[mgalusha]

What do you get when you combine one dead Carver amp, 2 large Signal power transformers, 1 small Tamura transformer, some fast recovery diodes and a slew of caps? Felicia Squeezemore... (aplogies to JoshK and Austin Powers)Squeezes more sound out of my Squeezebox 2 that is.  

A friend recently gave me a couple of old and dead Carver TFM-22 power amps. I thought about fixing them but I would never use them. A while back I bought three sets of the 175VA Signal transformers to build a Felicia balanced power supply but that was about the time I started using the SB2, which removed the need, or at least most of the need for Felicia as I wasn't using my CDP or Preamp any longer.

So having a few days off for Thanksgiving I decided to build a somewhat over the top SB2 power supply with a Felicia setup on board. Plenty of room inside the chassis for everything, so why not.

When I purchased the parts to build a Felicia supply I bought the AuriCaps as that was prior to the Jantzen coming into favor. I also had some of the big oil caps from Hosfelt but since those had since been replaced by the Jantzen as well I used a pair of 10uF metalized polypropylene bypassed with the AuriCaps.

For the DC board, I used 4 of the IRF HEXFred diodes into four 4,700uF Panasonic FC caps bypassed by another of the 10uF Dayton's. These feed a Linear Technologies 1086-5 regulator. This is followed by a 10,000uF Panasonic FC bypassed with one of the 10uF Dayton's and a .1uF Sonicap.

The output is a standard XLR. Most of the wire is the copper from inside a Volex cord except for the wire to/from the small transformer as I figured 14GA was just a bit excessive. I used some 16GA teflon insulated stuff I had on hand. Still way bigger than is needed but I had it.

I used bare copper for the bus on the DC board and for the output leads along with a little teflon spaghetti tubing. Is the teflon really better than plain old spaghetti? I dunno but it was on hand as I bought it for the heat tolerance for tube amp project.

I'm quite happy with the end result. The SB2 seems to have moved up to another level of refinement. Perhaps it's overkill but like most everything else in the audio chain, the better the power supply, the better the sound.

Mike

[JoshK]

Very cool!  I was thinking of doing a similar thing for my SB.  I rescued some old chassis and was going to put a nice wood face on them and stuff them with a similar gizmo.  I was thinking of maybe adding a battery option in there with a switch.  Who knows?  Where did you get the step down tranny?

P.S. I squeeze my Felicia plenty, so no offense.  

[Scott F.]

Very cool  

I just tossed an old Spectro Acoustics amp in the trash. Maybe I'll go liberate the transformer and have a go at the Uber-Supply myself

[mgalusha]

Very cool!  I was thinking of doing a similar thing for my SB.  I rescued some old chassis and was going to put a nice wood face on them and stuff them with a similar gizmo.  I was thinking of maybe adding a battery option in there with a switch.  Who knows?  Where did you get the step down tranny?

I helped Wayne at Bolder Cables with his PC and took the tranny in trade. It's the same one he's using in his SB power supplies. A Tamura PL30-16-130B, 8V secondaries @ 3.8A. Digikey carries them.

I thought about adding an SLA as well and I did add a jack and switch to the power supply I got from Wayne but after listening for a while to mine with the battery and Jerry's with the battery I decided I like the sound better without it. The sound seemed slower/heavier (best I can describe it) when running on battery. It has a blacker background and at first I liked it better but after a while I found I preferred the sound via the linear. Having the option of both is nice though and if you have the room, why not.

P.S. I squeeze my Felicia plenty, so no offense.  

Good man, keep it up!

Mike

[JoshK]

Oddly enough, what you say about the SLA versus AC powered I find somewhat true at this point with my UcD comparisons.  As you may know, I have built up a stereo pair (dual mono) using standard AC PSU and then also with the SLA powering caps in a similar PSU.  

I find that at this point the SLA powered version sounds darker, heavier and maybe slower.  Not sure if those are the right adjectives, but they sound about right.  I still have a couple of tweaks to take my SLA version up to current AC status, with the previous SLAs trumping AC.  Maybe the snubbers helped a lot, don't know.  But currently my AC version is less contained the speakers and seems to project beyond much better.

