capacitor

"ivan" wrote in news:eogaa8$kvq$ snipped-for-privacy@news.flashnet.it:

Do you really need to specify a capacitor by make? If you're looking for a replacement, the more you say about it, the more likely someone can point out one that is right.

"Lostgallifreyan" ha scritto nel messaggio news:Xns98B9ACADD2CE7zoodlewurdle@140.99.99.130...

I have found this capacitor that perfoms much better than the others I have tried, so I am interested to know what are its characteristics in order to confirm my guess about the type of capacitor needed. That triangle must be some kind of logo so I was hoping that someone knew where it is from.

If you can define your meaning of 'better' in your application it's quite likely someone here ( me even ! ) could help explain why that's so.

And I'm sorry, that logo's not familiar to me. There are *thousands* of capacitor makers worldwide.

Graham

Den Mon, 15 Jan 2007 17:32:26 +0100 skrev ivan:

Perhaps Vishay??

--
Vy73 de OZ1GNN
Christian Treldal

"Eeyore" ha scritto nel messaggio news: snipped-for-privacy@hotmail.com...

upside

for a

point

have

to

be

likely

ok.. I have a series resonant branch that works at its resonant freq. the frequency is genereated by two square waves in phase opposition at the two ends of the circuit. If I oscilloscope the voltage between the inductor and the capacitor I see that the voltage swing boosts very high (70-80Vpp) as expected. In fact I seem to observe that the higher the voltage that the capacitor can take the higher the voltage swing generated by the resonant circuit. Now I am quite sure of this, but then found this capacitor that boost the voltage as high as 120Vpp which I do not know its characteristics. This is why I am looking to find its Voltage spec, it will confirm my guess.

thanks ivan

The voltage in a resonant circuit is proportional to the Q of the tuned circuit, and how close the driving frequency is to resonance.

It's possible the capacitor you chose was at just the right capacity to resonate with the driving frequency and/or has a high Q. But Neither capacitance nor Q is all that special-- most any capacitor will have high Q, and of course you can choose almost any capacitance value, or trim it by adding exta parallel capacitors.

So don't focus on that triangle.

"Christian Treldal" ha scritto nel messaggio news:45acb685$0$144$ snipped-for-privacy@dreader2.cybercity.dk...

down

no it does not seem Vishay, but thanks!

"Ancient_Hacker" ha scritto nel messaggio news: snipped-for-privacy@51g2000cwl.googlegroups.com...

well I measure each capacitor I use to be as exact as possible so I exclude the capacity value hypothesis. Q depends on the values of R, L and C and none of them is changing. However the capacitor does seem to be very critical on the voltage when I change the type. I think is a question of losses of the capacitor...

That isnt a hypothesis. It's the basic equation of RLC circuits.

How are you measuring the capacitor? Resonance can really sharp, much narrower than the tolerances of your typical capacitor or simple cap meter.

--
The reason the voltage is being boosted is because of the "Q" of the
circuit, not because of the voltage rating of the capacitor.
However, I suppose that it's possible for the various capacitors
you're using to exhibit higher Q's as their dielectric voltage
ratings increase.  An easy way to find out would be to measure the
impedances of the various capacitors in order to get the real part
of their impedances.  The lower the R the higher the Q and the
greater the voltage at the LC junction.

BTW, are you sure the 'delta' isn't a VDE certification mark?

http://www.vde.com/VDE_PI_en/TestingCert/VDEMarks/

To what tolerance do you mtach these capacitors ?

Different dielectric types have different losses.

Polypropylene has very low losses. For high currents, metal film has an advantage over metallised film too.

Graham

I have R=22, L=737uH and C=2,2nF so Q is about 26.3. Now, if short circuit R, I get 15V increase, hence a change in C of less than 1% cannot give me such a voltage change. I remind you that I am selecting the value of the capacitor to be exact.

no it looks different, thanks

less than 1% and I use the same cap meter, I suppose that the measuring would be quite exact.

I

yes, this is my guess, it must be a question of capacitor losses. Unfortunaley do not have too many capacitors available to try, and the ones I have are not from a known provenance - hence no specs

As I wrote, resonance can be a lot narrower than 1%.

And in most LC circuits the Q of the circuit is more limited by the Q of the inductor. Inductors can have Q's from around 1 to 500. It's hard to find a capacitor with a Q under 500. So in most cases the exact Q of the capacitor is irrelevant.

If you write Delta, this *is* a triangle, but not upside down. Are you sure it is a capacitor and not a MOV? I have here a blue part from EPCOS 7mm dia. which has only a triangle with a longer left side. The bottom line is curved. something written like K275 and 05. All the other epcos parts I have, have actually written the company name in letters below the filled triangle.

--
ciao Ban
Apricale, Italy

"ivan" wrote in news:eoism2$9n6$ snipped-for-privacy@news.flashnet.it:

No, but I think John Fields was talking about a capactor with lower equivalent series resistance. That would allow a higher peak voltage during transitions.

It could be instructive to try testing the circuit with an additional high-quality air-variable capacitor wired in parallel with the cap under examination. An air variable whose maximum capacity is perhaps

2% of that of the fixed cap might be suitable. Try adjusting the air variable to tweak the resonant tank for peak voltage.

In roughly half of the cases, you'll probably find that an intermediate setting of the variable cap does create a voltage increase... in other words, the fixed cap didn't have quite enough capacitance to resonate with the inductor at the exciting frequency. I suspect that you'll find that when this is done, there will be little difference between the peak voltages using different fixed capacitor types (as long as they're decent film caps).

In about half of the case, you'll find that the presence of the variable cap in the circuit _decreases_ the maximum voltage you can achieve, even when it's set to its minimum capacitance. In most of these cases, it'll mean that the fixed capacitor has too much capacitance... it and the inductor are resonant _below_ the exciting frequency. You could add a bit of additional inductance in series with your existing inductor to compensate for this, and then re-trim the variable cap to peak the circuit.

Yup. And, if the circuit has a Q of (say) 200, and you only match the caps to within 1%, you'll see a big variation in peak voltage if you compare even "matched" capacitors of the exact same make and model.

1% matching just isn't good enough.

I think the original poster is going to _have_ to trim the circuit (using variable or fixed trimcaps) to achieve peak voltage.

--
Dave Platt                                    AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
  I do _not_ wish to receive unsolicited commercial email, and I will
     boycott any company which has the gall to send me such ads!

bottom

sorry about this guys, I meant an upside down triangle or upside down Delta. It think I have found it though... it seems to be a 2n2H100 capacitor which I suppose it means a 100V 2,5% cap (tell me if I am wrong). This proves my guess that the higher the voltage of the cap the lower the losses (of the dielectric) hence better performance, in this case.

thank you all

It proves absolutely nothing of the sort.

You're barking up the wrong tree.

Graham