Sorry for bugging you all, but may I post PDF version of this sch, after all this is not mine and I do not have right to decide how to ditribuite it, if frederic agree I can post hi res pdf of sch right now ?
regards
Zeljko Kaiser
regards
Zeljko Kaiser
Now I can see the schematics...
Some comments about the amplifier: you take feedback after the filter, but run it to the input of the comparator, that is, you don't use an error opamp. Mmmm, that's different to the implementation Charles and me were discussing...
You use a SMPS: is this the same Model-1200 advertised in your webpage? If so, that's not a resonant-mode SMPS, but a classic half bridge, isn't it?
BTW: I like the way you drive the HS mosfet, using a fast digital optocoupler instead of a GDT or IR211x.
Some comments about the amplifier: you take feedback after the filter, but run it to the input of the comparator, that is, you don't use an error opamp. Mmmm, that's different to the implementation Charles and me were discussing...
You use a SMPS: is this the same Model-1200 advertised in your webpage? If so, that's not a resonant-mode SMPS, but a classic half bridge, isn't it?
BTW: I like the way you drive the HS mosfet, using a fast digital optocoupler instead of a GDT or IR211x.
Hello zkaiser
You can add a PDF version....
Pierre
I dont need error amp because of the level of the triangle wave, open loop gain is hight, so error of the amplifier canot be higger than the level of the triangle wave.....That's the trick!
Yes the PSU of the 1200 is straight forward Switching. Cheaper than the resonant mode of the 1600 and 2000. Anyways, this arrangement give very good effiency! Thanks for the praise about the HS! Stupid but work well! As you know, if you can put a piece of wire at the place of somme electronics, it's better!
Bye
Fredos
You can add a PDF version....
Pierre
I dont need error amp because of the level of the triangle wave, open loop gain is hight, so error of the amplifier canot be higger than the level of the triangle wave.....That's the trick!
Yes the PSU of the 1200 is straight forward Switching. Cheaper than the resonant mode of the 1600 and 2000. Anyways, this arrangement give very good effiency! Thanks for the praise about the HS! Stupid but work well! As you know, if you can put a piece of wire at the place of somme electronics, it's better!
Bye
Fredos
What THD values do you achieve this way and what is your load dependancy at the upper audio range ?
Regards
Charles
Charles, I am curious about your reaction to my comment on B&O Icepower modules' zobel 😉
Deception? Surprise?
Deception? Surprise?
I think that something that is marketed as turnkey solution should be foolproof to some degree.
What happens if the snubber fails ? Is the rest of the amp going up in smoke as well ?
If you see how much effort they laid into this multiloop topology one could think they should manage to achieve stability without using Zobels with caps of almost 1 uF.
Regards
Charles
What happens if the snubber fails ? Is the rest of the amp going up in smoke as well ?
If you see how much effort they laid into this multiloop topology one could think they should manage to achieve stability without using Zobels with caps of almost 1 uF.
Regards
Charles
That was my impression, too, Charles.
A little off-topic, but these modules really **** me off: you need +80V and +/-12V external supply to power them, weird connectors and then there is the HF issue, not to talk about their price. (The more I see it the more I like other modules that I have tested, such as coldamp, much more self-contained and easy to use).
You have to see the coil: they use 0.8mm or so wire for it. And that's supposed to produce more than 500W?
Well. Back to the topic...
I am not sure if I understand Fredos' comment about triangle amplitude and that " error of the amplifier cannot be higger than the level of the triangle wave..." Could you please clarify this further?
A little off-topic, but these modules really **** me off: you need +80V and +/-12V external supply to power them, weird connectors and then there is the HF issue, not to talk about their price. (The more I see it the more I like other modules that I have tested, such as coldamp, much more self-contained and easy to use).
You have to see the coil: they use 0.8mm or so wire for it. And that's supposed to produce more than 500W?
Well. Back to the topic...
I am not sure if I understand Fredos' comment about triangle amplitude and that " error of the amplifier cannot be higger than the level of the triangle wave..." Could you please clarify this further?
The triangle-amplitude does indeed determine the gain of the comparator/output-stage combination. So far so good. But one can only get low feedback factors the way he is doing it.
And the part about "error of the amplifier cannot be ...." I don't understand either.
Regards
Charles
You mean that, removing the error amp you don't have that opamp gain, is that correct?
Then you will have less feedback and hence more estability, but also higher distortion, right?
(I am talking only intuitively)
Then in order to have some effective feedback, the triangle amplitude must be quite low, so the comparator+power stage gain is much higher than what we are used to (around 20-25dB).
In this case, Vrail/Vtriangle is in the 40dB range or so.
