Looking for a XYZ ARINC407 syncho explanation in simple terms

I have read a lot about this subject and can see how it works. Most of the material is 50+ years old and it is really mostly quite simple.

But I need some specific advice for something I am building.

It is a box which takes in serial data containing an angle, 0 to -180 (for western headings) or 0 to +180 (for eastern headings).

Yes, south can be both -180 or +180 but that's not a problem :)

I am using two DACs to generate the two output sinewaves, 400Hz. These are called X and Y.

I could optionally also generate the reference one, possibly called Z I think, also 400Hz but for reasons unrelated a lot more power capability is needed.

What I can't get my head around is how to arrange the phases.

I *think* the reference would be an INPUT to this box, and the angle is represented using the two other sinewaves, which I will be generating. The phase will be relative to the reference, somehow.

I can't work out whether X and Y carry the heading between themselves or whether they are shifted symmetrically around the reference, or whatever.

The signals will go to various instruments used in aviation, for testing.

Many thanks in advance for any info. I am sure the answer is very simple!

Reply to
Peter
Loading thread data ...

Is this a 2-phase system? Synchros are 3 phases.

"Phases" here doesn't mean continuous phase shifts in a synchro-like system, it means multiplication of the reference in the range -1,1.

Something like this might be it. I'm not sure of your application.

formatting link

We do that sort of thing in our synchro simulators, with ADCs and DACs and DSP and transformer coupling. We've done similar things with analog multipliers, at higher frequencies, like for oil debris sensor simulation.

This was fun, used when developing a digital synchroscope.

formatting link

--
John Larkin      Highland Technology, Inc 

The best designs are necessarily accidental.
 Click to see the full signature
Reply to
jlarkin

snipped-for-privacy@highlandsniptechnology.com wrote

I think that is probably it. It is widely used in avionics. For example on page 68 here

formatting link
you can see the reference coming in on H (Z,C being the GND for that) and the two sinewaves on X and Y.

However your diagram suggests that all three sinewaves are always in phase and it is only the amplitude of X and Y that varies.

A multiplying DAC would be a good solution but (a) needs the reference to be a good sinewave and (b) the DACs I have, inside an ARM CPU, have a limited Vref range.

So, if my theory is right about using the reference *phase* and generating the two X and Y outputs relative to that, I would just use the reference zero crossings for the timing and then generate two sinewaves of the appropriate phase and amplitude.

Reply to
Peter

That's how a synchro or a resolver or an LVDT works. The information is in the ratio of in-phase voltages. The 400 Hz excitation available on an airplane can be really ugly, "wild power", but shaft angles can still be accurately extracted from the voltage ratios. There are some clever tracking algorithms.

You probably need bipolar multiplication, too, in the range -1 to 1.

You could do that if the excitation frequency were constant. It's more normal to multiply the excitation waveform, analog or digitally.

PWM multipliers might be fun.

Reply to
John Larkin

John Larkin wrote

OK; I get it. I didn't realise this is an LVDT, but one which is bent around. An LVDT obviously cannot change the phase of anything.

I wonder how the system manages to output 180 degrees accurately.

Do you mean phase-locking an internal reference sinewave to the excitation?

Sure.

That's a very good point. And I know of systems which are ~500Hz e.g. the KFC225 autopilot.

That's one way to do multiplication.

Referring to your diagram, how is the input angle to be encoded into the two multiplication ratios?

Reply to
Peter

It's not. Just scale the input sine wave to the outputs.

For a synchro, there are three paths, A B and C. A resolver has two, X and Y. An LVDT or RVDT could have one or two.

Reply to
John Larkin

John Larkin wrote

You've lost me there :)

I have an input parameter, a decimal value from -180 to +180, coming via RS232.

I have the reference waveform, nominally 400Hz. This is supposed to be sine but is often pretty square.

I need to generate the two outputs X and Y.

Reply to
Peter

Suppose you want to simulate 45 degrees. Multiply the reference sine-ish wave by 0.707 and call that X. Multiply it again by 0.707 and call that Y.

