SOLUTIONS for Control signal crosstalk problems

This contains the scottyspectrumanalyzer yahoo group backup
Post Reply
Posts: 254
Joined: Sat Apr 28, 2018 3:29 pm

SOLUTIONS for Control signal crosstalk problems

Post by spyker49 » Sun Sep 19, 2010 4:08 am

Hi Guys

Suggestions for those with control line problems.

(1)Try other LPT cables.

BAD home brew or many commercial LPT cables are just bunched up normal flat ribbon cable and are NOT proper twisted pairs and will have excessive crosstalk. Normal printers have a lot of input debouncing and can work with crap cables.

GOOD LPT cables use either discrete twisted pair wires or TWISTED PAIR flat ribbon cable (Yes Santa it really exists) and those work much better.

NOTE: LPT cable price is NO indication of GOOD/BAD. Carefully slit a bit of your LPT cable sleeve open and have a look inside.

(2)Inside the MSA with the star topology of the main clk/strobe signals make a twisted pair with a new-extra OV ground line and the clk/strobe signal wire.

2-3 turns per inch, 1 turn per cm is adequate. The OV signal of the twisted pair needs to be connected to Ov at each end.

(3)Add a series 100R - 200R resistor on the major control lines.

With a CMOS chip with 50mA digital drive capacity and 100 pF load there is a 2nsec/volt slew rate this equates to a 30-80 MHz signal region that most CMOS logic is quite happy to clock with.

With a series R at the logic source the possible rate of change is lowered (de glitched),and the (unterminated) reflections are damped out (absorbed) quicker.

Len Spyker

Post Reply