There are at last three easy ways to kill a timer: The 5V and ground pins can be switched if the connection to the ESC is offset by one pin; if using the FM-9 system, by connecting the programmer incorrectly; and through static electricity. In addition, there may well be some cases of natural deaths (aka “infant mortality”). Until the advent of the Castle Creations’ ICE ESC and its use of a switching power supply (replacing the Phoenix’s linear regulator) for the BEC’s five volts (which you can program to more than 5 volts as another timer killer). But the important fact is that these acts all render the timer unable to even initialize an ESC, let alone turn it on. Unfortunately, a dead timer tells no tales, so autopsies are of little or no value.
The coming of the ICE seemed to change that; a number of failures appeared, relatively close together. At first, suspicion fell on the possibility of voltage spikes on the BEC’s 5-volt supply to the timer. In response, I added two passive components to the FM-9 timer. Then I received a couple of timers that weren’t quite dead (although they still didn’t initialize or run the motor); they had a damaged output driver on the pin that supplies the throttle signal to the ESC. The embedded program was still running the program properly but the output voltage was too low to be recognized by the ESC as a legitimate logic signal. In response, I added two more passive components to the FM-9 timer, as well as making firmware changes. These changes appeared to make the timer more robust, although most flyers (who don’t fly anywhere as much as Norm!) have had none of these problems.
But now Norm and others feel that we have a much more serious problem, one in which the ESC causes the timer to reset (to motor off) during a high-g maneuver, presumably because of a momentary drop-out of the BEC voltage. Apparently, this doesn’t damage the timer or the ESC but it can and has been disastrous for the health of the model! Ironically, there have been suggestions that the data recording that the ICE provides, thereby storing clues to problems, also makes it more likely to produce anomalies on its BEC output…
(I have speculated that a connection problem between the timer and the ESC could be involved, too. The male pins on the timer are gold coated and should be long-lived, but the ESC provides spring-loaded female contacts that may be subject to wear.)
If the BEC from the ICE is the source of the problem, obtaining the 5-volts directly from the battery with our own linear regulator on the timer board should solve it. The FM-9R, FM-9VR, and FM-9Rretract timers already do this; they were developed for the Schulze F2B ESC (no BEC) and the Phoenix (with 4S or 5S LiPo batteries). A linear regulator like the 7805 provides a very stable 5-volts but it is inefficient when working with a high voltage because it dissipates power equal to the voltage drop times the current. The FM-9 requires only about 0.7 mA, so this means that the power dissipated in the 7805, even with a 5S battery, is only (0.7 mA) x [(5 x 3.7V) – 5.0 V] = 9.45 mW, which the 7805 can easily do, even without heat-sinking. However, we did find that it couldn’t provide more than the current for one wheel retraction without overheating and shutting down when Paul Walker tried using it; switching to CC’s external BEC solved that problem.
Note that when using a timer with its own BEC generator on board along with an ESC that provides a BEC, it is vitally important to disable the ESC’s BEC so that there is no fighting over any small voltage difference, with one trying to charge the other. As CC advises, this disabling is accomplished by temporarily cutting the middle (+5V) wire in the 3-wire cable between the ESC and the timer; this disabling must be temporary, as Norm reminds me, because you will probably want to be able to re-program the ESC if you are using a Phoenix or an ICE ESC. Also, the on-board regulator has to have its own connection to the ESC’s battery leads; this can be accomplished by opening up a gap in the leads and soldering in the connections or by using E-Power’s E-Tap (which I also can provide).
I’ve just made some relevant measurements on and with my timers. For one thing, I measured the supply current with a 5-volt supply: 0.5 to 0.7 mA, meaning a maximum power dissipation of 3.5 mW (0.0035 W). Compare this with the 5A and 25W capability of the ICE’s BEC! I’ve also found that this FM-9 timer would continue to generate a pulse over its entire rated supply voltage range of 2.0 to 5.5 volts, as well as even lower. But the pulse width did increase from the proper 1.004 ms (throttle-off) at 2.5 volts to 1.046 ma at 2.0 volts, and 1.64 ms at 1.75 volts.
Then I connected this timer to a Turnigy Super Brain ESC-60A ESC that was powering an E-Flite Power 25 motor. To my surprise, the motor continued to run at flight power down to a supply voltage of 2.0 volts! Below that voltage the ESC and motor made painful noises and I terminated the experiment. Next, I substituted a CC ICE Lite 50 ESC for the Turnigy – and found the same, unexpected result. This suggests that, if a BEC dropout is responsible, the voltage must be dropping, quickly, below about 2.5 volts, at least momentarily. But surely the battery voltage can’t be dropping below the 8 volts or so needed by the 7805 and so an on-board 7805 should provide a steady voltage to the timer.
Another possible solution to a BEC problem is to add a large (memory back-up) capacitor on the timer board. I am designing a new circuit board to test this idea. A third approach, also suggested by hard-working Norm, is to use a separate coin-type battery on the timer board, but this is the hardest to implement.