Hotrodding the UCD modules

Status
Not open for further replies.
Jump to Latest
#981
polypropelyne is considered to have less of a sonic signature than polyester caps, and they also handle more current. So I agree with you Koldby.

The polyester based caps can be made physically smaller and I think is the reason they appear as the stock UCD cap.

Incidently I'm following the discussion closely relating to Mike and Gertjan's work. I've been thinking about bypassing the ferrites as well, I understand they often give the sound a certain thinness.

Seems replacing them with low value power resistors would be the right thing to do as opposed to bypassing them though.

Can anyone post the exact layout/circuit used in this area, values and types included? I'd like to see how my UCD 180 compares and will post my findings on it in due time.

I dont' see how an array could effectively be used in this area, it would mandate offboard component placement and may be counter productive, a snubber would most definatly be required, but even then EMI would be increased, how much remains to be seen.

I think the optimal solution may be a CRCRC type filter.

Regards,
Chris
 
#984
Mike2 said:
Any thoughts on replacing the ferrites with a choke/inducor?
Click to expand...


I think in UcD700 they use chokes (the toroids). I think the ferrites work well, they suppress the HF rubbish very well.

Using CRCRC may be possible, however, you want to use low values for the R. In the UcD 400, the last R (on the UcD itself) is only 5mOhm, on UcD180 it is 50mOhm. The series resistance of the caps should be ideally lower than those Rs otherwise you do not get much suppression of HF junk. That will be really tough. So I think using an LC as done now is a better way to keep the HF junk inside the UcD. Reducing it on the UcD itself maybe possible by putting two caps in parallel, one on each side of the PCB and keeping the wires as short as possible, this is not possible with anything added next to the UcD board as the lead wires of at least a few cm will add inductance which will probably show up in a nasty way (resonances). So using two low series resistance caps in parallel instead of the one 470uF cap may be the best solution. Assuming of course the snubber cap is still effective to prevent resonances due to the lower series resistance. At some stage, I plan to put another panasonic FC at the bottom of the PCB to see if it improves things and whether I do not introduce unwanted side effects.

Gertjan
 
#987
I dont' see how an array could effectively be used in this area, it would mandate offboard component placement and may be counter productive
Click to expand...

I was talking about the main PS, using an array of very low ESR caps in parallel to create a power supply that is able to supply large currents very quickly to the amplifier. As noted, you would want to have this array close to the amplifier or use high current cable to the amp, and it would need a carefully tuned snubber to tame parasitic HF resonances that the switching nature of a Class D seems to excite more than a standard amp.

If only I could find a place to source Rubycon ZL 1500uF 63v caps! :bawling:

Regards,
Dean
 
#988
Mike2 said:
Any thoughts on replacing the ferrites with a choke/inducor?
Click to expand...

Mike2,
Ferrite beads are made to be resistive (just like an inductor) at low frequencies, but very lossy at high frequecies. Inductors strive to store the energy and losses are concidered a design fault at best. Thus, I'd stick to ferrites over inductors or toriods in this position.

That is not to say that a toriod designed to be lossy at high frequencies couldn't be a good way to do it, but the point being you want very high core losses at high frequecies.

I'm almost 24hrs from doing some major mods on my boards.

I have a slug of wima MKP2 0.33uF's as well as more 470uF replacement caps than I'll likely get to.
I do wanna hear the FC's and coloring. But the M series seem interesting and have been said to be best. I think it would be interesting to tweak the dead time for minimal distortion for each cap tried.


:smash:
I have read in this thread that the bypass caps and the filter caps interact.
I also hate to do more than one change at once, although its much easier.
Any thoughts on what to do first?

Should I just do the filter cap change first and listen a bit?

Mike
 
#989
I've just fiited some more Rubycon ZA/L in my modified cdp so now its full of them, these caps are just brilliant:bigeyes: I also compared them against some well used Blackgate standard and the none polar N types, to my ears the ZA/L types are much better.
I really hope we can find a good source for the 470uf 63v ones, if they give the similar results in my UCD180 as in the cdp I'd be well pleased!

