Recapped one of my 3As today

[Mexicomike]

Today I recapped one of my AR3as and compared it to the other with completely original caps. I have to say that I was happy with the sound of the speakers and I undertook the recap more to ensure I didn't hurt a driver somewhere down the road.

After closing it back up I moved the speakers together and using my AR amp A/B'd the two for two hours. The difference was not instantly dramatic but was noticeable. The recap was more detailed in the treble and tighter in the bass. Both speakers went equally low as verified with some test tones in the 30-39 HZ range but the recap had better control of bass drum/guitar. In comparison, the uncapped speaker had a bit of boominess. In the upper end, the recapped version was more articulate, providing better detail - cymbals startlingly so!.

I used CDs of a variety of material. Although the difference were most apparent in the low bass and high end, overall the oem speaker sounded veiled in comparison.

All original caps remain in place with their wires unsoldered at the terminals and coiled so that the original lengths of wire remain attached to the old caps. The new caps can be removed and the originals re-connected with no indication that it was ever touched.

In one of my first posts here I talked about comparing the 3a to my pair of Stax SR30 headphones which I bought originally because they sounded like AR3a speakers. I compared the Stax to the recapped speaker - stunningly similar! The only notable difference was, of course the soundstage since they are headphones and I was using only one speaker. But the sound quality was essentially identical after fooling with the mid/treble controls. I found that the "match" came with the mid and tweeter at around 10 oclock - of course what that setting actually translates to resistance-wise, I don't know. So based on the ASSUMPTION that the since I bought the Stax's new because they sounded like AR3as in the 80's when AR3as were still relatively new, I'd say that this cap method essentially matches the sound of a pair of new AR3a's.

Of course, this could still be flawed because just as the speakers "aged" so have the headphones. Do THEY sound like they did when new? I can't say for sure but there are fewer internal components to degenerate so it seems logical to assume they are less affected by aging. In any case, I was very pleased with the recap results.

For anyone interested, I used caps per the following recommendation I received via email from one of the folks on this site who participated in the AR3a rehab paper. Since I didn't tell him I was going to repost a portion of his email to me, I don't think it would be correct for me to name names! But I want to make it clear that all of this credited to someone else, not me! I used OPTION B for the woofer.

Tweeter:

Add a 0.33-Ohm, non-inductive resistor in series with a 6-uF polypropylene capacitor

Parts Express: 6.2 uF, Dayton Brand, # 027-427, $2.15 each

Parts Express 0.33 Ohm, non-inductive resistor, $004-0.33, $1.38 each  ($1.08 for four or more)

Midrange:

Add a 0.33-Ohm, non-inductive resistor in series with a 50-uF polypropylene capacitor

Parts Express: 50-uF, Dayton Brand, # 027-443, $11.29 each

Parts Express 0.33 Ohm, non-inductive resistor, $004-0.33, $1.38 each  ($1.08 for four or more)

Woofer (a or :

(a) 150-uF non-polar electrolytic with a small, say, 0.47 uF polypropylene film capactior in parallel to deal with its inductance at high frequencies

Parts Express: 150-uF, non-polarized, # 027-364, $2.55 each

Parts Express: 0.47-uF, Dayton (house brand), # 027-447, $1.38 each

( 150-uF poly propylene capacitor with a 0.51-Ohm non-inductive resistor in series

Parts Express:   50-uF, Dayton, # 027-443, $11.29 each

Parts Express: 100-uF, Dayton, # 027-447, $19.51 each

  (wire the two capacitors in parallel to obtain (150 uF)

Parts Express 0.47 Ohm, non-inductive resistor, $004-0.47, $1.38 each

[Guest matty g]

Today I recapped one of my AR3as and compared it to the other with completely original caps. I have to say that I was happy with the sound of the speakers and I undertook the recap more to ensure I didn't hurt a driver somewhere down the road.

