Hi sfx,
congratulations for your development! The Thunderball looks nice, especially the simplicity is very interesting.
My first try with the Sörensen-design was not so succesful, I blowed the power stage two times.
I'm looking for a relatively simple solution as a first shot. The BCA-design seems to be interesting also, is it possible to see your schematic soon? I would try to extend it to the entire audio range.
Best regards, Timo
congratulations for your development! The Thunderball looks nice, especially the simplicity is very interesting.
My first try with the Sörensen-design was not so succesful, I blowed the power stage two times.
I'm looking for a relatively simple solution as a first shot. The BCA-design seems to be interesting also, is it possible to see your schematic soon? I would try to extend it to the entire audio range.
Best regards, Timo
)
Hi Johan,
unfortunately not, because my 15 years old sun begun to sing and play the electric guitar in a school-band some month ago. Therefore they need 'big' loudspeakers, which we are building now. It took five weekends fulltime and a lump of money until now and they aren't ready yet.
The amplifier has to wait a little. The Uni-project-PWM-amp diverges somewhat because of the different needs. This amp is still in progress also.
But I'm interested in all improvements regarding the class-d amp, even though on a more theoretical basis temporarily.
Regards, Timo
unfortunately not, because my 15 years old sun begun to sing and play the electric guitar in a school-band some month ago. Therefore they need 'big' loudspeakers, which we are building now. It took five weekends fulltime and a lump of money until now and they aren't ready yet.
The amplifier has to wait a little. The Uni-project-PWM-amp diverges somewhat because of the different needs. This amp is still in progress also.
But I'm interested in all improvements regarding the class-d amp, even though on a more theoretical basis temporarily.
Regards, Timo
tiki said:
I see. Myself, I am currently engaged in building a SMPS to drive my amp(s) with.
However, when I feel ready for it, I plan to revisit the feedback topology of my amp - the "leapfrog" design method seems to provide some new openings to incorporate feedback from the output filter...
Right now, I'm employing moderately "light" filtering with a 20uH coil followed by 470nF cap. This is enough to keep my TV-set and radio receiver happy, but might not pass CE EMI tests
I am very happy with the way my amplifier sounds, although I have been told by one auditioner that it seemed a bit lacking in dynamic response (treble, not bass!). So I guess the step response might be somewhat muted by the output filter; thus being an area for improvement.
Regards / Johan.
IVX said:
Well, that's the embarrasing part: I haven't gotten around to do any formal measurements on my latest channels. I did do some rudimentary tests on my "2.1" version, and found 0.024 %THD @1Khz and 10W. (It was more limited in terms of power ability.)
People keep bugging me about measurements, but since I'm not selling the amp, I never seem to get around it
Best Regards / Johan.
Thunderball
Hi Felix,
hehe, all the guys are keen on your design, mee too, of course. I would like to use it for this.
If you finished the schematic, will you post it?
Many Thanks in advance!
Timo
Off topic:
If anybody has to decide, which schematic capture/layot program should be used - be careful with Orcad (V10)!
Yesterday I lost my project, don't know if the support will be able to recover it. I love the DOS-Version from 1989 instead.
Hi Felix,
hehe, all the guys are keen on your design, mee too, of course. I would like to use it for this.
If you finished the schematic, will you post it?
Many Thanks in advance!
Timo
Off topic:
If anybody has to decide, which schematic capture/layot program should be used - be careful with Orcad (V10)!
Yesterday I lost my project, don't know if the support will be able to recover it. I love the DOS-Version from 1989 instead.
alive...
Hi all,
yes, I am still alive... I was on holiday in NZ, and after I came back, I was a bit busy at work, so I forgot about the schematic. I hope I can finish it within the next days. Sorry about that... I'll let you know...
My amp is also still alive, and it does a fantastic job in my subwoofer. I combined it with Elliot's EAS preamp (basically an equalizer, that compensates the response below the resonance frequency of a closed box), so now it has more or less flat frequency response down to 16 Hz or so (the whole speaker, measured with a Behringer Reference Microphone). Quite some tremendous sub-sonic stimulation
The speaker itself is a 30cm Woofer in a 60l enclosure, and at full volume, it moves for about 1cm. This is a displacement of approximately 600 ccm.
I carefully adjusted it to match my main speakers (some 25 year old Orbid Sound 4way speakers, which already go down to 25 Hz), so now I have a nice flat response from 16Hz to 20kHz. I was surprised, it does make a difference...
Cheers,
Felix
lkadar said:
Hi all,
yes, I am still alive... I was on holiday in NZ, and after I came back, I was a bit busy at work, so I forgot about the schematic. I hope I can finish it within the next days. Sorry about that... I'll let you know...
