Hotrodding the UCD modules

Status
Not open for further replies.
Jump to Latest
#141
Bruno Putzeys said:


Something that's on the edge of being too large but still doesn't produce too much adverse effects (and sounds great) is BC components MKP379, 0.68u, 160V. You will need to mount it on the bottom side of the board.
Click to expand...


Thanks for the tip, Bruno.
Unfortunately, the MKP379 caps are impossible to find in retail. I tried most electronic part distributors in the US listed on Vishay's site, and no luck. I also sent a few e-mails. DigiKey told me they don't carry these parts at all. I wonder if they're discontinuing these series... Do you know of a source for these parts or an alternative cap series? Anywhere in the world...
I'd really like to optimise the output of the modules
Thanks.
 
#142
Julien_M said:
Unfortunately, the MKP379 caps are impossible to find in retail.(...) Do you know of a source for these parts or an alternative cap series? Anywhere in the world...
Click to expand...
Hmm. They have indeed been taken out of production. It looks like the MKP479 is the same capacitor element in a dipped epoxy housing. Bend the leads straight and mount as close to the pcb as possible. On the signal side you can use the hole of the original cap. On the gnd side you can solder to one of the ground faston lugs. If you can manage to mount the cap on the top side (so it will sit between the 2 fastons marked GND), do so, because in that way the ground plane will be closer. It's really this kind of detail we're looking at.
 
#147
Bruno Putzeys said:

No idea. I got my MKP379's as samples while both myself and the capacitor factory were still part of Philips. Try nagging your local Vishay sales department.
Click to expand...


Yes, these parts are very hard to find, if not impossible.
I sent an e-mail to Vishay's US corporate contact and apparently they will try to give me samples.
The part number for the MKP379 0.68uF/160V is: 2222 379 34684
Lets hope I get some so I can try them out and report back.
 
#148
Julien_M said:
Yes, these parts are very hard to find, if not impossible.
I sent an e-mail to Vishay's US corporate contact and apparently they will try to give me samples.
The part number for the MKP379 0.68uF/160V is: 2222 379 34684
Lets hope I get some so I can try them out and report back.
Click to expand...


Note that the 379 are out of manufacture, but I think the 479 are comparable.
 
#149
Bruno Putzeys said:



Note that the 379 are out of manufacture, but I think the 479 are comparable.
Click to expand...

Bruno,
I found a Vishay cap series that might work well, but it is of a higher voltage and might be too big.
The MMKP 383 series are AC and Pulse Double Metallized Polypropylene Film Capacitors and might sound close to the 379.
I've attached a drawing comparing the 379 and 383 cap dimensions.
What do you think?
Thanks.
 

Attachments

  • mkp_caps.jpg
    mkp_caps.jpg
    30.5 KB · Views: 920
#150
Bruno Putzeys
Regarding the fitting of large capacitors in the output filter would you think that a REL multicap would have a sufficiently low inductance in view of the claim that their inductance value is never greater than a piece of wire the same length of the capacitor body?

Curious,
Bob Lewis
 
#151
BOB LEWIS said:
Regarding the fitting of large capacitors in the output filter would you think that a REL multicap would have a sufficiently low inductance in view of the claim that their inductance value is never greater than a piece of wire the same length of the capacitor body?
Click to expand...
I can't get my head around claims of "low inductance capacitors", made by audiophile companies. Pretty much all modern film capacitors have the terminations made using a metal spray (schoop) process. This contacts the electrode metallisations or foil across the entire length. The result is that the inductance becomes equal to that of a solid block of copper of equal dimensions as the capacitor. You can't go lower than that, even if you're a maker of audiophile caps. The only capacitor construction that does have somewhat more inductance (and I say somewhat) is the type that has the wires contacting the ends of the foil. This is rare.
Anyhow, if you take a typical audiophile grade 680nF MKP cap of 35mm long and 20mm diameter. Mounted on a board in the most obvious manner, this produces over 20nH of series inductance. Compare to this a cheap 5mm MKT cap, typically clocking in at 6nH or less.