After closing it back up I moved the speakers together and using my AR amp A/B'd the two for two hours. The difference was not instantly dramatic but was noticeable. The recap was more detailed in the treble and tighter in the bass. Both speakers went equally low as verified with some test tones in the 30-39 HZ range but the recap had better control of bass drum/guitar. In comparison, the uncapped speaker had a bit of boominess. In the upper end, the recapped version was more articulate, providing better detail - cymbals startlingly so!.

I used CDs of a variety of material. Although the difference were most apparent in the low bass and high end, overall the oem speaker sounded veiled in comparison.

All original caps remain in place with their wires unsoldered at the terminals and coiled so that the original lengths of wire remain attached to the old caps. The new caps can be removed and the originals re-connected with no indication that it was ever touched.

In one of my first posts here I talked about comparing the 3a to my pair of Stax SR30 headphones which I bought originally because they sounded like AR3a speakers. I compared the Stax to the recapped speaker - stunningly similar! The only notable difference was, of course the soundstage since they are headphones and I was using only one speaker. But the sound quality was essentially identical after fooling with the mid/treble controls. I found that the "match" came with the mid and tweeter at around 10 oclock - of course what that setting actually translates to resistance-wise, I don't know. So based on the ASSUMPTION that the since I bought the Stax's new because they sounded like AR3as in the 80's when AR3as were still relatively new, I'd say that this cap method essentially matches the sound of a pair of new AR3a's.

Of course, this could still be flawed because just as the speakers "aged" so have the headphones. Do THEY sound like they did when new? I can't say for sure but there are fewer internal components to degenerate so it seems logical to assume they are less affected by aging. In any case, I was very pleased with the recap results.

For anyone interested, I used caps per the following recommendation I received via email from one of the folks on this site who participated in the AR3a rehab paper. Since I didn't tell him I was going to repost a portion of his email to me, I don't think it would be correct for me to name names! But I want to make it clear that all of this credited to someone else, not me! I used OPTION B for the woofer.

Tweeter:

Add a 0.33-Ohm, non-inductive resistor in series with a 6-uF polypropylene capacitor

Parts Express: 6.2 uF, Dayton Brand, # 027-427, $2.15 each

Parts Express 0.33 Ohm, non-inductive resistor, $004-0.33, $1.38 each  ($1.08 for four or more)

Midrange:

Add a 0.33-Ohm, non-inductive resistor in series with a 50-uF polypropylene capacitor

Parts Express: 50-uF, Dayton Brand, # 027-443, $11.29 each

Parts Express 0.33 Ohm, non-inductive resistor, $004-0.33, $1.38 each  ($1.08 for four or more)

Woofer (a or :

(a) 150-uF non-polar electrolytic with a small, say, 0.47 uF polypropylene film capactior in parallel to deal with its inductance at high frequencies

Parts Express: 150-uF, non-polarized, # 027-364, $2.55 each

Parts Express: 0.47-uF, Dayton (house brand), # 027-447, $1.38 each

( 150-uF poly propylene capacitor with a 0.51-Ohm non-inductive resistor in series

Parts Express:   50-uF, Dayton, # 027-443, $11.29 each

Parts Express: 100-uF, Dayton, # 027-447, $19.51 each

  (wire the two capacitors in parallel to obtain (150 uF)

Parts Express 0.47 Ohm, non-inductive resistor, $004-0.47, $1.38 each

Mike -

Thanks for posting this. I'm about to re-cap/re-edge my AR5s and I'm going to follow your example on the re-cap. I was planning to use the Dayton caps also - they seem like a pretty good alternative to the expensive exotics.

Thanks also for the interesting and informative posts on the AR amp - an excellent reference for the future!

Matt

[Mexicomike]

Finished the other one - thought I'd include a photo of how I mounted the crossover caps in the speakers:

[Rich W]

Woofer (a or :

(a) 150-uF non-polar electrolytic with a small, say, 0.47 uF polypropylene film capactior in parallel to deal with its inductance at high frequencies

Parts Express: 150-uF, non-polarized, # 027-364, $2.55 each

Parts Express: 0.47-uF, Dayton (house brand), # 027-447, $1.38 each

It seems that I'm always asking another question on this forum! I've recapped my AR-3a using the same overall combination of capacitors and resistors, except for Solens caps. Of course, I'm very happy with the results.

