Infinity RS 2.5 crossover -- capacitor replacement

rdp91356

Member
So after getting a lot of good information about recapping, especially this thread: http://www.audiokarma.org/forums/showthread.php?t=78978 I decided I would recap the crossovers my RS 2.5s. I thought my speakers still sounded great so I wasn't sure that I would hear much of a difference, and also afraid that I might change the sound too drastically from what I was used to. So I planned out how everything would fit and what replacements I would use and ordered my parts.

But I also wanted to really know what was going on with the old caps so I bought this ESR meter (from an eBay seller): http://www.peakelec.co.uk/acatalog/jz_esr70.html. It can measure the caps in circuit, but I also measured them out of circuit to be sure. The measurements were only slightly different in most cases, except for that 3.83uF tweeter capacitor -- I got crazy numbers measuring that in circuit. Below are the numbers testing all the old caps out of circuit. Much better than I thought they would, which made me wonder if I should just replace the ones that were the furthest out of spec and just solder everything else back the way it was. But I decided to replace them all since all the new caps were within 2% of where they should be.

1981 Infinity RS 2.5 original 30 year old capacitor ESR meter readings December 2011.
Left:
1100uF = 1202uF .07 ohm
700uF = 735.4uF .06 ohm
700uF = 768uF .06 ohm
125uF (75uF + 50uF) = 143.2uF .06 ohm
125uF (75uF + 50uF) = 135.6uF .06 ohm
50uF = 56.26uF .10 ohm
3.83uF = 3.64uF .08 ohm

Right:
1100uF = 1190uF .08 ohm
700uF = 744uF .08 ohm
700uF = 726.2uF .08 ohm
125uF = 133.4uF .11 ohm
125uF = 128.8uF .10 ohm
50uF = 55.85uF .11 ohm
3.83uF = 3.59uF .08 ohm

*Note the differences of the 125uF caps between the left and right crossovers. The left uses parallel 75+50, while the right uses single 125. What's even stranger about this is that the serial numbers are sequential and I am the original owner, so it came from the factory like this.

Replacement capacitors (from parts-express.com) All measured within 2%:

1100uF =
two Parts Express 500uF 100V Non-Polarized Electrolytic Capacitors
+ one Parts Express 100uF 100V Non-Polarized Electrolytic Capacitor in parallel

700uF =
one Parts Express 500uF 100V Non-Polarized Electrolytic Capacitor
+ two Parts Express 100uF 100V Non-Polarized Electrolytic Capacitors in parallel

125uF =
one Dayton Audio DMPC-100 100uF 250V Metallized Polypropylene Capacitor
+ one Dayton Audio DMPC-25 25uF 250V Metallized Polypropylene Capacitor in parallel

50uF =
one Dayton Audio DMPC-50 50uF 250V Metallized Polypropylene Capacitor

3.83uF =
two Jantzen 1.5uF 800V Z-Superior Capacitor
+ one Jantzen 0.82uF 800V Z-Superior Capacitor
+ one Dayton Audio Film & Foil DFFC-0.01 0.01uF 400V By-Pass Capacitor

I also made a few other improvements:

1.) New binding posts to replace the original crappy ones.

2.) Added a Neutrik Speakon 8-pole connector to make an air tight connection from the top end and crossover to the bottom cabinet (to replace the hole filled with six single wires and a blob of putty). I saw where another AK'er used this method on his 4.5s. I used 13awg 8-conductor speaker cable from parts express. Also, the small hole in the top of the cabinet where the wires com through is a flimsy 1/8" piece of pressed hardboard glued and stapled over a much larger hole in the cabinet itself, this was full of gaps, so I reglued and sealed that up and added a piece of wood to fill the larger hole for a sturdier mounting point for the Neutrik connector.

3.) Caulked all the seams inside the lower cabinet, new woofer frame gasket, and added a front to back cross-brace.

4.) Added a toggle swith safety cover to protect that important "single-amp - bi-amp" mode switch on top. I always thought this was mis-labeled since you can biamp using the internal crossover and keep that switch in single-amp mode. It should only be in bi-amp mode when using an active external crossover with 2 amps. IMO it should be labeled "internal crossover - external crossover".

