Help in Class D amplifier designing - Page 3 - 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 12th March 2011, 06:02 AM   #21
diyAudio Member
 
Join Date: Feb 2011
I have attached the pics of breadboard circuit and will send you soon the oscilloscope pics of each point.

Other problem is that the design is not able to run more than 30 seconds, my p-MOSFET IC gets burnt. Power supply shows current of 0.5 A under full load condition but yet MOSFET IC get burnt.

You open the pics in this sequence which will make you easily understand.
1- Overall pic of breadboard
2- Operational amplifier circuit
3- 555 timer circuit
4- Inverter IC and Mosfets ICs
5- Inverter, BJTs and MOsfets in zoom
6- Mosfets ICs in zoom
7- capacitive load
8-my working cabin (for power supply showing)
Attached Images
File Type: jpg Inverter, BJTs and MOSFETs in zoom.JPG (840.0 KB, 146 views)
File Type: jpg Overall pic of breadboard.JPG (198.8 KB, 116 views)
File Type: jpg Operational Amplifier circuit.JPG (174.8 KB, 100 views)
File Type: jpg 555 Timer circuit.JPG (173.8 KB, 88 views)
File Type: jpg Inverter IC and MOSFETs ICs.JPG (217.7 KB, 85 views)
File Type: jpg MOSFETs ICs in zoom.JPG (175.6 KB, 11 views)
File Type: jpg Capcitive load.JPG (163.8 KB, 32 views)
File Type: jpg my wroking cabin.JPG (152.2 KB, 36 views)
  Reply With Quote
Old 12th March 2011, 07:54 AM   #22
Pafi is offline Pafi  Hungary
diyAudio Member
 
Join Date: May 2004
Location: Budapest
Do I understand it well? You drive capacitive load without filter choke? (At least I don't see the choke connected to output, however it is on the shematic.)

I don't see any decoupling capacitor near MOSFETs.

Usually the second problem would be enough to ruin everything, but the first one is even more serious!

Where is C23?????????

Last edited by Pafi; 12th March 2011 at 08:00 AM.
  Reply With Quote
Old 12th March 2011, 08:09 AM   #23
Pafi is offline Pafi  Hungary
diyAudio Member
 
Join Date: May 2004
Location: Budapest
The power supply rails have to be decoupled as close to the MOSFETs as possible! For this you have to use C22+C23!

Layout: the smaller is the better! I suggest to forget the breadboard!

You mentioned PWM freq of 40 kHz, but I don't think it really useful. First of all specify the load impedance and intended bandwidth!
  Reply With Quote
Old 12th March 2011, 08:21 AM   #24
Pafi is offline Pafi  Hungary
diyAudio Member
 
Join Date: May 2004
Location: Budapest
The modulator doesn't seem to be a PW modulator. It's more like a frequency modulator, PWM is only a side-effect. I don't know what do you really need.

Is this some kind of ultra-sonic experiment?

Last edited by Pafi; 12th March 2011 at 08:29 AM.
  Reply With Quote
Old 12th March 2011, 03:27 PM   #25
diyAudio Member
 
Join Date: Dec 2003
Location: Munich
Great that this thread is gaining attention.

Yes, as Pafi says.
The smaller the layout the better.
If I see it correctly, then the output choke is a drum choke which is in series to the speakers - connected from and to the speaker board. Basically this should work.

In an earlier posting I mentioned to put the supply rail cap close to the Fets and keep loops small.
On the breadboard I am not seeing any C22&C23 at all.
One of the most critical loops is:
Starting from upper MosFet through lower MosFet through C22, C23 back to upper MosFet.
In your set up this loop seems to be closed through the lab supply only!
Probably there is reasonably size cap, but far away with huge series inductances....

When your lab supply shows 0.5A current at full load, then this means that the average current is 0.5A.
That's already a lot and would normally fit to a proper working amp which has to drive full power normal music program from a 12V full bridge into a 4 Ohms load. 0.5A as average current means multiple amps peak, that's why the limitation has to be set higher. How much higher, depends on the internal structure of the supply. (For supplies with large output caps it might be not necessary to set the limit higher, because the peaks would be delivered completely from the caps.)
BTW which impedance does your speaker board have?
Probably you are not transferring the power to the speaker, but into heavy resonances of your bread board set up, because of giant loops ( means parasitic inductances) - resulting in catastrophic gate drive and defetcs.