My amp is also still alive, and it does a fantastic job in my subwoofer. I combined it with Elliot's EAS preamp (basically an equalizer, that compensates the response below the resonance frequency of a closed box), so now it has more or less flat frequency response down to 16 Hz or so (the whole speaker, measured with a Behringer Reference Microphone). Quite some tremendous sub-sonic stimulation
The speaker itself is a 30cm Woofer in a 60l enclosure, and at full volume, it moves for about 1cm. This is a displacement of approximately 600 ccm.
I carefully adjusted it to match my main speakers (some 25 year old Orbid Sound 4way speakers, which already go down to 25 Hz), so now I have a nice flat response from 16Hz to 20kHz. I was surprised, it does make a difference...
Cheers,
Felix
,The time has come
... and the waiting is finally over!
You can find the schematic to my Class D amp here:
http://rsise.anu.edu.au/~felix/iar/Projects/Thunderball/index.html
It's a pdf file, 57kBytes. I hope I got it all right - as I said, the design mostly happened with the soldering iron in my hand, and a few non-consistent scribbled schematics on paper. Since it already does its job inside my subwoofer, and I didn't want to get it out again, all I had were a few photos I took before I built it in. So I can't guarantee that all the values are correct, but it should be more or less right. Most values are not critical anyway. The important ones are the resistors and the cap in the triangle generator, to get the right frequency and amplitude (250kHz).
Don't be confused by strange values like 5.6k resistors or so everywhere else in the circuit - I just took the parts I had available. Slightly different values will also do.
The output filter is too conservative - it's ok for a subsonic subwoofer, but cuts off at about 2-5 kHz.
Unfortunately I couldn't do any THD measurements, since the power stage is completely bridged, and there is no common ground anywhere around. I checked the output on a scope once, by subtracting the 2 output signals, and it looked really nice - no visible glitches. I know, not very precise...
Now, since I use it every day, and boostet it up to 35V, the MOSFETs do get a bit warm, so now I added a small heatsink. It doesn't get proper ventilation inside the case (and inside the subwoofer), but now it's fine. Up to 25V a heatsink is not really necessary, but at 35V, especially at this high switching frequency, there is a bit of heat to get rid of . In my next design, the FETs will be at the edge of the board, that I can screw them onto the case. By the way, this nice waterproof aluminium case is an old case of a motor amplifier for our research submarine
I hope my schematic will be useful for some of you. Let me know your opinions and experiences with it!
Cheers,
Felix
... and the waiting is finally over!
You can find the schematic to my Class D amp here:
http://rsise.anu.edu.au/~felix/iar/Projects/Thunderball/index.html
It's a pdf file, 57kBytes. I hope I got it all right - as I said, the design mostly happened with the soldering iron in my hand, and a few non-consistent scribbled schematics on paper. Since it already does its job inside my subwoofer, and I didn't want to get it out again, all I had were a few photos I took before I built it in. So I can't guarantee that all the values are correct, but it should be more or less right. Most values are not critical anyway. The important ones are the resistors and the cap in the triangle generator, to get the right frequency and amplitude (250kHz).
Don't be confused by strange values like 5.6k resistors or so everywhere else in the circuit - I just took the parts I had available. Slightly different values will also do.
The output filter is too conservative - it's ok for a subsonic subwoofer, but cuts off at about 2-5 kHz.
Unfortunately I couldn't do any THD measurements, since the power stage is completely bridged, and there is no common ground anywhere around. I checked the output on a scope once, by subtracting the 2 output signals, and it looked really nice - no visible glitches. I know, not very precise...
Now, since I use it every day, and boostet it up to 35V, the MOSFETs do get a bit warm, so now I added a small heatsink. It doesn't get proper ventilation inside the case (and inside the subwoofer), but now it's fine. Up to 25V a heatsink is not really necessary, but at 35V, especially at this high switching frequency, there is a bit of heat to get rid of . In my next design, the FETs will be at the edge of the board, that I can screw them onto the case. By the way, this nice waterproof aluminium case is an old case of a motor amplifier for our research submarine
I hope my schematic will be useful for some of you. Let me know your opinions and experiences with it!
Cheers,
Felix
...
Thunderball
Hi Felix,
Thank you very much for your effort.
The design looks very straightforward, easy to understand and to "copy", hopefully.
Maybe, I will add some modifications, regarding to some ideas of the other thread in this forum. But it will take some time...
Thanks again and congratulations for the success.
Best regards, Timo
Hi Felix,
Thank you very much for your effort.
The design looks very straightforward, easy to understand and to "copy", hopefully.
Maybe, I will add some modifications, regarding to some ideas of the other thread in this forum. But it will take some time...
Thanks again and congratulations for the success.