You might try out a snubber too.  Typical snubbers are in the 220nf - 470nf cap followed by a 1R resistor (5w for amp typically).  They have a certain je ne sais pas (use of a recently learned pretentios phrase, like that?).

[Occam]

Mike,

Very very cool! I've been waiting for someone to smack Felicia on a SB2/3 linear, regulated supply.

If you're feeling masochistic, I'll send you a retro 2PDT center off toggle switch so you can bypass Felicia on the fly (really, the caps on the regulator board will stay charged so the switch isn't going to cause any glitches) and you can antagonize your family members by making them participate in your single blind test. Hours of fun and a true bonding experience!!!  

Gratuitous, trivial comments -
Your using ground (Green&Yellow) wire to connect your 'hot' to the fuse, had me momentarily quite flummoxed, evolving to a generalized, anticipated dread..... Some years from now, when AC color codes become second nature, and your (or more likely, someone else's) memory fades, you just might assume  that it is a ground wire and...

Iff'n your deriving your ground for the balanced output center tap at a hidden chassis connection by that transformer, run an actual wire from the center tap to that chassis 'star ground' by the inlet. You really want minimal impedance for that connection.

I think you now clearly have the biggest Squeezbox powersupply around.

[mgalusha]

You might try out a snubber too. Typical snubbers are in the 220nf - 470nf cap followed by a 1R resistor (5w for amp typically).

I thought about it but thought I'd see how it sounded without them first. I have the needed caps and resistors on hand, so I may warm up the iron this weekend and see how she sounds with some snubbers.

Gratuitous, trivial comments -
Your using ground (Green&Yellow) wire to connect your 'hot' to the fuse, had me momentarily quite flummoxed, evolving to a generalized, anticipated dread..... Some years from now, when AC color codes become second nature, and your (or more likely, someone else's) memory fades, you just might assume that it is a ground wire and...

Yeah. Truth be told, I didn't strip enough wire from the Volex cord so I used a chunk of the extra ground wire instead of stripping out some more. I have plenty and will likely replace it one of these days.

Iff'n your deriving your ground for the balanced output center tap at a hidden chassis connection by that transformer, run an actual wire from the center tap to that chassis 'star ground' by the inlet. You really want minimal impedance for that connection.

The center tap is bolted to the chassis with the transformer. I thought about running it back to the ground at the inlet but wasn't sure how much difference it would make since the little transformer and output circuit doesn't reference ground at all. But now that I think about it, I suppose that will help to better reduce the common mode noise. If I have a chance this weekend to work on it I'll probably run a new wire. I certainly have plenty of it, so I don't have a good excuse not to.

I have a couple of large DPDT switches I could use to switch Felicia in and out in real time but I think I'll just leave that part alone as I'm lovin' the way it sounds right now.

Mike

[mgalusha]

A little follow up on the Felicia Squeezemore power supply.

The new power supply has been in the system for a month now and seems to have improved a bit, mostly in the first week or so. I know that electrolytics take some time to fully form, so perhaps that it the reason.

Last weekend I decided to try a shielded cable from the PS to the SB. Even though it's DC I figured it was still susceptible to induced AC and noise and being that I have way too much junk in my shop I figured building up a new umbilical wouldn't hurt and might even help.

The original cable was a piece of fairly small coax with a 20ga stranded silver plated copper core and silver plated copper braid and teflon insulation. I don't know if the fancy materials made/make any difference but it was cheap, something like $0.20/ft off of eBay, so I used it. Even though this was a coax I didn't really think of it as shielded since the braid was not grounded. When I connected it to the scope, I could see additional noise if I put my hand around the cable, so it was obviously picking up something.

For the new cable I twisted up a pair of 19ga stranded copper wires, put those inside a piece of flexible tubing and then put a piece of heavy tinned braid over the tubing. The braid was really too big for the wires, so the tubing makes it fit just right. I covered with it Techflex to make it purdy... Since I was using an XLR for the PS connection, I just tied the braid to pin 3 of the XLR. I also had to open up the PS and tie pin 3 of the chassis jack to ground. While I was in there I finally got around to running a new wire from the center tap of the second transformer back to the IEC instead of to the chassis where I had it. I also put in the correct color of wire from the IEC to the fuse. This might help Paul to sleep better at nights.

