Erratic RPM Readings

[RockyMeLad]

While we had some high winds a few days back, I replaced my "dash" with new instruments. I now have a 3 1/2" ASI, the rectangular Taskem EFI, the rotor tach (from Ernie), and a compass. The transition went well except for the engine RPMs. Unlike the stand-alone gauge I was using, this one only connects to one side of the "lighting coil" with a "ground" reference. When I first put it on I was playing with props doing static test runs(too windy to fly), and it was fine. Then I flew with it and at full throttle (6300) it read randomly from about 1200 to 3600. :confused:
The engine sounded fine, but I landed and reconnected the original meter (also) and went up again. Same erratic readings on the EFI, but the analog meter (taped to my leg) was normal. Back on the ground it seems ok again.
There is a ground wire from the engine to the EFI. There is no discernable difference in the electronic "engine noise" on the radio. Suggestions???

[Al_Hammer]

Rocky, I just looked at the Taskem EFI manual online and it appears that
the tach does not connect to the lighting coil, it connects to the gray wire which is the dedicated tach wire on a Rotax.
-Unless you have a rectifier/regulator, in which case you can use one of the lighting coil wires as a source.


I wonder if certain types of regulators might cause erratic readings given the fact that they basically short out the coils many times per second. You might try switching to the gray wire if you aren't already using it to see what happens.


The tach input to the instrument is the gray harness wire. This must be connected to some suitable engine source. Most
Rotax engines have a mating gray wire on the engine for this purpose. The engine's lighting coil is another good source of a suitable
tach signal, but this can only be used if a simple rectifier/regulator is also connected since it provides the "ground reference" for the
tach signal. In this case, connect the harness gray wire to either of the heavy yellow lighting coil wires.
http://www.taskemcorp.com/pdfnew/EFI_Manual.pdf



the grey
lead provides six pulses per revolution

the grey lead is a center tap from one of the ignition power coils,

the pulses are created by the stator interacting with the flywheel,

the flywheel has twelve magnets, NSNSNS, etc

these produce six pulses per revolution, the grey lead pulling off as a
tach 'sender'

the trigger coils are energized by the magnets on the outside of the
flywheel, and there being two of them, we get two sparks per revolution,
one of the sparks being a waste spark as it occurrs just before bottom
dead center BBDC,

[RockyMeLad]

Thanks, Al
Methinks a Rotax is more common, but alas... I have an MZ-202. Very similar with a couple of variations.

1. It has 5 pulses vice 6 for Rotax. The Taskem EFI was easily configured for this.

2. There is no grey wire equivalent on the MZ-202. Your description did make me rethink the circuit though. The documentation with the original analog meter indicated either using the "grey wire & ground" or floating the "two leads from the lighting coil" to the meter. Polarity was only a concern with the former. Not having a grey wire, I used the latter originally. It is probable that the 2 "floating" leads don't really have a ground reference under operating conditions (both leads gave the same erratic results relative to ground). Need to dig through my junk-box to find a suitable 1:1 transformer, "float" the input to the 2 leads, then ground one side of the output with the other to the EFI. That should isolate the charging circuit and still give me a ground reference signal for the EFI.

Sometimes the obvious... isn't. :o
Thanks again.

[RICK MARTIN]

Damn Rocky,

I'd give a lot to be able to understand what you just said. Electicity is a real weakness for me.

Oh well, everyone can't know everything.

It sounds like you have a handle on it, so good for you!

[RockyMeLad]

Just a quick note that the center-tap transformer "fix" worked fine.
One of the other guys at the EAA hangar is also a ham and his junque-box was better than mine. He had a small transformer with a high impedance center-tapped winding. The "ends" were connected to the charging leads, the RPM lead from the EFI to one side only, and the center-tap of the transformer to ground.
After several hours, it's still working beautifully. It's nice to win one sometimes. :D

[Al_Hammer]

Rocky, It sounds like you worked out a good solution. That's great.
I was away for a few days and couldn't follow up on your 2nd post- not that you needed any help. :)

[RockyMeLad]

Al,
I think it's called "the blind hog finds an acorn syndrome". :D
My first try with just a transformer (balanced to unbalanced) went up in smoke after about 20 minutes. Didn't hurt anything but the transformer. Fortunately we have a great bunch at the EAA hangar, and one of the guys had the high impedence center-tap, and that worked. ;)

[scottessex]

Hey Rocky, I think I can help.

http://www3.telus.net/bc_triumph_registry/smoke.htm


:D :D :D :D :D :D :D :D

It's made for a triumph, but it might work in other foreign engines as well.

[PW_Plack]

Rocky, I was going to say something about a series resistor in the lighting coil side, but when you said all was working well, I figured you'd found a transformer with a high enough DC resistance. You did...but only on the second try!

Scott, that's a classic! Even better with the Lucas label!

[RockyMeLad]

Scott ol' Buddy...
Does it come in gallons? :D
or maybe 55 gal drums I could take to work? :D :D :D

[Adam H]

Well, I've got the MZ running and I ran into the same tach problems. I remembered this thread and used this transformer to create a 0v reference.
http://www.radioshack.com/product/index.jsp?productId=2103994&cp
It works now, but I see that Rocky had a problem with the first transformer he tried. Is what I used going to last, or do I need to add a resistor or something?

[Adam H]

Here is a schematic showing what I did.

[PW_Plack]

Adam, I got your PM.

The transformer is designed for isolating line level audio circuits, and will not live long without a resistor.

If the primary is 51 ohms, the primary winding will draw at least a 1/4 amp continuous current. In the audio circuits for which it's designed, it would be exposed to about .001 amp.

