UcD400 Q & A

Jump to Latest
#341
ClassD4sure

I am a little confused.

I am looking at the sw. freq after the filter. It is of course a sinus...
Without the cap. there is MORE noise and disturbances in the curve form than with the cap.

At small to moderate levels in the live channel there is still less disturbances in the disconnected channel.

At high levels just before clipping the modulation is detectable.

But of course a worsening of the distortion figure is NOT desirable .

And you are probably right - I am not an expert in ClassD.

😉


koldby
 
#342
koldby said:
ClassD4sure

I am a little confused.

I am looking at the sw. freq after the filter. It is of course a sinus...
Without the cap. there is MORE noise and disturbances in the curve form than with the cap.

At small to moderate levels in the live channel there is still less disturbances in the disconnected channel.

At high levels just before clipping the modulation is detectable.

But of course a worsening of the distortion figure is NOT desirable .

And you are probably right - I am not an expert in ClassD.

😉


koldby
Click to expand...

Ok,

So I'm now thinking maybe you're seeing a smoother ripple/residual because that extra cap will screw with the feedback loop/filter gain/phase in a way I'm too unqualified to predict.

Maybe you should throw a switch in there and have someone turn it on/off for you while you listen? Nice blind test. Maybe you'll feel differently about it.

Main point is though modulation does appear on the disconnected amp, therefore you know it is ruining your seperation and what not.

Regards
 
#343
Hi again,

Just wanted to add, any switching residual/noise you're seeing should be inaudible because it is at the switching frequency anyway right, so I wouldn't worry about trying to make it look smoother anyway, you won't hear it.

Side note:

There was a time when I got my first stereo, I was playing with mono/stereo back and forth, at first I strongly felt mono sounded alot better, because it was more centralized in the brain it seemed louder (and I guess two 15"'s working in unison etc), but after awhile I got tired of it and missed stereo imaging. Maybe there's a bit of that kind of thing happening here too?

Regards
Chris
 
#344
It does bring up a valid point of course. A common problem with conventional clock-driven Class-D amplifiers is that you can get audible heterodyning on a stereo amp if the two amplifiers are fed from separate free-running clocks with the difference frequency of the two clocks in the audible range. Say if one was running at 300kHz, and the other at 301kHz then you can get a 1kHz heterodyne tone. The application notes from Zetex on the SCD100 chips show how to lock the clocks together to stop this.

On a self-oscillating amplifier the effect should be much less apparent as the clocks are altering in frequency all the time. (As a function of the audio signals).

Have any measurements been done to see if any heterodyne signals are measurable on the outputs of a uCD180 or 400 if two amps are close together?
 
#348
classd4sure said:
Now there's a question! Forgive the ignorance, what would that sound like? A whistle?
Click to expand...

So the Zetex notes say. Now for a pair of self-oscillating amps where the switching frequency of each is varying, it may just increase the noise floor.

I suggest ferrite common-mode beads on the LS output leads and careful seperation of the two amps will be adequate to prevent any problems. (I'm OK, I'm using the uCD400 modules to replace the original Quad 405 modules in a Quad405 case, and they are at each end of the case, with the PSU between them).
 
#349
Hello, guys.
I am new to this thread.

What is the residual switching carrier amplitude at the speaker's output for UCD180 and UCD400 modules?

It would be nice if somebody could also tell me what that specification is for other modules, like LCAudio Zap Pulse.

And... what's is a good maximum amplitude (as a rule of thumb) to avoid EMI/RFI problems with other equipment?

Thanks!!!
 
#350
Ouroboros said:
It does bring up a valid point of course. A common problem with conventional clock-driven Class-D amplifiers is that you can get audible heterodyning on a stereo amp if the two amplifiers are fed from separate free-running clocks with the difference frequency of the two clocks in the audible range. Say if one was running at 300kHz, and the other at 301kHz then you can get a 1kHz heterodyne tone. The application notes from Zetex on the SCD100 chips show how to lock the clocks together to stop this.

On a self-oscillating amplifier the effect should be much less apparent as the clocks are altering in frequency all the time. (As a function of the audio signals).

Have any measurements been done to see if any heterodyne signals are measurable on the outputs of a uCD180 or 400 if two amps are close together?
Click to expand...