Best regards, Timo
lots of questions
sfx:
1. Have you measured the output power, or are you just guessing (4ohm, 150W)?
2. What trafo are you using for the amp?
3. Are you using it for 4 or 8ohm subwoofer?
4. If it is 150W in 4ohm, would you recommend a 225VA trafo, 300VA trafo or even bigger?
5. What's the max supply voltage when driving a 4ohm load?
6. Are the 1000uF caps the only caps or are there more in the powersupply?
7. What do you adjust the Vgnd to? (6V?)
8. Would there be any benefits in using better mosfets?
9. What are the specifications of the output chokes (brand, amperage)?
sfx:
1. Have you measured the output power, or are you just guessing (4ohm, 150W)?
2. What trafo are you using for the amp?
3. Are you using it for 4 or 8ohm subwoofer?
4. If it is 150W in 4ohm, would you recommend a 225VA trafo, 300VA trafo or even bigger?
5. What's the max supply voltage when driving a 4ohm load?
6. Are the 1000uF caps the only caps or are there more in the powersupply?
7. What do you adjust the Vgnd to? (6V?)
8. Would there be any benefits in using better mosfets?
9. What are the specifications of the output chokes (brand, amperage)?
some answers
1. Have you measured the output power, or are you just guessing (4ohm, 150W)?
Something in between... I measured the output with a scope, into an 8 Ohm load, and got an amplitude which was almost +/- the supply voltage. So I assumed this would be the same for 4 Ohm load, and multiplied. I used two 4 Ohm 10W resistors in series, cooled in a glass of water, and they got VERY hot. That's why I didn't want to torture only one of them with twice the power.
You are right, I really should do some more measurements... It's maybe a bit less than supply voltage, but approximately it is right. The power into resistors is calculated by amplitude^2 / (2*R_load). So (35V)^2 / (2*(4Ohm) =153 W
2. What trafo are you using for the amp?
My trafo is an ordinary 24V 200VA trafo, rectified and filtered with chokes and some big caps. Bigger is of course better, but I couldn't afford it.
The Modulator is driven from a separate 12V secondary, rectified and regulated with a linear regulator
3. Are you using it for 4 or 8ohm subwoofer?
My speaker is a 4 Ohm 30cm (12") Woofer, in a 55l closed box. To achieve a flat frequency response below 20 Hz, I used the Linkwitz transform circuit described here.
The final system is more or less linear down to 16Hz, and drops off a bit below that. The speaker has a displacement of 500-600ccm at full volume. It's very impressive - sometimes I think a car drove by, but in fact it was in the movie/music...
4. If it is 150W in 4ohm, would you recommend a 225VA trafo, 300VA trafo or even bigger?
I'd say, as big as you can get... I'd actually prefer a regulated SMPS, but I couldn't find any with 35V.
5. What's the max supply voltage when driving a 4ohm load?
The maximum supply voltage is limited by the MOSFETs (50V) and the Schottky diodes i used (only 40 V). And of course also by the temperature of these parts...
6. Are the 1000uF caps the only caps or are there more in the powersupply?
There are plenty of caps in the power supply. The two 1000uF Caps sit directly behind the MOSFETs to compensate resistance effects in the power cables. You can see them on the pictures. There is also a 100nF cap, which I forgot to draw...
The power supply has about 15.000 uF, I think. A bit more would be good, or better some regulation. Haven't done that yet... The amp is fine, but the preamp picks up a very slight hum - 50Hz is a frequency which is already boosted by the preamp, so it is a bit sensitive here. But it's not really disturbing.
7. What do you adjust the Vgnd to? (6V?)
Yes, around 6V. Basically, I watched the triangle wave and the inverted wave on the scope, and adjusted it, until they were symmetric. It's probably better to use a MAX for the wave generation - the OpAmp solution is a bit dodgy, and difficult to adjust.
8. Would there be any benefits in using better mosfets?
Better MOSFETs and better Schottkys means higher supply voltage means more power of course there is a limit, but there is definitely room for improvement.
9. What are the specifications of the output chokes (brand, amperage)?
The chokes are standard chokes for powerline filtering. Brand is unknown, I got them from Jaycar Electronics. I think they are rated 10A. So, nothing fancy...
Conclusion:
My next Amp will have a bipolar Power supply. I just did it this way because I wanted to use this transformer I had. Ultimately I'd like to have a switchmode power supply...
Also, I will use better diodes and mosfets with a higher rating, to have some more headroom. The chokes will be smaller (lower inductance), probably around 47uH, to get a full range response. They shouldn't be too low, since they have to store the energy of the 50% on-cycle in the BCA principle, and send it back over the diodes.