The 0.47 parallel film capacitor on the woofer has me intrigued, though. In one-third octave tests I've run on my 3a's, I've noticed a 3 db peak centered at 800 hz, spanning about a half an octave. It's easy enough to get rid of with an equalizer, but I'm wondering if the addition of a parallel cap would help mitigate this peak. I don't believe it to be a room acoustic issue as my AR-7's do not have this peak when placed in the same location. Is it possible that my woofer is not rolling off fast enough? I did upgrade to the #9 inductor.

Is follow-up surgery in order for my AR-3a's?

Best Regards,

Rich W

[RobertAR3A]

Today I recapped one of my AR3as and compared it to the other with completely original caps. I have to say that I was happy with the sound of the speakers and I undertook the recap more to ensure I didn't hurt a driver somewhere down the road.

After closing it back up I moved the speakers together and using my AR amp A/B'd the two for two hours. The difference was not instantly dramatic but was noticeable. The recap was more detailed in the treble and tighter in the bass. Both speakers went equally low as verified with some test tones in the 30-39 HZ range but the recap had better control of bass drum/guitar. In comparison, the uncapped speaker had a bit of boominess. In the upper end, the recapped version was more articulate, providing better detail - cymbals startlingly so!.

I used CDs of a variety of material. Although the difference were most apparent in the low bass and high end, overall the oem speaker sounded veiled in comparison.

All original caps remain in place with their wires unsoldered at the terminals and coiled so that the original lengths of wire remain attached to the old caps. The new caps can be removed and the originals re-connected with no indication that it was ever touched.

In one of my first posts here I talked about comparing the 3a to my pair of Stax SR30 headphones which I bought originally because they sounded like AR3a speakers. I compared the Stax to the recapped speaker - stunningly similar! The only notable difference was, of course the soundstage since they are headphones and I was using only one speaker. But the sound quality was essentially identical after fooling with the mid/treble controls. I found that the "match" came with the mid and tweeter at around 10 oclock - of course what that setting actually translates to resistance-wise, I don't know. So based on the ASSUMPTION that the since I bought the Stax's new because they sounded like AR3as in the 80's when AR3as were still relatively new, I'd say that this cap method essentially matches the sound of a pair of new AR3a's.

Of course, this could still be flawed because just as the speakers "aged" so have the headphones. Do THEY sound like they did when new? I can't say for sure but there are fewer internal components to degenerate so it seems logical to assume they are less affected by aging. In any case, I was very pleased with the recap results.

For anyone interested, I used caps per the following recommendation I received via email from one of the folks on this site who participated in the AR3a rehab paper. Since I didn't tell him I was going to repost a portion of his email to me, I don't think it would be correct for me to name names! But I want to make it clear that all of this credited to someone else, not me! I used OPTION B for the woofer.

Tweeter:

Add a 0.33-Ohm, non-inductive resistor in series with a 6-uF polypropylene capacitor

Parts Express: 6.2 uF, Dayton Brand, # 027-427, $2.15 each

Parts Express 0.33 Ohm, non-inductive resistor, $004-0.33, $1.38 each  ($1.08 for four or more)

Midrange:

Add a 0.33-Ohm, non-inductive resistor in series with a 50-uF polypropylene capacitor

Parts Express: 50-uF, Dayton Brand, # 027-443, $11.29 each

Parts Express 0.33 Ohm, non-inductive resistor, $004-0.33, $1.38 each  ($1.08 for four or more)

Woofer (a or :

(a) 150-uF non-polar electrolytic with a small, say, 0.47 uF polypropylene film capactior in parallel to deal with its inductance at high frequencies

Parts Express: 150-uF, non-polarized, # 027-364, $2.55 each

Parts Express: 0.47-uF, Dayton (house brand), # 027-447, $1.38 each

( 150-uF poly propylene capacitor with a 0.51-Ohm non-inductive resistor in series

Parts Express:   50-uF, Dayton, # 027-443, $11.29 each

Parts Express: 100-uF, Dayton, # 027-447, $19.51 each

  (wire the two capacitors in parallel to obtain (150 uF)

Parts Express 0.47 Ohm, non-inductive resistor, $004-0.47, $1.38 each

Mike,

I am in the process of recapping my Ar3a's. I just took out the 150uF cap, and one of its terminals is marked "+", whereas the replacement, which is Parts Express: 150-uF, non-polarized, # 027-364, has no such marking. Does it matter which wire on the new cap I connect to the wire that was connected to the old cap's "+" terminal ?

