647jc

647jc

Well Known Member
I’m in the final stages of an RV9A build, the panel has a Dynon D100 EFIS and D10 EMS, the engine is a new Lycoming O320 that came with a small light weight Sky-Tec starter and I also use the standard Odyssey battery. I have wired the starter circuit using Vans typical FW starter solenoid with diode suppression, also used Vans power cables from his wiring kit. I have the standard Vans 60 amp alternator also but do not have it connected at this time. The problem is, when I turn the ignition key and spin the starter (I don’t have the flywheel on the engine so when I engage the starter, only the starter motor spins), the Dynon D100 EFIS resets itself, screen goes blank for a few seconds and then everything is fine again. The D10 EMS does not show any problem during this time. I have several other microprocessor based controllers in the panel on at this time also and none of them exhibit any problem, only the D100 EFIS. Dynon sent me a refurbished D100 EFIS to help trouble shoot this problem and the refurbished unit has the exact same problem.

I have distilled the problem down to a fairly basic test situation trying to eliminate all possible suspects. I currently have the Dynon D100 sitting on a board on top of the forward fuselage skin, only two wires go to it, +12 VDC and Gnd directly from the fully charged Odyssey battery terminals. Problem still occurs when I turn the ignition key and spin the starter. When I turn the key and the starter first starts spinning is when the D100 reset occurs. If I continue to hold the key and continue to spin the starter, the D100 after resetting from the initial spin-up appears normal, the glitch only occurs when the starter first starts to spin but from then on the D100 is OK. I connected a digital ‘O’ scope on the battery leads and captured a pretty huge negative (down into the negative voltage) and positive spike on the battery positive terminal when the starter is first engaged and the spike rings a bit for a few hundred microseconds but then dissipates.

About all I can conclude is either the Sky-Tec starter is defective somehow or possibly my testing without the flywheel is problematic causing the power spikes or noise. I’ve checked all the usual suspects, power cable to starter, grounds, etc., everything is installed per instructions as near as I can determine. Maybe I should hook up the alternator in case its voltage regulator or noise suppression devices might mitigate the spike. I can’t explain why both D100 EFIS behave the same way yet the D10 EMS and everything else in my panel does not have a problem. If the D100 design is overly sensitive to short power spikes or glitches, I think others would have complained of this problem before me. Has anyone else observed this kind of behavior with a D100 EFIS or have others had problems with noise from the Sky-Tec starter?

If I can’t get to the bottom of this, some of the options I have are to just ignore the problem or possibly connect the EFIS to my nav/com power which is automatically turned off when the starter key is turned which will in fact reset the D100 EFIS when I start the engine also. I could also build a power filter for the D100 EFIS but I shouldn’t have to go to that extreme. I'm concerned once I start the engine with the alternator online and everything else, will the D100 EFIS go nuts or is this JUST an issue with the starter?

Comments, suggestions, ideas?
 
Last edited:
Okay, call me old fashioned, but I don't start my engine with my EFIS (D10-A) powered up. I have an avionics switch, and this is why. Any particular reason to start with the EFIS fired up?

Bob Kelly
 
I'm going to give you a really simple explanation which could explain this......it's simply called voltage drop. I guess before I say this is what's happening, I should ask if you have internal batteries in the units that are dropping off line - if so, then it is something else....

But if not, when you hit the starter, you put a tremendous load on the battery, and this drops the voltage to the point where the microprocessors reboot. this is essentially normal behavior (not desired, but normal). This happens with the GRT EFIS's, which is why many of us add an auxiliary battery, feeding power through one of the alternate power feeds, to keep the voltage from sagging during start.

This is just my experience with other units - and yes, this happens to the D180 in Louise's -6 if we forget to turn on the external Aux battery before hitting the starter as well.

Paul
 
647jc said:
I’m in the final stages of an RV9A build, the panel has a Dynon D100 EFIS and D10 EMS, the engine is a new Lycoming O320 that came with a small light weight Sky-Tec starter and I also use the standard Odyssey battery. I have wired the starter circuit using Vans typical FW starter solenoid with diode suppression, also used Vans power cables from his wiring kit. I have the standard Vans 60 amp alternator also but do not have it connected at this time. The problem is, when I turn the ignition key and spin the starter (I don’t have the flywheel on the engine so when I engage the starter, only the starter motor spins), the Dynon D100 EFIS resets itself, screen goes blank for a few seconds and then everything is fine again. The D10 EMS does not show any problem during this time. I have several other microprocessor based controllers in the panel on at this time also and none of them exhibit any problem, only the D100 EFIS. Dynon sent me a refurbished D100 EFIS to help trouble shoot this problem and the refurbished unit has the exact same problem.
Click to expand...

Yep, I concur with Paul. My D10-A has always dropped off-line momentarily when the starter is engaged. I asked some of the techies at Dynon if this would damage the Dynon and they said not to worry about it, it was plenty robust to take the voltage drop. I don't have an avionics switch or backup battery for the device and don't bother with shutting down the Dynon before engine start.
 
