Checking Tach

NeverBlue · Post by **NeverBlue** » Wed Dec 30, 2015 3:09 pm

You need to have a reference...the ignition makes the most sense...then you'll know for sure.

NeverBlue · Post by **NeverBlue** » Wed Dec 30, 2015 9:19 pm

I can say however that the measurements "on the bench" of the powerline frequency and the close agreement between the two units "in the field" give me an extremely high degree of confidence in their power to measure the prop speed sufficiently accurately to confirm the prop is well within the allowed 4% tolerance

Mearsuring a bulb is completely different than measuring a spinning prop through a windscreen with reflections from both and any other possible light in the area.

Completely different...

'kinda like testing snow tires for traction control in the mud...

NeverBlue · Post by **NeverBlue** » Thu Dec 31, 2015 9:42 am

The man with one watch knows the time. He with two is never sure.

...unless the second one is Atomic...

Nice spin though!

Troubleshot · Post by **Troubleshot** » Fri Jan 01, 2016 11:27 am

//Personal attack removed by Sulako

It's 2016, a time of new beginnings and resolutions. Maybe yours should be to play nicer, and get fewer strikes. Just a suggestion.

cgzro · Post by **cgzro** » Fri Jan 01, 2016 11:52 am

reflections

I suspect that reflections are unlikely to be a problem given the speed of light and the distance covered would mean they are arriving within a few nanoseconds of the direct pulse and would therefore be integrated into that pulse as far as the rest of the circuitry is concerned. If however you were trying to make something for extremely high strobe rates this I suppose could be an issue (as it is for radio and optical fibre communications at high data rates). Photofly can no doubt describe how the analog and digital parts of the circuit react (or not) to such closely spaced pulses.

photofly · Post by **photofly** » Fri Jan 01, 2016 7:59 pm

Really back of the envelope stuff: rise time on the input pin to the microcontroller is R-C limited to about 1.1 milliseconds, so it shouldn't be possible to trigger faster than about 900Hz (corresponding to 54,000rpm) - light travels about 180nm in that sort of time, so if you're worried about reflections from a mirror it would have to be that far away and you'd need a very bright light for it to be picked up by the sensor.

If I ever get me a decent waveform generator I can do some measurements.

NeverBlue · Post by **NeverBlue** » Sat Jan 02, 2016 9:49 am

Speed of light has nothing to do with it in the time frame in which we are measuring.

I will ask again...

If they work perfectly...and are soooo accurate...and never go out of cal...why the different readings?

None of you can answer that.

Suspect maybe....but prove...no

My methods leave nothing unknown...nothing.

photofly · Post by **photofly** » Sat Jan 02, 2016 12:19 pm

NeverBlue wrote: If they work perfectly...and are soooo accurate...and never go out of cal...why the different readings?

Which different readings?

cgzro · Post by **cgzro** » Sat Jan 02, 2016 1:13 pm

Speed of light has nothing to do with it in the time frame

You cant have it both ways, either a) reflections are a problem, in which case the reflected flickering signal has to arrive after a sample period is complete and therefore add errouniously to the count and therefore RPM, or b) Speed of light is not relevant because its so fast the reflection cannot ever arrive outside of the sample interval given the small distances involved.

I suggested a) , photofly put numbers to it confirming a) and it seems you are now also agreeing with a) contrary to your previous assertion.

NeverBlue · Post by **NeverBlue** » Sat Jan 02, 2016 3:18 pm

Please...I'll let you know when I agree...

You nor anyone else can tell me if any of what the photocell picks up are reflections or not.

NeverBlue · Post by **NeverBlue** » Sat Jan 02, 2016 3:35 pm

Which different readings

2 RPM difference in the air...5 RPM difference on the ground...which one is right? Is it within 95 RPM of the actual engine RPM? You can't answer I know. ..

Why is it outside the xtal tolerance?

photofly · Post by **photofly** » Sat Jan 02, 2016 3:42 pm

The measurement is not outside the crystal tolerance.

The engine speed isn't constant. Each unit measures the average speed over 16 rotations.

Since the measurement periods for the two units aren't synchronized the average RPM during the measurement period for unit 1 is different from the average RPM as measured by unit 2, so they show slightly different figures.

As they should.

Is it within 95 RPM of the actual engine RPM? You can't answer I know. ..

No physical quantity can be known exactly, except for the speed of light which is a defined quantity. Everything else has to be measured by experimental apparatus. All measurements come with an error range, based on the best endeavours of the person doing the measuring to assess the errors in the system doing the measuring.

