Recapping original Time Windows

[Jerry Koontz]

I have a original set of Time Windows.
I saw a video a couple of weeks ago of Steve Eberbach (time window designer) who said he would recap the Time Windows speakers.
The only schematics of the crossovers I could find where on Steve's patients. They look like they could be the design I have, not sure.
So digging into the crossover board, I found two values of electrolytics; four 47uF and one 10uF.

Here's my question;
When I think about recapping, I think about replacing the caps in the signal path. There are two pics showing Fig3 and Fig4 from the patient. Fig3 is the mid crossover and Fig4 is the tweeter crossover.
There are what appear to be poly type caps valued at 3uF. I was not planning on recapping these.
In Fig4, there are two 47uF caps that are more of a shunt filter. Do these need to be replaced? It seems like the only caps in the signal path are the 3uF caps? What are the advantages, if any, in replacing caps not in the signal path to the speaker?
The mid speaker, in Fig3, does have electrolytics in the signal path.

Thanks for the help,
Jerry

 

[slipperybidness]

I have a original set of Time Windows.
I saw a video a couple of weeks ago of Steve Eberbach (time window designer) who said he would recap the Time Windows speakers.
The only schematics of the crossovers I could find where on Steve's patients. They look like they could be the design I have, not sure.
So digging into the crossover board, I found two values of electrolytics; four 47uF and one 10uF.

Here's my question;
When I think about recapping, I think about replacing the caps in the signal path. There are two pics showing Fig3 and Fig4 from the patient. Fig3 is the mid crossover and Fig4 is the tweeter crossover.
There are what appear to be poly type caps valued at 3uF. I was not planning on recapping these.
In Fig4, there are two 47uF caps that are more of a shunt filter. Do these need to be replaced? It seems like the only caps in the signal path are the 3uF caps? What are the advantages, if any, in replacing caps not in the signal path to the speaker?
The mid speaker, in Fig3, does have electrolytics in the signal path.

Thanks for the help,
Jerry

Are you trying to solve a particular problem? Or do you just want to recap these as an insurance policy?

 

[Jerry Koontz]

Probably more of insurance. I have heard some high freq distortions in a small amount of listening, but it's not driving me crazy.
Recapping in the past has always provided positive results, so it's just something I do.
Jerry

 

[Swerd]

I'm not going to try to persuade you that there is no audible benefit in replacing the caps of a crossover that works. You seem to be convinced otherwise. But I have a few other thoughts.

That schematic leaves me scratching my head. I see what look like transformers, but no drivers.

A loudspeaker gets an AC signal. All components in the crossover are functioning parts of the network. It isn't as simple as defining a DC signal path.

If you are determined to work on this, why not also spend about $30 on a simple hand-held LC meter? You can use it to test the capacitors, before dismantling it, and see if they're OK. And if they measure within 10% of their marked value, I say…

If It Ain't Broke, Don't Fix It.

 

[Latent]

If you are determined to work on this, why not also spend about $30 on a simple hand-held LC meter? You can use it to test the capacitors, before dismantling it, and see if they're OK. And if they measure within 10% of their marked value, I say…

If It Ain't Broke, Don't Fix It.

Wish it was that easy to but you can't measure the value of a component like this while it is still soldered into a circuit board as then you are just measuring the board as a whole and you just get a rubbish figure. It can be a good estimate if you know the circuit well enough and measure the right component. To find the value of a component you would have to de-solder one end of it and then measure the component like this. The other end can remain in place. At least that would reduce the damage to one solder joint before you rip and replace a good capacitor. Once you measure a couple you will know what needs replacing.

 

[Jerry Koontz]

The drivers in Fig3 are 46 and 48. The drivers in Fig4 are 64 and 66. All the drivers are represented by inductors. The signal path I was referring to is the AC signal path, not DC.
When I recap any crossover, I look for the old electrolytic caps that probably have dried out dielectrics that are in the direct AC path to the driver.
My question was more about the caps that shunt to GND, like the higher order crossovers, zobel networks, notch filters and any other cap that would reduce frequencies to the driver.

