Albert H wrote: ↑Sun Feb 05, 2023 11:13 pm Here's the initial mods I did to the NRG Limiter to make it work properly in stereo. I added the left and right together to make the feed to the sidechain common, so there was only one sidechain needed. There were a couple of issues - the FETs had to be matched for pinch-off voltage, and it really needed an anti-overshoot clipper after it, with a further audio filter to shave off the upper harmonic clipping products.

We prototyped it, and it sounded great, but Stephen decided to go with two sidechains so that it could be separated (with minor modifications) for use as two separate mono limiters.

The circuit given below works well (with the above caveats), and I have a Veroboard layout that I did for the prototype and a PCB layout that I etched (by hand, as a one-off) - both worked well.
NRGlimmods+.png

You'll also notice that i used separate op-amps for each channel (NE5534) for the lowest noise and distortion. I also planned to replace the TL074 with an LM324, because it's a much cheaper part, and switches more slowly, making it much less likely to put switching noise into the audio as the LEDs light.

I also did a replacement op-amp based 15 kHz lowpass filter to replace the now unobtainable Toko Block Filter Modules.

Cheers Albert!

I do like the release jumper which i should have added to the design? But i guess modifications could be done to the RC combinations if needed?

The Stereo expander is very interesting? I still do not understand fully how it works?

I have some further previews of the PCB!

Albert H wrote: ↑Sun Feb 05, 2023 11:13 pm Here's the initial mods I did to the NRG Limiter to make it work properly in stereo. I added the left and right together to make the feed to the sidechain common, so there was only one sidechain needed. There were a couple of issues - the FETs had to be matched for pinch-off voltage, and it really needed an anti-overshoot clipper after it, with a further audio filter to shave off the upper harmonic clipping products.

We prototyped it, and it sounded great, but Stephen decided to go with two sidechains so that it could be separated (with minor modifications) for use as two separate mono limiters.

The circuit given below works well (with the above caveats), and I have a Veroboard layout that I did for the prototype and a PCB layout that I etched (by hand, as a one-off) - both worked well.
NRGlimmods+.png

You'll also notice that i used separate op-amps for each channel (NE5534) for the lowest noise and distortion. I also planned to replace the TL074 with an LM324, because it's a much cheaper part, and switches more slowly, making it much less likely to put switching noise into the audio as the LEDs light.

I also did a replacement op-amp based 15 kHz lowpass filter to replace the now unobtainable Toko Block Filter Modules.

Looks good .Yup albert ,i believe some Compressors and Limiters operate on the same principles, but their effects and the reasons for using them are completely different.A compressor compresses, it reduces the dynamic range of its input signal. This means as the input amplitude varies over a certain range, the output amplitude varies over only a fraction of that range. In this case with every change in the input amplitude the output changes only half as much.a limiter passes its signal unaffected till the input amplitude reaches its threshold. at this point the limiter prevents the output increasing much by compressing its input much more strongly than in compressors,so its important piece of equipment for any pirate ,once I saw that launched on the nrg website I bought it ,when I set it all up for a test ,it worked and did its job .

I also had this boxed ,stereo version also ,

Based on theory and simulations, The PCB should give some decent results?

From experience of prototypes some adjustments may need to be made? And i see no reason why the current PCB cannot be customised easily with a few component value changes!

Do you also want to add an outboard lowpass filter?

Here's the 15 kHz lowpass I designed to replace the Toko block filters on the PRO IV stereo coder. It needs a bipolar supply (+/-12V), and each IC needs supply filtering capacitors from pins 4 & 11 to ground. We powered the ICs through 47Ω resistors with 22µ electrolytic capacitors at each power pin. I originally used the LF347 (I had lots of them) but the later prototypes used TL074 with exactly the same results: This filter is >55dB down by 19kHz, so you're very unlikely to interfere with the pilot tone.

All the resistors are standard values from the E24 range and are 1% metal film types. The 2n2 capacitors we used were 2% ones , but we measured little change in results if we used 5% parts. However, accuracy is a good thing, and I got the best results with a set of hand-selected parts that were accurate to below 0.5% - I selected them from a bag of 1000 capacitors!

We had our pre-emphasis stage on the way in to this filter (take a look at the PRO IV Coder circuit for details).

Nice one Albert!

