Felicia Balancing PowerConditioner Constructor's Thread

Occam · « **Reply #380 on:** 30 Oct 2006, 10:53 pm »

BRN,
I don't know the specifics of your SET amp, but the Felicia has successfully powered a Baby Sophia, as well as other low powers SET amps. Your use of a A41-175-36s, as opposed to the A41-175-28s would be more efficient with slightly lower losses, so I'd respond with a qualified yes, if the total draw was less than about 35-45va.

Regards,
Paul

HAL · « **Reply #381 on:** 14 Dec 2006, 02:20 am »

I am very interested in this idea for line filtering. Does anyone in the Southern MD, Northern VA or DC area have a Felicia BPT setup that I can make measurements on with my Fluke 43B Power Line analyzer? Can also use CLIOwin to measure the frequency response of the system.

I tried building an isolation transformer, similar to one I read about using two RS 120VAC to 25.2VAC, 75VA, E-I core units and a Solen 3.3uF/400VDC cap. Already had most of the parts on-hand. The RS transformer's secondary went into saturation without any load and just the capacitor. Looks like a L-C resonance phenomena.

Have started a PSPICE model of the system, but it is very limited. Will try measuring the RS transformers with CLIOwin to get the inductance parameters for PSPICE.

Any feedback greatly appreciated.

Occam · « **Reply #382 on:** 15 Dec 2006, 12:15 am »

Hal - The transformers going into saturation with a 3uf cap as the 'betwixt and between' cap is very strange.... The obvious follow up question is whats happening when you hook the secondarys back to back with no caps across the 25.2v secondaries, and are you sure you've hooked the cascaded transformers secondary to secondary?
I've generally found that when I get anomalous results, I find I've done something silly, which unfortunately, I find happens all too often.

HAL · « **Reply #383 on:** 15 Dec 2006, 05:48 pm »

Occam- I tried just the two transformers back to back without the 3.3uF cap, and the waveform looked like the input. No additional distortion was measured. I checked to make sure that they were secondary to secondary. The color codes are different for primary and secondary. I also checked the output voltage level matching of each transformer and they were within 0.01VAC on the secondaries.

Once the cap went across the secondaries, the distortion level increased about 50% over the standard line. Core saturation is about the only thing that can explain this that I know.

exxtreme · « **Reply #384 on:** 6 Jan 2007, 10:50 am »

Hi Occam,
First let me apologize for spamming ur PM

A few questions that have been bugging me:

As i mentioned before i'm from Singapore and the voltage here run at 230v - 240v hot.

1) You mention several times in ur threads, that we (e dudes living 240v countries) don't need to step down but run it straight thru? I presume you're talking about using the A41-X-230 series where the pri and sec can run at 230v? Is there any downside for doing this or would u rather have the voltage step down to 15v-25v and stepping back up again?

2) Alot of mentions of Jantzen caps and Auricaps. Anyone tried other varieties from BG, Murdorf, Rubycon etc?

Many thanks!

Occam · « **Reply #385 on:** 6 Jan 2007, 05:53 pm »

Hello Exxtreme,

And welcome to AudioCircle.
Yes, building a Felicia for 240vac isn't any more difficult than building one of 120vac. Look at the Felica 'sticky' -
http://www.audiocircle.com/index.php?topic=18441.0
While the schematic is labeled 'unbalanced 120vac in' and the output is labeled 'Balanced 120vac output', the schematic would be exactly the same for 230-240vac operation, fed 240vac at the input, it gives you balanced and filtered 240vac at the output. (assuming the transformers are appropriately rated for voltage)

As you mention, one could alternatively use a non stepdown transformer for this purpose. You wouldn't have to reverse the output transformer. You could use A41-175-230 taking 230 in producing 230 out, put capacitors across that output and cascade to another A41-175-230, ground the center tap, and you get balanced output appropriate for your country. The values of the 'across the line' capacitors would have to be adjusted for the inductive characteristics of the transformer, as would the voltage rating. The last point is very important.

But it is important to note that the value of this project revolved around the fact that the transformers were available for less than $10 each. So for around $40+- one could build a very, very good balancing power conditioner. The fact that it was constrained in power, large and bulky, was offset by its cost and performance.
The equation changes radically when you have to start paying 'retail' for your transformers IMO, at that point, non-transformer based solutions become attractive, as do non DIY retail products.

