UcD400 Q & A

[classd4sure]

Originally posted by Golgoth



Splendid really! Maybe you could add a page with a bill of materials and details of fabrication, looks like a reference implementation to me!

I'm forced to agree! Layout is superb, the supplies all out... I'm turning green.

[koldby]

Michel

No the sync. is only reffering to the switching freq. and not to the modulation (the audio signal) . The 47nF is too small to have an impact on the audio signal.

Koldby

[Golgoth]

Originally posted by koldby
Michel

No the sync. is only reffering to the switching freq. and not to the modulation (the audio signal) . The 47nF is too small to have an impact on the audio signal.

Koldby

The switching freq and the modulation are the same in this case: the modulator in the UcD is basically a comparator which compares the input signal to a scaled down version of the output signal.

IMHO the two channels when synced via the 47nF should play the same sound, even with zero signal on one of the two inputs (XLR disconnected). Keep us updated

[classd4sure]

I agree, and that's the whole point to that cap, in bridge mode, so that both modules do play the same sound, by synching the modulation of them, while the frequency itself can be ignored.

If you reaaaally want the frequencies to be the same I believe you can mix a clock signal with the input (high frequency triangle) and that will force the frequencies to be matched without killing channel seperation. I've no idea how that would sound or what it would do for noise etc. Likely to be far more trouble than it's worth.

[Jan-Peter]

It has no sense to try to synchronize the two modules. In a way it will worsen it, because you are trying to synchronizing two modules who has two different signals. It has only sense if the two modules have the same signal as in a bridge mode.

It is possible to use a clocksignal at the input of the UcD modules, the oscillator will start directly to synchronize on this signal. The only reason to do this is to avoid mixing product or tones. But the HF emission of the UcD modules are low enough, so you normally wil not have this kind of problems.

And using a synchronize module will decrease the loopgain, so this will in a way WORSEN the whole situation .

Regards,

Jan-Peter

[koldby]

Michel

Try it!!

It playes stereo!

Dosen't even effect the stereo perspective negatively.

The switching freq. and modulation is IMHO not the same . It is like a modulated oscillator. And if you look at the output before the restoration filter, the switching freq. is pritty stable up to high modulation rates. In fact it is difficult to see, on this output, that it is not a PWM made from comperator with adiuo on one input and triang. switch freq. on the other.

Koldby

[Jan-Peter]

Kolby,

Let me explain;
In bridge mode stays the modules till a high mod. index in sync. In the case of two different signals the synchronize will loose quit fast, and in this case you will get a lot of IM products between these two signals. Supose CH1 1.2kHz and CH2 1.5kHz, you will see on both channels IM distortion on the boths channels of this two frequencies.

Regards,

Jan-Peter

[classd4sure]

Koldby,

Just try what Michel said if you want to prove it to yourself, disconnect/ground the input to one module, does "any" signal come out of the module with the disconnected input? hmmmm

[koldby]

Classd4sure

I would assume the output from the disconnected module would only show residual of the switching freq. !!


And just tried it. Only just before clipping in the other channel, there is some modulation in the disconnected channel If you look at the disconnected output with a scope you can only see it as small deviation in the freq. when he time base is set high enough to see the sw. freq.
If you look with the time base set to the audio signal, you can only see the noise from the sw. freq.

Jan-Peter

If you say so it must be so. But does it affect loop gain in the audio range? And how much IM and at what modulation level?
It is easy to see that the sw. freq. residual is much cleaner if you use the cap.
Maybe it is because the modules are very close in my setup?

If anyone has two UcD very close - try it and tell me if I am hearing ghosts??

Koldby

[classd4sure]

Originally posted by koldby
Classd4sure

I would assume the output from the disconnected module would only show residual of the switching freq. !!


And just tried it. Only just before clipping in the other channel, there is some modulation in the disconnected channel If you look at the disconnected output with a scope you can only see it as small deviation in the freq. when he time base is set high enough to see the sw. freq.
If you look with the time base set to the audio signal, you can only see the noise from the sw. freq.

Jan-Peter

If you say so it must be so. But does it affect loop gain in the audio range? And how much IM and at what modulation level?
It is easy to see that the sw. freq. residual is much cleaner if you use the cap.
Maybe it is because the modules are very close in my setup?

If anyone has two UcD very close - try it and tell me if I am hearing ghosts??

Koldby

Right, so that's the argument case and point.

You see it does indeed affect channel seperation/IM/etc. As with zero input, should have zero output, any switching residual you see on it should be 50% modulation and never changing, anything else is noise and distortion. If you really prefer the sound of it like that, who's to say we're wrong, you're the final judge after all. Throw a switch in there to turn on/off and call it an effect.