I do not have high hopes in respect to all Limiters! The concept is great? But in practice they ALL perform horrible!

From a component and design point of view? It should be pretty robust! The TL074s are tough and can handle plenty of abuse!

Some tweaking of my LPF will be inevitable! Thats part of the fun!

Limiters are (generally) essential. The big problem with limiters that handle the whole frequency response is that a big bass pulse will "dip" the treble as well. This sounds horrible (though it's used as an "effect" on some EDM records)!

The only way to get a good sounding limiter is to split the audio by frequency. Most commercial broadcast limiters use three or four bands (mine use three in one version and four in another), but you can get acceptable results with just two bands with a changeover from one to the other at about 250Hz. I've built a few that way (using the old NE571 fast stereo limiter trick from the Philips databook), and they sound pretty good. I'll put the circuit up here if anyone's interested.

The other thing that most commercial broadcast limiters include is a clipper or two to handle overshoots (since ANY overdeviation isn't allowed!). These are usually followed by lowpass filters to get rid of the worst of the harmonic distortion they cause....

Over the last 30 years or so - since I've learned how to do it - I can get limiters to sound OK!

Another approach is to introduce a delay line into the audio, so that the limiter "sees the peaks coming" and pulls the gain down pronto - yoou can get a "zero attack time" limiter that way, but the delay circuits either include unobtainium parts or a hell of a lot of op-amps or LC "allpass" filters. This technique is probably beyond the resources of the average radio pirate!

sinus trouble wrote: ↑Fri Feb 10, 2023 12:13 am I do not have high hopes in respect to all Limiters! The concept is great? But in practice they ALL perform horrible!

From a component and design point of view? It should be pretty robust! The TL074s are tough and can handle plenty of abuse!

Some tweaking of my LPF will be inevitable! Thats part of the fun!

Something I came across for the blv5085, which is coloured in green ,that's all I could find .

Cheers Electronically!

The circuit looks great but still leaves unanswered questions?

I just noticed a small error i made the PCB design!

One of the traces i made on the schematic failed to connect? Therefore the PCB ignored it!

As it is such minor issue, I will not bother going into any further details unless requested?

This issue will also not affect the final Gerber release! So there is no need to panic!

sinus trouble wrote: ↑Sat Feb 11, 2023 1:20 am Cheers Electronically!

The circuit looks great but still leaves unanswered questions?

True very difficult to source ,probably your best bet is to redesign that area ,or if you find a substitute for blv5085 helical filter,closest match to that is blr3157 but different pinout,

Both those filters haven't been made since about 1998. There's no current equivalent.

I gave the circuit of the filter we've been using for a while. Before anyone asks - the component values ARE critical, and it has to be powered from a dual-rail supply (+/-12V). The capacitors we use are Wima brand polyester types with 2.5% tolerance. The resistors are 1% E24-series metal film, and the op-amps are TL074. The passband response of the circuit is excellent, with a roll-off at almost exactly 15kHz. By 19kHz the response is >40dB down.

Incidentally - the circuit above that "Electronically" posted is my Pro IV stereo coder, as supplied by NRG (until Steve Moss died). It's a good compromise design - reasonably cheap to build, but with superb performance!

Yea I came across that circuit online ,so thought if I show it here then it is correct that the blv5085 was used on those circuits ,but that's good to know its the pro IV Albert,I do know that they did used to call those filters, helical filters and notch filters back in the day ,
And also the 15khz filter was also on yours and Steve's NRG website, in the components list section,in the quartz crystals for sale section though .from what ive read about those filters I can understand why they call them brickwall filters .

I found a bode plot from the datasheet of the BLV5085. It came in paper form when I bought the last 2 filters in stock (the last of the whole world, I think) from Barend Hendriksen in 2007.
Unfortunately it is a copy of a copy of a copy and so on. I have it scanned at 600ppp and increased contrast but it still is crappy.

Enjoy.

