Jump to content
The Classic Speaker Pages Discussion Forums
redpackman

AR-3a Recap - Is this heresy

Recommended Posts

I'm looking to recap a pair of AR-3a's I have. I understand that involves 50 uf and 150 uf caps as well as a smaller ones. Frankly, those larger sizes can really get expensive with the foil and other types of caps.

Is it pure heresy to use an eletrolytic NPE cap in the crossover? I realize they don't have the longevity the other caps have, but when the cost is 10% or less of the other options and funds are tight, I'd like to know what the ramifications are? I understand that originally AR used NPE electrolytic caps in some of their speakers.

Will it work...for awhile at least? Will the sound be improved over the originals that now inhabit the cabinet? I'm sure a pair of $50 or $150 caps would be marvelous, but at some point the budget just won't bear that.

This is a NPE electrolytic I'm considering:

http://www.parts-express.com/pe/showdetl.cfm?partnumber=027-364

Comments?

Share this post


Link to post
Share on other sites

That's fine--lots of guys use NPEs for larger values. The sole drawback is that your children will have to replace them in 20 years or so.

Another possibility: Use a bunch of 10uF Surplus Caps from Madisound. 30 @ 0.60 each will cost you $18.00. Or spring for 50 (spares will come in handy for another project) at 0.50 each, or $25.00.

Maybe I should stop recommending these. They may be all gone next time I want some! <_<

Kent

Share this post


Link to post
Share on other sites

Even though I am not looking at an AR-3a schematic at the moment - - - which means I do not know offhand all the appropriate cap values for this speaker - - - I am inclined to both endorse and repudiate JKent's recommendations. I wholly agree with his comment that NPE caps are both fully suitable and a very good value for the larger cap values - - - sheesh, even if these caps are good for "only" twenty years, that timeframe far surpasses the longevity of most contemporary consumer products we might buy today.

However, as much as I do like these 10 uf surplus caps and am pleased that Kent has directed us to them with high endorsement, I would be totally uncomfortable combining this many capacitors into one cabinet for a single driver - - - many of you are certainly superior to me at circuitry assembly, but this recommended scenario would be nothing short of a wiring nightmare in my hands.

My limited knowledge recommendation: simplify and cut expenses for large value woofer caps, but if you are able, spend a bit more for higher quality caps for tweeter and midrange.

Share this post


Link to post
Share on other sites

The 150 uF cap is in parallel with the woofer to make it's electrical response second order. It's not a critical signal path. So, there's not a critical need to use a film cap on the woofer.

I recommend a film cap replacement for the 50 uF midrange series cap. It will brighten up it's sound vs an NPE.

Share this post


Link to post
Share on other sites

I think Carl's compromise makes sense. The 10uF caps are easy to work with and putting 5 together is no problem: They are flat on 2 sides so they are easy to stack, or just lay them out in a row. Attached is a photo of an AR-5 Roy did, with the 10uF surplus caps used for the 72uF and the 24uF values.

Putting 15 of them together would not be a nightmare. They could be laid out the way Roy did his, but in a double layer: 8 in a row on the bottom and 7 on top with a strip of double-faced foam tape holding them together. I'm hardly "superior in circuit assembly." My knowledge is limited and my fingers are arthritic. But cobbling together this Franken-pacitor in a nice big AR-3a cabinet would not be a problem.

But as I said--new NPEs are fine.

Kent

post-101828-0-96436000-1372079770_thumb.

Share this post


Link to post
Share on other sites

"However, as much as I do like these 10 uf surplus caps and am pleased that Kent has directed us to them with high endorsement, I would be totally uncomfortable combining this many capacitors into one cabinet for a single driver - - - many of you are certainly superior to me at circuitry assembly, but this recommended scenario would be nothing short of a wiring nightmare in my hands."

Hello:

Wiring ten 15-uF caps in parallel is really simple as the ten can be combined outsidet the cabinet and installed as one unit. It does have one advantage. When connecting ten in parallel, the total capacitance becomes 10x15=150uF, as expected, but the equivalent series resistance becomes ESR/10. Levenson used ten small film caps in his cello amati likely for this reason--reduced ESR. That said, there is absolutely no problem with using non-polar electrolitics other than giving one's grandchildren a headache in some years.

The reason most old electrolyics failed was a poor end seal that allowed water in the dielectric to evaporate. First C increases as the aluminum etches faster, then it becomes very small, as the device dries completely. The worst cpacitors were encapsulated in PVC tubing (usually black), and then end sealed with pvc (usually red). The plug did not stay sealed to its shell. The best were made by Sprague and Cornell-Dublier; both firms used a high quality elastomer with a strong crimp seal--many are still within spec (both C and ESR) thirty-five years after fabrication! Still, its a good idea to replace all at this age. Today inexpensive electrolytics are rated anywhere from 2000--7000 h. I have fixed many a modern widget for a few cents worth of small electrolytics!