Thanks.

[RoyC]

Woofer (a or :

(a) 150-uF non-polar electrolytic with a small, say, 0.47 uF polypropylene film capactior in parallel to deal with its inductance at high frequencies

Parts Express: 150-uF, non-polarized, # 027-364, $2.55 each

Parts Express: 0.47-uF, Dayton (house brand), # 027-447, $1.38 each

It seems that I'm always asking another question on this forum! I've recapped my AR-3a using the same overall combination of capacitors and resistors, except for Solens caps. Of course, I'm very happy with the results.

The 0.47 parallel film capacitor on the woofer has me intrigued, though. In one-third octave tests I've run on my 3a's, I've noticed a 3 db peak centered at 800 hz, spanning about a half an octave. It's easy enough to get rid of with an equalizer, but I'm wondering if the addition of a parallel cap would help mitigate this peak. I don't believe it to be a room acoustic issue as my AR-7's do not have this peak when placed in the same location. Is it possible that my woofer is not rolling off fast enough? I did upgrade to the #9 inductor.

Is follow-up surgery in order for my AR-3a's?

Best Regards,

Rich W

Rich,

Adding a .47uf cap will not make a bit of difference regarding your bass response issue. How much, and what type of cabinet stuffing did you use?

We debated the inclusion of the added resistors and small by-pass cap tweaks in the 3a restoration guide when we were putting it together. While there *may* be some very slight (unmeasurable) benefits to the addition of these components, they are not necessary for a very satisfactory restoration.

Roy

[RoyC]

Mike,

I am in the process of recapping my Ar3a's. I just took out the 150uF cap, and one of its terminals is marked "+", whereas the replacement, which is Parts Express: 150-uF, non-polarized, # 027-364, has no such marking. Does it matter which wire on the new cap I connect to the wire that was connected to the old cap's "+" terminal ?

Thanks.

It makes no difference. What caps were in your 3a's?

Roy

[genek]

I have to say that I was happy with the sound of the speakers and I undertook the recap more to ensure I didn't hurt a driver somewhere down the road.

Does recapping a classic AR really do that much to protect drivers? I know some other brands have crossovers with steep enough slopes to almost completely cut off frequencies outside intended ranges, but classic AR crossovers seem to be meant more to shape natural driver rolloffs. The most common scheme I see for protecting MR and HF drivers usually involves parallel inductors, which is what AR did with the 10pi and 11 in the ADD series.

[RoyC]

Does recapping a classic AR really do that much to protect drivers? I know some other brands have crossovers with steep enough slopes to almost completely cut off frequencies outside intended ranges, but classic AR crossovers seem to be meant more to shape natural driver rolloffs. The most common scheme I see for protecting MR and HF drivers usually involves parallel inductors, which is what AR did with the 10pi and 11 in the ADD series.

Hi Gene,

Capacitors always offer some protection.

Even if a high frequency driver cannot adequately reproduce lower frequencies, it still can be damaged by them. The greatest advantage to re-capping is to replace worn caps which have drifted out of spec (typically higher), and can no longer provide the same protection from lower frequencies. The sound will also be affected to varying degrees by changing capacitor values. I agree that the 3a tweeter's natural steep roll-off will probably be less affected sound-wise with its crossover cap value increased (or even removed) than any other tweeter on the planet....but it would still be more vulnerable to damage.

Parallel inductors do provide steeper crossover slopes, and offer more protection, but work in tandem with capacitors to do so.

Roy