DIYHiFi.org

For the sake of audio
It is currently Thu Nov 23, 2017 4:28 pm

All times are UTC




Post new topic Reply to topic  [ 150 posts ]  Go to page Previous  1 ... 6, 7, 8, 9, 10
Author Message
 Post subject: Re: 1kW Gen2
PostPosted: Wed Jan 02, 2008 7:43 am 
Offline
Sheep
User avatar

Joined: Wed Nov 15, 2006 2:35 pm
Posts: 59
Wonderful Results and happy new year choco!!!!!!!!

One Question........
You are using floating supply for your side high driver, have you faced any difficulty with it when countering hi dv/dt .............??

_________________
The Power Fanatic


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Wed Jan 02, 2008 11:38 am 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
Hi Kanwar,
the floating supply from my flyback did not make any trouble at all.
But I have no measurements for radiated EMI, in this regard I would theoretically expect the boot strap or charge pump solutions to be fortunate.

And here the updated schematic...


You do not have the required permissions to view the files attached to this post.

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Wed Jan 02, 2008 11:39 am 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
...updated overcurrent protection....


You do not have the required permissions to view the files attached to this post.

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Fri Jan 04, 2008 2:05 pm 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
Today I started to use this wild proto in my HiFi system.
Even without any speaker relay, the turn on 'pop' is unexpected low. The turn off is just a silent 'tick'. Nice.
The listening comparison is going vs the analog Rookie.
Please remember the analog Rookie amp ended up in two designs. A MosFet and a bipolar version. Originally I preferred the bipolar, it was more alive. The MosFet version was to decent. Two friends preferred the MosFet version, because they felt the bipolar version to be slightly intrusive. In the mean time we also went on with listening tests to different tweeters. And found my new love! Ribbons. In combination with ribbons also the very decent MosFet Rookie turned really alive. And in this combination I cannot say anymore which Rookie I prefer, they are both lovely in their own way.
My new class D proto again has its own character, it is something inbetween, but not exactly inbetween...
I am listening to it now since two hours and there is no annoying effect. It plays quite relaxed on the same level as both Rookies. So I will keep the proto in my system and go on with listening. ... going to gain more experience in the musical pros & cons...

Overall ...aloha.... class D seems to be by far more audiophile than what I was assuming two years ago....

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Sat Jan 05, 2008 3:28 am 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
I am considering the IRS20955. There is a unscientific concern in my stomach regarding the long propagation delay.
Currently the overall propagation delay from comparator input to half bridge output is roughly 120ns.
The IRS20955 level shifting would increase the delay by 50ns. On the other hand then I could use a simple discrete buffer with some Zetex bipolars instead of the UC3710 MosFet drivers. This would save around 20ns.
Another possibility to keep propagation delay short is the comparator. The LT1016 instead of the LM160 could save me another 10ns.
So I could get even with IRS20955 an overall delay around 140ns-150ns.
...hm, this is temptation for a low-component-count-junky... ...even the shut downs integrated..

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Sat Jan 05, 2008 8:45 am 
Offline
Account disabled on owner's request

Joined: Wed Dec 29, 2004 8:31 pm
Posts: 899
Hi Markus,

If your delay is already short enough, who cares? If anything I would look at how well it's matched from high to low under various conditions and see if that can be optimised, and a way to account for it in the control loop, self oscillating perhaps?

Cheers


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Sat Jan 05, 2008 3:18 pm 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
Yes, that's why I say for a unscientific reason. It is just my stomach, who tells me to keep the delay as short as possible.
I matched the switching events HIGH==>LOW and LOW==>HIGH with roughly 5ns accuracy. But I did not examine the delay difference. Interesting point.

Regarding self oscillating I do not like the lock ups, the frequency modulation and the beating. Nevertheless I am digging a little bit for fixed frequency & synchronizable selfoscillators....

In any case I am currently of the opinion that the modulation method is less important compared to the behavior of the power stage.
My modulator is just so so, but my power stage is quite well behaving. The overall results are pretty fine. Many designs with much better modulators seem to perform less good. From this I am concluding that the power stage is more critical than the modulator.

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Sat Jan 05, 2008 9:07 pm 
Offline
Account disabled on owner's request

Joined: Wed Dec 29, 2004 8:31 pm
Posts: 899
Hi, I was agreeing with ya completely up to a point.

I'd want it minimal as well, I mean, it's probably just a natural thing to want optimised whether it needs it or not, but I think there are perhaps more important areas in the delay worth looking at. You probably already took care of it with your high idle current :)

Synchronisable self oscillators? As far as I know, all you need in order to synch them is a clock signal mixed in. So by that they all should be synchronisable. However it's a foolish thing to do, for as the second you synch them you give up most of all the advantages of self oscillating. You'd have to build it double duty otherwise and there goes the part count.

Once you've handled the EMI there should be no issues with heterodyning, and the only real real concern becomes if you want to use it within a receiver with FM radio, at that point you'd need the clock to avoid interferance with the reception by shifting Fs away from the band of interest, whatever it happens to be. Probably just a high and low Fs setting when clocked and quality be damned. There are of course patents covering this.

