FSBUS Strange behaviour

[Guest GeorgeDorkofikis]

Hello everybody,work is progressing on my initial stage of my small cockpit setup. I have finished my MCP (hopefully I will be able to post some pics once I get my hands on a digital camera), and I'm almost finished my pedestal's radios panel.However, I noticed some strange behaviour of FSBUS I haven't noticed before. If I turn the MCP on, start FSBUS (but NOT FS), and leave it alone, the displays on the MCP start to change over time like getting erroneous input from the rotaries. This happens in random times. Sometimes after 5mins, other after 20mins...I have the following setup :a. FSCOM3 on a separate box along with the PSU providing 9V to the COM3 and 5V to my redecs (my own design PSU, solid voltage, 2A max current). Measured voltage 8.98V and 5.04V. The voltmeter I use has a 10% tollerance so the reading I get are perfect.b. MCP using Alps EC11B rotaries decoded by Redec circuits.c. The MCP communicates with the FSCOM3 via a 9pin serial cable which has been altered to accomodate both the FSBUS signals and the 5V to the redecs in the MCP. In particular I have connected together the pins that carry the same signal (like GND and the 5V of FSBUS).I think the problem may be the redecs but can't be sure. It may be the software too (ver 2.2)Any ideas what could be interfering with the key card or the redecs?Anyone else noticed such behaviour?Thanks for your help.George DorkofikisAthens, Greecehttp://www.precisionmanuals.com/images/forum/betaimg.jpg

[Guest PoRrEkE]

What happens if you turn on FS ?Still a problem ? Because I don't see why you would not have FS on at the same time since the vlues for the displays are read from FS ...Other thing .. where did you get your redecs ?

[Guest Glenn Weston]

George, something simple you could try is, if you can get access to a knitter switch or similar, connect it up & see if the problem still exists and/or try disconnecting the rotary from the redec & see if it still happens if it does disconnect the redec from the FSbus input & see what goes, just a thought.Glenn.

[Guest GeorgeDorkofikis]

I don't have FS running because I am just assembling the hardware and I only want to setup the FSBUS.Edit : I'm also redirecting the rotaries output to the displays for fast updating. So even if FS is not connected, if I turn a rotary the displays are updated.I made the redecs myself and they worked perfectly before. Never had such a problem. Now that I think of it, I accidently reversed the power on those redecs due to a sortcircuit... So maybe I half fried the ICs! However they do work ok when I use the rotaries.Unfortunately Knitter are out of my reach! But the idea to leave the redecs without connected rotaries is good. That way I could see if the false pulses are created by the redecs or something else...Troubleshooting takes looong time, doesn't it?!Thanks guys.George DorkofikisAthens, Greecehttp://www.precisionmanuals.com/images/forum/betaimg.jpg

[Guest PoRrEkE]

"Unfortunately Knitter are out of my reach! But the idea to leave the redecs without connected rotaries is good."Who needs knitters :D when it is for testing. Just hang 2 push buttons and press them repeatedly. 1 is CW and 2 is CCW. Simple as that.Also, even when just testing .. why not try with FS on ? That's how you'll use it in the end anyway.

[Guest GeorgeDorkofikis]

You have a point there, but it's because I didn't have FS connected that attracked my attention to this behaviour.Otherwise it may have been FS that fooled.George DorkofikisAthens, Greecehttp://www.precisionmanuals.com/images/forum/betaimg.jpg

[Guest PANTU]

Hi George,it looks that you

[Guest GeorgeDorkofikis]

Hi Peter,Thanks for advises!I do have RF cutout capacitor on the output of the power supply, after the regulator (7805). I haven't put such on each redec board.The input is indeed pulled down by 100K resistors.The supply is pretty stable, I don't have an oscillator but I'm sure the 7805 is doing it's job.As for the length, none of the cables that connect the redec's output to the key card is more than 30cm long, and the serial cable that connects the MCP to the COM3 is the standard 1.8mI'm sure I have some 100nf capacitors home... I'll give that a shot too.Thanks for all your replies. Hopefully I'll find the solution with your help.George DorkofikisAthens, Greecehttp://www.precisionmanuals.com/images/forum/betaimg.jpg

[Guest GeorgeDorkofikis]

Hi all,I'm re-bumbing this as it's still current and I have some news about it.The 'ghost' changes came back! Well... Not in the MCP as I think I found a temporary solution for that, but now in my Pedestal.I'm now 99% convinced that this is a software related problem. Either FSBUS route or maybe the PIC program itself.When it occurs:When you connect the redec output to subsequent bits on the same key bank. (0+1, 2+3, 4+5, 6+7 ), i.e. 4 decoders per bank.It does NOT occur:When you leave 2 bits empty between each decoder output in the same bank. (0+1, 4+5, 2+3/6+7=empty), i.e. 2 decoders per bank.That's what I did in my MCP and the 'ghost' changes went away.Unfortunately I have to build another key card to do the same for the pedestal so it will have to wait for a while.I have measured up everything, tested everything and can confirm theabove, so I'm convinced it's a software problem.I have already sent an email to Dirk and I'm sure he will look at it when he finds the time. He never let us down! (thanks again Dirk!)George DorkofikisAthens, Greecehttp://www.precisionmanuals.com/images/forum/betaimg.jpg