Group Design: PWM Limiter Project

[verbos]

Here are some 555-used-for-PWM circuits...

http://musicsynthesizer.com/Corpse/VCEOM.html
http://musicsynthesizer.com/DIY/Grant/CVtwister.html

This isn't a 555, but it uses PWM to do some tricks to control voltages. The PWM "VCA" is used for crossfading in this case. PWM is good to get vary linear voltage control of a control voltage. The problem with this circuit is that the 21kHz triange will only give you about 10kHz bandwidth. If you would run audio through this, it wouldn't sound too nice. This is a Buchla control voltage processor, from the 200 series modular synthesizer.

http://musicsynthesizer.com/Buchla/b257.jpg


GET BACK TO WORK!!!

I'd hate to see this thread die.

mark

 

[verbos]

hey maybe using a DG202 or other FET switching chip would be a better idea. You know that they can pass audio from rail to rail at +/-15 volts. And they switch from anything, so you could use a 5 volt signal for the control path. something like the Buchla above would work, just tweak it way up in frequency (the CMOS 4016's are WAY slower than DG200, DG202, DG442 etc)

You could steal the sidechain from your favorite compressor and have something to build right away. OR, steal the envelope follower from the Buchla 230 just to be funny.

Mark

 

[Svart]

I totally forgot about this thread.. Lets get back to it..

 

[SSLtech]

Well it was a good while ago, but it worked as posted pretty much okay.

I didn't do the R/C thing, and I only fed it audio frequencies, so I was aware of PRR's admonition that my filters wouldn't be much protection against RF aliasing...

Other than that, if you build it, it should compress. -I'd definately add some passive filtering to the input and (especially) the output though.

Keith

 

[Rob Flinn]

Svart

I hadn't forgotten about this !!

I've been doing a bit of work on it.

I used the balanced input from the GSSL coupled to a 4 pole filter that I designed using "the active filter cookbook"

I got the fet working with a pulse generator!

I was working on the PWM part of it with a circuit like Keiths using a 555 etc. but I found it to not be very stable & was getting dc shifts when the pulse width changed.

So I've started looking at the SG3524 PWM chip, but haven't really done anything with it yet. I've been trying to get some application notes, but most of them seem to relate to using it for a switching PSU.

 

[Svart]

Rob and all, I really like the 3525 IC. much easier to work with and fewer parts than a decent 555/556 setup.

I was thinking that you could drive the error amp with a rectified and conditioned version of the signal..

 

[Rob Flinn]

Svart

I'm having problems trying to understand how to get the chip working. The data sheets seem to asssume prior knowledge

 

[Svart]

which manufacturer's datasheets are you looking at? Some of them are really lacking in info.

How are you attempting to hook it up? What is it driving? there are some misleading circuits on the internet also.

the easiest way to hook it up is this:

pins:

1(INV in) hook to 9 (comp)
2 your voltage input, 0-4v
3 not used
4 not used
5 cap to ground, try 4n7
6 resistor to ground, try 10k
7 resistor to pin 5 between cap and IC, try 100R
8 cap to ground try 10uf
10 reset/shutdown
11 out A
12 gnd
13 VCC
14 out B
15 V in
16 V ref

That should get you working. Another thing they don't tell you is to load the outputs or else your waves will be deformed.

You can see the triangle ramp wave on the OSC pins.

for 100% duty cycle you can do two things, run each output through a signal diode and load it with 200R that will clean the squares up nicely.

OR

Use two switching FETs, each with 50% duty. this is the preferred method of power switching.

Let me know if you need more info.


:thumb:

 

[Rob Flinn]

Svart

Thanks for that, I will try it tomorrow when I have located the chips !!!

 

[CJ]

I saw at the AES that DW Fearn still has his PWM tube compressor for sale.
But don't bootleg it, the guy is too cool.

 

[bcarso]

This reference has what I usually think of when the term Wilson current source (or sink) is used:

http://www.kettering.edu/~bguru/CM/CM-04.pdf

 

[gyraf]

[quote author="CJ"]I saw at the AES that DW Fearn still has his PWM tube compressor for sale.
But don't bootleg it, the guy is too cool.[/quote]

Yep. But the PWM/VCA/sidechain part in it is not actually Fearn, but directly borrowed from Cranesong.

Jakob E.

 

[bcarso]

[quote author="mediatechnology"]bacarso - your cite is a better one and the toplogy I've used before. Camenzind in his original book had that one too and it dissappeared in the later pdf.[/quote]

That came up on google as I looked for something with an image. I was amused to see that the author's name is Guru (watch out!). I thought it might be a gag name at first, but he's a real guy and in fact touts his introductory semi textbook nearer the home page*.