From the schematics I can't find the triangle amplitude. The RC filter values where it comes to are not given. Besides, I think that the output of the clock divider is wrong: following the schematics the Class-D amp frequency is 62.5KHz meanwhile the PSU freq. is 125KHz, it makes no sense.
Fredos, could you check that?
Best regards,
Pierre
In this case, Vrail/Vtriangle is in the 40dB range or so.
From the schematics I can't find the triangle amplitude. The RC filter values where it comes to are not given. Besides, I think that the output of the clock divider is wrong: following the schematics the Class-D amp frequency is 62.5KHz meanwhile the PSU freq. is 125KHz, it makes no sense.
Fredos, could you check that?
Best regards,
Pierre
pierre,
from what i have seen: clock for amp is 250khz, smps 125khz.
the sg3525 uses 2pulses for full-wave, so 65k eff. smps frequ.
regards
alf
from what i have seen: clock for amp is 250khz, smps 125khz.
the sg3525 uses 2pulses for full-wave, so 65k eff. smps frequ.
regards
alf
You are right, I was reading the 4060 outputs backwards!
But the values of the lowpass R and C that determine the triangle amplitude are still missing.
But the values of the lowpass R and C that determine the triangle amplitude are still missing.
Pierre said:
phase_accurate said:
Seems to be an open loop amp, with multiple localized loops to compensate for delays and maintain sych. for both halves.
"U" and "U9" aren't LM319's, pinouts aren't even right, they're CA3080.
Anyone wanting to post a PDF version should note there's a few discrepencies.
Regards,
Chris
phase_accurate said:
I laughed when I saw your reaction to that.
The mixed feedback seems to require the zobel.
I'm a little stunned myself that's what they settled on given the level of research that went into it, and that they've yet to improve on it.
No harm done to the DIY community I guess, we've got far better options anyway 😀
Regards,
Chris
Hello guy
Sorry for long reply, lot of job! Yes I keep the triangle wave parts secret, sorry but easy to find the stability point....If you remove Zobel, the amplifier is still stable with load, but go into oscilation without load, no smoke, sorry! At long time the only bad thing is that the output coil beacame very hot, anyways, anything can fail in any amplifier! The lo gain of the power stage avoid the change in frequency response with variation on the load, but 2 ohms is the limit. Under that load (ex short circuit) the dead time controler avoid full conduction of the output fet and amplifier go near the open loop limited current mode. Yes the amplifier run at 250Khz and switching at 62.5Khz. A your right too, the limiter is a CA3080, unfortunately, not in the parts list of circad....But every one know this in house, so....No change for now, just have to know! Sorry I have forgot to tell, but anyways, that's just the limiter! Yan can have an idea about the triangle wave with the error detector for the limiter, it was setted to a max of 2% clipping...Take your calculator!
Expect that's awnser all your questions.....Anyways, schematics was their to help anyone who whant to built a simple and stable class d! Just for fun, chines have just began to copy my HVI serie...
Not nice!
http://www.stunitedsupply.com/webboard/open_webboard.php?id=973
They have done this with my first serie, but I know that the copy was not stable...! I will see if they have learn something of that!
Good day!
Fredos
www.d-amp.com
Sorry for long reply, lot of job! Yes I keep the triangle wave parts secret, sorry but easy to find the stability point....If you remove Zobel, the amplifier is still stable with load, but go into oscilation without load, no smoke, sorry! At long time the only bad thing is that the output coil beacame very hot, anyways, anything can fail in any amplifier! The lo gain of the power stage avoid the change in frequency response with variation on the load, but 2 ohms is the limit. Under that load (ex short circuit) the dead time controler avoid full conduction of the output fet and amplifier go near the open loop limited current mode. Yes the amplifier run at 250Khz and switching at 62.5Khz. A your right too, the limiter is a CA3080, unfortunately, not in the parts list of circad....But every one know this in house, so....No change for now, just have to know! Sorry I have forgot to tell, but anyways, that's just the limiter! Yan can have an idea about the triangle wave with the error detector for the limiter, it was setted to a max of 2% clipping...Take your calculator!
Expect that's awnser all your questions.....Anyways, schematics was their to help anyone who whant to built a simple and stable class d! Just for fun, chines have just began to copy my HVI serie...
Not nice!
http://www.stunitedsupply.com/webboard/open_webboard.php?id=973
They have done this with my first serie, but I know that the copy was not stable...! I will see if they have learn something of that!
Good day!
Fredos
www.d-amp.com
Wouldnt' worry about it, given their typical build quality it'll be in for constant repair. They've just no dignity it seems.
The problem is that it looks exactly like a naughty 1:1 copy which might give the original a bad reputation amongst the unaware.
🙁
Regards
Charles
🙁
Regards
Charles
- Status
- Not open for further replies.