XO ESR resistors with capacitors in parallel

I've been re-capping my Celestion Ditton 44's and followed to advice on the "Crossover nightmare" and "Celestion 66 needs mid-range" threads.

For the bass section of the filter the recommended ESR value is 1R with each of the 68 and 75 uF caps.

However, I made up these values as follows:-
68uF: two 22uF Jantzen Cross Caps, one 2uF Cross Cap and one 22uF Mundorf Ecap.
75uF: one 27uF Cross Cap, one 1uF Z Cap and one 47uF Ecap.

The resistors are Mundorf MOX, 5watt.

For neatness and space reasons I put all the caps and resistors on a separate board which hangs off the XOs (which are still mounted on the rear of the speakers till I've finalised everything).

Bass sounds very good. Not boomy but very deep and detailed.

So the question is should I have fitted a separate ESR resistor in series with each cap? If so, what values should I use?
Maybe it wouldn't make any difference, but just want to clarify the correct method.
Thanks Gordon.
 
I'd be very surprised if anyone could answer this question.

To know the exact values you'd need - you'd need to know the crossover points, and the difference between the ESR of the original capacitors, compared to your new ones.

Another thing to consider is if the drivers are still in spec and if the crossover was spot-on in the first place.

It wouldn't do any harm to try a small value resistor in series with the caps to see if it improves the sound (say 1/2 ohm to 1 ohm). Try it, and have a good listen.
 
However, I made up these values as follows:-
68uF: two 22uF Jantzen Cross Caps, one 2uF Cross Cap and one 22uF Mundorf Ecap.
75uF: one 27uF Cross Cap, one 1uF Z Cap and one 47uF Ecap.
I don't see the point in your choice. If you want to get rid of the electrolytic caps in the original crossover, I'd choose only MKP or MKT as a substitute. Using the Ecaps you are re-inserting the ageing problem. The other logical move would have been to use only new electrolytic caps, so to match more or less the original ESR. Having said that, I expect the ESR of a 68-75uF electrolytic cap to be very low, in the order of 0.5 Ohm or even less. If you want to try, use a 0R5 resistor and hear what happens.

Ralf

68uF: is a standard value, if not available for some reasons I'd have used 2x33uF in parallel (less than 3% error), but one bigger cap cost always less than 2 or 3 smaller value caps
75uF: not standard value, 47+27uf is a very good approximation
 
I decided to sim this. So I chose a 200Hz Butterworth slope, made with 9mH and 70uF. The pink trace is that slope, the brown trace is when I use 1 ohm in series with the capacitor. The response goes down 1/2dB at the cross, and 1dB lower than that.

So I looked at what it would take to return to the original response. I reduced the inductance to 8.5mH. I guess that if you didn't add the resistance and wanted to emulate that you'd go the other way and add 0.5mH.
 

Attachments

  • Screenshot from 2020-06-28 20-57-46.png
    Screenshot from 2020-06-28 20-57-46.png
    15.1 KB · Views: 128
rephrase question

Thanks for the prompt replies.
Let me just rephrase the question.
Is it better to use separate ESR resistors for each cap when you have to use different cap values in parallel?
Obviously one single resistor works fine but will separate ones be better?
Is there a convention?
Thanks
Gordon.
 
Is it better to use separate ESR resistors for each cap when you have to use different cap values in parallel?
No. You treat all the paralleled caps as a single cap, and after the bunch (or before, it is the same) you connect a single resistor.

But I'd like to know why you mixed MKP and electrolytic caps to obtain the desired value, I simply don't get the point.

Ralf
 
Now I see the question differently. In some circumstances (much less with speakers) parallel capacitors can interact creating the potential for resonance which separate resistors might help to damp. Of course this is a secondary effect (primarily as giralfino says, target the resistance at the crossover itself).
 
Thanks for the replies giralfine & allen, it more or less answers my query. I'm just going to leave things as they are.

Long story: The reason I used mixed MKP and electros is because I bought a bunch of both over a period of months to get the correct values for the MF section. The 44's have different value input cap on the MF section (seems Celestion experimented with two different mid driver magnets and cap values before settling on one arrangement). Mine were 28uF but this did not sound good when I first replaced the Elcaps with Cross Caps. Eventually I settled on 33uF as giving the best balance, along with trying different value ESR resistors there. I then read somewhere that the Mundorf Ecaps sounded very good in 44's so tried them and they were closer to the original Elcaps but much clearer.
They were not that expensive and I thought logically (or not) that electros would be a better match as that was what Celestion fitted, and would be more likely to retain the "Celestion sound". They did.

However, after a couple of months I thought maybe I can improve the mids even more and bought Clarity Caps CSA's, which were a big step up in SQ. And improving by the day as they burn in!

So I ended up with a bunch of Cross Caps and Ecaps. I had used a whole load of different value of mostly Ecaps for the bass section just to get it working in a reasonable fashion before settling on the values I mentioned in first post, mounted on the separate boards. This was a big improvement and I'm very happy with the bass now. I do agree that the sound is more integrated when you use only one cap of the required value (or multiples of the same to get the final value), but to be honest, I baulked at buying even more caps just to get them all the same type for the bass filter and just used what I already had.

I know electros don't last forever, but I've had these speakers since 1975 so new electros will out live me! lol. Maybe I'll treat myself to some better matching caps for the bass section as an xmas pressie.

Finally, just to complete the story, when I decided that the 44's were not sounding great I could have just bought new speakers, but these have been with me all my adult life so I couldn't just ditch them, they're part of me. I'm so glad I went down this road as I've learnt so much about XO's and speaker design. It's been a very satisfying experience, and a big thanks to everybody on this forum for the help in getting where I am. Also, I doubt I could get equally good sounding new speakers without having to spend a fortune. I'm blown away with how good they now sound, so much better than even when they were new. It's like having a new music collection, I am hearing things I never heard before and not just the micro details, but the sound stage and the timbre of instruments.

I'm sure it must be possible to improve them a bit more yet but it's the old case of diminishing returns, so I'm just going to stick with things as they are for the mean time.

Really, what I've just described is only part of the story but it will suffice for this thread.
Once again, thanks to everybody.
Gordon.