I'm still using 0.68uf 63v Evox SMR in the output of my UCD180 which don't seem bad
 
#990
Portlandmike said:

<snip>

I'm almost 24hrs from doing some major mods on my boards.

I have a slug of wima MKP2 0.33uF's as well as more 470uF replacement caps than I'll likely get to.
I do wanna hear the FC's and coloring. But the M series seem interesting and have been said to be best. I think it would be interesting to tweak the dead time for minimal distortion for each cap tried.


:smash:
I have read in this thread that the bypass caps and the filter caps interact.
I also hate to do more than one change at once, although its much easier.
Any thoughts on what to do first?

Should I just do the filter cap change first and listen a bit?

Mike
Click to expand...


Hi Mike,

Just wanted to make sure you're aware that it wasn't recommended to go below 470nF as minimal filter value. Tweaking the delay for each cap is an interesting idea, if it's affected by the change.

I don't think you'll get to experience the level of interraction to the same degree I had, simply because your stock filter cap is already an upgrade over what mine was, I'd think it would have better HF precision already.

I did try the stock decoupling caps with the upgraded filter cap later on and it brought back much the same characteristics.

I'd say try decoupling caps first, but doesn't really matter, long as you do it one at a time.

While I won't say the snubbers wouldn't have their own level of interractions as you've mentioned, I do believe each caps just have their own sound too, and by looking around the forum in other sections this seems evident. With the snubbers you can tweak the response somewhat, "probably" limited mostly by the PCB layout, but the cap will still impart its signature to a certain degree.

Regards
 
#991
classd4sure said:



Hi Mike,

Just wanted to make sure you're aware that it wasn't recommended to go below 470nF as minimal filter value. Tweaking the delay for each cap is an interesting idea, if it's affected by the change.

I don't think you'll get to experience the level of interraction to the same degree I had, simply because your stock filter cap is already an upgrade over what mine was, I'd think it would have better HF precision already.

I did try the stock decoupling caps with the upgraded filter cap later on and it brought back much the same characteristics.

I'd say try decoupling caps first, but doesn't really matter, long as you do it one at a time.

While I won't say the snubbers wouldn't have their own level of interractions as you've mentioned, I do believe each caps just have their own sound too, and by looking around the forum in other sections this seems evident. With the snubbers you can tweak the response somewhat, "probably" limited mostly by the PCB layout, but the cap will still impart its signature to a certain degree.

Regards
Click to expand...


Thanks Chris,

I did some swapping out yesterday. Started by putting FC's in first. I can't recall the value right now, but they were 16mm diam, and I think 35 high FC's. I wanna say 1200uF/63V FC's.
I did one channel. Couldn't really hear a huge difference on one channel. My speakers don't have perfect loading symetry in the room, so its tough to tell.
I then put 100uF 100V M series in the other side.
Then put two Wima 330nF's per board.

I must say I really didn't hear any major change, but I take awhile to develope an opinion. I really didn't find it that very different but it did seem like the bass was a bit puncher on the
FC amp, and the mids and highs were also not as good on that channel.

I really kind of wish I'd done both sides the same since I kind of already new that mono comparisons are pretty iffy when your at this level of amp.

I think I'll be putting my efforts into integrating a couple UcD400's into a Denon integrated amp I have. Should be pretty simple to get seperate regulated rails for the OA's there, and its got dual rail transformers, so I'll do a dual bridge supply.

I really want to bring these UcD's into the lab though. I think to truly gain all the benifit of the different caps one should adjust the dead time for minimum distortion for each supply cap.
It just seems like it would matter. Maybe Bruno or Jan can comment, although if I were them I'd likely try to keep people scared of playing with that. 🙂

I also need to make speaker grills for my JBL 4343's before the wife and kids get back, so I don't know if I'll get to lab testing the amps this week. 🙁

Regards,

Mike
 
#992
I then put 100uF 100V M series in the other side.
Click to expand...