I've got other filter designs, but this one is about as good as it gets. The weird resistor values ARE necessary, but I made them up in the Veroboard prototype from series and parallel combinations of E12 series resistors, and got the right result. I only had 5% 2n2 capacitors in stock, so I went through a bag of them with an Atlas™ tester and got the most accurate ones I could find - to better than 1%!

For anyone who doesn't want to download a zip of a pdf heres the graph from krakatoas post

Also you'll notice the blv5085 is in there also .

Krakatoa wrote: ↑Sun Feb 12, 2023 7:48 pm I found a bode plot from the datasheet of the BLV5085. It came in paper form when I bought the last 2 filters in stock (the last of the whole world, I think) from Barend Hendriksen in 2007.
Unfortunately it is a copy of a copy of a copy and so on. I have it scanned at 600ppp and increased contrast but it still is crappy.

Enjoy.

Excellent, yup I have information on barend on blv5085, your spot on krakatoa

Krakatoa wrote: ↑Sun Feb 12, 2023 7:48 pm I found a bode plot from the datasheet of the BLV5085. It came in paper form when I bought the last 2 filters in stock (the last of the whole world, I think) from Barend Hendriksen in 2007.
Unfortunately it is a copy of a copy of a copy and so on. I have it scanned at 600ppp and increased contrast but it still is crappy.

Enjoy.

Excellent, yup I have information on barend on blv5085, your spot on krakatoa

Thanks Krakatoa!

WOW!! That is some serious attenuation going on there! I can see why Stephen chose that option!

Whilst i cannot achieve that same performance with a basic active filter? The tech regarding digital media has improved somewhat? Infact, maybe worsened and less likely to retain any high fidelity these days!

There is a Maxim chip - MAX7415 that does it really easily:

https://www.mouser.co.uk/datasheet/2/25 ... 515904.pdf

If you clock the chip at 1.5MHz from a simple crystal oscillator, you have a very accurate 15kHz elliptic filter. I clocked two ICs from the same crystal oscillator. The "shape factor" means that it's >37 dB down by 19 kHz. When you look at the datasheet, you'll see that the circuit is trivially easy. Mouser has them in stock (though only in surface-mount packages it seems) for around a fiver each.

The performance of the IC LPF above isn't quite as good as my homebrew elliptic, but it's a lot quicker to build, and might be cheaper in the long run.

Sinus keep it as simple as yo can , so maybe as an amateur i can build one to me .Thanks , but i will do it single side and put a link , because i see it is double side , and i face problem in production , for lack of few materials in my hands .

Sinus - you really should include a 15 kHz lowpass filter. If you go with the Maxim (Analog Devices now) MAX7411CPA - it's an 8-pin DIP package and runs from 5 Volts. It really couldn't be simpler. You'll need a 5V regulator (78L05 will be fine) and a simple crystal oscillator running at 1.5MHz (or you could derive it by dividing your PLL reference crystal to 1.5MHz if you wanted). For stereo, your crystal oscillator and your regulator can feed both channel's filters. The filter circuit is simply:
You'll need DC blocking capacitors on the way in and out, but the circuit really couldn't be simpler! The specification is good enough for most broadcast purposes, but if you're really particular, you could cascade two of these filters per channel, which would give you a filter characteristic that would exceed any specifications worldwide!

radium98 wrote: ↑Tue Feb 14, 2023 6:33 am Sinus keep it as simple as yo can , so maybe as an amateur i can build one to me .Thanks , but i will do it single side and put a link , because i see it is double side , and i face problem in production , for lack of few materials in my hands .

That is my objective!

The original PCB by NRG was single sided and is easily possible with very little modifications!

The top side of my PCB are only DC power connections! All the signal tracks are on the bottom!

Sinus - you will have seen my adaptations of the NRG circuit for stereo. It only adds a few extra components to each channel, and you can have a single sidechain for both channels (so that you don't compromise the stereo image). The only minor issue is that you have to match the FET pinch-off voltages, but other than that, it's plain sailing!
Laying out the PCB so that it's just one-sided is tricky, but can be achieved if you're prepared to accept a couple of jumper wires (or "Zero Ohm Resistors" like NRG used to supply in some of the kits).