My very first Class D pwm (switching) amplifier. - Page 32 - diyAudio
Go Back   Home > Forums > Amplifiers > Class D

Class D Switching Power Amplifiers and Power D/A conversion

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 1st October 2003, 07:45 AM   #466
IVX is offline IVX  Russian Federation
diyAudio Member
 
IVX's Avatar
 
Join Date: Jul 2003
Location: Everywhere (Buddhist's context)
Hi Jaka Racman,
<output inductors on the same core> it is interesting.. And now it is the transformer too.. how to phase wires (beginning / end) & what advance?
Best regards
__________________
Best regards,
E1.
  Reply With Quote
Old 1st October 2003, 09:02 AM   #467
diyAudio Member
 
Join Date: Dec 2002
Location: Slovenia
Hello,

Quote:
What would be the effect of common mode voltage on the output capacitor?
Your feedback circuit would have hard time rejecting 100V switching transients and also possible common mode ringing of output capacitor and filter inductor.

Quote:
how to phase wires (beginning / end) & what advance?
One dot at output of half bridge, another dot at filter capacitor.

Sorry, but I currently have no time for a more detailed answer.

Regards, Jaka Racman
  Reply With Quote
Old 1st October 2003, 09:09 AM   #468
diyAudio Member
 
Join Date: May 2002
Location: Switzerland
Quote:
Your feedback circuit would have hard time rejecting 100V switching transients and also possible common mode ringing of output capacitor and filter inductor.
That's a clever consideration I never thought about personally.
Because capacitor related distortion is increasing with DC bias it might be a good idea doing a mixed approach like some app notes, by TI and the likes, show: Connect one part of the filter capacitance between the two inductors and another from the two inductors against ground.

Regards

Charles
  Reply With Quote
Old 1st October 2003, 03:08 PM   #469
diyAudio Member
 
Join Date: Dec 2002
Location: Slovenia
Quote:
Connect one part of the filter capacitance between the two inductors and another from the two inductors against ground.
Yes, this is how it is usually done. Another possible approach is to split output capacitor into 2 series capacitors and then have midpoint connected to series RC damping network to ground.

Best regards, Jaka Racman
  Reply With Quote
Old 1st October 2003, 07:14 PM   #470
IVX is offline IVX  Russian Federation
diyAudio Member
 
IVX's Avatar
 
Join Date: Jul 2003
Location: Everywhere (Buddhist's context)
Jaka Racman,
Whether correctly I have understood you?:
2 inductors on the single core?
The magnetic stream should be opposite?
Probably so done for avoidance of saturation of the tiny and closed core?
__________________
Best regards,
E1.
  Reply With Quote
Old 1st October 2003, 09:01 PM   #471
diyAudio Member
 
Join Date: Dec 2002
Location: Slovenia
Hi IVX,

no, it is a differential mode inductor. Imagine a single layer toroidal inductor. Now cut existing winding in half and insert filter capacitor. You now have effectively two series connected inductors with filter capacitor in the middle. Transformer action of the windings will center capacitor voltage in the middle of the H bridge and prevent any common mode voltage swing of the output capacitor.
Coupling factor should not be to high, to allow some imperfection of switching times of both halves of the H bridge. This is usually achieved by aforemetioned winding technique on toroidal inductor or one winding on each leg of gaped UU ferrite core.

Best regards, Jaka Racman
  Reply With Quote
Old 2nd October 2003, 05:17 AM   #472
IVX is offline IVX  Russian Federation
diyAudio Member
 
IVX's Avatar
 
Join Date: Jul 2003
Location: Everywhere (Buddhist's context)
THX Jaka Racman,
<The magnetic stream should be opposite
for avoidance of saturation of the tiny and closed core>
seems it is true too & core can be smaller? wrong?
Bect regards.
PS: UU=E?
__________________
Best regards,
E1.
  Reply With Quote
Old 2nd October 2003, 05:47 AM   #473
subwo1 is offline subwo1  United States
diyAudio Member
 
subwo1's Avatar
 
Join Date: Jan 2002
Location: North American Continent
Hi all,

Now the amp has been changed and simulated to 20kz response. The feedback is now applied to both transistors in the differential pair. The performnce of the circuit has improved noticebly. The differential feedback permits the inductor and capacitor values to inadvertantly be of slightly different values, which is likely even if the nominal values are the same. Otherwise, the oscillation will not work correctly. The oscillation frequency is doubled from the prior subamp version, the power supply may need to be lowered to maximum +/-45 volts with fan cooling for full power output.


The new circuit can be done easily by stacking one IR2113 on top of the other. This way, the circuit paths are very short making the size small and EMI less. The pins not sharing the exact same
circuitry like the bootstrapped power supplies and the input and
output pins can be bent outward. Then the components can be soldered directly onto those pins. The entire amp, not counting the heatsinks, output chokes, and filter capacitor should take up an area of no more than 9 square inches. I think the amp should be able to put out 500Wrms if the mosfets are cooled with a fan, the power supplies are +/-45V at, and the mosfets are mounted on a 3X3 inch microprocessor heatsink. The link at the bottom gives ideas for building high power switch mode power supplies.

