Suggestions for Capacitor

Another "welcome back!"

Sounds like what is tailor made are the run/start cap combinations used in air conditioner systems. They get fairly large, though, because the voltage range is higher than what you need.

Another possibility are these new ceramic caps used as bulk storage in front of dc/dc converters. They replace electrolytics, and come in DIP packages. Because they are ceramic, there is no dc limitation, plus the inductance and ESR are very low. They will cost more than the electrolytics they replace, however.

Hi TVP45,

Did you get upset with RP ("Bows in direction to Albany")? If so, I know the feeling. I have seen you here before and I respect your input - welcome back.

Do you have any schematics? How do you know that the eCaps are the problem? You could try tantalum caps. Here is the DigiKey link:

http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=131082&keywords=tantalum%20cap

Hope this helps,

Mike

P.S. Where's the donkey?

Tantalums are like electrolytics - they are only good for dc.

Yeah, yeah - filtering apps; so why were eCaps used in the first place?

Well, TVP45 did allow as how they didn't work all that well...

OK - so what AC-applicable caps can you find in the 10μF to 100μF range?

I disagree. For the application I would recommend tantalums. Big value in a small size. It's for students you know.

But, as noted previously, tantalums are polarized; they are not supposed to be used for ac.

You can use polarized caps in non-critical applications like this, to obtain non-polarized values:

Use two polarized caps of twice the desired non-polarized cap's value.

Connect the positive terminals of the two caps together.

Use the negative terminal of each cap for each lead of the required non-polarized C.

Of course, this does not change the other characteristics of the caps used; in fact, some are made worse. For example, the ESR doubles, more or less. But I've used this method many times, including audio systems, speaker cross-over networks, etc. Again, I wouldn't use it for a critical app, though I can't really think of why. I guess it is just my reluctance to use any kludge for a critical app.

regards,

Kos

Non-polarized electrolytic capacitors are used to created resonant tank circuits for speaker cross-overs. They will not be as low in leakage as a poly cap, but they are smaller and much cheaper. Places that specialized in speaker building will have a good selection, like this place......

Parts Express

Hope that helps,

Tom

Thanks so far.

What about the super-metallized polycarbs? Anybody have experience with these? They're a bit bigger than I had hoped, but???

The problem is that the electronics instructors have little practical experience and didn't know the problem with electrolytics. Often, you can get by with putting several volts ac across them for short periods - perhaps only one in five goes to ruin quickly. Since they already have the inductors, they don't want to start from scratch.

I personally would have used a couple hundred turns of 16 AWG around a low μ torroid, added a good polycarb 1 μF cap and used properly shielded leads for measurement. That's not where they're at.

You can always use the electrolytics by connecting them back to back (eg + to +, or - to -) if you consider the new combined capacitor to be half the value of the individual capacitors.
E.G. Two 100uF connect back to back is a 50uF non polarised capacitor.

Each must have a voltage rating exceeding AC the peak to peak.
Ok for audio... don't do it for power circuits, terms and conditions apply etc.
Del

(PS, If you need any help eating crows I'm here)

Thanks for the offer. I thought you guys had Rooks?

I've used the old trick of back to back, paralleled by diodes "just in case". I would be OK doing that for something like audio coupling where ESR doesn't matter much (I've got a tin ear and can't hear distortion anyway), but this is a resonant circuit lab where they want to then measure "Q".

Welcome Back.

Newton and Goethe had a hell of a polemic going on color theorie.

Newton won. At the time.

Current knowledge (today)has established that A) Newton Fudged data; B) Goethe was closer to todays knowledge than Newton and 3) (deliberately!) All of humanity will suffer newtons folly as a result of the darkside of the moon album cover from pink floyd.

's Okay to have differences.

In a couple hundred years the knowledge always refutes the supposed winner.

's called continuous improvement.

Welcome back

milo GOETHENEWTONPOLEMIC

Neat commentary. I had never looked critically at the cover. Is there some message intended or just a screwy artist at work? Interesting aside: On the way to class last week, I was listening to early morning alternative music (Before about 7 am, there is apparently no format rule at all) and I couldn't stand what they were playing, so I hit SCAN and sighed with relief when the next station had some classic rock...and then I realized with a start it was Pink Floyd. I think I might be getting old.

At www dot parts-express dot com you can find
high quality bi-polar (non-polarized) electrolytic capacitors that may fit your need. Perfect for crossover design.
They feature a 5% dissipation factor and are rated at 100 vdc which are effective to approximately 200 watts.

Dimensions
Part # Value Dia x L(mm)

027-340 10.0 uF 10 x 24

027-342 12.0 uF 10 x 24

027-344 15.0 uF 13 x 27

027-346 17.0 uF 13 x 27

027-348 22.0 uF 13 x 27

027-350 33.0 uF 13 x 32

027-352 47.0 uF 13 x 32

027-354 50.0 uF 13 x 32

027-356 68.0 uF 16 x 38

027-358 80.0 uF 18 x 38

027-360 100.0uF 18 x 38

Thanks, Guest.