Having never done this before I was a little nervous when I hooked up and tested the first speaker. It's all pretty straightforward, but still there is room for making mistakes. Especially getting the 6 wires in the right sequence between the male and female parts of the Neutrik connector. A mistake there could be very bad. There was definitely a difference between the new left and the original right so I was anxious to get the right crossover done.

I am happy to say it was a complete success! I am somewhat surprised that everything sounds that much better -- especially since the old caps weren't that bad. More detailed cleaner highs, cleaner mids, tighter, deeper bass (probably partly due to sealing up the bottom cabinet) Everything good about these speakers just got better. I don't think they sounded this good when they were brand new. I just finished them yesterday and the more I listen the better they sound.

Larger before and after images here:
crossovers before
crossovers after
 

Attachments

  • rs_2_5_crossovers_before_labeled.jpg
    rs_2_5_crossovers_before_labeled.jpg
    77 KB · Views: 280
  • rs_2_5_crossovers_after.jpg
    rs_2_5_crossovers_after.jpg
    70.7 KB · Views: 248
  • orig_wires_through_cabinet.jpg
    orig_wires_through_cabinet.jpg
    78.5 KB · Views: 204
  • neutrik_connector.jpg
    neutrik_connector.jpg
    60.6 KB · Views: 233
  • switch_cover.jpg
    switch_cover.jpg
    66.9 KB · Views: 229
Last edited:
Honestly, the large non-polar electrolytic caps used by Infinity are specifically designed to take huge currents, and I'd be surprised if the Parts Express replacements could reliably deal with 1/2 the current that the originals can. Part of the reason I'm still using them on my 2.5's. The big film caps, however, are on my radar. I'll do it eventually.

I do like the Speakon connector. Nice.
 
Another tweak to do, that Speakon connector is nice. Although my RS 2.5s have
had some crossover work, I have not done much else.
 
Honestly, the large non-polar electrolytic caps used by Infinity are specifically designed to take huge currents, and I'd be surprised if the Parts Express replacements could reliably deal with 1/2 the current that the originals can. Part of the reason I'm still using them on my 2.5's.

That was a concern of mine too, but how do you know what the current handling rating is on the old 100v caps? Or the new ones? They are 100v rated at 200 watts each.

From PartsExpress:
"100V Non-Polarized Crossover Capacitors
These high quality Bi-polar (non-polarized) electrolytic capacitors are perfect for your next crossover design.
They feature a 5% dissipation factor and are rated at 100V which are effective to approximately 200 watts."


I don't know if this is correct for crossovers, but it's all I had to go by:

AMPS (current) = Watts / Voltage
I = P ÷ E
Example:
200w divided by 100v = 2 AMPS each
Would 3 in parallel give you 6 amps?
Is this totally wrong?
 
Last edited:
Congrats on your first speaker refurb/upgrade (it's an addiction from here on out!:sigh:).

Geoff in Seattle
 
That was a concern of mine too, but how do you know what the current handling rating is on the old 100v caps? Or the new ones?
You have to look at data sheets. The stock caps were likely custom made for Infinity, so you're unlikely to get much there. All you can do is make a guesstimate based on the physical size of the caps.
They are 100v rated at 200 watts each.

From PartsExpress:
"100V Non-Polarized Crossover Capacitors
These high quality Bi-polar (non-polarized) electrolytic capacitors are perfect for your next crossover design.
They feature a 5% dissipation factor and are rated at 100V which are effective to approximately 200 watts."


I don't know if this is correct for crossovers, but it's all I had to go by:

AMPS (current) = Watts / Voltage
I = P ÷ E
Example:
200w divided by 100v = 2 AMPS each
Would 3 in parallel give you 6 amps?
Is this totally wrong?
No, but can you trust the ratings given? A sales blurb is not a data sheet, and hell...we don't even know the manufacturer (who is probably is China, and we all know about the marvelous quality of electronic parts coming from that country).
 
Thanks echowars -

The original caps are marked T.I. (Texas Instruments I assume) made in Mexico

The only non-polar capacitors I came across that listed a ripple current specs were Nichicon 100uF (425ma) or 220uf (720ma) and they are even physically smaller than the Parts Express caps of the same value. 11 of the 100uF in parallel would give you about 4.7 amps, or 5 of the 220uF would give you 3.6 amps -- that seems like plenty to me.

