Active speakers with UCD400

[rlv]

Hi folks.

My first thread for my first UCD project...

I've just purchased 6pcs of UCD400ST from www.frequence.dk (Great service) for my 3-way active speaker setup.

My plans are pretty simple... one amp per loudspeaker driver, directly connected. Filtering is 4'th order Linkwitz-Reiley with additional Linkwitz transform for the woofer in a closed box.

I do have some questions...

- How much transformer power is needed? I reckon that the tweeter and the mid amps wouldn't need so much power. I plan to use only one transformer for all 3 amps in a speaker.

- I have 10pcs of 10.000uF/80V caps for each speaker. Is there anything wrong in supplying the 3 amps from the same capacitor bank? Or should I make 3 seperate PSU's (One transformer with 3 sets of diodes and caps)

- I am a little concerned about connecting the speakers directly to the amp. Should I implement some kind of DC protection or place a capacitor between the the speaker and the amp?

- How is DC protection usually implemented? I am designing a microcontroller circuit for soft start and remote power-on. The microcontroller has a lot of A/D converters that could be used for DC measuring. But I find it quite stupid to waste a lot of time on the issue if it has been done a 1000 times before in an easy way or if it simply isn't necessary...

Any comments and inputs to my project are higly appreciated.


Regards,

Rasmus

[tbr]

- How much transformer power is needed?

Depends on the efficiency of your speaker..., but if it's nothing unusual, 200-300VA should be way enough.

Is there anything wrong in supplying the 3 amps from the same capacitor bank?

No. But it helps keeping the ps's as seperated as possible.

Should I implement some kind of DC protection or place a capacitor between the the speaker and the amp?

Well, the cap would get really big for the woofer... and would also play into the x-over calculation. For the price of those caps you can easyly get replacement chassis...
The UCDs use coupling caps (elkos !) between the input op-amp ( again!) and the UCD pcb (the vertical one), so it's protected from DC of your pre amp.

- How is DC protection usually implemented?

If you use a microcontroller (with A/D converters) feed a DC coupled into one A/D input and an AC coupled into another and compare them by the software.

... and remote power-on.

If you don't choose the voltage for the UCDs too high, leave them on all the time. Takes normally hours for semiconductor electronics to sound good... and the amps will draw next to nothing (especially compared to conventional amps).

TB

[rlv]

Depends on the efficiency of your speaker..., but if it's nothing unusual, 200-300VA should be way enough.

That seems a kind of small to me... I need all the power from the bass amp that I can get. This is due to the Linkwitz transform where i amplify the lower frequencies in order to enhance the lower end, as I have a rather small cabinet.

No. But it helps keeping the ps's as seperated as possible.

What helps? And how to separate when I use the same PSU for all 3 amps?

The UCDs use coupling caps (elkos !) between the input op-amp ( again!) and the UCD pcb (the vertical one), so it's protected from DC of your pre amp.

This prevents DC at the input of the amplifier. But I was thinking of protection against a faulty FET, resulting in high DC voltage directly into the loudspeaker driver.



Another question: I normally use ceramic capacitors and metal film resistors for filtering, but are there more "audiophile" solutions to this?

Also, I have a bunch of AD8620 and OPA2134 that i plan to use for the filters (OPA2134 for bass section). Any good/bad experiences with these?

Regards,

Rasmus

[tbr]

I need all the power from the bass amp that I can get.

Then I would use seperate trannies for the bass and the rest.
Keep also in mind, that the efficiency of class-D amps is much higher, so you can expect less than 10% loss between trannie power and output power.

What helps? And how to separate when I use the same PSU for all 3 amps?

...the sonics, and sometimes even intermodulation problems. Use seperate rectifiers and caps, probably decoupled with low ohm Rs.

But I was thinking of protection against a faulty FET, resulting in high DC voltage directly into the loudspeaker driver.

As long as you don't overheat the module there is a little chance for that. The caps will be quite big and expensive, and the thing with activating speakers is that you DON't have passives between amp and chassis.

I normally use ceramic capacitors and metal film resistors for filtering, but are there more "audiophile" solutions to this?

Sure, film and foil caps (for example Mundorf ZN, now cheaper because of non ROHS conformity) and non-magnetic Rs (like Dale CMF-55-143, available from Schuro). Another good sounding cap and available in small sizes suitable for an active x-over are the SMR and SMC series from Evox-Rifa (polypropylene sulfide type). But you have to match them then as tolerances are quite high.

Also, I have a bunch of AD8620 and OPA2134 that i plan to use for the filters (OPA2134 for bass section). Any good/bad experiences with these?

Any op-amp out of the signal path is a good op-amp! But if you want to go for it, they are not the worst...

[rlv]

Then I would use seperate trannies for the bass and the rest.
Keep also in mind, that the efficiency of class-D amps is much higher, so you can expect less than 10% loss between trannie power and output power.

I'll prepare the construction for this. I'll start with a single 500VA, but two sets of rectifiers and capacitors. Then it's easy to add the extra transformer.

As long as you don't overheat the module there is a little chance for that. The caps will be quite big and expensive, and the thing with activating speakers is that you DON't have passives between amp and chassis.

I don't like the idea of a large cap in front of the speaker - thats why I thought of some intelligent DC detection. But I still need to figure out exactly how...

Sure, film and foil caps (for example Mundorf ZN, now cheaper because of non ROHS conformity) and non-magnetic Rs (like Dale CMF-55-143, available from Schuro). Another good sounding cap and available in small sizes suitable for an active x-over are the SMR and SMC series from Evox-Rifa (polypropylene sulfide type). But you have to match them then as tolerances are quite high.

Nice :-) I'll have a look at them.

Any op-amp out of the signal path is a good op-amp! But if you want to go for it, they are not the worst...

I found a nice little circuit by Lars Clausen (se attachment). Perhaps I should try it out...