Nelson doesn't in the factory Firstwatt amps. Nor are the big electrolytic caps bypassed in the Pass Labs amps.
But if you want to try it, it's not going to hurt anything. Just don't expect there to be any difference.
A PSU connected by umbilical isn't scary. (Although the heat generated by 2 BA-3 in the same chassis is, in fact, somewhat scary if you haven't done your calculations for required heatsink....
)Inductance from a few feet of straight wire? Completely ignorable.
I'm in that camp as well. There's no need.
If you want to make a difference, add 15,000uF caps near the rails in the Amp chassis. Or more. 0.1uF won't do anything on a power amp.
What's really scary for me is that because I want to put two BA-3 boards in one chassis the psu needs to go in another chassis.
I have built the BA-3 preamp using the super-regs (really great stuff) boards which are designed with a sensing line which runs separately to the bA boards. The voltage sensing if you go with this configuration works really well at providing regulation at the preamp, worth checking out.
I have built the BA-3 preamp using the super-regs (really great stuff) boards which are designed with a sensing line which runs separately to the bA boards. The voltage sensing if you go with this configuration works really well at providing regulation at the preamp, worth checking out.
I've bypassed the power caps on my F5 with 10uf poly caps. I have 6x33,000uf caps in the psu, each one bypassed with a 10uf poly cap. Actually, the last caps on the rail are bypassed with a 10uf poly, a 4.7uf poly and a .1 polystyrene cap.
I thought the sound improved with the bypass caps. No joke...it got smoother and more fluid. I don't have measurements to support this, just my ears.
I thought the sound improved with the bypass caps. No joke...it got smoother and more fluid. I don't have measurements to support this, just my ears.
Attachments
Last edited:
Nice to see the BA-3 still chugging along.
Bypassing the electrolytics is a non-issue. There are some circuit dependencies,
but problems are rare. If the ps is in another chassis, you will very likely
want local capacitance in the gain stage chassis, even if it's only electrolytics.
As DIYers, you are in the privileged position of being able to try it both ways
and see what you get.
Bypassing the electrolytics is a non-issue. There are some circuit dependencies,
but problems are rare. If the ps is in another chassis, you will very likely
want local capacitance in the gain stage chassis, even if it's only electrolytics.
As DIYers, you are in the privileged position of being able to try it both ways
and see what you get.
Nice to see the BA-3 still chugging along.
Bypassing the electrolytics is a non-issue. There are some circuit dependencies, but problems are rare. If the ps is in another chassis, you will very likelywant local capacitance in the gain stage chassis, even if it's only electrolytics.
To compensate for the 3 foot cable inductance? And the film caps installed extremely close to-or better still-on the BA-3 boards? But if, say, 3 film caps per rail would be best, what value (s)? 100nF, 10nF and 1nF?
Last edited:
Not because of the inductance... Because the the wire's resistance it turns into another R in the RC filter.
As this is a power amp, you need real caps in the amp chassis, 15,000uF electrolytics or more would be suitable.
If you would like to try a film bypass, across that cap would make the most sense.
As this is a power amp, you need real caps in the amp chassis, 15,000uF electrolytics or more would be suitable.
If you would like to try a film bypass, across that cap would make the most sense.
So you're saying to install all of the E-caps in the BA-3 chassis?Aha! So make the electrolytics smaller. 3300uF. Easy.
Any risk of induced noise by feeding pulsating DC from bridge rectifier over the 3 ft cable to the BA-3 chassis?If the ps is in another chassis, you will very likely want local capacitance in the gain stage chassis, even if it's only electrolytics.
If yes, then keep the E-caps in the supply chassis and install 100nF, 10nF and 1nF film caps per rail on the component side of each of the BA-3 boards?
A PSU connected by umbilical isn't scary. (Although the heat generated by 2 BA-3 in the same chassis is, in fact, somewhat scary if you haven't done your calculations for required heatsink....
If you want to make a difference, add 15,000uF caps near the rails in the Amp chassis. Or more. 0.1uF won't do anything on a power amp.