I have not done any measurements but I can tell you that if there would be any heterodyning, then it is far below the (very low) noise level of the UcD180 modules that I have. To hear the noise of the UcD modules, you really have to put your ears almost in the tweeter. When I switch off the modules, during the discharging of the power supply caps, I can hear some heterodyning whistling sound), however, again only when I put my ear in the tweeter. Completely inaudible at the listening position. My modules are mounted pretty close to eachother, only a few cm seperation, no problem at all. Just very good sound, no noise and whistless.

I think Jan-Peter has some measurements posted on his web-site where he shows measured noise spectra in case multiple modules are built in one case.

I just checked it, it is in here:

http://www.hypex.nl/classd/classd.pdf

It is measurable but at an extremely low level, I can`t hear it. It s also measured on an experimental setup, not on the final UcD180. In the final version, it may ne masked by the input noise of the opamp.

Best regards

Gertjan
 
#351
I can fully agree to all what Gertjan wrote. The modules are approx. 30mm away from each other in my setup and still driven by the same supply. I do not know, if the light whistling during switch off is a mixing product of both carriers or simply the carrier itself (if it's getting so much lower at shut off). I already posted my impressions from first listening.
Chrisb03, I'm getting envious, your setup looks really good.
Timo
 
#352
Hi Timo,

I have a feeling it isn't heterodyning at switch off. You could try to disconnect one module and see if you still experience the same whistle? I've a feeling it will remain, some sort of "byproduct" if you will, of the feedback gain and therefore modulation collapsing as things discharge. That's my theory 🙂
 
#353
Hi Chris,

I wonder a bit, don't you sleep ever? When my daughter stayed in Canada (Cornwall/Ontario) last year, we had to wait 7 hours doing our phone calls (so it was crossing the day margin most of the time 😱). We picked her up last summer, going through the whole east, but we did not meet you?! 😀

I will try to find out something on Sunday hopefully, the weekends are occupied increasingly by music arrangements with both, daughter and son (choir, piano and e-guitar). I'll keep you posted.

My modules are UcD180s, not 400s, but due to the identical modulator...

Timo
 
#354
Hi Timo,

Sleep? Sleep isn't always restful.. Non the less, I do get a few hours here and there when I burn out.

I don't think you would have run into me, did you go to Elliot Lake? Kind of out of the way of anything, people mostly come here to die, um, I mean, retire.

It's funny, normally people go fishing to just get away from it all (or dump bodies), but here, people go fishing for something to do.... (or dump bodies).

It'd be nice if someone did do a few simple and quick tests to see if that turn off whistle is heterodyning, maybe it'll help burry any paranoia caused by it. One module, as mentioned, would be an easy test, and just seeing if the whistle changes/or how bad it does, the closer you move them together might be another. Who knows maybe we'll get a rule of thumb of no less than so many inches apart from each other out of it.

This last bit is way off topic but I feel compelled since I just lost my hard drive:

If you have a Fujisu drive

The HDDs at issue bear a model number composed of or starting with the sequence
MPF3102AH, MPF3153AH, MPF3204AH, MPG3102, MPG3153, MPG3204, MPG3307, or MPG3409, which may be followed by
alphabetical character(s).

Read this:
http://www.pcbuyerbeware.co.uk/HDDProblems.htm#fujitsu

Regards,
Chris
 
#355
ssanmor: We have around 1.66 V RMS of residual switching signal @ 460 kHz.

Most tweeters have at this frequency an impedance of > 30 Ohms, so power dissipation in the tweeter (from residuals) is something like 0.1W.

btw. i did some further listening tests, and compared the UDC400 with our 2.1 and 2.2SE. In the initial tests i had used some too small loudspeakers. Now i have used our bigger ref. speakers, i had hoped to pinpoint the findings from the first tests to the small speakers. But i think the difference found in the first listening test just stood out clearer with better speakers.

Which is better is hard to say, some like the mother, others like the daughter, and some even like both. I can recommend making this comparation, it clearly shows the difference of feedback schemes work very clearly on the sound, even in PWM amplifiers.

In dec. we will make demonstrations in our showroom with ZAPpulse vs. UcD400 in a similar setup.

😉 all the best from

Lars
 
You must log in or register to reply here.
Top Bottom