As to appropriate capacitors for any powerconditioner - They MUST be metalized film capacitors. No electrolytics, even Black Gates; they're not appropriate. Ideally, they should be X2 certified to be used across a mains voltage, to 'guarantee' that they will fail either completely open, or completely shorted (so that the mains breaker will trip). Yes, there are X2 ceramic caps, and IMO, they 'sound' horrible. I don't know why. If you choose not to use X2 caps, as I do, you must fuse each Felicia as shown with fuses appropriate for the component to be powered. I use 600vdc rated Auricaps across 120vac because they 'sound' better than X2 caps, but only with appropriate fusing. Audience recommends 1200vdc Auricaps for use on 230vac. It starts to get expensive. Those same 'rules of thumb' should also be applied to other film caps, Mundorf, Audyn, SCR, Bennic, etc...., if not rated for mains use.

exxtreme · « **Reply #386 on:** 7 Jan 2007, 06:40 am »

Hi Occam,
Thanks for ur reply. Sure cleared up alot of qns that i had in mind. Hope this info help the ppl living the 230v countries as well. In Singapore, its kind of weird cause we have a lot of varieties components available, trans, caps, resistors, etc, but ready-made retail balanced power conditioner cost a whipping bomb; almost in the range of 2-3k (USD). Last i check the A41-175-230 cost ard $20-30 (USD) here.

Two quick qns:

1) For the fuses, at output section, i take that for 230v the appropriate fuse would be roughly 0.7-0.8A? Which is the max amp draw out of a A41-175-230 at 230v. Or should i be lowing it?
2) I'm attempting to power a valve amp off (300b) which rates about 200-230W. Any constraints with that? Or anyone ever tried to power such a thing off this conditioner? My only worry is that the above trans is insufficient to power such a load.

Once again thanks!

Edited: There isn't 18uF for Auricaps at 1500V for C1. Any advice what others could possible fit here?

Occam · « **Reply #387 on:** 7 Jan 2007, 03:31 pm »

Exxtreme,

The fusing of a Felicia conditioner should be the same fuses as those in the component to be powered.

A transformer rated for 175va is no going to be sufficient to power an amp drawing 280 watts. You can check the actual power consumption with instrumentation like this -
http://www.maplin.co.uk/Search.aspx?criteria=L61AQ&menu=0&WorldSearch=Y&doy=7m1&ShowPics=true&Stock=2&SD=true

The Felicia project using the specified surplus transformers was intended for powering source components drawing 50watts maximum. This was due to the voltage drops associated with using transformers back to back. Using a non stepdown transformer eliminates this specific limitation, but you still would want to operate any transformer well below its maximum rms VA rating for powerconditioning use.
Unless you're able to source large transformers 500va+, inexpensively (meaning surplus), you should consider non-transformer, inductively based conditioners -
http://www.triode-systems.com/modules/wfsection/viewarticles.php?category=1
I have built similar conditioners (in process), but I can't offer any specific advise regarding the above project. Non transformer based, inductive/capacitive filters are easier to build successfully for your higher mains voltages than in the States, as your currents are proportionately lower.

Regards,
Paul

exxtreme · « **Reply #388 on:** 7 Jan 2007, 03:41 pm »

Thanks Paul. Guess i ran the conditioner for my CD and pre for now. ~~As i stated my previous post, Auricaps dun have 18uf. Would u recommend i parallel them? Or easier to just step down the voltages.~~

As for my amp, going to grab one of these ampere meter off my company and monitor it peak value.

Edited: Revised my statement. Off to grab some generic cap from some motors and try them out.

Hopefully i dun blow something up...

Occam · « **Reply #389 on:** 7 Jan 2007, 05:55 pm »

