pulse width amplifier - UK Vintage Radio Repair and Restoration Discussion Forum

Trevor · 8th Jun 2008, 7:39 pm

After many years of audio equipment building success wth class d has always eluded me
I can remember as a small boy reading of the new leap forward the sinclair X10 well after a period of 40 or so years I have finally evolved a simple design that features a 100 khz clock a supply of 12-30 volts dc .while the design is not hifi it provides a rewarding experiance in both building and listening
please see attachment
Devices are alll very low cost and easy to obtain triange generator is with a TL072 pulse width comparitor is via a LM 339 chip
output devices are with a pair of TP31/32

Steve Smith · 9th Jun 2008, 8:14 am

I made one of these a few years ago.

I used a 555 oscilator running at 50kHz but replaced the resistor from positive to pin 7 with a constant current source. This gives a linear triangular wave at pin 2.

This triangular wave goes to one input of a comparator, the suitably buffered audio goes into the other input.

The comparator output then drives a 'totem pole' output stage similar to an analog audio amplifier.

I didn't bother with any filtering out of the output (although I should for the reason stated above!) as the speaker impedance at 50kHz is so high, the power draw from the speaker at this frequency was negligible.

It worked very well and had a good quality sound. I had planned to develop it a bit further but like the hundreds of other things I start.....

Steve.

Trevor · 9th Jun 2008, 8:19 pm

My amplifier is rather similar to yours Steve ! However I used a dual op amp comparitor integrator for the clock and then a comparitor LM 339 there are 4 in the chip for about 25 pence this was followed by a complimentary pair to give current boost
I know that this type of design will have limits due to poor filtering due to the low clock frequency but started at 100 Khz in an effort to use available output devices As explained this is to get started and I have in mind a faster output stage so that I can increase the sampling frequency and thus improve the filtering of the carrier
However I know that this is a worthwhile project in that a result can be obtained !
This should take but a few hours for any one to build and I feel sure that they will like the result for so little effort
Regards Trev

Steve Smith · 10th Jun 2008, 4:04 pm

I tried mine at both 50kHz and 100kHz and couldn't really tell any difference.

It's definitely worth experimenting with. I had plans to incorporate mine into a guitar amplifier which would not need to be totally 'hi-fi'.

Steve.

8th Jun 2008, 7:39 pm	#1
Trevor Octode Join Date: Aug 2006 Location: Stoke-on-Trent, Staffordshire, UK. Posts: 1,082	pulse width amplifier After many years of audio equipment building success wth class d has always eluded me I can remember as a small boy reading of the new leap forward the sinclair X10 well after a period of 40 or so years I have finally evolved a simple design that features a 100 khz clock a supply of 12-30 volts dc .while the design is not hifi it provides a rewarding experiance in both building and listening please see attachment Devices are alll very low cost and easy to obtain triange generator is with a TL072 pulse width comparitor is via a LM 339 chip output devices are with a pair of TP31/32 Attached Thumbnails Last edited by Dave Moll; 10th Jun 2008 at 2:26 pm. Reason: pictures now attached

9th Jun 2008, 8:14 am	#2
Steve Smith Retired Dormant Member Join Date: May 2008 Location: Isle of Wight, UK. Posts: 24	Re: pulse width amplifier I made one of these a few years ago. I used a 555 oscilator running at 50kHz but replaced the resistor from positive to pin 7 with a constant current source. This gives a linear triangular wave at pin 2. This triangular wave goes to one input of a comparator, the suitably buffered audio goes into the other input. The comparator output then drives a 'totem pole' output stage similar to an analog audio amplifier. I didn't bother with any filtering out of the output (although I should for the reason stated above!) as the speaker impedance at 50kHz is so high, the power draw from the speaker at this frequency was negligible. It worked very well and had a good quality sound. I had planned to develop it a bit further but like the hundreds of other things I start..... Steve.

9th Jun 2008, 8:19 pm	#3
Trevor Octode Join Date: Aug 2006 Location: Stoke-on-Trent, Staffordshire, UK. Posts: 1,082	Re: pulse width amplifier My amplifier is rather similar to yours Steve ! However I used a dual op amp comparitor integrator for the clock and then a comparitor LM 339 there are 4 in the chip for about 25 pence this was followed by a complimentary pair to give current boost I know that this type of design will have limits due to poor filtering due to the low clock frequency but started at 100 Khz in an effort to use available output devices As explained this is to get started and I have in mind a faster output stage so that I can increase the sampling frequency and thus improve the filtering of the carrier However I know that this is a worthwhile project in that a result can be obtained ! This should take but a few hours for any one to build and I feel sure that they will like the result for so little effort Regards Trev

10th Jun 2008, 4:04 pm	#4
Steve Smith Retired Dormant Member Join Date: May 2008 Location: Isle of Wight, UK. Posts: 24	Re: pulse width amplifier I tried mine at both 50kHz and 100kHz and couldn't really tell any difference. It's definitely worth experimenting with. I had plans to incorporate mine into a guitar amplifier which would not need to be totally 'hi-fi'. Steve.