If some time later you want to change the simulated angle, change the two multipliers. The Y factor is cos(angle) and the x factor is sin.

Reply to
John Larkin

If you want to simulate 45 degrees you need to add equal parts of an in-pha se sine wave and a quadrature sine wave (which is to say a cosine wave) and scale the sum back by 0.707 (which will leave you with the same peak ampli tude as the sine and cosine waves - though it won't peak at the same time a s either of them peaks.

John Larkin's exposition seems to have left out the phase shifts.

Not the clearest exposition ever posted.

--
Bill Sloman, Sydney
Reply to
Bill Sloman

Bill Sloman wrote

OK - super many thanks. I get it!!

I wonder how this is done in typical avionics instruments, given the reference is usually a crappy waveform. Typically it comes out of a crude two transistor oscillator driving a transformer, which enables it to be done without +/- supplies. So the XY outputs should be the same crappy waveform. And filtering won't do it because you will get a massive phase shift. Perhaps the answer is that it just works...

Doing it properly would involve a PLL. That would solve both the crappy reference waveform, and the frequency which is only roughly

400Hz.

One recent avionics product (from Garmin) no longer works with the usual reference; they insist on a pure sinewave :)

Reply to
Peter

Sloman is confusing synchronous motor with synchro. The only deliberate phase shift from the reference in a synchro/resolver/CT/LVDT/RVDT system is 0 or 180.

There are no intentional phase shifters inside a synchro or resolver.

In a real airplane, the 400 Hz can come from all sorts of places: ground power, one of the main engines, the APU, batteries, or "the thing you never want to see used", a RAT. So naturally, it's not very stable on frequency or voltage. There are some special alternators that make a decent 400 Hz over a range of shaft speeds.

Synchros work to arc-seconds precision with junky references. The angular data is encoded in amplitude ratios. I don't understand your application, but it sounds like you need two MDACs or equivalent.

Tacky. Possibly dangerous.

--
John Larkin      Highland Technology, Inc 

The best designs are necessarily accidental.
 Click to see the full signature
Reply to
jlarkin

I am just trying to generate the XYZ signals for testing avionic instruments.

Yes.

Reply to
Peter

Buy one of these!

formatting link

--
John Larkin      Highland Technology, Inc 

The best designs are necessarily accidental.
 Click to see the full signature
Reply to
jlarkin

You have been very productive in your life! :)

I need something small, and I will build 5-10 of them. It is a part of something else.

Reply to
Peter

Short attention span.

Let me know if I can help, design review or something.

--
John Larkin      Highland Technology, Inc 

The best designs are necessarily accidental.
 Click to see the full signature
Reply to
jlarkin

snipped-for-privacy@highlandsniptechnology.com wrote in news: snipped-for-privacy@4ax.com:

First off, there is no ARINC 407.

IN the 400 Series, there is 404, 424, and 429. It is all ATR comms.

There is a 604 for built in test equipment.

407 is only in interface spec. 3 wire.

Study ATR and the 400 series and answers about synching interlinks will emerge.

Reply to
DecadentLinuxUserNumeroUno

Get the book from , read it and come back with the remaining questions.

--

-TV
Reply to
Tauno Voipio

Tauno Voipio wrote

Thank you.

I've just gone through it and struggled to find something which directly tells me how this system is implemented.

Maybe some of the maths was applicable but maths is not my strong point :)

The diagram posted above by John, with the two multiplying DACs, seems very clear.

Reply to
Peter

A resolver transmitter is a shaft with an armature winding. The 400 Hz excitation powers that coil. The case has two stator windings, mechanically 90 degrees apart. The AC mag field from the armature induces AC voltage into the stator windings, voltage ratio ballpark

1:1, minimal phase shift. As the armature changes angular position, the coupling into the two stator coils changes in the obvious way.

It's just a pair of variable transformers.

A transmitter synchro has three stator windings.

Connect two units, and they can act as if an invisible coupling connects the two synchro shafts, but they are not generally used that way.

Reply to
John Larkin

I think that was how they were originally used, back in the 20s to ~70s, mainly in antenna rotators.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.