Here's a diagram from 2010 showing how measurements of G, the universal gravitational constant, have been updated and narrowed over the years. What's the true value? Who knows - although the people doing the measuring have given their assessment of the range of values within which they are each "sure" it lies:

So it is, even with something as ephemeral as the engine RPM at any given moment. What's the real value? Who knows? Who cares? We care if we have a value that's close enough to check the tach.

NeverBlue · Post by **NeverBlue** » Sat Jan 02, 2016 4:02 pm

It is outside the xtal tolerance.

Who cares?

What the heck do you think we've been debating here?

If you used the ignition as a reference instead of the sun there would be no error...period.

If you checked the actual tach for accuracy in the same situation you're actually using it you would not be guessing.

...and guessing has never been "good enough" for me....

5 tachs like yours ...not one 'll read the same.

cgzro · Post by **cgzro** » Sat Jan 02, 2016 4:06 pm

You nor anyone else can tell me if any of what the photocell picks up are reflections or not.

The photocell most certainly can pick up reflections however as we have shown It does not matter because a reflection arrives so quickly that it cannot have any effect unless either the speed of light is reduced or the reflection distance is increased (to kilometres). Neither of these are possible thus the reflection cannot have any effect. This is true irrespective of the wavelength(s) of the light used or the directionality of the light.

If you wanted to confuse the tach you would need another source of flicker within the 20 degree arc or so that was not a reflection of the propellor flicker and of equal or greater magnitude.

photofly · Post by **photofly** » Sat Jan 02, 2016 4:08 pm

NeverBlue wrote:It is outside the xtal tolerance.

Who cares?

What the heck do you think we've been debating here?

If you used the ignition as a reference instead of the sun there would be no error...period.

If you checked the actual tach for accuracy in the same situation you're actually using it you would not be guessing.

...and guessing has never been "good enough" for me....

5 tachs like yours ...not one 'll read the same.

How do you know *your* test equipment gives the right answer?

Troubleshot · Post by **Troubleshot** » Sat Jan 02, 2016 4:15 pm

Troubleshot wrote://Personal attack removed by Sulako

It's 2016, a time of new beginnings and resolutions. Maybe yours should be to play nicer, and get fewer strikes. Just a suggestion.

So there is a moderator in the maintenance forum! lol

I'm gonna take a self vacation from this thread seeing how there is one guy absolutely knockin it outta the park with technical information and a guy that knows nothing about it questioning his every post...hardly worth reading anymore because it has become a special ed class.

NeverBlue · Post by **NeverBlue** » Sat Jan 02, 2016 4:27 pm

Because measuring ignition is easy and exact....you said yourself "frequency is one of the easiest quantities to measure"...or something like that.

The I-A 1790 tach tester uses an optical pick-up with it's own light source in an enclosure measuring a rotating shaft that the actual aircraft tach is connected to...it uses a xtal for the processor and display.

...and it will without a doubt tell you if the aircraft tach is measuring RPM correctly or not.

They generally have ZERO error...0.....zilch...none...

...put 5 calibrated 1790 units together reading the same instrument and they'll all read exactly the same...guaranteed.

photofly · Post by **photofly** » Sat Jan 02, 2016 5:38 pm

Reading the same is not the same as reading accurately. They can very easily be giving you all exactly the same wrong answer. How do you know they're accurate?

Besides which- have you put five units side by side to check? Otherwise it's just idle boasting.

GyvAir · Post by **GyvAir** » Sat Jan 02, 2016 5:49 pm

Let’s say you took the workings of your 5 freshly calibrated 1790 units and rigged them all up in a lightproof enclosure large enough to house the aircraft assembly on which we wish to check the tach accuracy “on site”. Have them arranged such that they are all focused on the prop shaft.

Place a piece of reflective tape on the prop shaft, run the engine as per normal, take readings from all 5 1790s during the same time period.

Will they all have the exact same readouts, all at the exact same time, as the engine/prop speed fluctuates?

photofly · Post by **photofly** » Sat Jan 02, 2016 5:57 pm

The 1790 is a unit into which you insert the tach after removing it from the aircraft. You couldn't do 5 simultaneous tests if you wanted to, so it is just an idle boast after all.

Additionally it only claims accuracy to +\- 1RPM just like the other systems. That also use a crystal.

Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach

Re: Checking Tach