Serbian "Most Wanted"

[Davor]

I understand what you say, yet this solution produces a proportional tone, and the blow is not so obvious as with binary sound. I know how to make a superb sound without DC blow to the speaker, but it would require some more tweaking of the circuit. I'll make a circuit as per my ideas, and not as an upgrade to the existing circuit.

There are many "skeletons in the wardrobe" of nowadays detectors, and most of them are results of incremental upgrades. I tend to do incremental upgrades for sake of the existing builds. Most of my mods are applicable on the existing PCBs, and I believe these can make a notable difference. Real advance would be making a new circuit, just as you did.

I personally do not have a SMW PCB, but Ivconic promised to try my mods on his already finished board.

Here are my sims:

[Davor]

I started analysing the SMW gain blocks. For a start I'm not a bit happy with the gain of the first stages - they are prone to saturation. Maybe it seem a bit drastic, but I'd slash the gain of these quite considerably. Guess someone thought pumping up the gain here would make a better rig. Well ... it wouldn't. Also the Fisher-like phase shifting is not as wrong as I thought. Provided there is no saturation in the first stages, the Fisher discrimination shifting is working nice.

Low pass capacitors in final stages are not required, in fact they roll the phase too far. Point is that for a neat group delay of a pulse you need a smooth phase transition within a frequency range of interest. Even with those capacitors that remains the truth, but the pulse becomes delayed beyond the 90° point and hence goes into counter-phase.

I'll make my proposal of the corrected gain blocks soon.

[Davor]

Almost there. I tried some principle that I thought of some time ago that would significantly improve impulse response of the gain block. It seems I am going to do some redo in my IGSL after this as well.
It is about phase stretching. When complex filters are used, and all corner frequencies are set at the same spot, a phase gradient is promoted, and you have a phase reversal close by. If you detune the corner frequencies by some factor you get a filter that has a somewhat broader corner, and skirts of very similar slope, which is good.
So if constructed carefully, a finished filter will have a neat phase response in a much wider range, while the unwanted frequencies will still be well attenuated.
Such detuning is meaningful to a factor of about 10. I'll play with this a bit more and let you know. It is promising.

[mukadder1957]

can I use 2N4403 / KSP2907A = BC327 ?

[Davor]

Yes you can. These transistors are not critical.
__________________________________________________ __

And now something a bit different ... modified gain blocks. What you see is a schematic where discrimination potentiometer (together with 4k7 resistor) is moved to the very inputs of the gain blocks. It is the only place where they can't screw up differentiators' response. Ideally they'd be removed completely, and discrimination would be steered at the phase shifters.

Some component values are changed, and 2 resistors are added. They promote differentiating function of the device, and are critically optimised with other components to maintain a flat differentiating response. To check how beautifully they perform, see how the gain blocks react to an exotic input signal (in blue, exaggerated 100x) with straight slopes ... flat tops, no overshoots. Gains are slightly pumped up against the original SMW, but as I don't actually have a SMW to see if it is too much, I just can't tell. If it is so, just reduce the R5 and R10 resistors.

I'm convinced this solution will work nicely, and most probably I'll fix my IGSL this way too.

[dalgic35]

could you please upload here all the changes you

[dalgic35]

SMW Do you have the latest version of

[Davor]

This is still work in progress. The mod above was checked by Ivconic, and it settles too slow. There is a small deviation from the above gain blocks layout that performs correctly, but the work with SMW is not done yet. There are still some quirks with sensitivity. Proportional tone works as anticipated.

Don't hold your breath over these changes as both Ivconic and I are a bit preoccupied with some other things at the moment. It will happen soon enough.

[eduardo1979]

Here is my modified SMW. I added the tone Bulgarian threshold detectors. If anyone is interested in this project, here is the entire scheme. Greetings, Eduardo.

[emoxristov]

PCB

[eduardo1979]

PCB

No PCB for this modification, sorry.

[walkman]

hi;
from which the PCB can. Completing the project, add pictures and video?

What is the latest situation regarding the depth and distinction?

[walkman]

knock knock.
Is there anybody in there?
Add photograph the front and back, please.

[kkyawhtet2010]

Errata and question

[ATTACH]21986[/ATTACH]

Yeah....thanks for all Freddy

[emoxristov]

Is not there some colleague to make PCB