JohnO

Share this post


Link to post
Share on other sites

Remember when you put capacitors together when put in parallel the value of the caps ad and when in series it like resistors in parallel. Also caps in parallel loose voltage handling ability and in series the increase. I t is jus the recipical of resistors.

Share this post


Link to post
Share on other sites

Remember when you put capacitors together when put in parallel the value of the caps ad and when in series it like resistors in parallel. Also caps in parallel loose voltage handling ability and in series the increase. I t is jus the recipical of resistors.

Can you explain "caps in parallel loose voltage handling ability"., if all caps have the same voltage rating.

Share this post


Link to post
Share on other sites

Remember when you put capacitors together when put in parallel the value of the caps ad and when in series it like resistors in parallel. Also caps in parallel loose voltage handling ability and in series the increase. I t is jus the recipical of resistors.

Huh?

Share this post


Link to post
Share on other sites

I think he means if you have say, 2 20uf caps rated at 200V in parallel, that is equivalent to 1 40uf cap rated at 100V....but I don't think that's right, as far as the voltage handling....I thought caps in parallel, you add the voltage capacity as well (so it would be equivalent to a 40uf, 400V cap)....but I'm a mechanical engineer, not an electrical one ;)

Share this post


Link to post
Share on other sites

The voltage ratings must be additive when paralleled. Otherwise Peter Snell's designs would have failed miserably in the marketplace. Just look at one example of a few cap bundles he put in his flagship Model A.

post-100237-0-98670100-1373285580_thumb.

Share this post


Link to post
Share on other sites

My understanding is the voltage rating of a bundle of paralleled capacitors is determined by the lowest voltage rating of the capacitor in the bundle. This is analogous to the weakest link theory. However, in the series connected capacitors, the overall voltage rating is additive of the individual capacitors. There must be an EE major among us who can give a definite answer.

Share this post


Link to post
Share on other sites

There must be an EE major among us who can give a definite answer.

Member Johnieo is your man! Not a mere "EE major" but a bona fide EE professor! Hope he chimes in to settle this.

Kent

PS: or just google it: http://www.aikenamps.com/AddingComponents.htm

As I read this, ligs is correct about the weakest link in paralleled caps. Series caps seem to be more complicated but we're not putting caps in series in our speakers.

Edited by JKent

Share this post


Link to post
Share on other sites

Can you explain "caps in parallel loose voltage handling ability"., if all caps have the same voltage rating.

Can you explain "caps in parallel loose voltage handling ability"., if all caps have the same voltage rating.

Hi Ligs:

Caps in parallel:

Add the individual C values to find total C, as you know. The max voltage of the group is that of the capacitor with the lowest voltage rating---they all have the same applied voltage, and the unit with the lowest rating will fail first. However, there is an issue if we are using electrolytic caps-- in case of electrolytic caps, one usually uses caps at some max voltage near to, but below their rated value. For example I might choose caps with a 50-V rating to use with say, 25-35 V max applied. So in the case of this max applied voltage I would likely not use an electrolytic with a 450-V rating in parallel with one having a 50-V rating, as one would operate at a very low fraction of its rating and not be formed properly. All electrolytic caps used in a parallel group do not all have to have the identical max voltage rating but should be similar-- perhaps, say, between 50--100 V for that example.

Caps in series:

As you and others know, the total C is found from the reciprocal rule. 1/Ctotal = 1/C1 + 1/C2 + etc.

It is not prudent to ever put caps in series because there is an added complication: capacitors have leakage resistance, and it affects the way the voltage is divided between the two devices. Capacitors have large (parallel) leakage resistances. Lets for example say that we have two 100 uF caps in series. Ideally we have 50 uF total, with half the voltage across each device. However, that assumes ideal capacitors with infinite parallel resistance. Nothing is ideal in real life!

Let's assume that one of the 100-uF capacitors has a leakage resistance of 2 MOhms and the other has a leakage resistance of 4 MOhms. The DC voltage across the two capacitors in series divides according to their resistance values, not their C values. So in this case 2/3 of the applied voltage would appear across the unit with 4MOhms leakage and 1/3 of the voltage would appear across the device with the 2 MegOhm leakage. Were we to apply a large voltage to these units, one of them would fail well before the other! Since leakage R can change with age, it is not wise to ever connect caps in series.

So the practical answer to what is the voltage rating of series caps? Don't do it! :)

hope this helps,

john

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...