[time-nuts] Lucent KS-24361 GPSDO arrived today several questions

Bob Stewart bob at evoria.net
Sat Nov 8 11:47:07 EST 2014


Hi Bob,
I've been using the "cheat" method that Stewart posted.  It's mostly working, but I do see occasional errors.  So, I ordered the following USB to RS-422 adapter.  Now your post makes me worry that there's still going to be a problem.  I'll post results back to the list.  It is a FTDI device.  Hopefully it's a "real" FTDI device and not subjected to getting bricked.

USB 2.0 to RS-485/RS-422 RS485/RS422 DB9 Serial Adapter Cable FTDI FT232 FT232R

And I was also wondering if anyone has put together an adapter using LTC-1485 or similar chips?
Bob

      From: Bob Camp <kb8tq at n1k.org>
 To: Discussion of precise time and frequency measurement <time-nuts at febo.com> 
 Sent: Saturday, November 8, 2014 10:10 AM
 Subject: Re: [time-nuts] Lucent KS-24361 GPSDO arrived today several questions
   
Hi

You make another good point, that is worth repeating. 

With RS-422, it’s not just the usual RX to TX and TX to RX confusion. You can have all that right and still get it wrong. The + and - of each must be properly identified and connected. 

That sounds easy, but there is a catch. If you are used to RS-232, you are used to a negative voltage being a “one” and a positive voltage being a “zero” data wise. It’s upside down from the conventions on say, CMOS logic (one = more positive voltage). 

RS-422 uses a convention where the + output is higher than the - output when you have a “one”. That’s the normal idle state for a serial output. Again, it’s a simple DVM check on something it’s VERY easy to get wrong. Simply put - at idle, the + output is the one that should be the higher voltage (3.5V in this case). 

Of course one should beware - It’s very easy to run though the archives and find *lots* of places I get this kind of stuff mixed up. Like I said, it’s not as easy as it seems.

(It’s even more confusing when you start talking about the control lines … thank goodness we don’t seem to have CTS and RTS involved on these boxes).

One further disclaimer, this is all for RS-422. That’s what the HP / Lucent boxes have on them.  Mil-STD-188-114B signaling is very similar, but without the 2.5V offset on the levels. They *should* talk to each other, troubleshooting a mixed setup like that is a bit more complex. If your adapter is actually a 114B adapter, things will be a bit different when you mate it all up. 

All of this differential signaling stuff is designed to be used in noisy environments. You *can* have a 2V sine wave driving one end’s ground compared to the other end and still get perfect data transmission. With some driver / receiver combos you can have a lot more than 2V and still do 100Kb/s error free. It’s also great for setups where you want to isolate the grounds of two systems. It’s not as good as opto isolation, but it’s a lot better than a hard ground wire. 

Bob

> On Nov 7, 2014, at 9:09 PM, Alan Hochhalter <alanh137 at gmail.com> wrote:
> 
> I got in a hurry and made mistake #1 listed by Bob a few replies back.  I
> hooked one up to the RS422/1PPI port transmit pins (same as on J8) and was
> receiving data fine in the PC so I just hooked up the other pair of
> terminals to the receive pins for J8 and couldn't get any response to
> commands.
> 
> I finally found that I had initially connected the Lucent box ouput pins to
> the wrong terminals on the converter.  The Time code data on J6 was still
> getting into the PC even though it was connected wrong.  So I just wired
> the Lucent box input pins to the other pins on the converter that are
> really inputs and couldn't get J8 sto work either.
> 
> What I ended up with to make it work is: J8-9 to RX-, J8-5 to RX+, J8-8 to
> TX- and T8-4 to TX+.
> 
> Alan
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.



_______________________________________________
time-nuts mailing list -- time-nuts at febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.


  


More information about the time-nuts mailing list