Transponder/Encoder troubleshooting

Hey all. Any avionics advice for troubleshooting my latest
gremlin? It's a Narco AR-850 altitude encoder connected vi 1N4007 diodes
to a KT-76 transponder and (no diodes) to a KLX-135A VFR GPS/COM. Trouble
is, once in awhile the ModeC doesn't work (center says only
ModeA). Every once in awhile, they report back an incorrect altitude.
Most of the time, however, it works fine.

I figured the encoder was crapping out, but the GPS reports the
correct altitude, even when center says it's off. I thought initially
that the diodes were causing too much voltage drop and corrupting the data
to the transponder.

Trouble with that theory is that if the diodes were too much drop,
then a bit in the grey code would be a '1' instead of being a '0' (i.e.
the encoder not pulling the line down low enough). When I checked the
error, however, the bits were flipped in the wrong direction. For
example, after a climb from 7500' (0x642 grey code) to 9500' (0x672
grey code), the center reported 7500'. The only way this could
happen (that I can think of) is that the grey code lines are negative
assert (i.e. '1'=0v, '0'=5v).

Any thoughts? I'm kinda at a loss here, since I can't isolate any
one issue, and I certainly don't want to start replacing components that
aren't broken!

Thanks in advance,
-Cory

--
************************************************** ***********************
* The prime directive of Linux: *
* - learn what you don't know, *
* - teach what you do. *
* (Just my 20 USm$) *
************************************************** ***********************

wrote in message ...
The only way this could
happen (that I can think of) is that the grey code lines are negative
assert (i.e. '1'=0v, '0'=5v).

Isn't that the case. The encoder is an open collector. It has low impedence (i.e. it grounds)
the pin when the value is 1 and a high impedence at 0. The connected device (transponder
or gps) has to pull up the value if it wants a positive logic value (it's this pull up voltage
that requires the diodes, if it was just sinking current they'd not be necessary).

Ron Natalie wrote: : Isn't that the case. The encoder is
an open collector. It has low impedence (i.e. it grounds) : the pin when
the value is 1 and a high impedence at 0. The connected device
(transponder : or gps) has to pull up the value if it wants a positive
logic value (it's this pull up voltage : that requires the diodes, if it
was just sinking current they'd not be necessary).

It could go either way... it's just a matter of convention.
*Usually*, a logic value of '1' means high voltage, and a '0' means
ground. In some circumstances, this is reversed. Yes, the encoder is
open-collector outputs, but the reason for diodes is to isolate multiple
devices connected to it. I know from a previous installation on another
plane that the AR-850 is particularly bad about not pulling a "low" low.
Rather than a Vce_sat of 0.3v or so, the guy's AR-850 could only pull down
to 0.8 or so. That, along with the 0.6 on the Si diodes made his Narco
AT150 not recognise any of the grey code bits.

I was hoping someone on here would know for sure if a low (i.e.
open collector for that bit at low impedance) is a logic '0' like normal
(my problem's still unsolved), or logic '1' (problem could be solved with
different encoder, or perhaps Ge diodes).

(I've already thought this through a bunch)
Thanks,
-Cory


--
************************************************** ***********************
* The prime directive of Linux: *
* - learn what you don't know, *
* - teach what you do. *
* (Just my 20 USm$) *
************************************************** ***********************

wrote in message ...

It could go either way... it's just a matter of convention.
*Usually*, a logic value of '1' means high voltage, and a '0' means
ground.

I'm specifically talking about encoders, NOT open collector outputs in
general. All the ones I am familiar with ground on the logic 1.

Yes, the encoder is
open-collector outputs, but the reason for diodes is to isolate multiple
devices connected to it.

The diodes may or may not be required depending on what the devices
interface looks like. Believe me, I've had tons of crap in repeaters tied
to a single open collector transitor for the COS and there were no diodes
present.

wrote in message ...

It's not for isolation when all the devices are on, but rather for
when one of them is off. If there's more than one device hanging off a
single encoder, then they must all be isolated. Otherwise, turning off
the transponder could pull all the lines low, and kill the encoder signal
for the GPS as well.

Depends on what the design of the interface on the various units is.
There's no reason why shutting off a transponder should cause it
to sink current enough to hose up the GPS even without diodes.

wrote:
: I'll take a look at that, although I've already looked through the
: harness pretty carefully. I had a guy at Narco tell me to use signal
: diodes, too. I used the 1N4007, because that's what the KLX-135 install
: manual called for. Funny thing is, silicon is silicon, and has silicon's
: voltage drop characteristics. The 1N4xxx will have more or less the same
: drop as a signal diode at these low currents. Only way to decrease the

Some 1N4007 diodes (400 VRR) get a little weird at low If, and have funky
transfer characteristics until you're pulling 10 mA or so. Depends on the
manufacturer, not all 1N4007's are as equivalent as they should be!
Take a look at Bob Pease's works for more info. 1N4007 also have a slow
recovery time, because they're made for 60Hz applications.

: total drop (that I can think of) it is to use Germanium diodes and/or new
: (read: NOT crap AR-850) encoder. I happen to know from taking an old one
: apart that the ACK A-30 encoder uses discrete 2N4401 diodes. Datasheet
: says Vce_sat = 0.4v at Ic=150mA ,Ib=15mA (probably 50 times more current
: than on the transponder)
: I'm just not excited about tearing the wiring harness apart to put
: in Ge diodes if that's not the problem.

If I were changing the diodes out and I was looking for a low Vf, I'd
probably choose a Schottky like the 1N5819 over a germanium. You'd get
a better recovery time to boot.

: I also know for a fact that *without* the diodes, there is
: interraction. When I put the KLX-135 in, I had wishful thinking that I
: wouldn't need them. Sure enough, without the transponder on, the GPS
: didn't receive encoder data.

: -Cory

Yes, I got this same result with my At-150/M3 Approach/Transcal D120P2t
setup. Needed diodes for the transponder. I used 1N4148, and it works
great!
--
Aaron Coolidge (N9376J)

For encoder units I've seen, their open collector outputs ground the signal
for a logic 1.

Cheers
Pete
Europa builder.

wrote in message
...
Ron Natalie wrote:
: It could go either way... it's just a matter of convention.
: *Usually*, a logic value of '1' means high voltage, and a '0' means
: ground.

: I'm specifically talking about encoders, NOT open collector outputs in
: general. All the ones I am familiar with ground on the logic 1.

If this is true, then it implies negative assert, no? If it's
grounded=='1', then that's opposite convention. Not saying it's not
true, but that makes (troubleshooting) sense.


: Yes, the encoder is
: open-collector outputs, but the reason for diodes is to isolate
multiple
: devices connected to it.

: The diodes may or may not be required depending on what the devices
: interface looks like. Believe me, I've had tons of crap in repeaters
tied
: to a single open collector transitor for the COS and there were no
diodes
: present.

It's not for isolation when all the devices are on, but rather for
when one of them is off. If there's more than one device hanging off a
single encoder, then they must all be isolated. Otherwise, turning off
the transponder could pull all the lines low, and kill the encoder signal
for the GPS as well.

-Cory



--
************************************************** ***********************
* The prime directive of Linux: *
* - learn what you don't know, *
* - teach what you do. *
* (Just my 20 USm$) *
************************************************** ***********************