Arduino directed & auto KAP rig: servo swap

At the January 2021 KAPIZoom I was asked to demonstrate my rig and controller and I was a tad embarrassed to have the thing misbehave on the day. The rig has been plagued with a jittering tilt movement from day one and I have put up with it for a while as ‘one of those things’. It was suggested the jitter was down to errors in the Arduino code, something I was sure was not the case as, when paired to other rigs, the Arduino controller performs perfectly.

A shot ruined by servo jitter.

The Ricoh GR rig has misbehaved now and then and I have put it down to low battery level or range issues with the radios (I use a tiny stub antenna on the trnsmitter) but to share the jittering rig with the world shamed me into investigating the issue. Work, weather and restrictions on exercise have reduced my flying time and my tolerance of duds is getting thin. The Ricoh GR is my KAP camera of choice now as it has the best compromise between resolution and weight. The time to fix the twitch has come.

I’d assumed vTx/ rcRx cross frequency interference was the issue. Most of the time a bit of tilt tweaking would kill the jitter but under auto control I have been getting a stack of nadir duds so a fix was in order.

I decided to separate the Vtx and RCrx which is a pain because it shifts the centre of gravity, I have been putting this off as it means finding the new centre and rebuilding the pan gear, a fiddly business at the best of times.

With the rig rebuilt, rebalanced and rewired the jitter on the tilt was still there. Hurruph!

Googling ‘servo jitter’ threw up 4 ideas:

1. fit a capacitor across the servo. The size (250mfd) and power draw of the capacitor recommended to smooth the voltage is hefty so not too keen on that and it’s discounted as overkill by most RC folk.

2. Separate and re-orient Tx and Rx. Did that, shifted the pan centre- no dice.

3. Fit a ferrite ring to the servo lead. Not too keen on adding yet more weight.

4. Improve voltage condition. My servos are powered from a 5V outlet on the VTx so this may be an issue, as the Tx draws power the 5v supply might be compromised. I checked by substitution of a 3.3v cell direct supply to the Rx: no change.

Went to bed a bit miffed after the rebuild business and the folowing morning more Googling found…

…5. Some servos are better than others at jitter. Apparently analogue ones are better than digital. I have a few spares and swapped out the super light Futaba for a nasty cheap ‘Tower Pro’ job. No change. The last one to hand is a heavy duty Swiss Savox SH 0256 which I don’t like because it’s very noisy (as in loud) and that fixed it.

Re-wiring threw up more issues (bad joints, movement clashes etc etc) but I think I have come out the the other side with a better rig.

Long story short: there is NO problem with the code as was suggested at KAPiZoom 2021- my tilt jittered because it was using the wrong kind of servo for the task. Quite why this is the case I don’t know but the Savox is now buzzing away on auto taking pics of my window cill, the office, the front garden and the floor while I wait for delivery of a more compact 50g 2s battery to fit the newly re-centred set up.

This clip at 2x speed shows the auto end points are stable:

This rig (and controller) is designed around two conflicting constraints: durability and light weight. The rig has to be compact and rugged enough to survive bike hauling without damage and yet be light enough (450g including camera) to fly in reasonably light winds (Bft 3 or 8-12mph ) as I’m not keen on riding into anything much tougher than that.

The Arduino control is a winner because autoKAP has a way of picking up framing that may not be obvious from the ground, no matter how carefully planned the shoot, there is always a keeper or two in the auto set. The control allows toggling between directed shooting and auto with a flick of a switch. The auto mode comes into its own during positioning moves, particulally when slack line is possible.

A short test flight confirmed the stability of the auto sequence end points.
To add to the fun the pan gear locked up as I’d not tightened the lock nuts to torque.

The controller and its sketch are described here:

How I set about building the controller:


About billboyheritagesurvey

Heritage worker
This entry was posted in KAP and tagged . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.