Next SINUS project??

[mikroman]

Sounds good! Any eq, stereo expander and similar actions before the limiter?

[hailstorm]

No. Everything flat, just PCM wav files ripped from CD. No pre-emphasis either. They're not stereo linked yet so each limiter acts independently. I'll try the stereo expander mod as well.
I'm also looking at building another two and making them dual band.

[radium98]

Thank you very much dear freind hailstorm , anyone could convert to me the gerber file to pdf , thanks .

[radium98]

excited to see avideo also . thanks sinus as well .

[ronald001]

Also a cool project: https://groupdiy.com/threads/texar-pris ... st-1089372

A guy that has cloned the legendary Texar Prism

[sinus trouble]

I'll work on getting a video done, meantime this is what the processor does.

audio input.png

audio output.png

This is a recording of the output, although Soundcloud have reduced the bit rate of my wav file so it sounds much better in real life.



Even the Cola Boy track survives without being mangled - surprising for a single band processor.

Nice work Hailstorm!

Also a wide variety of test tracks was great to see!

As you mentioned, Two units are needed for stereo and the "Expand" feature is simple to implement! Just a couple of resistors and maybe a toggle switch if you wish to disable the feature!

For those who may not have seen? Here is a quick video of stereo operation!

[Albert H]

The problem you'll have - unless you link the sidechains - is that the stereo image will "wander". If you get something loud in one channel, only that side will limit.... It usually sounds weird and is quite unsettling to listen to for much time!

[rigmo]

The problem you'll have - unless you link the sidechains - is that the stereo image will "wander". If you get something loud in one channel, only that side will limit.... It usually sounds weird and is quite unsettling to listen to for much time!

how to marriage

[Albert H]

NRGlimmods+.png

These are the modifications for stereo for the NRG limiter. You'll see that I used a common sidechain with samples from both channels combined into the rectifier section. The FETs have to be matched for pinch-off voltage - I'd usually buy a bag on 2N3819s from CPC, Farnell or Rapid, and pair them up on a simple test jig. If I bought 50 FETs, I'd usually get about 20 accurate pairs, and the ones that didn't match to any others were used for the mono limiters!

If you're really critical you could use separate sidechain rectifiers for each channel and combine them at the diodes (as a sort of "wired OR" gate), but there's only very marginal benefit in a single band limiter like this.

Obviously, the Sinus version of this limiter doesn't have the Toko 15kHz filter. It's not too difficult to build a replacement for that filter, and I'll put up a good reliable circuit for one if anyone's interested. My filter is an "elliptic" type, uses 10n capacitors all the way through (ideally 1% types) and 1% resistors, and it's pretty accurate, providing an accurate 50µs pre-emphasis from 3kHz upwards (that's the actual specification you want incidentally) and it turns over at 15 kHz and is about 56dB down by 19kHz. Optionally, you can add a tunable notch filter at 19kHz if you really feel the need....

[Albert H]

Perhaps the next project should be a two or three-band limiter, perhaps using transconductance amplifiers for the gain control. (I like to use them in the negative feedback loop of a good quality op-amp to cancel their noise and distortion characteristics). The LM13600 (or 13700) is still widely available and contains two accurately matched transconductance amplifiers in a single package. I've used pairs of cascaded op-amp / TCA gain cells for wider dynamic range, and got great results.

The band filters need to be phase linear types, and the recombination of the bands is easily achieved with a simple "virtual earth" mixing stage. The beauty of split band limiting is that you can easily apply different time constants to each band, so that you can minimise bass distortion, whilst keeping really tight and punchy-sounding control of the mid-range and the treble. When you look at the circuit, it looks like a complex design with lots of parts, but in reality, it's the same circuit (with differing values at some points) repeated either 4 times for a stereo two-band job, or 6 times for a three-band stereo limiter.

It occurred to me that if we're going to add lowpass filtering to the circuit, some of it could be done after the limiting stages, giving the option of adding clippers to each band (or the combined output) if they're wanted. I found a convenient bi-directional green LED type that goes into conduction very accurately at PPM 6 (1.9V RMS), which is ideal for final limiting prior to a transmitter!

[radium98]

to not duplicate the thread , i will ask here for some help .In 2006 when it still veroxxxx i bought as a kit the comp lim ver3 . and i copied it perfectly , not used till today i was trying to see how superb is it . It only differ that it does not have the 15khz filter and for my test i only applied a well known technics tuner , and also a 1khz tone generator both not seeing the led flickering , only this happen at power on the 3 led lit for seconds and one by one goes out . if i tune the trimpot slowly i see that volume increase or decrease but not autolevel and not hitting in leds . the only one component and i did not fit is a regulator 78l82 precision 8.2 v does this affect the functionning for the threshold ?
i attach the original comein schematic and my pcb , some help please .nb i fit a 7809 regulator instead of 78l82 and how if get drease the voltage in a way ?

[radium98]

Any help . My question in other way the regulator used in this model was 78l82 wich is precision 8.2v , and i have used instead a 7809 could this be a critical if yes how possible from 7809 to drop the voltage down to 8.2 other than put a diode 4007 in series
thanks

[rigmo]

Any help . My question in other way the regulator used in this model was 78l82 which is precision 8.2v , and i have used instead a 7809 could this be a critical if yes how possible from 7809 to drop the voltage down to 8.2 other than put a diode 4007 in series
thanks

if you have IC 7809 9.0 8.9v out with two 1n400x in series connection you can reach approx 8.3v depend of diode type and characteristic..

[jvok]

Theres no way 8.2v vs 9v makes a difference in a circuit like that. Not when the threshold voltage tolerance on the fet can be +/- several volts.

Are you staying you never fitted the 78L82 at all? If so no wonder it never worked right. Fit a 78L09 there and see what it does

[radium98]

Jock the 7809 is fitted already . The circuit seems to work .but the audio coming from a descent tuner on all frequency never hit even -6 db .I am sure everything is correct and original this was made years ago .only what differ is the reference

[Krakatoa]

Jock the 7809 is fitted already . The circuit seems to work .but the audio coming from a descent tuner on all frequency never hit even -6 db .I am sure everything is correct and original this was made years ago .only what differ is the reference

If the audio is too low now, you may have the J-Fet attenuating too much. Try adjusting the biasing potentiometer a little bit to make the J-Fet open and pass more audio, but make sure that it acts on loud audio passages, so that the limiting effect takes place.

[radium98]

i will make little video when possible , keep in touch

[radium98]

The only different components .an demonstrating video later . The 3led lit for fewseconds at start and settle . the audio barely lit for -6db led 1 ,if i vary the pot the audio go out or raise up , but i never notice an agc work . the sound i took for test a 0.775mv 1khz , and a cd player kenwood , and a tuner technic

[radium98]

i came this time with what i did in video , i hope it is now clear and you will help me.





thanks.

[hailstorm]

Here is my latest test of the Sinus Limiter set up as a dual band processor. An active crossover is used which splits the audio at 250Hz.



The single band version sounds great on it's own so this was only an experiment. I'd got enough PCBs to make 5 boards. However with separate bands you can tweak the mix of high and low, and have different release rates per band to suit your audio preference.

As before the source audio file is the same, no EQ or pre-processing but Soundcloud have compressed it - so ignor any digital artifacts.