From what I've heard Bennic is the manufacturer for Parts Express's non-polar electrolytics, and they all measured within 2% of their stated uF value, but I wasn't able to get any ripple current specs from them.

I've been trying to find info on whether the physical size of the capacitor matters, and everything I've come across says no, as long as voltage ratings are the same as the original.

In any case I kept those big old coke can caps if I need to put them back in.
 
Physical size does matter, as it directly relates to ESR and the ability to dissipate heat, which in turn go to overall current handling ability.
 
Physical size does matter, as it directly relates to ESR and the ability to dissipate heat, which in turn go to overall current handling ability.

I appreciate the info.
Considering the 700 and 1100uF caps were within 10% I will probably put them back in.
 
Hi rdp

The current handling capacity of a cap is given using the parameter "ripple current". The bigger computer can caps have ripple ratings in amperes compared to the small signal caps. Maybe you can check the datasheets for the values. I will be wary of the cap used in the 2 ohm vc circuit. Hope this helps :) Film caps that have big enough electrodes and are metallized may also work.

PS: Echo is right, like he always is. Also why I am choosing to "reform" caps in my RS4.5 - see my 4.5 thread.
 
Last edited:
if you use multiple smaller caps in general, you should be alright. for example, i would look at dividing large values into 5 smaller, pretty even values. that is effectively increasing the ripple current capacity by at least a couple factors.
 
The current handling capacity of a cap is given using the parameter "ripple current".

That's what I assumed, but then I read this at Erse Audio:

"The 3% df caps will exhibit better acoustic properties than ... and handle larger amounts of power due to it's lower ripple current."

-- which sounds like the opposite to me.

The Search Function is your friend. Here is a post by member PacificStereo wherein he details how he replaced the 1100uF caps with (11) 100uF caps in parallel, achieving high ripple current capability in the process.

I read that thread (wish the photos were still up) and that's what got me looking at the Nichicon 20% caps -- they were just about the only ones I found that listed ripple current in the specs. But then ended up going with the larger Parts Express caps.
 
I read that thread (wish the photos were still up) and that's what got me looking at the Nichicon 20% caps -- they were just about the only ones I found that listed ripple current in the specs. But then ended up going with the larger Parts Express caps.

Send PacificStereo a private message and ask him if he has copies of the pictures. From what I remember he arranged them pretty neatly in a circle with a buss running around the top.
 
So is it absolutely true had I used 7 200uf capacitors instead of 2 500uf and 1 400uf for the 1400uf on my QLS-1s, I'd have a better discernable bass response????
 
Send PacificStereo a private message and ask him if he has copies of the pictures. From what I remember he arranged them pretty neatly in a circle with a buss running around the top.

Ahh, not sure PacificStereo is on this board anymore. Search for him over at tapeheads.net.
 
Honestly, the large non-polar electrolytic caps used by Infinity are specifically designed to take huge currents, and I'd be surprised if the Parts Express replacements could reliably deal with 1/2 the current that the originals can. Part of the reason I'm still using them on my 2.5's. The big film caps, however, are on my radar. I'll do it eventually.

Could you please tell me if you are of the opinion that the use of 2 500uf and a 400uf Bennic capacitors in parallel to simulate an Infinity 1400uf cap is insufficient???

Thanks
 
I cannot say with any degree of authority without seeing data sheets, and that's unlikely to occur. My gut feeling (which is all I have to go on) is that the original large-can bipolar capacitor is probably superior unless it shows signs of leakage.

Too bad the big 100µf (and up) film caps are so damned expensive (and so big). They'd almost certainly be superior to any electrolytic solution if you could find a way to mount them.
 
Too bad the big 100µf (and up) film caps are so damned expensive (and so big). They'd almost certainly be superior to any electrolytic solution if you could find a way to mount them.

I drew it to scale just to see what the size (and cost) of all those 100uF film caps would be....
Yikes!
 

Attachments

  • 25x100uF.jpg
    25x100uF.jpg
    54.5 KB · Views: 150
Well, Solen has a 200µf. ;)

If I could afford the film caps, I'd build a crossover and put it in a separate enclosure...no more parts mounted to the back of the speaker.
 
Back
Top Bottom