Parrot MKi 9200 remote control Help Please

[Augustus Elm]

Hello All,

I come seeking advice for a project I want to do.

I have a Parrot MKi 9200 remote control, and my plan is to take the board out and wire it into the steering wheel buttons on my car, then hide the board behind the dashboard and just use the steering wheel buttons.

In the main this is straight forward, however there is a jog/scroll knob on the remote so I need to get two buttons to emulate the behaviour of said knob. I have had a look at said knob and it is entirely mechanical, through testing I have established that when it is turned CW the two outputs change state on each click with one leading the other (say A leading B). When it is turned CCW the two outputs change state on each click with B leading A. The other observations I have made are:

Number of clicks in per 360 degrees: 30
Minimum realistic time for 360 degree turn: 1 second

From that I have decided that a 5ms signal delay should work quite nicely, but ideally need to be able to fine tune it.

I would be using relays to interface the two circuits and have found some relay ICs (more later).

I would be using the 12v DC supply from the car and would like to use 5v TTL although the relay ICs are CMOS. I will also need 3v DC for the remote control circuit board.

So attached is what I came up with.

Questions are:
Will it work?
Are the JK inputs held low when unconnected?
Can you tell me which Dual JK Flip Flop IC to buy? (I've looked and it's very confusing to me).
Can you suggest a delay circuit?
The relay ICs I have found are ECE EPR2 (http://www.ece.com.tw/product-list.php?uID=6&cID=7), they seem ideal but I cannot find anywhere to by less than 500,000 of them! Can you suggest an alternative that I can buy one of?
Should I use TTL and CMOS or just CMOS, then I will only need a 3.3v supply and the remote board is happy at 3.3v (normally operates from a CR2032 battery)?

 

[Bluejets]

The A and B signals are quadrature encoder signals spaced 90 degrees apart to determine direction. If you Google that , say on Wikki, you will find more explanation.

 

[AnalogKid]

1. A nicely formed question set with a schematic - excellent start.

2. I recommend all CMOS logic, lose the 5 V regulator, add some transient protection.

3. Logic chip inputs (TTL, CMOS, whatever) usually do not have internal pull up or pull down elements. CMOS in particular is very sensitive to static. Each clock pin will need a resistor to GND.

4. BUT - switch contacts "bounce". That is, they do not just make or break cleanly. You can get 100 very fast pulses in a few milliseconds, and the flipflops will respond to every one. You need to add debouncing circuits. These can be as complex as a 2-gate flipflop or as simple as a capacitor in parallel with that pull-down resistor.

5. The CD4013 is a very common CMOS dual D flipflop. Connecting the -Q output to the D input makes it a toggle ff, like your JK's.

6. Panasonic, Clare, and others make photomos or optomos optical couplers, and are available in ones from hobby and industrial distributors. Sometimes they are with the optocouplers, and sometimes with solid state relays.

7. Since you will have your interface circuit and the controller on the same GND, you might not need an optical isolation barrier to protect the controller inputs. You might be able to drive the inputs with 5-cent transistors, or even direct drive from the delay circuits with some attenuation.

Is the idea is that each button press equals one tick of the jog wheel in either the up or down direction depending on which button is pressed?

The jog wheel probably is a rotary quadrature encoder: https://en.wikipedia.org/wiki/Rotary_encoder#Incremental_rotary_encoder

If the remote board 3.3 V current requirement is low, you probably can eliminate the other 317 and replace it with a resistor and zener.

ak

 

[Augustus Elm]

The A and B signals are quadrature encoder signals spaced 90 degrees apart to determine direction. If you Google that , say on Wikki, you will find more explanation.

OK, thank you.

 

[Augustus Elm]

1. A nicely formed question set with a schematic - excellent start.

2. I recommend all CMOS logic, lose the 5 V regulator, add some transient protection.

3. Logic chip inputs (TTL, CMOS, whatever) usually do not have internal pull up or pull down elements. CMOS in particular is very sensitive to static. Each clock pin will need a resistor to GND.

4. BUT - switch contacts "bounce". That is, they do not just make or break cleanly. You can get 100 very fast pulses in a few milliseconds, and the flipflops will respond to every one. You need to add debouncing circuits. These can be as complex as a 2-gate flipflop or as simple as a capacitor in parallel with that pull-down resistor.

5. The CD4013 is a very common CMOS dual D flipflop. Connecting the -Q output to the D input makes it a toggle ff, like your JK's.

6. Panasonic, Clare, and others make photomos or optomos optical couplers, and are available in ones from hobby and industrial distributors. Sometimes they are with the optocouplers, and sometimes with solid state relays.

7. Since you will have your interface circuit and the controller on the same GND, you might not need an optical isolation barrier to protect the controller inputs. You might be able to drive the inputs with 5-cent transistors, or even direct drive from the delay circuits with some attenuation.

Is the idea is that each button press equals one tick of the jog wheel in either the up or down direction depending on which button is pressed?

The jog wheel probably is a rotary quadrature encoder: https://en.wikipedia.org/wiki/Rotary_encoder#Incremental_rotary_encoder

If the remote board 3.3 V current requirement is low, you probably can eliminate the other 317 and replace it with a resistor and zener.

ak

Thank you AnalogKid, a very comprehensive reply. I have a few questions that I hope you won't mind answering:

What is transient protection?

What size resistor should I use to ground the clock pins?

If I try a capacitor to eliminate bounce, what size should I try? Ceramic or Electrolytic?

Regarding your question "Is the idea is that each button press equals one tick of the jog wheel in either the up or down direction depending on which button is pressed?" ... Yes, that's exactly it.

 

[Augustus Elm]

Hi All, Can you help with my additional questions above?