Thoughts on this limiter?

[MiXiN]

I've had a search about but can't find any reviews on it, so what's the opinion of this from those who've used it or are familiar with the circuit? http://www.ebay.co.uk/itm/STEREO-LIMITE ... SwB09YJcSr

[Albert H]

It's the bog-standard NE571 & LM339 limiter from the Philips compander cookbook. Unless there are surface-mounted parts under the board, it appears to have been "simplified". Don't believe the "no pumping" bit in the advertising and don't believe the bit about "fast attack" either.

This was originally suggested by Philips as a cheap stereo tape recorder limiter. If you go "over" on tape, it doesn't matter too much - you get a bit of tape saturation distortion but nobody gets hurt. On FM, it's a different issue - you have to limit, but you just can't go beyond 75kHz deviation or you cause interference - overshoots aren't acceptable. This thing will overshoot horribly! I used to use something a bit like this, but with a diode clipper after it. It sounded OK, but couldn't really be called "high quality".

[Banus_radio]

I have one here, im not impressed with it, it distorts before it limits, so you have to turn the input down to stop the distortion but then its not limiting and you get 'highs' coming through causing transmitter to over modulate. Luckily it only cost me £20 so its sitting in the junk box.

[MiXiN]

Nice one guys, that's £20 saved then!

[radium98]

try veronica compressor limiter or nrgkits are better

[yellowbeard]

You need level controls on the input and output with that sort of yoke, you set the input so it is acting on the peaks then turn down the output to suit the transmitter. It is only slightly better than nothing.
http://sbpcb.blogspot.ie/2012/12/stereo ... ne571.html

[NOYB]

My advice, just get an old laptop and run either Stereo Tool or Breakaway on it - you'll get results that approach the sound of a £10k optimod. They are so good that they can deliver totally legal, Ofcom compliant performance. Ideally use their ability to generate the MPX signal - processing is always best if audio and MPX is done in same box (the sound card needs to be able to run a sample rate of 192kbps, which most can nowadays)

There is lots of advice that you must use an expensive sound card, yet I've found that standard on-board sound card in a Dell laptop sounds great and when measured on an analyser meets all the specs you'd want.

The trick will be finding the right preset. Quite easy with Breakaway with its simple interface. Quite difficult with Stereo Tool - easy to adjust it so it sounds horrible! But adjusted correctly, Stereo Tool can sound great.

Less is more is what I always advise for processing!

[Albert H]

Banus

If it's running into distortion before limiting, someone has got the resistors wrong around the LM339. Lots of stations used this limiter back in the 80s. It was only acceptable if you pre-emphasised before it, and added a clipper after it. I found that it was better to do a 4 : 1 compression (using the NE571 and a dual op-amp) followed by a simple diode clipper. As long as the peaks were brief, you'd not really hear the clipping (the same principle that Orban uses!). This was about as much as you could do with a simple single band processor without a delay line.

The best thing that you can do is to use multiple bands, with differing recovery times according to frequency. You need to have a delay line in the audio path (before the channel selection filters) that is the same as the attack time of the compression elements, so that you can (effectively) get rid of the attack time. This delay only needs to be 500µs, but it's enough to allow the sidechains to detect the levels and begin the gain reduction. This completely removes the overshoots and pumping, and is the way that the BBC used to do their processing.

Audio processing for broadcast isn't often done properly, and if you get it wrong, it sounds horrible - most stations - illegal and legal - sound really nasty!