Large number of flickering/ghost inputs on Master Cards

[Boeing 747 Flyer]

Hi Les,

I have received another response from Alejandro, he says that the new version of the mastercards are unlikely to fix the issue. He thinks that the problem is "an excess of charge connected to J3", I assume "charge" has been lost in translation somewhere, I think he must mean current. What I don't get is, J3 is an input connector, not an output connector (J1 or J2), so how can J3 be drawing excess current (he also mentioned "outputs connected to J3" in a previous message).

Still working on neatening everything out. If that doesn't work I think I'll just have to work around the problem, perhaps just not use J3 (although like I say, sometimes I can get it to happen on J4, nothing is consistent here).

One thing I will mention, the 25-core cable that connects the overhead panel itself to the master cards has both inputs and outputs within it (ie inputs and outputs are bound within the same 25-core cable), perhaps this could be a factor. Even so, if this is the case, why doesn't it happen with every single input?

[iwik]

Hi Jack,
Yes I don't think translation helps greatly.
I would definitely not mix inputs and outputs in the same cable. If the least you do is to separate these.
I have been in Electronics for near on 50 years and many a time what happens electronically defies all logic.
Even had Design engineers unable the help me fix a problem on equipment they have designed.
For all we know there maybe this inherent weakness in this design that your particular layout highlights.
As I can produce your effects on my setup with a little rearranging of my cabling and power supply.
It's a shame if you cant use J3. Even using another make of I/f card maybe a simple work around.
Les

[iwik]

Hi Jack,
Yes I don't think translation helps greatly.
I would definitely not mix inputs and outputs in the same cable. If the least you do is to separate these.
I have been in Electronics for near on 50 years and many a time what happens electronically defies all logic.
Even had Design engineers unable the help me fix a problem on equipment they have designed.
For all we know there maybe this inherent weakness in this design that your particular layout highlights.
As I can produce your effects on my setup with a little rearranging of my cabling and power supply.
It's a shame if you cant use J3. Even using another make of I/f card maybe a simple work around.
Les

[Boeing 747 Flyer]

Hello everyone,

Unfortunately I am still experiencing the problem and am still in the process of rebuilding my wiring stack.

I am liasing with Opencockpits' support and they have asked my to produce a video of SIOC monitor to see what is happening. I have done so, and so I thought I would share the video here too so that you can all have a look at what I'm referring to.

For simplicity, only ONE mastercard was connected to the USBExpansion card for this video (MC 1), so just ignore all of the IO readings for 2/3/4. The part of the cockpit connected is column 1 of my 767 overhead panel. You may have to look very carefully at the SIOC monitor display for MC 1 in order to see the "doubling" and flcikering of inputs.

https://www.youtube.com/watch?v=6GwQh7cN_ts

Jack

[Boeing 747 Flyer]

A further update.

I have spent today unscrewing every single output from the screw-in terminals connected to the master card (MC 1). In other words, there were ONLY inputs connected to the master card. After commanding ALL ON, the problem still occurred.

I then proceeded to unscrew every single input connected to the screw in terminals connected to the master card. In other words, there was now nothing at all connected to the master card. After ALL ON was commanded, the problem did NOT occur.

So then, I proceeded to wire the inputs back in one-by-one. After wiring the first, no problem. Then the second, no problem, and so on, until I had wire the first batch of 9 inputs (ie one ground connected). Then, after intensively testing these switches, I eventually got the "two at once" problem to reoccur whereby two inputs are activated by a single switch press. However, there was absolutely no input flickering whatsoever when repeatedly commanding ALL ON.

I think this must surely conclude that the outputs, although perhaps making things worse, are not the root cause of the problem? Then again, the wiring of the inputs has been quadruple checked, there are absolutely no shorts/poor connections.

[iwik]

Jack,
Good to see you are still plugging away at it. Can you pay a visit to someone in the UK and have a look at how they have wired their one up. Still think its how the inputs are wired. Maybe you have'nt got any shorts but layout may still be a problem. You also say you had to test extensively to show the fault. Now it is possible that you may experience something like that in extreme cases but it would be rare and may not cause a problem. As you said earlier it only
became a problem when you did a ALL DISPLAY test. So be careful not to follow something that maybe a red herring.
This interface may not be perfect. With some things you try hard enough and you will find fault.
As I said earlier in one of my posts that I can get what you experience.
Maybe David would allow you to visit and see how he wired his one.
Les

[Boeing 747 Flyer]

Just an update:

Opencockpits have agreed to send us the new PIC - I assume this swaps out with the old PIC in the black chip socket on the USBExpansion card?

Jack

[iwik]

Yes that's correct, just make sure you put the new one in the same way round the old one was. The chip has a little
notch or spot.
Les

[Boeing 747 Flyer]

Alright, thanks very much Les.

[SteveL]

Hello Jack,
You might try adding a .1mfd ceramic capacitor across each of your switches to reduce switch bounce and noise. Also, a small resistor in series with each switch will limit the current draw. Experiment with values between 20 and 200 ohms. The largest useable resistance will provide the greatest current reduction, but if the resistor is too large the input will have trouble interpreting it, most certainly you don’t need zero ohms to recognize a switch input.

EDIT The resistor advice is based on accepting the premiss that you need to reduce the current draw. If that's not the case, only do the capacitor.