PDA

View Full Version : Is there any benefit in replacing electrolytic capacitors with PP?



lofi-ear
12-06-2009, 12:17 PM
My speakers are about 10 years old. Here's the crossover network http://manuals.harman.com/JBL/HOM/Technical%20Sheet/Ti10k%20ts.pdf and let's assume the following capacitors 33uF, 39uF, 120uF, 47uF, 100uF and 120uF, are electrolytic (I'm not sure if they actually are, so if anyone knows, I'd be happy to know :) ).

The question is, do I gain anything soundquality-wise by replacing those capacitors with polypropylene ones?

jcrobso
12-07-2009, 03:23 PM
The only one that is not is the 6.6uf for the HF driver, but it is a small value. Most likely it is a PP.

whizzer
12-08-2009, 08:01 AM
The question is, do I gain anything soundquality-wise by replacing those capacitors with polypropylene ones?

One double-blind test after another has shown that NPE's cannot be reliably differentiated from other capacitor types by ear, regardless of how "golden" the ear believes itself to be. However, one consideration is longevity. Electrolytics dry up and fail with age--about 20 years or so. You have some time left to make a decision.

Ian Mackenzie
12-09-2009, 04:01 AM
These capacitors are already bypassed by selected smaller values so it may not be worthwhile on sound quality terms.

If the previous reply were accurate JBL would not a bother bypassing.

These non polarised electros can however vary by as much as 50% on the stated value so you may wish to buy a bunch of each value. Then measure and match.

lofi-ear
12-29-2009, 04:42 AM
Thank you for your responses.

What wonders me is that all the electrolytic capacitors are rated as high as 400V! Do they really have to be that high? Doesn't 100VDC suffice? (pretty much all hifi-shops here sell only max 100VDC bipolar capacitors)

All the polypropylene caps (bypass and the ones in the tweeter network) are average SCR/Solen capacitors, could it be worth upgrading at least those in the tweeter network to something better, such as Mundorf M-Cap Supreme? I'm not saying that the speakers sound bad, just that because the cabinets and drivers are of first class, there must be some room for improvements ;)

Does it matter if I select a smaller/larger value for the bypass caps, as long as the total capacitance stays the same?

Allanvh5150
12-29-2009, 12:43 PM
The capacitors dont need to be 400 volt, 100 volt is fine. The bypass caps need to be small as thier job is to cancel the inductance of the larger capacitors. 0.01uf is good. Or you could change the whole network to charge coupled.

Allan.

lofi-ear
12-31-2009, 06:39 PM
The capacitors dont need to be 400 volt, 100 volt is fine. The bypass caps need to be small as thier job is to cancel the inductance of the larger capacitors. 0.01uf is good. Or you could change the whole network to charge coupled.

Allan.

OK, here's what I have decided to do. First, I will keep the original values for capacitors. I trust the engineers at JBL are better than I at crossover design
(network schematics again: http://manuals.harman.com/JBL/HOM/Technical%20Sheet/Ti10k%20ts.pdf )

The tweeter network
- add a 0.10uF Audyn Cap KP-SN 250V bypass cap to each 6.8uF SCR polypropylene cap(http://www.riviera-acoustics.com/catalog/default.php?cPath=46_35_57_56&language=en). Whether or not this any effect on sound will remain to be heard.

The midrange network:
- Replace the 18uF electrolytic cap with 10uF+8.2uF=18.2uF polyester caps (all polyester cap tolerances are +-5) and leave the 1.0uF SCR bypass cap in place.
- Replace the 33uF electrolytic cap with a 33uF polyester cap.
- Replace the 39uF electrolytic cap with 22uF+10uF+6.8uF=38.8uF polyester caps and leave the 1.0uF SCR bypass cap in place.

By combining several smaller caps (with varying tolerances) I have better chances of getting the desired final cap value

The midbass network:
- Replace the 120uF electrolytic cap with 100uF+22uF electrolytic caps and leave the bypass cap alone. It would get too expensive to replace this with a non-electrolytic cap.
- Replace the 47uF electrolytic cap with a 47uF polyester cap.
- Replace the 33uF electrolytic cap with a 33uF polyester cap.

The lowbass network:
- Replace the 100uF electrolytic cap with another electrolytic cap of the same value. Again, a polyester version would be too expensive. (Maybe later?)

So, how does this sound? Should I also get some extra resistors due to lower ESR of the polyester caps, compared to electrolytic?

Allanvh5150
12-31-2009, 07:11 PM
Hi there. The caps in this crossover will definately benefit in replacement. Larger values do get expensive though. I am thinking that there are no actual bypass caps in this schematic. 1uf is too large to be a good bypass. Instead I would take the total value of capcitors, i.e. 18+1, 39+1 etc, and get some 0.01 polystyrene capacitors and bypass everything. The smaller the bypass cap the better. This value will not add significant value to the circuit. As for the resistors, I would leave them as they are or you could replace them with non inductive varieties.

Allan.

lofi-ear
01-08-2010, 08:35 AM
Got all the replacement capacitors today. I went to measure them, and all the polyester MKT's were within 1.5% tolerance. Pretty good, even though they have +-5% marking on them.

The electrolytes, on the other hand, left me a question - they are all about 15% over their marked capacitance. Does this mean that when they slowly burn in, they lose some capacitance and settle to their nominal values?

Dan Elleson
07-21-2010, 05:41 AM
Thank you for your responses.

What wonders me is that all the electrolytic capacitors are rated as high as 400V! Do they really have to be that high? Doesn't 100VDC suffice? (pretty much all hifi-shops here sell only max 100VDC bipolar capacitors)

All the polypropylene caps (bypass and the ones in the tweeter network) are average SCR/Solen capacitors, could it be worth upgrading at least those in the tweeter network to something better, such as Mundorf M-Cap Supreme? I'm not saying that the speakers sound bad, just that because the cabinets and drivers are of first class, there must be some room for improvements ;)

Does it matter if I select a smaller/larger value for the bypass caps, as long as the total capacitance stays the same?

Just passing this and thought I may add.. consider the kick back voltage spikes caused by the voice coil inductance etc and also consider that the high voltage caps may have been chosen for their size, which is inversely propotional to ESR, maaayyybe, or could just be due to bulk ordering of caps at the time.

jcrobso
07-21-2010, 07:58 AM
Got all the replacement capacitors today. I went to measure them, and all the polyester MKT's were within 1.5% tolerance. Pretty good, even though they have +-5% marking on them.

The electrolytes, on the other hand, left me a question - they are all about 15% over their marked capacitance. Does this mean that when they slowly burn in, they lose some capacitance and settle to their nominal values?

Given sufficient time electrolytic caps can decrease to a -90% of the original value!:eek:

tomt
10-19-2010, 05:08 PM
read this here -


http://www.reliablecapacitors.com/pickcap.htm