advertisement


Replacing caps in speaker xover

Zombie

Zombie

pfm Member
Hello! I've acquired a pair of Sony APM-22ES speakers and I thought I'd replace at least the bipolar elcos in the xover.
There's one 15 uF on each filter.
I see from the service manual that there are also two polyester film caps, 6.8 and 10 uF, - probably MKT and two coils, 2.2 MMH and 0.47 MMH. What on earth is MMH? Is it milli-milli-Henry - mH or uH?
Anyway, replacing the bipolar is obvious, it always clears things up around the bass.
But what to do with the film caps and coils?
I'm tempted to replace them with smth sexy, like Mundorfs or others.
I appreciate the clear non-hifi but undistorted sound from these speakers, very much like my stats, and I don't want to make them hifi edgy in the treble.
What do you thinks a good MKP or a paper in oil would achieve?
Will replacing the coils with smth exotic make any worthwhile improvements?
And what goo should I use to decrease impact from vibrations? (Toying with idea to move the filters out of the box, too).
Thank you for any input.
 
Last edited:
MMH would normally mean milli-milli-henry, i.e. microhenry. But those values would not make sense if they were uH, and seem perfectly typical for mH.

Vintage Knob says there are three versions of the crossover for that speaker, one of which sounds good. A quick search did not reveal any details of the differences between the good one and the not-so-good ones.
 
Yes, maybe that's related to the second version with a dome tweeter. There are two addendums to the original service manual, but they don't mention any cap/coil changes apart from the tweeter, another circuit breaker and "network assy". So there is obviously other information floating around, but I haven't found any.
 
I wouldn't change the inductors. Their DCR forms part of the electrical damping factor for the woofer. Lower DCR reduces Qes, and will lean out the bass response. If your loudspeakers are bass-reflex, that could throw the port alignment as well.
 
One more silly question: Would combining bipolar elcos (ex. Jantzen EleCap) with a smaller premium MKP (ex. Jantzen standar Z-cap) - for instance a 10 uF capacitance with 8,2 uF bipolar + 1,8 uF MKP - be inferior to just using a 10 uF MKP?
 
Regarding your inductors, it's important to position them relative to each other to minimise crosstalk. The magnetic fields interact with each other raising the noise floor. With a lot of speakers this is overlooked to save money and speed up assembly by allowing all coils to lie the obvious easy way which is flat on the XO board. One of them need to be repositioned to stand on its rounded side such that if they were to be moved close together they would resemble two links of a common chain. Hope I haven't confused you.

The existing coils will be fine but you could coat them by painting them with a slow cure epoxy to minimise vibration. Also I think you mentioned moving them outside the box which will improve things further. Helps to place the XOs on some isolating material. I have recently discovered just how effective Vibrapods are for this.

I rather like the Jantzen Superior Z cap for the tweeter series cap. Something else often overlooked are resistors. I find the white sand cast resistors noisy, most have steel end plates. Replacing with Mills brings nice things to the party. Another resistor is a more affordable Mundorf for about 2 bucks each, I think.

If you're going this far perhaps consider rewiring internally with solid core wire.
 
Thank you for the advice. The coils are positioned like you describe. One is a BLM, German made I think, quite heavy, but probably not iron core, the other one is a smaller air core and is standing on its side. HP is third order, while LP is second order. I ordered Standard Z caps for the HP and a Jantzen EleCap for the LP. No resistors to worry about.
 
I just wonder what effect it'll have to replace the coils.
The 3rd harmonic distortion around 1 kHz is supposed to come from smth called cone breakup of the bass speaker.
APM-22ES distortion
How can this be remedied or made lower in a simple way? Are the coils involved in the speaker distortion, too, or is it neglible?
A similar model looks like this:
APM-33W-EU-Restored-Fabric-gen1_@-50cm-1.415V_In-between-TW_Waterfall-2.png
 
Zombie said:
The 3rd harmonic distortion around 1 kHz is supposed to come from smth called cone breakup of the bass speaker.
APM-22ES distortion
How can this be remedied or made lower in a simple way? Are the coils involved in the speaker distortion, too, or is it neglible?
A similar model looks like this:
APM-33W-EU-Restored-Fabric-gen1_@-50cm-1.415V_In-between-TW_Waterfall-2.png
Click to expand...
Cone breakup is a physical phenomenon, which occurs when the stiffness of the material is overcome by the acceleration forces on it. You can dampen the effect by applying doping, but that also changes the parameters of the driver because doping adds moving mass. The other way is to lower the crossover frequency, but that will depend on whether the tweeter/mid can cover the lower ranges. Most tweeters will struggle below 1kHz. This also requires a redesigned crossover network, which isn't simple or trivial.

The simple solution is to find a better loudspeaker.
 
The AP-22ES is actually quite good. I've got freshly refurbed ESL-988s behind them to compare with. ;)
I did a primitive measurement with REW and could replicate the 3rd harmonic distorsion around 1 kHz, but no trace of the breakup in the waterfall. This can mean two things, my waterfall is incorrect (but based on the same measurement data as the distortion) or it has no problem with cone breakup like the AP-33W on the pic above.
 
You must log in or register to reply here.


advertisement


Back
Top