Now what you might do with yours if for some reason you want to stick with clocked is account for the delay within your modulation and remove it as a source of error (which self oscillating does intrinsically). Minimising it overall would be one stop towards making it less of an issue, but I tend to think it should be corrected as well, and such a thing will really only be so useful, so I'd want it done simply.

Baybe have a look at Fredos from D-amps old posted design, if I recall he showed a fairly simple method (low part count) of removing the delay error.

So I guess you're probably right with what I found myself disagreeing with, but one should strive for a point where the output stage is considered "ideal" in its behavior, within reason, and things like modulation will have the greater impact on performance. In truth they both need to be at a certain level where neither one is the bottleneck, and the bar in this respect is getting set higher.

Hey, it's cool that you're listening to it already.

Cheers


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Thu Jan 10, 2008 3:36 am 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
...agreed... probably modulator and power stage both have to satisfy the proper level.
May be my power stage is already better than my modulator.
Here we go with my signature: The best overall performance is derived by the right trade off...


And ...hi ho... this is looking also nice:
http://datasheets.maxim-ic.com/en/d...2-MAX15013B.pdf
(May be with a an additional PNP for fast gate pull down)....

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Fri Jan 11, 2008 3:03 am 
Offline
Account disabled on owner's request

Joined: Wed Dec 29, 2004 8:31 pm
Posts: 899
Hi,

That link is broken. I saw the pdf though.

I'm impressed with their claim of matched delays between drivers.

However, in the past I have read of Maxim having issues with supplying customers with certain components that looked a little too good on paper, and how their needs were ignored. Turned a lot of people off wanting to use them, mostly because how things got handled I guess.

Here is one isolated example of someone discussing this (last post), it's just the very first that google turned up, but there are scores of examples of such things mentioned in sci.electronics.design if you care to dig for them:

http://forums.freescale.com/freescale/b ... ad.id=8583

As such, personally, I wouldn't bother with them for such a critical component, but decide for yourself.

Why maintain the same thread on two forums when one showed zero interest until you had something to show? Rhetorical question.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Sat Jan 12, 2008 3:48 am 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
...the spec looks nice, but just at first glance.
There is a minimum pulswidth requirement, which should for classD not be overcome by the Maxim proposal. That RC-delay proposal is OK for SMPS, but not classD IMHO.
To use this IC properly in class D we would need a duty cycle limitation at the comparator input...
Thanks for the hint about their customer service. Nothing new under the sun. Such acting is a normal habit of many companies, when it comes to 'worthless' mini customers.

TragicallyDistorted wrote:
...
Why maintain the same thread on two forums when one showed zero interest until you had something to show? Rhetorical question.

May be: '...never give up...'
Well.. OK. Simple reason. I restarted when diyhifi was down. And then went on, because of feedback on both chanels.

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Sun Jan 13, 2008 12:17 pm 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
Hi Chris,
you are mentioning Fredos' delay error correction.
I searched for this now roughly for one hour...
Do you have a link?
Or may be Fredos, do you remember your posting?

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Tue Jun 17, 2008 7:23 pm 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
I started to play with a triangle modulator followed by a fast level shifter.

The comparator was the LT1711 and the isolator/level shifter a HCPL9030.
The HCPL is known as fast, but critical with noise.
AD has made a comparison and is stating that with noise you might cause the HCPL to go for wrong HIGH signal at the output.
http://www.analog.com/UploadedFiles..._IsolatorsF.pdf

Up to now I was not able to force this misbehavior. HCPL9030 acts like a nice guy so far in my breadboard set up.
More difficult seems to be to settle a blameless triangle modulator for pulses below 10ns. In myexperiment the pulses are becoming instable and jittering below 10ns.
Please note that my screen shots show the triangle and the output of the HCPL, the output pulse of the LT1711 happens a few ns earlier,

First screen shot unzoomed with 50ns/Grid.
Triangle is shown with 100mV/Grid
Output of the HCPL with 2V/Grid
Attachment:
PWM_50ns.jpg


You do not have the required permissions to view the files attached to this post.

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Tue Jun 17, 2008 7:24 pm 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
...and now with 5ns/Grid
Triangle is shown with 100mV/Grid
Output of the HCPL with 2V/Grid

Attachment:
PWM_5ns.jpg


You do not have the required permissions to view the files attached to this post.

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
 Post subject: Re: 1kW Gen2
PostPosted: Tue Jun 17, 2008 7:25 pm 
Offline
Cow

Joined: Sun Jun 11, 2006 3:36 am
Posts: 112
In fact I am not convinced that it is really necessary to squeeze down the last few ns of delay.
With a IRS20955+NPN/PNP-driver I can go for roughly 200ns delay between comparator input and halfbridge output.
If I spend a lot of efforts with a HCPL level shifter and additional drivers - and have to settle all shut down functions in discrete... then I could reduce this to roughly 120ns.

...so now compare this to the group delay of the output filter of an class D amp. The typical step response is 10us-20us delayed to the input signal...
Is it really worth to squeeze the overall delay from 15us down to 14.92us and run into a complex circuit with high component count?

_________________
Best overall performance is derived by the right trade off.


Top
 Profile  
Reply with quote  
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 150 posts ]  Go to page Previous  1 ... 6, 7, 8, 9, 10

All times are UTC


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group