No power amp; a pair of BA-3 boards (for bi-amping my two F4 amps) and a supply to drive them. As for ventilation, with the (matched) JFETs, Nelson said “……you want Idss around 6 to 12 mA.” Presumably, higher bias would sound somewhat better, albeit with more MOSFET heatsinking. But with this Idss,would two of the BA-3 boards biased like that still pump too much heat into that Galaxy chassis 210 square mm interior? Worst case, I guess I could leave the top cover off; or swap it out with a screen door mesh cover.
However, like Nelson cautioned, with a separate chassis, I’d then have to add bypass caps on each rail, as close to (soldered directly on) each BA-3 board as possible, in that BA-3 separate chassis.Nice to see the BA-3 still chugging along. Bypassing the electrolytics is a non-issue. There are some circuit dependencies,
but problems are rare. If the ps is in another chassis, you will very likely want local capacitance in the gain stage chassis, even if it's only electrolytics.
And Hikari likes the sound (and Walt Jung likes the distortion measurements) of using polypropylene caps for that. I’ll try for metalized; they’re physically smaller than the regular versions. Survey says any film cap performs way better than any E-cap. Do I use 100nF, 10nF and 1nF? One of each per rail?
As for that sketch, I was half through the it until I realized there’s no way that a set of 2000uF or 3300uF CRC e-caps per rail, two BA-3 boards, two mono GoldPoint attenuators per board, input selector switch, RCA jacks, DC input connector and LED power lamp could possibly fit into one of these Galaxy 288 (3mm Front) - Compact with Quasi Heatsinks - ChassisHang on a second... Before trying to answer your questions could you please draw/sketch (literally, and then take photo of it, and post here) a diagram of what you intend. Think that will greatly assist all parties involved.
Sure, I could fit everything into one 4U or maybe 3U chassis, but my unbalanced I/O cables would have to be WAY to long to reach the outputs of Nelson’s B4 crossover. Or if I stacked the B4 atop this chassis, again too much unbalanced cable from the B4 to my two F4 amps. I’m so insanely limited with space in my room; I just don’t have the room for any hardware consolidating rack. That’s the main reason why I need to use a separate supply chassis. Two rails of CRC E-caps won't fit into the BA-3 chassis; they’ll have to go into the supply chassis.
Maybe if I use one of these
http://www.diyaudio.com/store/ampli...asi/1gxa288-medium-electronics-enclosure.html per BA-3 FE board. Then I might be able to fit the E-Caps and the rest of the parts. And from the psu chassis run an umbilical cable to each BA-3 chassis. But as each board is 70mm x 140mm, it still
won't leave a whole lot of space.
Last edited:
It looks like the best bet is to use two of those chassis for each BA-3 board and its hardware-including the pair of 2000uF or 3300uF CRC filters for each rail. Thus, the PSU chassis will have two output cables to feed each BA-3 chassis.
However, will there be any EMI and/or reactance issues with sending pulsating DC from the two bridge rectifiers over two 3 foot cables to the CRC filters in the two BA-3 chassis? Any special precautions?
However, will there be any EMI and/or reactance issues with sending pulsating DC from the two bridge rectifiers over two 3 foot cables to the CRC filters in the two BA-3 chassis? Any special precautions?
Last edited:
I wouldn't worry to much about the heat. Look at the heatsinks on Nelson's BA-3 FE, not a lot of heat being put out by the mosfets.
Put as much capacitance as you feel like (probably 4,700uf or so is all you need, how much you WANT is a different story) run the umbilicals to each channel and put as much cap as will fit in there (probably 470 uf minimum) and you will be fine. I wouldn't worry too much about bypassing...it's not a major concern and it's not going to be a game changer. Of course all this could be bettered by the use of a regulator in front of each BA-3 board (preferably in the same case as the BA-3 boards, closer is better) but that's a whole different discussion...
Would you suggest a regulator like this for each BA-3
http://diy.beyarn.com/bigdata/BIBguideRev2.pdf ? However, a + & - pair of those reg boards and a BA-3 board won't fit into a http://www.diyaudio.com/store/ampli...asi/1gxa288-medium-electronics-enclosure.html box. But if I put the reg boards for both BA-3 boards into the supply chassis and run the umbilicals to each BA-3 box, will I still need to add E-caps and/or film caps to the BA-3 boards? And how long can I make the umbilicals? Could they be as long as 5 or 6 feet without consequence, IF it were really necessary?
Last edited: