Rig Interfacing (was: [RTTY] Yaesu FT-920 setup - HELP!!)

Tue Jul 13 20:20:03 EDT 2004

I have been following these threads for a couple years now and it seems 
"we" have enough information to consolidate both the experiences and some 
engineering to be able to provide some direction when these situations arise.

Ian, John, Chen, et al are on the right track and each have pieces of the 
story.  But not the whole story yet.  I can shed a little light on each of 
the tracks here but I think I will "volunteer" to write up a more complete 
summary and picture of what might be happening.  I am reviewing some of 
AA5AUs new work on his web site and maybe that is a good place to put some 
of this as was done with the USB information and the sound card information 
(tnx WA9ALS) in the past.

For this current thread, the quick and dirty "fix" for Kyle is either the 
direct resistor drive to the FSK line through the resistor or a single 
transistor inverter as John has suggested.  There is no time for 
"elegance"...  the contest approaches!!  Whatever he chooses, Kyle needs to 
make sure that his resulting FSK output is not inverted and it needs to be 
checked before the contest.   Chen is right on the money with the circuit 
analysis I certainly would like to capture the rest of that story.

Ian is somewhat on the right track.  The devil in the "details" is that you 
can run the rigs through the CW keying line and the rig's own PTT will be 
internally actuated in most rigs.  So that does not establish the your own 
PTT circuit is working (separately).  Ian may have more info on this than 
in his recent post mentioned so I can't go much further with it.  However, 
I do prefer the term "voltage offset" to describe the output of these 
keying transistor systems because darlington structures are not "saturated" 
and their output offset voltages are significantly higher than the single 
transistor structures.  This applies whether driven by an optical diode for 
directly.  This characteristic seems to be at play when one person succeeds 
with an opto isolator (single transistor) and another person does not 
(darlington output).

Optical devices must have significant "light output" to switch their output 
structures.  These numbers are in the various spec sheets.  The "current 
transfer ratio" is a good measure of it but Ian is quite correct in that 
something like 5 to 10 mA is needed to meet those specs.  As Chen observed, 
the "wimpy" serial output ports might not do it for the newer 
computers.  The darlingtons have a higher "current transfer ratio" so they 
"seem" better.  But that offset voltage is the killer that seems to ruin 
their performance in the Yaesu rigs.

Chen is right on track with the circuit analysis of the FT990.  I would 
also like to know if the open collector output of Kyle's interface pulls 
that FSK line down (by accident) and thus prevents the interface from 
actually switching.  But Kyle needs success now, my engineering curiosity 
needs to wait.

The Yaesu rigs also seem very sensitive to this offset voltage on the PTT 
lines.  That is why John has seen that bad behavior on the FT1000s.  Most 
of those devices we discussed over the past couple years were darlington 
output devices and many did not have any success with them.

With all the words above, each of the rig manufacturers use different 
circuits, and may have changes in circuits between models.  And there are 
differences in the CW, PTT, and FSK control circuits.  So it is a matter of 
getting more complete information on what the "switch to ground" really 
needs to look like for each rig and input in order to make it work 
reliably.  And these electronic switches we use need to match up with those 
requirements...  and we know some schemes can't do it for particular rigs.

So I will volunteer to pull as much of this together as I can.  I recall 
problems with Yaesu, and Ten Tec most recently and will add Icom and 
Kenwood to the matrix of radios that we use.  As a start, any info any of 
you can send (direct email) about your interface that "works" would be 
appreciated.  If you are so inclined and interested, any analysis you have 
done about the input to your radio as Chen has done would also help.  I 
will talk with Don and see what we could do on the web page to capture this 
information and other data from this reflector archive as might help direct 
people in the future.

BTW, if this has already been done and I just don't know where it is, let 
me know.  I have no ego in this and no desire to reinvent wheels.  I just 
like to put engineering circuit analysis on these problems so they get 
solved and can be operated "unattended"...  I used to design this stuff for 
satellite and space work and service calls were not an option...  Hi Hi...

73, George .. W1ZT

At 06:46 PM 7/13/2004, Ian White, G3SEK wrote:
>John Fleming wrote:
>>
>>>To repeat a previous message, I've had no problems with two - no, make
>>>that three - different types of regular non-Darlington opto-isolators
>>>keying the PTT, CW and FSK inputs of an MP. It would be very surprising
>>>if the 920 required different levels, since all that generation of Yaesu
>>>HF rigs are re-using a lot of the same circuit technology.
>>
>>Ian, what "regular" optos did you use?  I've heard of some successes, but
>>not with the 4N-series, e.g. the 4N25.
>Devices that I either know personally, or know of, that have successfully 
>keyed various inputs of the MP from a PC's RS-232 ports include:
>
>CW :
>4N33 (Darlington, CTR 500% @ 10mA) with 560R input resistor
>4N36 (transistor, CTR 100% @ 10mA) with 470R input resistor
>MCT5211 (transistor, CTR 110% @ 1mA) with 1k input resistor
>FSK :
>SFH618-A3 (transistor, CTR 100% @ 1mA) with 2.2k input resistor
>TIL117 (no details handy) with 2.2k input resistor
>
>CTR = current transfer ratio, which is the opto-isolator's version of "gain".
>
>The CW keying circuits are in G4AXX's and my own PC-MP interfaces, at:
>http://www.granta.digital-crocus.com/equipment.php3
>
>The opto FSK interfaces came straight out of the WriteLog Help files. They 
>are stuffed inside DB9 plugs... and please don't *ever* ask me to open 
>them up again...
>
>G'night all.
>
>--
>73 from Ian G3SEK
>_______________________________________________
>RTTY mailing list
>RTTY at contesting.com
>http://lists.contesting.com/mailman/listinfo/rtty