The first thing you should try, is to place a cap right from the source pins of the upper MosFets to the source pins of the lower MosFets.
Since your breadboard has huge wire inductances between upper and lower MosFets anyway, the type of capacitor will not be very important.
Even a medium size electrolytic cap will have a comparably low series L.
So for first step simply try to place a 220uF e-cap (anything between 100uF-470uF) from the source pins of the upper MosFets (pin 1&3) to the source pins of the lower MosFets (again pin1&3) - try to have this cap connection as short and direct as possible.
If this helps, you may add 1000uF...3300uF + 1uf ceramic in the same way.
Let's hope that this would already bring the required improvement.
If not, you probably you have to rework the entire set up with an optimized bread board or better a proper PCB (as Pafi already mentioned).

For better understanding.
Every wire connection has an unavoidable inductance. Also every wire inside components.
As a rule of thumb you may anticipate approx. 5-10nH for every cm of wire.
Furtheron there are parasitic capacitances inside the components and also in the bread board.
If you add all the inductances of your bread board in the schematic and start calculations
...and if you then start to add also the parasitic capacitances
That's why people call the layout: Hidden schematic.
  Reply With Quote
Old 14th March 2011, 04:27 AM   #26
diyAudio Member
 
Join Date: Feb 2011
Hello Pafi, thank you very much for your responses.

Yes you are right it is ultrasonic experiment. I have used inductor (choke) to balance the capacitive load reactance (which you can find in figure no.3)

Inductor =56uH
Capacitive load = 142.5nF (50 ultrasonic transducers are used in parallel and each has capacitance of 2.85nF)

Bandwidth is audio frequency around 20KHz.
  Reply With Quote
Old 14th March 2011, 05:31 AM   #27
diyAudio Member
 
Join Date: Feb 2011
Yes Pafi and ChocoHolic, I am missing C22 and C23 capacitors. ChocoHolic you are right you mentioned about that but I did not get your point correctly.

I used:
C22 = 2200uF
C23 = 271pF

but when I observed 12V rail again it had some impulses yet, so I increased little capacitance

C22 =2200uF + 220uF (both electrolytic)
C23 =271pF (non electrolytic)

and observed 12V supply under full load picture is attached along with output of MOSFETs under full capacitive load, we can observe that there are no impulses or very less impulses means good rectangular waveform.
Attached Images
File Type: jpg WEA0001.JPG (20.7 KB, 25 views)
File Type: jpg WA004001.JPG (18.4 KB, 21 views)
  Reply With Quote
Old 14th March 2011, 06:16 AM   #28
diyAudio Member
 
Join Date: Feb 2011
@ChocoHolic, thank you very much for your detailed explanation for my understanding.

@both Pafi and ChocoHolic,
Small Layout: you are right for best performance I need to have PCB, but at this moment I need to experiment to replace this 555 timer IC with my DSP based algorithm so I need to stick to breadboard for few experiments.

Yes, to reduce the stray capacitance and parasitic inductance I will try to redesign on breadboard with optimized way.

@Pafi, you were saying that it is frequency modulation with side effect of PWM, so please can you share this point in detail, means you are saying that this 555 timer IC is working in the fashion of frequency modulator!!!....please elaborate this.
  Reply With Quote
Old 14th March 2011, 06:23 AM   #29
diyAudio Member
 
Join Date: Feb 2011
It looks like that I have got the solution of perfect rectangular wave at the output (which was the cause of burning of MOSFET IC), but the SPL (Sound pressure level) has not increased, even small; it is exactly same as previous.

my next step will be to make the circuit compact on breadboard with less loops, any other suggestions regarding increasing the sound level it is very small.

Thank you very much.

Regards
Dileep

Last edited by dileepkella85; 14th March 2011 at 06:28 AM.
  Reply With Quote
Old 14th March 2011, 11:41 AM   #30
Pafi is offline Pafi  Hungary
diyAudio Member
 
Join Date: May 2004
Location: Budapest
56 uH and 142 nF results a 56 kHz resonant frequency. Q factor is unknown yet.

20 kHz: from where? 20 Hz to 20 kHz? The amplifier can pass it through, but transducers can't!

I don't know what do you want, but it reminds me an experiment which is very interesting, but works completely different from the normal amplifier+speaker. The very high level ultrasonic signal have to be amplitude modulated, and on the target it is demodulated by non-linear effects. Is this what you want?
  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


Similar Threads
Thread Thread Starter Forum Replies Last Post
Help with designing differential amplifier krish2487 Solid State 12 24th February 2009 02:07 AM
help me in designing an amplifier nightcatz Chip Amps 4 22nd June 2007 06:57 AM
Designing power supply for class D amp darw82 Power Supplies 1 3rd December 2006 07:24 AM


New To Site? Need Help?

All times are GMT. The time now is 12:13 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