sr1200
Well-known member
Didnt abechap do the PIE (pye) clone and use a chip for the "chopper"? Yep, its sitting next to me and sounds great. why not borrow that section from him?
PWM compressor
- Thread starter ruffrecords
- Start date
Help Support GroupDIY Audio Forum:
ruffrecords
Well-known member
sr1200 said:
I checked out the schematic on his web site and I did not see a chip for the chopper; just a transistor as in the original.
Cheers
Ian
Rob Flinn
Well-known member
ruffrecords said:
He used a 555 to create the oscillator pulse, but the actual modulation & pulse width is not done with the chip. The chip only provide a 250KHz pulse.
jasonallenh
Well-known member
Getting into some murky water here, but SKnote has a PWM compressor as well. I have one. Ask away. I think the PWM is 500kHz?
ruffrecords
Well-known member
jasonallenh said:
Have you had the lid off? What chips does it use?
Cheers
Ian
I think PIC18 will do 1MHz PWM with some sacrifice of resolution?
ruffrecords
Well-known member
MicDaddy said:
It just occured to me there may be a fundamental problem using PWM in a micro. Assuming we want 8 bit resolution in our PWM signal, this means the clock feeding the PWM counter needs to be 256 times the the required PWM frequency. Thismeans even if we want a PWM frequency as low as 100KHz, we need a clcok of 25MHz.
Cheers
Ian
Still possible.
I have not check for other manufacturer but from microchip, you can found in the PIC24 or dsPIC33E line high speed PWM with claim to have 7ns resolution. This means that at 10 bits you still have a PWM frequency of 130kHz.
DSPIC33EP32MC202 is at 2€30 at Mouser, so it is still cheap ...
I have not check for other manufacturer but from microchip, you can found in the PIC24 or dsPIC33E line high speed PWM with claim to have 7ns resolution. This means that at 10 bits you still have a PWM frequency of 130kHz.
DSPIC33EP32MC202 is at 2€30 at Mouser, so it is still cheap ...
ruffrecords
Well-known member
Chris_V said:
Interesting chip, and it can be programmed with my PICkit3. No SDIP proto baord that I can find but that is easy to fix.
Cheers
ian
Good to know Ian, I too have the PICkit3 and having a lot of fun with these ucontrollers.
jasonallenh
Well-known member
ruffrecords said:
I just pulled it out and checked. It has one of these:
http://www.vishay.com/docs/70054/dg444.pdf
I'd be interested to know what you think.....................
jasonallenh
Well-known member
ruffrecords said:
DG444/5 are well known and respected audio switches. Perfect for this job. Were there any other (none audio) chips near them?
Cheers
Ian
[/quote]
ruffrecords
Well-known member
jasonallenh said:
The TL5001CP is a basic PWM controller with built in oscillator. Although intended primarily for switched mode power supplies you can use it for generating a PWM signal just by feeding a dc voltage inoto its feedback input. The dc voltage is derived by the compressors side chain in the normal manner.
Cheers
ian
usekgb
Well-known member
ruffrecords said:
I use an Atmega328P to generate a 200kHz PWM signal at 50% duty cycle to drive an analog delay line. It can go faster, but 100kHz is the maximum frequency of the BL3208a chips that I used (each chip see fs/2 with original polarity and inverted polarity). The microcontroller generates any frequency that I need between 26kHz and 200kHz, depending on the delay setting, or the tempo tapped in. So, this is well within the ability of a microcontroller.
I have been investigating the use of a microcontroller for a PWM based guitar compressor for some time. I think the biggest limitation will be attack time, due to the time it takes the micro to do the a/d conversion and process the signal.
ruffrecords
Well-known member
usekgb said:
getting a hi enough frequency is not to hard. The problem is being able to alter the pulse width to the necessary resolution. If you want say a resolution better than 0.5% then you need 8 bits. This means the internal oscillator in the micro driving the PWM modules needs to be 256 times the PWM frequency. So if the PWM frequency is 100KHz, the oscillator needs to be 25.6MHz.
Cheers
Ian
> I think the biggest limitation will be attack time, due to the time it takes the micro to do the a/d conversion and process the signal.
A) A burst of overload in the front is not always bad. It tickles the ear into "hearing" the missing transient.
B) That initial milliSecond does not have to be *precision*. Scale a basic 1-bit analog pin into flipping when signal is large. Turn-down without much computation. (If still high, turn down again.) As the ADC path settles, now calculate where you want to gain-ride.
A) A burst of overload in the front is not always bad. It tickles the ear into "hearing" the missing transient.
B) That initial milliSecond does not have to be *precision*. Scale a basic 1-bit analog pin into flipping when signal is large. Turn-down without much computation. (If still high, turn down again.) As the ADC path settles, now calculate where you want to gain-ride.
usekgb
Well-known member
ruffrecords said:
Good insight Ian. I will need to take a look at the PWM library that I used, and see what the resolution is as the frequency changes.
usekgb
Well-known member
Maybe an ARM chip, like that used in the Arduino DUE, running at 80MHz, could do the job?
Sure a better micro helps, the ARM M3 in the due has a lot more than just faster clock, a closer look to the manual would help but the timing between the timer and the output pin I think is also something you may benefit from the M3 architecture, while in the 328 you miss a few clock cycles. Something important with PWM compressors is to have the gain range limited to the usable range, if you allow for 40dB range but never using more than 15dB attenuation you are loosing precious bits in the way, which takes with it speed, filter complexity, etc.
Maybe some auto range could help, making adaptive to swap between 12dB range cell and 36dB range, just swapping a resistor in the cell. Then is the fact of when use one or the other, how and when to switch, etc. Better specs up to 12dB range, when you need to go further you probably are having other problems to be looking at.
JS
Maybe some auto range could help, making adaptive to swap between 12dB range cell and 36dB range, just swapping a resistor in the cell. Then is the fact of when use one or the other, how and when to switch, etc. Better specs up to 12dB range, when you need to go further you probably are having other problems to be looking at.
JS
usekgb
Well-known member
It looks like the M3 has a 12 bit PWM resolution, but I would have to do some digging to see how this may change with a usable PWM frequency.
