BCD-996T tap problems redux

[DaveNF2G]

My setup works fine with UniTrunker via audio input. It's slicer-to-serial port that is giving me trouble.

 

[pro92b]

I guess I'll have to reverse engineer the damned thing because I can't find the exact diagram I used when I was building it. I don't think there are too many different ways to build a 741-based 2-level slicer, though. Like I said, I built the thing years ago so the design has been around for a while.

There is one basic circuit topology but the component values vary depending on where you got the schematic. I know of three variants and there may be more. Eventually you will need to reverse engineer the circuit if you want to build a second slicer. For now if you post just the resistor values, I may be able to guess at what should be changed.

 

[DaveNF2G]

I still don't understand the focus on the slicer. The variables here are the scanner (PRO-2067 vs. BCD-996T) and the discriminator tap components (1.0 uF cap vs. 1.0 uF cap + 10k resistor). The slicer and computer are constants.

 

[SCPD]

I still don't understand the focus on the slicer. The variables here are the scanner (PRO-2067 vs. BCD-996T) and the discriminator tap components (1.0 uF cap vs. 1.0 uF cap + 10k resistor). The slicer and computer are constants.

Sorry Dave. I missed that. The PRO-2067 works but the '996T does not, right?

1. you've tried shorting out the coupling resistor (reduced to zero ohms) so try a larger value for the coupling resistor (like 50k or 100k).

2. increase the gain on the '741 - you may need to change out a different resistor to accomplish this.

 

[pro92b]

I still don't understand the focus on the slicer. The variables here are the scanner (PRO-2067 vs. BCD-996T) and the discriminator tap components (1.0 uF cap vs. 1.0 uF cap + 10k resistor). The slicer and computer are constants.

I'm not looking for variables, just system interface problems. Each component may 'work' by itself but still produce a non-functional system when used in combination. Humor me and post the resistor values in the slicer.

 

[DaveNF2G]

OK, now I understand. I'll diagram the circuit later today. The whole assembly is pretty small and I don't want to break any of the wires.

 

[DaveNF2G]

my ugly slicer diagram

OK, I have attempted to generate a circuit diagram from the actual slicer. Below should be either the picture I drew or a link to a PDF. Not sure what's gonna happen here.

I think I accounted for all of the components.

 

[pro92b]

Try bypassing the 1K resistor with a piece of wire. Test with 996 and 2067 and let me know what works if anything.

 

[DaveNF2G]

It worked! Slight performance improvement on the PRO2067, but it improved the BCD996T from not working to working fine.

Thanks!

 

[pro92b]

Bypassing the 1K resistor removes all hysteresis from the circuit. If the slicer performs to your satisfaction, you can just leave it that way. I would probably change the 1K to 100 Ohms on the theory that a little hysteresis is a good thing. The slicer had excessive hysteresis originally and was on the border of not working with the 2067. The 996 has less signal amplitude so it didn't work at all. The attachment explains all the details if you are interested (or if you can't get to sleep some night).

 

[SCPD]

Mr. pro92;

I've always had problems with parasite powered slicers - particularly ones using '741s. The only circuit that worked well for me was 4 level slicer with external power (a pair of 9 volt batteries).

I built the circuit you describe in the PDF posted directly above. It works well. 100% decode of a strong signal. This is the first "parasite" powered slicer that has worked well for me. Easy to put together. All components are standard values. I could not find the MC33172 readily available over-the-counter but I did have some MC34072s in my parts bin. I used an 8 pin socket so I can pull out the '072 and drop in a mail order '172 once it arrives. The '072 is probably a little to "hot" (slew rate comparable to the LF411) but the mild filtering seems to keep it in line.

Thank you for putting this all together - both the "how" and "why".