Once back in the system I did some comparing between the two cables. The first thing I noticed was that even with the original cable it seemed to have improved. The only thing different in the PS was the new wire from the transformer to ground instead of the chassis. Must have knocked out a little additional common mode noise. The sound had a bit more low level detail and the imaging was a little better. I've observed similar effects with other noise reduction efforts.

As for the differences between the umbilical cables. They are not hugely different but they are different. The most noticeable was improved bass. A little bit more prominent and deeper. Not sure why this was the case but it most certainly is. Also more low level detail and improved soundstage/imaging. The music was less attached to the speakers. I suspect this can be attributed to a small but apparently meaningful reduction in noise. The differences are most noticeable on vocals like the Fairfield Four or Manhattan Transfer. The performers just have more body and richer harmonies.

I put the PS back on the scope to see if I could measure the difference in noise at the end of the cable. It was visible on the scope but there was not a lot of it. With the "old" cable I could see about .6mV p-p of very high frequency noise which would increase if I put my hand near the cable. With the "new" cable, this was reduced to about .2mV p-p, so I guess a 3x improvement is good. I don't have a very high end scope and it's most sensitive setting is .1mV/div, so this is a close as I could get. I have a wide-band HP AC voltmeter and it too showed about .2mV of AC on the cable.

As for power line ripple. I can barely see a slight 60Hz wave in the scope trace, approximately .01mV p-p. One last item of interest. I see much more noise on the scope with the power supply unplugged than when it's on. I'm pretty certain this is due to the lack of ground and the pickup of noise. My lab/shop has three PC servers running and overhead fluorescent lights, so it's not exactly the quietest environment electrically. (I do turn them off to measure preamps/amps).

I have to say I'm delighted with the results of combining Felicia with the SB power supply. I have my amps on balanced power as well and the system is extremely quiet and that makes for some very nice music.

Almost forgot to mention. I tried plugging the PS into my BPT and the wall and I find I prefer the sound slightly better direct from the wall. Perhaps since it's already balanced internally feeding it balanced power doesn't help and maybe there is some kind of odd interaction occurring. I don't know and the difference is very small.

Mike

[Occam]

While I was in there I finally got around to running a new wire from the center tap of the second transformer back to the IEC instead of to the chassis where I had it. I also put in the correct color of wire from the IEC to the fuse. This might help Paul to sleep better at nights.

Whew..... after a full month, now I can finally catch some sleep. I was getting a bit fried  

Many thanks for the update. I'm pleased as punch that the results were roughly equivalent with and without the BPT. I was terrified that the level of performance was actually due to the BPT, with no actual contribution from Felicia. I really do appreciate your efforts, especially as you'd probably gotten the same performace iff'n you'd skipped the Felicia altogether and fed the SB ps directly from your very fine BPT. As I started my post, I'll end it with the same, Whew!

[04dgmsilv]

Bringing this one back from the dead...

Mike,  I saw some of your posts/reactions in regards to the new bolder PS for the SB.  You mentioned you were thinking of upgrading some of your components.  I am interested in building a PS with a LT1086 and some of the Jensen caps.  What kind of values are you planning on using?  Are you thinking of using them on both the input and output side of the regulator or just leave the panasonics on the input?

Thanks for any input you may have!

Dan

[mgalusha]

Hi Dan,

After hearing how the Jensen 4 poles worked in the Bolder Cables supply I ordered a pair and replaced the Panasonic and film caps in my power supply. I used one Jensen between the rectifiers and the regulator and the second after the regulator, locating it as close to the regulator as possible.

IMO the improvements were substantial. Richer harmonic structures, deeper and tigher bass and greater clarity in the midrange and treble. Very noticeable on cymbals and very enjoyable.

I used the 47,000 25V variety. It was an expensive undertaking but the outcome, to my ears, was well worth while.

Mike

[04dgmsilv]

Interesting, no more bypass caps, that makes for a simple (although expensive) circuit, thanks for the info.  This will go on my to do list, after some of my other changes settle it.

Thanks Again!

Dan