Tuesday, 21 February 2017

Getting inside the Horus

The Horus is a sweet piece of kit, and becoming increasingly popular. Unfortunately my production unit arrived with a faulty gimbal - one axis was double-centring and the right end point was inconsistent. After sending a video of the issue to FrSky, I received a replacement under warranty, so today I cracked open the transmitter to do the swap.

Opening the case

Some transmitters are easy to open up, for example my MPX Profi 4000 is a model of accessibility - just spring open a couple of latches and bingo, all is revealed. The Horus isn't so easy as (a) it's held together with six self tapping screws and (b) the front and back halves are spanned by various cables.

Before proceeding, I made sure to unplug the battery.

Separation

The first job was to undo the six case screws, easily done with a cross head screw driver. The two halves of the case are a good fit, but after a bit of prising and wiggling, they came apart easily enough. (Tip: don't remove the textured side cheeks - it's not necessary and they are a PITA to reassemble.)

However, there were still some cables to deal with, so care was needed not to put any strain on these, in particular the RF cable.

Next job was to free the external antenna connector from the rear moulding. Access is via a flap at the rear. A socket head spanner made quick work of the nut, although the locking washer needed a bit of persuasion before it could be removed.

I now had more room to play with, and was able to unplug cables for the charger port and sliders.

Showing disconnection points for access

All that remained now was the RF cable, one end of which plugs into the module via a UFL connector, the other end goes to the antenna on the back. I decided to leave it alone, as these connectors are intolerant of repeated disconnection (and in any case the connector was sealed to the board with goo).

The solution turned out to be simple: with all the other cables now out of the way, I was able to flip the rear moulding over the top of the unit, without disturbing the RF cable. All the components were now easily accessible.

RF cable should not be unplugged

I used a scrap of foam  and a weighted box to prevent the two halves from sliding around on the workbench.

Next task - swapping the gimbal!

Replacing the gimbal

The gimbal assembly is held in place by four self tapping cross-head machine screws, accessible from within the case. The screws are surprisingly tight, and it helps to have a screwdriver with a substantial grip. Underneath the gimbal sits the bezel, a simple aluminium ring which is keyed to prevent rotation.

The gimbal itself is a neat unit with Hall sensors mounted on tiny circuit boards in place of the usual potentiometers.

Gimbal MC-12

Mechanical design of the gimbal is thorough, for example the contact area of the metal rocker arm is clad with a teflon-like material.. However, I do have some concerns about the wiring: the loom for the aileron sensor was too short for comfort and not very supple. I had to bend it into a tightish 'U' in order for it to reach the main board, and even then it was straining at extremes of movement. Furthermore, at extremes of movement, the loom was brushing slightly against the rocker arm.

As a precautionary measure, I would recommend applying a little hot glue where the wires exit the sensor boards.

Loom arrangement

[Aside: in contrast to the production unit, my prototype Horus has the sensor leads exiting vertically instead of horizontally - this is a better arrangement for the RH gimbal, though worse for the LH unit. It seems that the manufacturer has compromised in using the same design for both gimbals.]

With everything reassembled, I switched on and calibrated the sticks, and was able to verify that the new gimbal was working correctly.

All in all, working on the internals isn't difficult, but it's worth planning each move, and taking it slowly. In the meantime, FrSky really need to take another look at the gimbal wiring.

No comments: