Hitec's Aurora 9 has a small issue

CALIBRATION A PROBLEM?

Dated: 28 Apr 2010

It can't be denied that Hitec's new Aurora 9 radio has hit all the right buttons for a lot of folks.

Existing Hitec fans are delighted with this huge step forward from the older Eclipse 7 and Optic 6 radios, while its features and performance are winning over more than a few previously staunch JR and Futaba fans.

So far, the Aurora 9 has had an enviable record when it comes to "issues" -- insomuch as it's been pretty much trouble-free.

Yes, some have moaned about the standard battery and the miserable charger that comes with the radio but by and large, it's hard to find serious fault with the design or construction of this transmitter. There have also been a few very minor software "fixes" but nothing that's going to rain on the average user's parade.

What's more, the 2.4GHz section of this radio has proven to be pretty bullet-proof too, capable of withstanding almost all the interference I threw at it in my testing and, unlike other systems, able to adapt to noise on the band through its "scan" mode.

Unfortunately, I haven't been able to get my hands on an Aurora 9 transmitter, so much of the above is simply hearsay obtained by listening to the comments of others. When funds (or possibly the generosity of brand-name suppliers) allow I plan to do a full shootout of the available 8-12 channel radios (including the A9) presently on the market.

However, it would appear that the A9 isn't perfect and a small number of users are reporting that they've encountered a "dead zone" at the extreme of stick movement in some cases.

The symptoms

Apparently, some users have found that on at least one stick, movement near the end of travel produces no corresponding movement of the servo. In effect, there is a "dead zone" there.

Now this isn't the first modern computerized radio that I've seen exhibit these symptoms. In fact there have been a number of folks reporting similar problems with the FlySky/Turnigy/iMax 9X radio and I've encountered exactly the same symptoms on one of these radios I had here.

It's not the end of the world of course and, in normal flying, it's most unlikely that anyone's going to notice if the last fraction of an inch of stick movement does nothing...

Unless, that is, it happens to be the lower part of your throttle stick that is affected (as it was on the 9X I encountered. It is actually very frustrating if the throttle servo doesn't track the first part of the stick movement precisely.

In the case of the Hitec, performing the built-in recalibration process doesn't seem to fix this problem -- and I think I know why.

What's the cause?

I'm speculating here, based on many years experience with microcontroller-based systems mind you...

At the heart of a modern computerized radio transmitter is a microcontroller -- a kind of "whole computer in a chip". This component does a huge amount of work but the first thing it does is convert the voltage which comes from your radio's stick units into a digital value.

To do this, a system called an Analog to Digital Converter (ADC) is used.

Most low-cost microcontrollers have 10-bit ADCs, which means that they have a maximum resolution of 1024 different values (hence the reason we've had 10-bit/1024 resolution in our computer radios for so many years.

In order to provide the maximum resolution for stick movement, it's important that the sticks are fed with a precise voltage and that the center position of the stick corresponds with exactly the mid-range of those 1024 steps that can be measured.

Now, if the potentiometer that measures the stick position (by returning a voltage) is not positioned so that the center-position is the mid-voltage then we have a potential problem.

In such a situation, it is possible that the voltage returned from the stick when it is moved to an extreme position, falls outside the maximum or minimum that the ADC can measure. In that case, the ADC simply returns a value of 0 or 1023 (depending on which end of the stick is affected).

In effect, the ADC can't report anything below 0 or anything above 1023 so when a misaligned stick tries to deliver a voltage outside that range, the value remains unchanged and therefore no servo movement is produced.

What's the fix?

The ideal fix is to move the potentiometer so that the stick center-position accurately represents the center-voltage. Unfortunately, the way most radios are designed, such tiny changes to the alignment of the stick and the pot are not possible.

Although some computer radios (like the Hitec Eclipse 7) are designed to allow easy centering of the potentiometers so as to avoid an endpoint dead-zone, it appears that some other radios (and that includes the A9) are not.

In fact, more often than not, these days the stick gimbal is keyed to the potentiometer shaft by a flat on that shaft so no change in the relationship of these two parts is possible. Also, the potentiometer itself mounts on the stick system such that it is keyed by a small metal tab on the body of the pot.

So, if a mechanical realignment of the parts isn't easy, what are the alternatives?

The pot could be replaced -- since it's likely that the one causing the issues is simply slightly out of spec. This will probably do the job, unless it's a design fault in the stick system alignment that's causing the problem.

Another possible solution is to add a small-value resistor in series with one of the pot-wires so as to slightly reduce the total voltage that appears across the pot and thus ensure that the stick movement doesn't result in values that exceed the ability of the ADC to track it.

In theory, this will also produce a small reduction of the total servo-throw on that channel but a recalibration of the transmitter and a small tweak to the EPA values should compensate for that.

Of course the best solution for most people would be to simply send the radio back under warranty -- and I recommend, wherever practical, that this is what people do. Fortunately for A9 owners, Hitec's after-sales support is the stuff that legends are made of so you'll be looked after -- once they figure out the cause and remedy.

However, if there are people out there who'd rather give this a go themselves then I'm happy to provide a blow-by-blow explanation of how to calculate the necessary resistor value and where to put it so as to eliminate an end-point dead-zone.

And remember, if you want a full RCModelReviews review of the A9, just let Hitec know ;-)

It's a bit of a shame that Hitec opted to use a cheaper microcontroller with just 10bits of resolution (1024 steps). If they'd gone for an 11-bit ADC they could have allowed for any pot misalignment and achieved almost twice the resolution (like the equivalent offerings from JR and Futaba) as well.

Still, it's worth remembering that the A9 is a very well-priced product and when establishing a competitive price-point in the market, some compromises must be made.

Related links