Mounted on a PCB, what matters is the loop area between the capacitor and the ground plane, as encircled from the "hot" terminal, through the cap, to the other terminal, through the ground plane and back to the hot terminal. A problem with axial capacitors is that the leads stand off from the capacitor body, adding substantially to inductance. A pitfall with radial capacitors, especially potted ones, is that the capacitor body may be several millimeters above the rim of the pot, covered in a thick layer of epoxy.

So, the most reliable way of getting to a low inductance is to use compact caps. I already think that the MKP379/479 is a bit large (the RF that makes it through the series inductance is visible on the scope) but at least the amp still works and remains within legal EMC limits.
 
#152
Julien_M said:
I found a Vishay cap series that might work well, but it is of a higher voltage and might be too big.
The MMKP 383 series are AC and Pulse Double Metallized Polypropylene Film Capacitors and might sound close to the 379.
I've attached a drawing comparing the 379 and 383 cap dimensions.
Click to expand...
I expect that in many applications it would even sound better than the 379, but the size makes use in class D output filters unattractive. I can't stop you from trying them out, but if your neighbours' radio goes mute when you turn the amplifier on remember I warned you :angel:
 
#153
Bruno Putzeys said:

I expect that in many applications it would even sound better than the 379, but the size makes use in class D output filters unattractive. I can't stop you from trying them out, but if your neighbours' radio goes mute when you turn the amplifier on remember I warned you :angel:
Click to expand...

Bruno,
I think I'm starting to grasp the concept of what to avoid in the output LC filter cap.
I think I found some Panasonic ECWF caps that might work well. They are just slightly larger than the 379's, and they are not encapsulated. There's also the ECQE type that's actually smaller than the 379 but it is specified as a "general purpose", not high frequency.
Is there a preference of which lead to make shorter, the ground or the "+".
Anyway. I think I should be able to find something else to try, but who knows if it will sound better... Right the amp is really good.
 
#156
Julien_M said:
There's also the ECQE type that's actually smaller than the 379 but it is specified as a "general purpose", not high frequency.
Is there a preference of which lead to make shorter, the ground or the "+".
Click to expand...
You have to make both leads short. This means that the lead that goes to GND will have to connect to GND somewhere else (near the power/output connectors!) You can use one of the 2 faston tabs that are connected to GND, or better scrape off the solder mask over the ground plane exactly where the component lead is. Here it is best to have a short lead to anywhere on the ground plane, rather than use the leads to route it back to the original capacitor pin. Again, orient the cap such that the grounded end of the capacitor is closest to the connectors.

Capacitors rated for "high frequency" are actually optimised for handling large currents. The currents here are not very large (few 100mA rms), so you can try the "general purpose" type as well.
 
#159
Auricaps

The 400's replaced a very respectable amp of my own design. This was a conventional A/B type amp with a great deal of refinement and excellent sound.
The new modules have quite a bit more residual carrier than stock but this is causing absolutely no known problems. This new amp has absolutely beautiful sound from top to bottom. It is the most transparent and uncolored amp I have experienced. It is now our new reference.
There might be some problems with the long cap leads but they must be very minor for the amp to sound this good. I am currently building a bunch more and will take my time to get useable measurements. I don't have an AP but I do have a decent older Tektronix analyzer as well as a 300MHz scope.
Much more when all the parts get in and as I do the tests.
Roger
 
#160
I finally ordered some caps for the output LC filter in my UcD400AD modules.
I got 3 types:
2 from Vishay: MKP 416 (63V) and MKP 418 (250V), 417 (160V) is not available from DigiKey, these appear to be very similar to the ones recommended by Bruno.
416 might be better for this application since it's slightly smaller than 418. 416 is the same size as 379.
The other type is a Panasonic ECWF(B) at 250V.
They are all metallized polypropylene film caps.
Bruno, what's the type of the original cap in the UcD400AD module? Is 63V enough?
Thanks.
J.


PS: Mouser sent me a quote for the MKP 379 0.68uF/160V:
"222237934684, Min 2000 @ .36ea, 14-16 Week LT, Non-Cancelable, Non Returnable"
 
Status
Not open for further replies.
Top Bottom