I saw those. I'm a little wary of the 5% ESR, but might have to live with it. I'll keep those as my "Ace in the hole".

How much are you worried about costs?

Digikey meet all the requirements in multilayer ceramics.

If you use multilayer ceramics, you won't need to overate them (voltage wise).

They are mostly surface mount but you could solder some leads onto them for student use. In fact you could keep the cost down by making up parallel combinations with some of the lower values. (The popular components are fairly cheap.)

Thanks.

I don't have experience with these so was unsure. Film caps I know, so was looking there first. I'll go back to DigiKey and look at the ceramics. I trust your advice. Again, thanks.

Randall's suggestion of the Multilayer ceramics is a good one if you can tolerate the surface mount, leaded ones in large values are very hard to come by. I haven't seen any bigger than 4.7uF in leaded.

Your space constraint is the real problem....therefore as far as film capacitors goes, metalized mylar is your best bet there, both by size and cost. They will more than meet your requirements for your application. Polycarbonate, unfortunately, has become very expensive, the best film for your money would be very good quality polypropylene but you pay a volume penalty for it as even metalized are fairly large. If you want to look at some reasonably priced polypropylene, try the EPCOS MKP, page 828-829 in Mouser, Allied also stocks them and may be a little cheaper. These are available up to 60uF. Some polypropylene cross-over capacitors from Parts Express, their Dayton label, are also available in large values and reasonable prices but they are necessarily big.

None of the electrolytic or tantalum capacitors are suitable for your use, mainly too high a D.F., leakage current can be minimized but will not affect the D.F.

Multilayer Ceramic Caps are the solution.

Small and come in uF ranges and voltages that exceed 20V

Most distributors carry them Vishay has a good line VJXR dielectric

I know that they have at minimum 1uF @ 50V

Check them out Vishay is not the only mfgr with multilayer caps

Here is a PS

These are chip caps so don't be affraid to parallel as many as you need to get the higher values. The chips are very small so they should fit you space requirment.

Check your text for caps in parallel and series. You can work out what you need.

First, welcome back to the grist mill of engineering debate.

This seems like a perfect lab on the effects of parasitics in capacitors. As you've noticed, a student using just one aluminum electrolytic capacitor for the resonant capacitor, does not work as well as theory predicts. Depending on the circuit topology this happens because for long periods of time the capacitor is reverse biased for the DC bias markings. An insightful student might then put two capacitors in series with the DC blocking polarity set to assure one capacitor is always properly biased. Hopefully this student will not rely on the manufacturer's values and actually measure the capacitor's values with a LCR meter. In redoing his calculations with the now more precise LCR values he should find that the resonance numbers still do not agree with the series capacitors. They will be considerably better than the single capacitor circuit but differences should still exist. Component tolerances will no longer be an excuse in the error write up, for the component values will be more precisely known than the tolerance. The parasitic series resistance, and inductance along with the parallel leakage resistance of an aluminum electrolytic capacitor will be effecting the resonant tank circuit.

I'm not sure if you should remove this scenario from a truly gifted student finding these characteristics. You need some gradient to sift the brilliant from the above average, while not crippling the ones who have just not yet grasped these concepts.

But as far as an ideal capacitor for this circuitry, look at the metallized polypropylene capacitors of Illinois Capacitor. The 250WVDC family goes up to 50 uF. But as you probably guess, they're not as tiny as an aluminum electrolytic nor are they as cheap. For an intermediate solution, tantalums have significantly less of a series inductance so the resonant frequency will stay the same. They still will exhibit less prominent parasitic resistance problems.

Thanks, redfred.

I'm not familiar with Illinois, but I'll check it out. ITW down in Lynchburg makes something like that, but I can't get my hands on them. Any ideas about distrubutors?

There's a host of suppliers available. The easiest might be Newark at http://www.newark.com/

Again, thanks.

RE your earlier comments on good students learning from bad caps. These are Physics students who don't have the background or the instrumentation to really look at this. Sad, but true. I suppose a sign of the times, eh? I went into an electronics parts store the other week and started chatting with the owner (who has got to be over 90) about how hard it is to find coil winding materials. He pointed out that his is probably the last parts store between Philadelphia and Cleveland (not counting Radio Shack, which isn't really that kind of store). This in Pittsburgh, once home of Westinghouse Electric, where it was said you could go into the Turtle Creek Valley and have anything imaginable manufactured, from a toaster to a nuclear reactor, without having anything shipped in from outside other than raw materials!! Ach!

These people (usual disclaimer) specialise in supplying the education market. They might have something appropriate, perhaps?

Thanks to all who posted. Excellent help! I have forwarded my recommendations to the Professor, who has blanched at the time and money it takes to fix what could have been done correctly the first time. I'm awaiting his pleasure.