Posts by Administrator

OK mystery solved. I didn't notice what my esteemed colleague was actually seeking to draw attention to. Here's an enlargement.

Can you expand on the issue and the cause? Looks good to me.

Just remember it's not the same order in the RJ45. I tend to put shared wires on the LHS, ground in this case, often +ve in throttles.

OK, so we're making progress. You say the board now appears but is that something random or as a result of something else you did since your last post? I have to say that these RJs look just as worrying (purple rings) as the control ones, so before we isolate circuits could you please redo them with nice short neat exposed wire ends?

i.e. more like this

Yes, can you show some sharp pictures of the RJ45 wiring in the hub and also the some of the wiring from the stick RJ45 to the bodnar board? After that we can break this down a bit by disconnecting some of the bridge wires and isolate just some of the axes.

These photos are very helpful and I see a problem straight away. Your wires into the RJ45 are way too long and you're very likely shorting across channels. See the length of wire I use and see this close up of an RJ45 and how the teeth of one channel come through the plastic to adjacent channels. This means a bunch of crushed wire from one channel can touch the teeth of an adjacent channel.

I suggest taking all the wires out, restripping a short 4mm or less length, twist the wires tightly and refit.

This is the the problem

Do it like this

Prepare wires like this

Fortunately it is possible to examine the pitch sensor without major disassembly. The steps are:

1. Remove the cover plate keeping the wires to the side
2. Remove the 6 screws holding the pitch maghall on one side and the opposite bearing case
3. Carefully slide out the pitch assembly

As you can see the problem was very apparent once I had done this. A poorly connected pin which had come loose. At this point I expect some will say "ahh, you should have soldered that" however I still don't feel that this is essential and it has downsides such as the difficulty of doing this without melting the adjacent plastic. Better to become proficient at crushing the ferrules and use a jewellers magnifying glass if you need it IMHO.

Once I had refitted the ferrule I tested the connection by reconnecting the pitch axis sensor wires with a simple twist before reassembling the pitch axis unit and to my relief everything was working as expected. The movement/deflection of the crosshair was not calibrated of course and that needed to wait until I had put everything back to gether again.

ISSUE RESOLVED

The Outcome

Note that you should always disconnect the USB cable between the hub and your PC apart from the brief moments of testing. If you do not then any short circuit could burn out a sensor. Also the bodnar board will often not correctly detect axes if they are not present when the board is powered up, so you might correct the problem and find the board doesn't detect it.

The wiring on the bridge was fine as was all the RJ45 wiring and the wiring to the board. In fact I took out all wires and gave the RJ45s a good shake. Some users have found that a tiny stray strand of broken off wire can drop into the RJ45 and short across connections and in fact a tiny piece did fall out but it didn't solve the problem. So we needed more investigation.

At this point more technical folks are likely to reach for their multimeter to test voltages and continuity but my aim is to ensure that AuthentiKit stays within the realm of the Lego Technical world rather than the amateur electrical engineer. So as a next step I decided to break down the issue to determine if

1. It was a failure of the wiring and connectors up to the point of the pitch axis hall sensor, or
2. It was a failure of the pitch axis sensor

A simple way to do this is to disconnect the 3 wires from the pitch axis hall sensor and fortunately all 3 can be separated at the bridge. Then as a second step apply a voltage to the yellow pitch sensor wire as per this diagram.

After disconnecting the hub, removing the pitch axis wiring and reconnecting the hub (see below), I found that everything was behaving exactly as before which was the outcome I expected. So after that, I disconnected the USB cable again, touched the yellow wire to the ground wire joint on the bridge (as per the white arrow below) and plugged in the hub again. This time I registered a Y (pitch) axis as well as the X (roll) axis as shown on the game controller image.

What this means is that the fault is not in the wiring/RJ45s but in the hall sensor. It could be a stray wire crossing over and shorting some pins or a wire come loose. Not the best outcome actually as I had to do more disassembly but I least I'd isolated the problem.

The Testing Plan

Firstly it's worth pointing out that it seems unlikely that the wires are in the wrong order as all the other controls are fine. An axis requires 3 working wires :

• 5 Volt
• Signal
• Ground

Could it be the 5v signal? Well as the 5v signal is shared by all controls on the stick we know that 5v is getting to the post on the bridge (as per this picture) so it seems likely that the 5v is getting all the way to the axis unless the wire has come loose from the sensor leg which is actually very rare. So the most likely thing is that the ground or signal is not connected all the way through. In this design the ground for the pitch is shared by the brake axis and as the brake is working we know that a ground signal is also reaching the bridge which is not far from the hall sensor. So the first assumption is that the yellow signal wire is not getting through - it connects to the white wire on the right hand side of the bridge.

So these are the steps to test.

1. Check the bridge - is the white connecting to the yellow securely?
2. Check the wiring on both RJ45s. Make sure contact is being made and that no wires are straying. This is a very common problem. Use a bright light and magnifying glass if you have one.
3. Check the pins to the Bodnar board. It's possible that a plug is stuck in between two pins rather than properly over a pin

To help share the learnings arising from assembly troubleshooting I have created this forum and will kick it off with an example from my own experience. To date we have been doing this in Discord but good knowledge is easily lost and buried in that environment.

So this incident occurred today. I rebuilt a flight stick as well as a universal hub and wired it up for the first time. I plugged the stick into the hub and connected it to my PC then fired up the game controllers application (joy.cpl) to see that everything was working apart from the pitch (forward backward) axis. You can see that the X (roll) axis is displayed with the cross dead centre (I've already run a calibrate which is why its dead centre) but no Y (pitch) axis is showing. So the question is why? In this thread I'll take you through my testing plan and troubleshooting.

NB: If the Y axis was showing but not behaving correctly that would be a different scenario. The key thing here is that the Y axis is absent from the game controller.

By the way, if you don't know how to access this app, it's standard windows. Just type joy.cpl into the search bar.