As before, the power for the control circuitry on the lower rail can
be regulated by the second differential transistor for more possible control over the DC offset.

As with the older design, the oscillation frequency can be
lowered by increasing the output filter choke values or the output
filter capacitors. The oscillation frequency can also be lowered by
increasing the value of the base resistors, R1 and R2 of the
differential BJT pair. The frequency can be increased by decreasing the values of those same components. The frequency should idle at 200khz and drop down to around 60khz at peak output. This variation of frequency raises the efficiency a lot because the mosfets have to switch less at full power.

I would like to be able to build the circuit to test it out. But, in case I don't get the chance, I am sharing the design beforehand even though I cannot say if the circuit works in real life. At the least it would need some adjustments.

Thanks everyone for your help. It makes things a lot easier.

Best Regards
Attached Images
File Type: gif n-champ5.gif (12.9 KB, 1857 views)
__________________
USMPS
http://groups.yahoo.com/group/switchmode/
  Reply With Quote
Old 2nd October 2003, 06:15 AM   #474
diyAudio Member
 
Join Date: Dec 2002
Location: Slovenia
Hi IVX,

wrong, magnetic flux cancels in common mode inductor, in differential mode inductor it is summed together. So single core inductor is larger than than the separated inductor used before. Do not forget that the core must be gapped or have distributed gap.

UU core is like E core without central post. TV set flyback transformers tipically use UU cores.

Best regards, Jaka Racman
  Reply With Quote
Old 2nd October 2003, 06:39 AM   #475
IVX is offline IVX  Russian Federation
diyAudio Member
 
IVX's Avatar
 
Join Date: Jul 2003
Location: Everywhere (Buddhist's context)
Hi all,
subwo at +/-45v is 1000w@4ohm will be( in US as i know 110v AC, maybe..4diodes&1cap as PS? ). What Micro-cap talked about THD?
Jaka Racman,
OK,THX
Best Regards
__________________
Best regards,
E1.
  Reply With Quote
Old 2nd October 2003, 06:55 AM   #476
subwo1 is offline subwo1  United States
diyAudio Member
 
subwo1's Avatar
 
Join Date: Jan 2002
Location: North American Continent
Hi IVX,
Ah, that power figure you gave is right. That should cause about 30 watts peak power waste in each mosfet with 50mohm channel resistance. That should work with fan cooling.

I don't know how to check distortion. But it does increase with increasing frequency. One thing that should lower it some is that the differential pair should not experience crossover distortion, the worse sounding type IMHO.

I would use a SMPS and +/- PS because that is the easiest way I know how to get the amp to idle at 0 Volts. That would give you isolation from the power line because of the switching transformer.
__________________
USMPS
http://groups.yahoo.com/group/switchmode/
  Reply With Quote
Old 2nd October 2003, 04:12 PM   #477
ingab is offline ingab  Italy
diyAudio Member
 
Join Date: Oct 2003
Location: Rome
Default Board class-D ampli

Hi Sergio,
I am nterested in the board. Can I have a copy?
Thanx in advance

Andrea
  Reply With Quote
Old 8th October 2003, 12:21 PM   #478
ssanmor is offline ssanmor  Spain
diyAudio Member
 
Join Date: Jan 2002
Location: Madrid
Default New design

Hello all.
I have finished the re-design of my PWM amplifier. I have designed a PCB board. Good news! it is single-sided! so very easy to make. The size is smaller and it has much less components.

The circuit has been greatly simplified. It is based on a MAX4295 chip, which implements a 250KHz PWM modulator itself and provide complementary outputs. They are level-shifted exactly the same way I did in my previous design, with the same dead-time control, but the XOR gates have been eliminated, thus attacking directly the inputs of the IR2110 chip. The rest is exactly the same. This should work at the first attempt, but the only issue is feedback: I will try to implement it the same way I did before, that is, before the output filter.

Have a look at the schematics and tell me what you think! I have no time by the moment to test it. I am willing to send the PCB design / schematics / construction notes to anybody who has the intention to etch the board and do some tests and share them with us. What do you think?
Attached Files
File Type: zip amplif_clased_max4295.zip (18.2 KB, 687 views)
  Reply With Quote
Old 8th October 2003, 12:24 PM   #479
ssanmor is offline ssanmor  Spain
diyAudio Member
 
Join Date: Jan 2002
Location: Madrid
A preview of the board:
Attached Images
File Type: png max4295_board_preview.png (7.3 KB, 1177 views)
  Reply With Quote
Old 9th October 2003, 07:20 AM   #480
IVX is offline IVX  Russian Federation
diyAudio Member
 
IVX's Avatar
 
Join Date: Jul 2003
Location: Everywhere (Buddhist's context)
Hi ssanmor,
i have some IMHO's about amplif_clased_max4295:
A) Rfilter and Cfilter value is are no successfully chosen.
B) R14, C18, R6, Q1, Q2 is can be eliminated.
Best regards.
__________________
Best regards,
E1.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 03:27 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2