I will see how the PCB performs and go from there! I cannot find ANY data on those TOKO filters! Which is frustrating cos i have no baseline to go by?

I have heard the term "Brickwall" associated with these filters? So i guess they have a substantial gain reduction above 15Khz?

If anyone does have any archived data on the TOKOs? I would love to see it!

sinus trouble wrote: ↑Wed Feb 08, 2023 12:00 am Nice one Albert!

I will see how the PCB performs and go from there! I cannot find ANY data on those TOKO filters! Which is frustrating cos i have no baseline to go by?

I have heard the term "Brickwall" associated with these filters? So i guess they have a substantial gain reduction above 15Khz?

If anyone does have any archived data on the TOKOs? I would love to see it!

There discontinued ,from what I've heard .15khz brickwall filter,I've not got the data for the tokos .

Simple 15khz passive filter and standard components, according to the author it works quite well, if this diagram has already been published here I am sorry for the error.

garada wrote: ↑Wed Feb 08, 2023 10:48 pm Simple 15khz passive filter and standard components, according to the author it works quite well, if this diagram has already been published here I am sorry for the error.

Cheers Garada!

No! That has not been posted!

However, I did consider passive filters in the design? See the problem i was asking myself was?

How can i reduce high frequency interference with what the circuit has already built in?

How can i reach a compromise between complexity and performance?

The TL074 has a spare OP AMP in the circuit that is not used? So the obvious answer was to use this as an active filter!

Granted it does not have the high spec of "Brickwall" filters! But may offer a solution which would reduce interference without degrading the high frequency audio response!

Infact! I suspect Stephen realised this problem many years ago!

It just so happens that the TOKO filter did a better job than an active filter which is why he went with that option!

Unfortunately today, The TOKO is NOT an option!

Anyways, I have received the PCBs today!

There are a few parts that i do not have in my stock currently!

The PCBs look Sh*t hot!! and more updates to come soon!

sinus trouble wrote: ↑Wed Feb 08, 2023 12:00 am Nice one Albert!

I will see how the PCB performs and go from there! I cannot find ANY data on those TOKO filters! Which is frustrating cos i have no baseline to go by?

I have heard the term "Brickwall" associated with these filters? So i guess they have a substantial gain reduction above 15Khz?

If anyone does have any archived data on the TOKOs? I would love to see it!

The Toko filters were either BLR3157N or (less favourably) BLR3107N, and turned over at 15 kHz.

Both needed to be sourced from and terminated with 4k7Ω (the "Ambit Catalogue" was wrong about the '3157 termination - these filters were symmetrical!) .

The '3157 gave better results at the expense of 3dB loss in the passband. Both were "good enough" for our stereo coders, and would prevent the "blinking stereo light" problem with mono transmissions. They were manufactured with pretty tight tolerances, and were well matched, so we never had issues with channel mismatches.

Beneath the modules, there were about 10 pins, with all but two of them grounded. We used them in the limiters and in the Pro IV Stereo Coder.

The op-amp circuit I gave above was designed as a replacement for these block filter modules, and actually had much better performance than the passive parts (though they took up more board space and cost more!).

I redesigned the op-amp filters several times, for various capacitor values, but the circuit given earlier in this thread was just about the optimum, and didn't need any special component values. I will re-design it shortly for a single-rail supply, though the dual rail option is much better in terms of hum, noise and headroom.

The Pro IV Stereo Coder used +/- 12V for the op-amps, and +/-7V5 for the CMOS logic ICs (derived by zener droppers), and the whole board was powered from a 12V AC adaptor, feeding two half-wave rectifiers to derive the dual rails. It's still the best stereo coder kit !

Albert! You are a ledge!!

The original version seemed to be triple cavity helical filters? I did find the actual part number, But no info was available?

Its interesting that you mention the resistors involved with the operation of this type of filter?

From simulations? I have found these values need to be higher when using a substitute active filter!

Those resistors seem to play a big part in DC offset parameters?

When removed completely, The circuit no longer functions!

The P/N of the 3 section low pass filter was BLV5085.

Krakatoa wrote: ↑Thu Feb 09, 2023 7:45 am The P/N of the 3 section low pass filter was BLV5085.

Haha! Yes thats the little buggerz!!

Started assembling the PCB this evening!

Got most of the parts, But a few obscure components to order over the weekend!