You need a series resistor in the primary, or the tiny wires with which it's wound won't live long. It's hard to know the optimum value without knowing the input impedance of the tach, and the minimum voltage it likes to see to trigger. Can you find those numbers?

[Adam H]

Here is the tach I'm using: http://www.mglavionics.com/EMS503.pdf

It has some info on the values it will take, I hope something here can be of use. It looks like the minimum voltage it needs to read correctly is 2v per pulse. Thanks for the help!

[PW_Plack]

Adam,

I reviewed the published specs, and there's no clue as to the circuit's input impedance.

I called MGL, and discussed the problem with a tech rep. He said the MZ's are having this problem everywhere, and Grand Rapids Technology has a regulator which solves the problem. He said it's very expensive.

I explained to him the transformer/resistor fix,and he suggested 2200 ohms would be a good start for the series resistor.

I'd stop by Radio Shack and pick up a pack of 2.2k resistors. (Whichever they have, 1/4-watt or 1/2-watt, will be fine, overkill for the application.) They come in multi-packs; start with one in series with the transformer. If it works reliably, great! If not, use a second one in parallel with the first resistor, which makes the combination 1100 ohms, add a third if necessary to get 733 ohms, etc, until it works.

I have to caution you, this transformer is a low quality piece, and external resistors will be a physically awkward package. Do the best you can to keep it all neat and out of harms way. If you have to go this way, consider sealing it all up with a glob of epoxy, or otherwize "unitizing" the parts for mounting. Your new gyro is too nice to be a rat's nest, and loose wires are always bad news for reliability.

[Adam H]

Wow, Paul! Thank you so much, you really went above and beyond calling the company and everything! So, just to make sure I understand you right, I attached an updated schematic. The dotted part would be the added resistor(s) if it doesn't work with only one.
One of the things I don't like about the MZ is that instead of simply having color coded wires coming out of the engine with a schematic showing you what they do and where they need to go like Rotax does, they give you this big, ugly wiring harness. I'm trying to keep the rat's nest look under control, but that harness isn't helping. I will take your advice and put the transformer and resistors in epoxy and maybe shrink a big piece of heat shrink tubing over it too.

[PW_Plack]

Yep - that's how it would look schematically!

[Brent_Brown]

I used a tiny tach worked great.

[Adam H]

Yes Brent,
I think I got a tiny tach along with the rest of old peg. I'm not opposed to using one, but I already bought this all-in-1 gage and without an accurate rpm reading, the hobs, flight timer, and needless to say, tach features of it would be worthless. I will install the resistors tomorrow and report back so future MZ users won't have to experiment.

[RockyMeLad]

Adam,

Won't deny it working, but I don't see how.

It appears that you have a "floating ground" through the transformer (and I use the term very loosely).

It might even be that in your installation that there is enough "common" ground through the frame that connecting the "+" line to one side gives
enough voltage to trigger the rpm circuitry. If so you might try just removing
the transformer completely, but leave the "+" line connected to one side.

If you do use the transformer, you definitely want a low current path through it or it may short out and then possibly short your charging circuit. Not good.

[Adam H]

I'm confused Rocky, I got the idea to use a transformer from you! Would you draw a schematic of how you used the transformer in your installation? From my meager understanding of electronics this is how I understand what I'm doing, the alternator is produucing ac current, there are 10 coils in an MZ alternator, so each lead is making 5 pulses per revolution. The tach needs to count one of those leads' 5 pulses. The tach also needs a 0 volt reference. The transformer is allowing me to seperate some of the ac current from the rest of it which is going to the rectifier regulator so I can mess with it and not affect anything else. By shorting the ac current coming out of the transformer, I'm combining the highs of the ac wave with the lows and getting a net 0 voltage which the tach needs to work properly.

[Adam H]

Ok, I'm really getting tired of messing with this tach!!! I tried using the resistor and it would not read. It would flash a really low rpm right when I start up the engine, and then nothing. I put 3 resistors in the parallel and still nothing, but it will work with none. I'm thinking that instead of messing with this junk transformer and protecting it with a resistor, couldn't I just get a better transformer? A Mouser part# would be great! Me confused, please help.

[kc0iv]

Ok, I'm really getting tired of messing with this tach!!! I tried using the resistor and it would not read. It would flash a really low rpm right when I start up the engine, and then nothing. I put 3 resistors in the parallel and still nothing, but it will work with none. I'm thinking that instead of messing with this junk transformer and protecting it with a resistor, couldn't I just get a better transformer? A Mouser part# would be great! Me confused, please help.

Adam,

I don't see how this circuit would work if it is wired as shown on the circuit you showed.

Looking at the transformer on the output it shows the leads connected together. I don't see how this would do anything. If it is wired as shown the output would be shorted out.

Leon
(kc0iv)

[PW_Plack]

Adam, sorry, Leon's right. I don't know how I missed that. I was focused on the area of the resistors.

The two wires from the secondary of the transformer go to the tach's sensor input. Eliminate the direct lead from the tach to the alternator.

(And start again with a single resistor.)

[GyroRon]

If someone took the time to figure up the right ratios, you could use a 7 dollar bicycle speedo to give engine rpm....

[Adam H]

Thanks so much for the clarification, Paul! So, this is how it would look?

[RockyMeLad]

Adam,

Now you're cooking. This one follows "good engineering practice". :peace:

My mod to your drawing really doesn't change the circuit function, but in my installation the "negative" lead for the tach was actually "ground". So I stuck it into the drawing for reference.