I think the 100uf cap is too small. You should stay around the 470uf or higher. The 1000uf 100v Panasonic M will fit with no problems. I find also the caps seem to take about 50 Hrs or so to settle in. Once you have a feel for the sound of the caps try adding a .01uf MKP or FKP across the two .33uf caps. I find they take a little of the grain (if thats the right word) out of the highs. The FKP also seems to thighten up the low end some.
 
#993
Mike2 said:


I think the 100uf cap is too small. You should stay around the 470uf or higher. The 1000uf 100v Panasonic M will fit with no problems. I find also the caps seem to take about 50 Hrs or so to settle in. Once you have a feel for the sound of the caps try adding a .01uf MKP or FKP across the two .33uf caps. I find they take a little of the grain (if thats the right word) out of the highs. The FKP also seems to thighten up the low end some.
Click to expand...


Mike2,

Thanks. I'll try that. I was typing fast, they are 1000uF/100V Mseries. They are playing now, but I think I will take out the FC on the one side tonight.

I am also interested in trying a 470uF 63V M series just for kicks, but given the lack of huge difference, and the long burn in... its not trivial. Believe it or not, I'd rather just listen to music then screw around lately. Don't know what's into me 🙂

Thanks again,

Mike
 
#994
Music!

Mike,
Yes and when you notice your self paying attention to things like type of microphones used and EQ or echo added instead of listening to the grain or distortion you are getting there. It gets frustrating to live with the fact that the systems used to master the music were not as revealing as our UcD based systems and consequently a lot of garbage is produced.
Roger
 
#995
Re: Music!

sx881663 said:
Mike,
Yes and when you notice your self paying attention to things like type of microphones used and EQ or echo added instead of listening to the grain or distortion you are getting there. It gets frustrating to live with the fact that the systems used to master the music were not as revealing as our UcD based systems and consequently a lot of garbage is produced.
Roger
Click to expand...


Your there when your listening to the music and not worrying about technical issues 😀

Mike
 
#997
Dean:

Did you have any success finding the source of Rubycon ZL caps? If so, maybe we could organise a group buy?
Click to expand...

I was about to propose the same thing as no distributor seem to be too impressed by my mails 🙁

If there's enough interest I think it could be done. I have no problem with buying BLINDLY 20 units of 470uF/100V Rubycon ZL, or whatever the model preferred 😉
(I buy blindly allways)
I have never been dissapointed with the ZL, wright t.?

Maybe a member with the right contacts or who owns an established bussiness could be interested in organizing a GB :angel:

Dear Mike:
Good luck with your mods and analysis. Give the mods an appropriate burn-in time, though.
(the JBL will be there for a while; no need to hurry with the grills 😉 )

Cheers
Mauricio
 
#998
Mike2,

I put 1000uF 100V M's in both amps. 2X0.33 Wima's MKP2's?

Burned them in a bit. Listened.

Interesting affect. Pretty smooth overall. Bass is fuller than stock, more solid, but I wouldn't say ballsy. Just effortless.
I really can't comment vs the FC's since I never did both channels, but I will say this. Its way better than one of each which would lead me to leave the FC's to switching regulators rather than audio amps, at least in this position.
I'm interested in trying the 470uF M series in this place, but the ripple current is iffy.

Sometimes I think its got some issues on "sssss's" its not sizzling, but it seems almost like its trying to. Undecided on that.

precussion sounds more back I think....

Highs seem better than the stock rev 6.1 400's with the wima's.
Haven't tried the 10nF accross the filter cap yet as I don't have them in hand.

I've still got the stock amps to drop in sometime, but I wanna listen to these for alot longer before going back.

Mike
 
#999
maxlorenz said:



(the JBL will be there for a while; no need to hurry with the grills 😉 )

Cheers
Mauricio
Click to expand...


Contrare my friend. I just bought a 60 inch sony XBR. Its not in my listening room. Although it does have one of the better built in speaker systems for a tv imo, two 4 4 inch drivers can't compete with the JBL's. AND, my kids are still can be little terrors.
"Don't touch the foam (Dad)"..... And what does that make them just not be able to resist?

I"ve been recording several live HD concerts that I'm looking forward to being at. The wife needs to be pleased too you know!

Mike
 
Status
Not open for further replies.
Top Bottom