Extreme,

Excellent edit!!! If you're looking for large capacitance to put across an AC line, motor run caps are excellent choices for their typically built in protection, and designed specifically for across the line use. Please do not use caps for 'motor start' as they're often electrolyics and inappropriate for your use.
Note that the 18uf capacitance was specified for use with those surplus Signal A41-175-28 transformers, used back to back. If this is your intent, as the 18uf cap would be used across 28vac, the voltage rating of the capacitor can be relaxed substantially. I use Jantzen 400vdc rated caps for this, as they sound very, very good, and they're available inexpensively in the States from PartsExpress.com . Solen/SCR 400v also work very well, as do Bennics, though not as well as Jantzens and Auricaps. Using Auricaps or Mundorfs in this role would be very expensive. If you're planning on using a non stepdown transformer, those caps are going to be subjected to full mains voltages. if you choose to, against all rational advice, to use caps that are not X rated appropriately or motor run caps minimally rated at or above the Vac to be used at, (I do for sonic reasons), you need high voltage rated caps, typically 600vdc+ for use on 120vac and 1200vdc+ for 240vac, and you must fuse appropriated as shown in the schematic. Also note that 18uf was specific for the original transformers. Using other transformers, or different voltage secondaries, would call for resizing the capacitors, which I can't help you with other than advising it will take substantial experimentation.

Regards,
Paul

exxtreme · « **Reply #390 on:** 9 Jan 2007, 07:05 pm »

Paul,

Thanks for ur prompt response. I was pretty much ready to go with the A41-175-230 straight thru but after much consideration, the first 3 caps will be a bumper for cost. There are plenty of 400vda motor caps here. My current plan is to go with the step down A41-175-28 back to back with the current advised caps. Also stepping down the voltage will give more choice for the first 3 caps and room to experiment. Finding caps that are rated 1200vdc+ is gg to put a hole in anyone wallet... off shopping i go.

Thanks!
Mark

exxtreme · « **Reply #391 on:** 9 Jan 2007, 07:16 pm »

One quickie question, anyone tried to parallel both the output of the power conditioner before? Might be dumb, but reasonable question as in theory it shares the load.

Occam · « **Reply #392 on:** 9 Jan 2007, 07:53 pm »

Exxtreme,

You only need those high voltage capacitor ratings on caps that are used accross the full AC mains voltage unless they are 'X' or 'Y' rated. Properly rated caps -
http://www.justradios.com/safetytips.html
X2 and Y2 rated caps are rated to operate accross AC mains voltages of up to 275VAC, and their failure modes are 'known'. An X cap is expected to fail either full open (removing it from the circuit) or dead short (tripping your mains fuse/breaker). Protected motor run caps can also be used.

Any other cap, regardless of their voltage rating, or your assumptions about its safety, should be fused when used on mains voltages 'accross the line', X, or 'to ground', Y. This includes the (in)famous "Auricap Tweak".

As to paralleling transformer's output.... The variability in windings generally preclude this unless 'load balancing center tapped chokes' [this is an old technique from radio amatuers used when surplus 'iron' was plentiful and cheap] are used. Alternatively, if you've precision wound bifilar secondaries, windings from each can be paralleled. But that implies custom wound transformers, and if you were commissioning that, you simply have them make a bigger transformer.

ctviggen · « **Reply #393 on:** 9 Jan 2007, 08:30 pm »

Quote

Any other cap, regardless of their voltage rating, or your assumptions about its safety, should be fused when used on mains voltages 'accross the line', X, or 'to ground', Y. This includes the (in)famous "Auricap Tweak".

Dang! I better get working on putting a fuse in the box I built. Do you use a slow-blow fuse rated for the amperage (in my case, 20 Amps)? Wouldn't the breaker also trip if the cap fails? (In my situation, I have about 6 feet of wire going from my circuit breakers directly to my box, with the Auricap tweak and four outlets. My entire system, save one sub, is run using that box. This is a single 20 amp circuit.)

Occam · « **Reply #394 on:** 9 Jan 2007, 09:11 pm »

Bob,

The Felicia is designed to power one component, about 60watts max. It can be viewed as a powercord with built in conditioning, and I actually build mine as that. I take a Volex powercord, cut it in half, and wire the Felicia in the middle. The fusing should be exactly that of the component being powered, not the amperage rating of the line powering it. Because non X or Y rated caps have a 'via Standards' possibility of partially failing, neither completely open or completely short, we fuse with same value as the components fusing for 2 reasons -

1. A partially failed cap can draw less than the mains breaker and potentially cause a fire, but probably more than the 1amp (max) fuse used Felicia with its intended transformer.

2. If there is a internal fault in the component being powered, connecting 'neutral' to the chassis, that component might well not fuse the heretofore 'neutral' line, and this line now has 60VAC (in N.A.). If the fault is on that 'neutral' line, we want to provide fusing via Felicia.