I’d certainly be surprised if the problem is a simple voltage drop. With the ring gear off the engine, the starter motor isn’t doing any real work and not drawing any significant current except possibly to overcome the initial startup inertia and activate the inbuilt solenoid / bendix. When I measure the voltage with the motor spinning I don’t see any appreciably drop. As I noted, I see a loud voltage ring (spike) of several hundred micro seconds right when the starter starts spinning, this isn’t from my FW starter solenoid but definitely from the starter itself, perhaps the bendix / inbuilt solenoid on the starter causes the ringing. I would also think other devices would be affected if in fact a true voltage sag was occurring, I did mention the D10 EMS and several other microcontroller based devices (flap controller and 2 trim controllers) are not affected, JUST the D100 EFIS. My test setup, with the EFIS directly connected to the battery should also eliminate any voltage drop from my wiring so the sag would have to be directly at the Odyssey battery terminals which is capable of providing cranking current in excess of 2250 Amps for 5 seconds. Perhaps we are talking the same thing, voltage spike, ring, drop, sag I guess could all be considered similar phenomena but to me, a voltage 'drop' incident typically lasts longer than a few hundred micro seconds.

When I spoke to Dynon they sounded a bit surprised to hear of the behavior and probably would not have offered to send a refurbished unit for testing if they thought the behavior was normal. I did NOT go with the internal backup battery option so that may also be affecting the problem. I personally think the EFIS power supply is overly sensitive to voltage transients of short duration but that’s just my opinion.

Whether the EFIS should be on a switched avionics bus or not is a valid argument. I use an EXP II Bus setup that has two main power lines that are on when the Master switch is on and that is where I connected my EFIS and EMS. I have my SL30, GTX327 and GPS connected to the avionics bus that is automatically switched off when the starter is cranked but I wanted the EFIS and definitely the EMS on line during startup and the Dynon manual says the unit ‘Will not be adversely affected by engine cranking’, not sure what that means now, perhaps it just means the unit will not be damaged.

Well, now that I know other people have / are experiencing similar behavior during engine start, I guess I will not be that concerned and just accept it as is. My biggest worry was how sensitive the unit was going to be once the engine was started, the alternator on line with the fuel pump, strobes, etc. Apparently that should not be an issue, it’s just when the starter motor is first spun up.

Thanks all for the input.
 
Last edited:
Possible cure.

In my opinion what you have observed is not unexpected. Short answer is I expect you could install a series diode and capacitor on the Dynon power supply input to prevent the shutdown.

A PM motor will look like a direct short across the battery at the instant of startup, 'a few hundred microseconds' would be pretty reasonable to see a collapse in your battery voltage. The current would be limited only by the impedance of the cabling. If you have a screen capture from the CRO that would be informative.

The ringing is probably the back emf from the windings mechanically vibrating due to the solenoid hitting the stop or some other mechanical resonance.

WRT the shutdown I am speculating here so I may be totally off beam but: Many if not most switching power supplies (the internal Dynon supply probably a switching DC/DC converter) of any significant size will have a soft or delayed start circuit the purpose of which is to delay or reduce the start up load placed on the supplying curcuit. This soft start prevents the massive overload that switching supplies can also put on the supply side. The soft start is not necessary for a battery supplied DC/DC converter but the integrated packs used in electronics have a more general application.

The soft start is usually triggered by the fast transient input voltage swing usually only associated with powering on. The circuit holds off the switching oscillator for a pre-determined period, unfortunately you are probably seeing this effect as a reset in the Dynon with the transient from the starter falsely triggering the soft start.

It would be informative if you could also capture on your CRO (using a second trace) the voltage of the Dynon supply output. This might be available on the Dynon interface, if not find somewhere inside if you're brave. You should see when the Dynon supply shuts down relative to the Starter load.

As I mentioed, to resolve the problem (if my hypothesis is correct) you might consider a series diode and capacitor in the Dynon power supply circuit. The capacitor might need to be only several 100 micro farads, but experiment to find a value sufficient to hold up the supply for the period the Starter chokes the battery volts. This will prevent the Dynon from seeing the transient.

HTH,
Doug
 
Last edited:
To answer a couple questions, yes the battery is fully charged. The capacitor / diode filter is what I alluded to in my earlier post; I would also use a much smaller cap in parallel to the filter cap to help knock down the high frequency ring. Yes, I did capture the voltage trace but can only describe it. It starts with a negative spike about 80 usec going down to about -10 volts. Then it rings back up to about + 20 volts. The ringing only lasts about 3 cycles or 400 usec total and after the first cycle is only a few volts in amplitude. Not sure what type of supply the Dynon D100 EFIS uses and I don?t want to break it open and start poking around. I?ve built quite a few embedded controllers over the years and I?ve never used a DC to DC supply, just simple IC regulators with input filtering so I don?t have any experience with them. I have 3 of my own embedded controllers in this airplane and they have no problem with the startup power spike, neither does the Dynon D10 EMS so I suspect the D100 EFIS power supply design is just more sensitive to the power spike. If the D100 is in fact interpreting this very short spike as a loss of power and generating an internal reset then I think its power supply is overly sensitive compared to other devices including their D10 EMS, but that?s just my opinion. Regardless, I can live with it the way it is, its just annoying. Now that I know others experience this same behavior with their Dynon?s, I?m not overly concerned about it.
 
You must log in or register to reply here.