I've done some sims with Wilsons and modified Wilsons and they look reamrkably fast and low distortion with matched components. With some emitter ballasting R they are also reasonably low noise, and contriving to make each corresponding transistor balanced thermally will minimize that distortion mechanism.

You can also cascade them "vertically" and combine some outputs to get integer values of current gain. You can do that in a single composite stage without incurring the voltage burdens of the stacked stages, but Wilson's magic of base current error cancellation begins to get lost.


*http://www.kettering.edu/~bguru He does appear to be devout in some tradition, based on some of the peripheral material on the site.

 

[Balijon]

Hi, I realize that this is a very old topic.

BSS used PWM protection limiters with 4066's on the outputs of their FDS-340 / FDS320 electronic crossovers.
I recall they worked very well and it was a relative simple circuit to build.
I worked on it some 25 years ago adding a noise-gate function to it.

I lost all schematics, can someone help me with a copy/scan?

 

[JohnRoberts]

Seeing this old thread makes me nostalgic for some of the talent that used to regularly post here.

Sorry I do not have schematics for a PWM limiter, but as an item of general interest, there is a relatively new speaker protection device based on a PWM chopper between the power amp and driver. Actually a pretty clever approach as it sends less power to the speaker and draws less current from the amp, when protecting. I think Eminence (raw speaker manufacturer) is selling these.

JR

 

 

[gyraf]

.. also, take a look at the Studer 961/962 mixers, they have a really good-sounding PWM comp/lim in output section.

all studer schemos available at ftp.studer.com

Jakob E.

 

[Balijon]

Thx, Jacob.
Actually the Studer 961/962 PWM comp/lim triggered me to start looking for the BSS schematics again.
I love it's characteristics.

I am looking for a simple adjustable fast FeedForward 'brick-wall' peak-limiter that I would like to add to a FeedBack optical-compressor.
BSS designed the FDS320/340 (and likely MCS200) PWM-limiter around a 4066 and 555, doubling it as a output-mute-switch function.
I added a simple circuit to the RMS rectifier to double the 'mute' as a noise-gate. I build a couple of rack-modules in the (long ago) past, for live concert on drums.
I have attached a picture of a BSS FDS320 crossover with 4 of those PWM-limiters inside.
I would be most grateful if someone could send me a copy of the schematics.

Theo

 

[Balijon]

Did anyone test the attached PWM signal-generator?
This is build around a SG3525A PWM-control-chip and runs up to 400kHz. http://www.onsemi.com/pub_link/Collateral/SG3525A-D.PDF
The original design comes from Clive Sinclair and Mark Case. http://robots.freehostia.com/Circuits/PwmGenerators/PwmGenerators.html
Extremely simple circuit with few components, the supply voltage can have a wide range (need not be +12V)
The smart part is the way the 2 outputs are tied to the 10k pull-up resistor, delivering the full duty-cycle control (0-100%).

Theo

 

[JohnRoberts]

In my (day job) product I generate a sine wave, using two PWM outputs from a micro. This is cheaper than adding a dedicated D/A convertor (PWM is free in the micro). I actually alternate the two PWM streams for less noise at low levels, and offset them for additional resolution. I the feed this through a few poles of passive LPF and into a small class D audio amp.  I get some low level birdies from the PWM clock, interfering with the class D clock, but it is more than clean enough for my application (only 12 bit A/D).

If that PWN signal generator makes 400kHz tones it's max clock rate would need to be extremely high for any kind of usable resolution, OTOH if the max clock is 400kHz that isn't much resolution at 20 khz (one part in 20 or just over 3 bits).

JR

 

[Balijon]

There is no micro in this section, it should perform as a 'stand alone' module.

Studer uses 300kHz as clock frequency in their 961/962 PWM comp/lim design.
Other brands range between 200-250kHz.
The PWM output signal drives the 'chopper' switch that passes through the audio.
It all depends on the duty-cycle what you 'chop-out' of the source signal. The high-pass filters 'smooths' the signal back together again, similar to DA converter filters.
Though it may look crude if look at a 'chopped' 20kHz signal-picture at the Fet/Switch port, after the filter section it is very clean.
It certainly helps if you have a fast switching element, a series+shunt 4066 setup is a simple and effective chopper and is designed to run up to 4MHz.
The advantage of PWM limiters is that they can work FeedForward, are extremely fast, can attenuate deep and are keeping the HF alive compared to alternative designs.
PWM-limiters are great for 'brick-wall' AD application and they saved me a lot of re-coning drivers in the 'PA-touring-days'.

Theo