Regards,
Paul

ctviggen · « **Reply #395 on:** 9 Jan 2007, 10:07 pm »

Thanks, Paul. What about the tweak where you use the Auricap from hot to neutral of (across?) an outlet? That's what I was specifically referring to.

The Felicia sounds like a great idea, but I just haven't had time to create one. I will reread this entire thread before doing so, however. The "power cord with built in conditioning" is an excellent idea. I could power my transport with one, my preamp with one, maybe the power supply for my modded SB with one, etc.

Occam · « **Reply #396 on:** 10 Jan 2007, 01:37 am »

Bob,

Re: the 'Auricap Tweak', we've been down that road before -
http://www.audiocircle.com/index.php?topic=24939.msg221401#msg221401

Quote

Well, I'll state up front that I think Auricaps do an excellent subjective job as accross the line caps. As to why they sound better than my favorite real X2 qualified cap, the Wima MP3X2, I don't have a clue. It might well be the goober dust.

That being said, I don't see any reason for inviting trouble, from either a legal or electical perspective. If you do have a fire, the insurance adjuster is going to be interested in your DIY hotbox, as its obviously going to look like a DIY project. If they open it up, and see a cap that isn't X2 certified, and its obviously damaged (it doesn't matter whether it was the cause of the fire, just that it appears damaged), you've just given your insurance company an excuse not to pay the claim. And Lord help you if someone was hurt (or worse) in the fire. I'm certainly no lawyer, but I've dealt with an insurance adjuster in the recent past as a result of flooding due to a frozen pipe, and simply say that they're paid to be pricks. (apologies to any insurance adjusters out there, I know you're just doing you job)

From an electrical perspective, I'd suggest you put whatever non-X2 caps behind fuse/breaker of whatever component has the smallest fuse. (and prefferably that component's switch, so that when turned off, its out of the picture). Ignoring the minimal inductance of your powercords, the cap is accross the line at the mains circuit level, providing 'cleanup' for all components plugged into that mains circuit. Installing caps on different outlets without intervening inductive components on the same mains circuit, is really the same as putting them all one the same outlet, save for the fact that its easier to mount multiple components in multiple places.
(the Audience Adept has intervening inductive components in the form of CMCs, common mode chokes isolating each outlet).
The reason I suggest this is the danger for ANY cap is not if it fails as a direct short or open. In the former, the mains circuit breaker blows, and in the later, the cap is out of the circuit. The danger is a leaking partial short, which is exactly what real X caps are 'guaranteed' NOT to do. This is going to heat the cap, and possibly cause a fire...... but possibly not cause your 15/20amp mains breaker to trip. But if its behind the 1/4-1 amp fuse of your DVD player (or Felicia or whatever) chances are that fuse will blow before that cap gets hot enough to cause a fire. And besides, the adjuster is far less likely to find that non-X rated cap if its buried inside a commercial component.

exxtreme · « **Reply #397 on:** 10 Jan 2007, 01:21 pm »

Hi Paul,
Fully understood. Because if i were to step down the voltage, C1-3 would be able to have a much lower voltage requirement compared to 230v straight. Cheaper and more choices. C5-6 still run at full AC voltage so no choice with these two buggers. BTW that page link was very informative. Thanks!

By running it at 60watts, u giving a headroom for safety i suppose? Because a A41 series trans runs approximately 174 watts max. (based on 230v 0.76A). FYI i alr placed an order for a few pieces of A41-28-175. Oh the joy!

Occam · « **Reply #398 on:** 10 Jan 2007, 02:06 pm »

Exxtreme,

The downrating of 60watts maximum from the 175va ratings of the 2 cascaded transformers is an issue of transformer regulation. Because one of the transformers is hooked up 'arse backwards', you'll find that the output voltage falls with increasing loads, more so than with transformers hooked up normally.
You might find that when fed 230vac from the wall, when drawing 60watts, the output voltage from Felicia is 200vac, or less, or more..... Somewhere within this morass of a thread, a number of folks have posted their measured results,..
As to whether this will work, that would depend the implementation of the actual components to be powered. Some components with low voltage shutdown circuitry will do just that. I'd like to be able to give you a more definitive answer, but I've not implemented or measured 230vac Felicias.

Regards,
Paul

exxtreme · « **Reply #399 on:** 10 Jan 2007, 04:51 pm »

Thanks for clearing that up. I'll try and see if there is too much drop in voltage.