.jpg would otherwise read at least between 4 and 10 gauss. Yes, a real DIY nightmare.
I'm not sure if there's enough interior space to stack two stuffed BA-3 boards, and still have room for 4 mono Goldpoint pots, RCA jacks, et al, into one 288mm box. What's the total height of your stuffed BA-3 board? Recall, I cannot exceed the above mentioned dimensions-because I have to put either one box with two BA-3s or two stacked BA-3 boxes inside a TV cabinet (LOL!, I know).
BUT the BA-3 boards are 70mm wide x 140mm long; that Galaxy 288mm box has 210mm interior width. And those boards are only 5.5 inches long. So there's
enough space to put them side-by-side and install the rest of the parts-but not any reg boards. Instead, scratch the reg boards, and maybe enough space for those >470uF E-caps, like Zen-Mod said I'd need to use.
Or, in another chassis 288mm (no bigger), I could use the reg boards and probably keep the cable length to between 12 and 18" long-IF I can fit both pairs of + & - reg boards, both + & - rails of ~ 4700uF CRC filters, both bridges and the ~95mm diameter, 100 VA (shielded!) Sum-R transformer inside of one 288mm chassis.
But then what's the advantage of going with a pair of reg boards for each BA-3 board, versus a pair or CRC filters for each rail (plus the 470uF caps in the BA-3 chassis)? Nelson said that you want IDss to be between 6 and 12mA. Presumably, higher biasing would give better sound. I don't know very much; someone here agreed to do all of the calculations and building for me. But at that IDss and with a 100VA transformer, can supply voltage from the 5000uF CRC filters sag enough to cause problems?
Last edited:
Let me do some measurements. Much will depend on the spacers you use and how much clearance is available. With the BA-3 boards, the thickest part will be the 4 caps, and or the coupling caps used. Mine are kind of big, but many use a silmic and a smaller film bypass. As to the 4 caps on the board, I ordered several with the same dimension for the "leg hole", and a couple types were substantially shorter...all were 1000 uF 25 volt as called for.
Russellc
OK, my board is mounted on standoffs that as near as I can measure are just over 1/2 inch. This could be shortened to 1/4 inch, as long as nothing shorts to case floor. Measuring from case floor to top of film cap is right at 1 1/2 inch. Bear in mind, I have the same Audyn cap 6l6 used and it is a fat one. Using a different cap, like a small silmic layed down in the electrolytic hole combined with a much thinner film cap, then combined with a shorter standoff, would reduce the height of the package.
As is, it would take an inch of space MINIMUM for the top board to clear the bottom parts, then another 7/8s to 1 inch for the top boards parts to clear the cover of the case...so 1 1/2 for bottom board, 1/8 inch clearance between top of film cap and top board...1 5/8s, plus another inch for the components of the top board, another 1/8 inch clearance from the film to case top cover, total of 2 5/8 inch....including board thickness likely a smidge under 3 inches....the pre discussed "flatter" components and bottom board stand off should keep it shorter yet...giving an extra 1/16 to 1/8 inch here or there, I think it could easily, with careful construction be under 3 inches.
Russellc
As is, it would take an inch of space MINIMUM for the top board to clear the bottom parts, then another 7/8s to 1 inch for the top boards parts to clear the cover of the case...so 1 1/2 for bottom board, 1/8 inch clearance between top of film cap and top board...1 5/8s, plus another inch for the components of the top board, another 1/8 inch clearance from the film to case top cover, total of 2 5/8 inch....including board thickness likely a smidge under 3 inches....the pre discussed "flatter" components and bottom board stand off should keep it shorter yet...giving an extra 1/16 to 1/8 inch here or there, I think it could easily, with careful construction be under 3 inches.
Russellc
Last edited:
On re measure, the Audyn cap is slightly smaller than 1 inch in diameter...no big deal, you can find smaller parts.
Russellc
Russellc
Thanks, Russellc. I'll be using the same polypropylene PWA series ClariityCaps or Q4 Audyn for the output caps and the Elna silmic at 25V. Would 16mm diam for the 35 or 50V overshoot their pc board pads? (heights are .5 and 1cm; minimizes R and DA). Don't want to crowd out any parts, of course-or make the boards impractically tall.
http://www.elna.co.jp/en/capacitor/pdf/catalog_15_16_e.pdf ; page 169.