And it will be on a proper PCB, with consequent use of SMD parts. The MOSFETs will be somewhere on the edge of the board, that I can fix them on a heatsink, i.e. the back wall of the final amp.
The IR2112 are great, fast enough and anything you want to have. The KA319 is also fine. Maybe the OpAmp circuitry can be replaced by a more precise MAX chip.
1. Have you measured the output power, or are you just guessing (4ohm, 150W)?
Something in between... I measured the output with a scope, into an 8 Ohm load, and got an amplitude which was almost +/- the supply voltage. So I assumed this would be the same for 4 Ohm load, and multiplied. I used two 4 Ohm 10W resistors in series, cooled in a glass of water, and they got VERY hot. That's why I didn't want to torture only one of them with twice the power.
You are right, I really should do some more measurements... It's maybe a bit less than supply voltage, but approximately it is right. The power into resistors is calculated by amplitude^2 / (2*R_load). So (35V)^2 / (2*(4Ohm) =153 W
2. What trafo are you using for the amp?
My trafo is an ordinary 24V 200VA trafo, rectified and filtered with chokes and some big caps. Bigger is of course better, but I couldn't afford it.
The Modulator is driven from a separate 12V secondary, rectified and regulated with a linear regulator3. Are you using it for 4 or 8ohm subwoofer?
My speaker is a 4 Ohm 30cm (12") Woofer, in a 55l closed box. To achieve a flat frequency response below 20 Hz, I used the Linkwitz transform circuit described here.
The final system is more or less linear down to 16Hz, and drops off a bit below that. The speaker has a displacement of 500-600ccm at full volume. It's very impressive - sometimes I think a car drove by, but in fact it was in the movie/music...
4. If it is 150W in 4ohm, would you recommend a 225VA trafo, 300VA trafo or even bigger?
I'd say, as big as you can get... I'd actually prefer a regulated SMPS, but I couldn't find any with 35V.
5. What's the max supply voltage when driving a 4ohm load?
The maximum supply voltage is limited by the MOSFETs (50V) and the Schottky diodes i used (only 40 V). And of course also by the temperature of these parts...
6. Are the 1000uF caps the only caps or are there more in the powersupply?
There are plenty of caps in the power supply. The two 1000uF Caps sit directly behind the MOSFETs to compensate resistance effects in the power cables. You can see them on the pictures. There is also a 100nF cap, which I forgot to draw...
The power supply has about 15.000 uF, I think. A bit more would be good, or better some regulation. Haven't done that yet... The amp is fine, but the preamp picks up a very slight hum - 50Hz is a frequency which is already boosted by the preamp, so it is a bit sensitive here. But it's not really disturbing.
7. What do you adjust the Vgnd to? (6V?)
Yes, around 6V. Basically, I watched the triangle wave and the inverted wave on the scope, and adjusted it, until they were symmetric. It's probably better to use a MAX for the wave generation - the OpAmp solution is a bit dodgy, and difficult to adjust.
8. Would there be any benefits in using better mosfets?
Better MOSFETs and better Schottkys means higher supply voltage means more power
of course there is a limit, but there is definitely room for improvement.9. What are the specifications of the output chokes (brand, amperage)?
The chokes are standard chokes for powerline filtering. Brand is unknown, I got them from Jaycar Electronics. I think they are rated 10A. So, nothing fancy...
Conclusion:
My next Amp will have a bipolar Power supply. I just did it this way because I wanted to use this transformer I had. Ultimately I'd like to have a switchmode power supply...
Also, I will use better diodes and mosfets with a higher rating, to have some more headroom. The chokes will be smaller (lower inductance), probably around 47uH, to get a full range response. They shouldn't be too low, since they have to store the energy of the 50% on-cycle in the BCA principle, and send it back over the diodes.
And it will be on a proper PCB, with consequent use of SMD parts. The MOSFETs will be somewhere on the edge of the board, that I can fix them on a heatsink, i.e. the back wall of the final amp.
The IR2112 are great, fast enough and anything you want to have. The KA319 is also fine. Maybe the OpAmp circuitry can be replaced by a more precise MAX chip.
Thanks for answering all my questions!
Here we go again :
1. What MAX chip would be a good replacement to make the triangle wave?
2. When are you going to try a bipolar power supply?
3. Can this amp be converted to a bipolar power supply? How? (Is it just to exchange the VGnd with real gnd, and change the gnd to -12V? What to do in the 35V section?)
Here we go again
:1. What MAX chip would be a good replacement to make the triangle wave?
2. When are you going to try a bipolar power supply?
3. Can this amp be converted to a bipolar power supply? How? (Is it just to exchange the VGnd with real gnd, and change the gnd to -12V? What to do in the 35V section?)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.