 

[TLS Guy]

The drivers in Fig3 are 46 and 48. The drivers in Fig4 are 64 and 66. All the drivers are represented by inductors. The signal path I was referring to is the AC signal path, not DC.
When I recap any crossover, I look for the old electrolytic caps that probably have dried out dielectrics that are in the direct AC path to the driver.
My question was more about the caps that shunt to GND, like the higher order crossovers, zobel networks, notch filters and any other cap that would reduce frequencies to the driver.

I think the caps are the least of your problems.

I suspect a number, if not all, of those old pots are your problem.

I know those speakers have their advocates, every lousy speaker ever produced has, and that one is no exception.

That crossover is nuts and causes phase and delay changes that is supposed to make you think there is more ambiance than there is. First they had lousy Philips tweeters that blew, and then they has slightly better Peerless tweeters.

Those speakers had very low production runs, as most were not fooled. I regard those speakers as more of a contraption then a serious sound reproducer, but to each his own.

 

[Swerd]

Wish it was that easy to but you can't measure the value of a component like this while it is still soldered into a circuit board as then you are just measuring the board as a whole and you just get a rubbish figure. It can be a good estimate if you know the circuit well enough and measure the right component. To find the value of a component you would have to de-solder one end of it and then measure the component like this. The other end can remain in place. At least that would reduce the damage to one solder joint before you rip and replace a good capacitor. Once you measure a couple you will know what needs replacing.

Thanks for pointing that out. I readily admit my ignorance on this. I do have a small hand-held LC meter, but have never tried using it in a soldered network as I (too) optimistically suggested above. It, however, does work well at checking capacitors and inductors before I assemble networks.

The meter allows me to unwind larger inductors to achieve a smaller desired value. And, it has convinced me that inexpensive polypropylene caps, such as those sold by Parts Express, Madisound, or Solen, are almost always exactly as their printed label says. It doesn't seem to matter if the tolerance is ±10% or ±1%. The cheaper non-polar electrolytic caps may also be within 10% of their printed value, or not, depending on the source. For what its worth, I have never found an older NPE cap in a speaker crossover (in the range of 8 µF or less) that actually was out of spec, even if they were over 30 years old.

That's why I don't see any benefit from recapping older crossovers.

 

[Jerry Koontz]

I certainly agree that if I compared a new electrolytic cap to a new poly cap, I would not be able to tell any difference by the ear.
I have recapped a couple of different crossovers, replacing 20 or 30 year old electrolytics with poly caps and the difference was very noticeable. I'll typically recap one crossover, then compare the two speakers side by side to the verify results.
The old electrolytic's have been explained to me as having dried out dielectrics that produce harsh sounds in the tweeter, having the cap directly in the signal path. I've only worked with first and second order crossovers. I can't remember anyone saying that they recap due to caps changing values. The sonic quality is the only reason I've experienced.
I would also agree that I could just replace the old electrolytic with a new electrolytic and get positive results, but if the dielectric are going to dry out, why not just fix the problem. I have not heard of anyone having to replace a poly cap for any reason?

 

[Swerd]

I have recapped a couple of different crossovers, replacing 20 or 30 year old electrolytics with poly caps and the difference was very noticeable. I'll typically recap one crossover, then compare the two speakers side by side to the verify results.
The old electrolytic's have been explained to me as having dried out dielectrics that produce harsh sounds in the tweeter, having the cap directly in the signal path.

When old NPE caps do go out of spec, it is because they dry out and their capacitance value drifts, usually to lower values. These out-of-spec capacitance values affect crossover function to the point where it doesn't work as intended.

Don't assume an old cap must be out-of-spec because it is old. Measure it and know for sure. Not all NPE caps dry out.

If a crossover capacitor works as originally intended, then there is no reason why exchanging it for a new one will make an audible difference.

I do agree with TLS Guy when he said the pots (or variable L pads) in your speakers are much more likely to go bad with age.

I have not heard of anyone having to replace a poly cap for any reason?

None that I know of.

If I am building a new crossover network for a DIY speaker, I always choose inexpensive polypropylene caps over NPE caps. But if the question is, like yours, about replacing the original caps in an older speaker, the answer is a bit different. Find out what's producing the noise or distortion you hear, and then address it. It could be the tweeters, or other drivers going into breakup, the aging pots, or something else completely different.