AR-90 "Hi Board" Capacitor Replacement Done

[johnfalc]

Yesterday I got to replacing all the caps on the LMR, UMR, HF crossover board in my AR-90s.

I did this with the crossover boards in place. I found it quite easy to get adequate access by removing both woofers and using a small flexible neck lamp to give me great illumination.

I purchased the caps by mail from ERSE. They arrived promptly and all are 250V but in spite of this their sizes were suitable for replacement in the same locations occupied by the original caps. ERSE didn't have a 40 uF cap so I bought their 39 uF caps and parallelled with a couple of old GE nominal 1.2 uF caps I had already. The values in the attached screen clip include the 39+1.2 uF caps in parallel.

You'll see that the original caps that measured significantly differently from their nominal values all did so on the high side. One of the black and red caps, an 8 uF one, had started leaking a white substance, otherwise all the original caps looked fine.

I had planned to leave one speaker as it was and then measure for before and after comparison but decided to just get the project done and listen to music instead. I may get around to measuring the results eventually but have numerous other projects.

Do they sound different? Maybe, but I'm suspicious of my being susceptible to wanting to hear an improvement ... the truth is I enjoyed them before and do now also. If anything, based on a few hours of listening, I sense a bit more stable positioning of soloists, either vocal or instrumental. Anyway, I'm pleased that the replacement was easy and that I don't have to worry about capacitor failure for the tops of these speakers.

I'm driving them with a Sansui AU-D11 integrated amplifier I recently rescued (multiple issues but I really like the Sansui engineers' pushing of the envelope on this and the AU-D9 with their incorporation of output stage feed-forward and consider these two amps to be really lovely to listen to - I have one of each). Source material comes in directly from a Thorens+SME+Ortofon MC20 or via an Emotiva XDA-2 from CD, server, or MOG.

Now to decide how to thin out my AR inventory - I've got two pairs of 3a, a pair of 2ax, a pair of 92, a pair of 4ax, and or course, the 90s.

[Kuja]

Hello!

I have just stumbled upon this topic.

Thanks for posting this info!


I also have replaced all the original electrolytic caps in my AR90s with modern polypropylene Mundorfs (not the fancy expensive ones).


After reading many "electrolytic vs poly caps" discussions,
I have learned that I should have added resistors in series with my new poly caps, in order to mimic the behavior of the original caps.

The explanation is that the original electrolytic caps had much higher ESR in the old days and that the modern poly caps have much lower ESR.

The advice was to use 0.5 ohm audio grade resistors in series in order to come close to matching the ESR of the old caps.

Otherwise, the voicing of my restored speakers might be changed and they might sound too bright...

Others said that I shoud not worry.

So I did the restauration without the resistors, but I continued to worry if I did it wrong...

My AR90s are much less forgiving with bad recordings now.

They can sound really shrill with some poorly produced material,

but when the recordings are good, I'm in heaven!



Then I saw your measurements and I was very surprised to see that there is not that much difference in ESR between the old caps and the modern poly caps!






According to your measurements, the difference in ESR between the old elco caps and the new poly caps is only around 0.05 Ohms in average,

...and the "standard" advice that is repeated on many forums is to add 0.5 Ohms in series.


So, after all, I'm good?!



.

[Pete B]

Electrolytic ESR is strongly dependent on frequency and it takes some judgement

as to where to measure the ESR depending on where and how it is used in the crossover.

I've seen ESR as high as 1.2 ohms in high DF NPE caps.

Those numbers do not look right, or they might have been measured at 10 KHz or higher.

[DavidR]

Hello!

I have just stumbled upon this topic.

Thanks for posting this info!

I also have replaced all the original electrolytic caps in my AR90s with modern polypropylene Mundorfs (not the fancy expensive ones).

After reading many "electrolytic vs poly caps" discussions,

I have learned that I should have added resistors in series with my new poly caps, in order to mimic the behavior of the original caps.

The explanation is that the original electrolytic caps had much higher ESR in the old days and that the modern poly caps have much lower ESR.

The advice was to use 0.5 ohm audio grade resistors in series in order to come close to matching the ESR of the old caps.

Otherwise, the voicing of my restored speakers might be changed and they might sound too bright...

Others said that I shoud not worry.

So I did the restauration without the resistors, but I continued to worry if I did it wrong...

My AR90s are much less forgiving with bad recordings now.

They can sound really shrill with some poorly produced material,

but when the recordings are good, I'm in heaven!

I've read on several forums that after 100 or so hours of use the new caps will 'settle down'.

[RoyC]

I've read on several forums that after 100 or so hours of use the new caps will 'settle down'.

More likely, your ears will get used to the sound....

[Carlspeak]

More likely, your ears will get used to the sound....

What he said!

[DavidR]

Another myth busted

[johnfalc]

I used the little AnaTek BLUE ESR meter and, out of curiosity just took a look at its output - the pulses it emits for measurement occur at two (2) KHz.

Electrolytic ESR is strongly dependent on frequency and it takes some judgement

as to where to measure the ESR depending on where and how it is used in the crossover.

I've seen ESR as high as 1.2 ohms in high DF NPE caps.

Those numbers do not look right, or they might have been measured at 10 KHz or higher.