Sperry MK 37 Gyro

[koch]

Hi,

I am encountering a problem in interfacing a Sperry MK37 gyro to a Magnavox MX 2400 satellite communication.It seems that the satcom does not follow the gyro heading reading.I found out that there is a interface junction box in between them.Inside the box were three zener diodes in series with the "S1","S2" & "S3" signals.These zener diode were rated as 18volts/500mW.There are also a cermaic cap.0.1mF across each zener diodes with respect to the return line (S5).
I measured each signal lines (S1,S2 &S3) with respect to the S5 and got a voltage fluctuating about between 0.6 volts to 30 volts at the gyro side,while at the satcom side the voltage is between 0.6 volts to 2.2 volts.Is it normal?
Note the vessel is anchored along side the wharf.
Need help on this


Thks

 

[BrianG]

Hi there koch,

I don't know if this was once a working system and you are now trying to fix it, or is it a new system?

I have no data on the Magnavox mx2400, but it sounds like the outputs from the Sperry MK37 gyro might be "synchro" signals on the s1, s2 and s3 lines. If this is the case they will be a.c. signals of differing amplitude, relative to an a.c. reference signal, which is decoded to give the positional readout. Does the Maganvox have the right kind of interface to deal with this? If not this is fairly easily handled by using a synchro-digital interface circuit (various chips exist for s-d conversion).

Hope that helps a bit, but come back to me if you have any more questions.

Regards, Brian

 

[koch]

Hi Brain,

Thks for responding.
This system HAD been working some time ago according to the user(approx about 4 years ago).Ever since then this system has been idle till now.
So some ALEX decided to activate this system.
The MK37 gyro is a minus 35volts dc stepper gyro.That's why they have included a junction box with zener diodes on the S1,S2 & S3 lines.
I have some info on the Magnavox MX2400 satcom unit and it does have the necessary interface circuit to handle for this type of gryo.
I should have mention this on my earlier msg is that they (user) have a spare gyro heading pcb board available for the MX2400.I have replace the board on the magnavox MX2400.After the ship set sail for the trial, the test is still shows negative (Failed). The Magnavox MX2400 did not follow the gyro heading.
What type of of signals should i expect to see from the output of Mk37 minus 35volts steeper gyro.
This there a circuit which i can build to simulate this type signals.


Regards guilngan

 

[BrianG]

Hi Guilngan,
I'm slightly puzzled by your reference to "stepper" outputs. The only kind of stepper signals I can think of are for stepper MOTORS, and these are not normally high precision devices, having a resolution of perhaps 0.25 degrees at best. As far as I know the phased signals you would use to DRIVE a stepper motor would not be used for an OUTPUT signal to another device, in this case the MX2400.

I note you mentioned an s5 output previously. My colleagues and I are convinced that the sort of precision heading output you would expect from a gyro would be produced either from a synchro output (medium voltage levels, 3-wire s1, s2, s3) or a resolver (sin & cos outputs, low voltage, 4-wire s1,s2,s3,s4), plus a reference phase in each case.

Can you describe all the signal lines you have, plus any power supplies, and can you see what sort of waveforms are on each line with an oscilloscope?

Regards, Brian

 

[koch]

Hi Brain,

Sorry for the delay.

I have measured these signals lines (S1,S2 & S3) with respect to S5 using the scope .It is DC pulses.It vary between 0volt and 35volts at any one time when the gyro changes course.At no given time does all these signals shows the same voltages.


Regards,Koch

 

[BrianG]

Hi again koch,

I have to say that from your measurements I don't have any experience of an angular transducer giving out signals like these, however it may just be that they are a form of pulse-amplitude coded version of the analog 3-phase synchro signals I mentioned in my earlier post.

Going back to you earlier post - it seems probable that these signals are barely in the right kind of range for a microprocessor interface and might be on the high end at the 6V maximum you measured. Depending on the input circuitry, if there is no additional voltage limiting inside the mx2400 it is quite likely that one or more inputs are being overdriven by the lines which have 6V pulses. I don't know what kind of interface chip you have there, but on many sorts of chip if the inputs are over driven to the point where the substrate diodes start to conduct then all sorts of unpredicatble results can occur.

I think if you can find some more details about the mx2400 satellite unit - particularly the interface specification -it should be possible to work out what may be going wrong.

Regards,

Brian

 

[CapnGene]

Try here:

http://www.sperry-marine.com/mk37vt/specifications/Index.asp

Gene Augustin
TECHNICAL SYSTEMS ASSOCIATES

http://www.tsa-orlando.com

 

[boxxy]

I am not going to disagree with any of the above all of which is good information from engineers clearly more experienced with synchros and resovers than me - last used when I was a much younger man.

If at one time this system did work no matter how poor the resolution, your task is to get it back to that state. Persuing that, judging from your measurements I reckon the zeners, put there as a crude method of limiting the voltage transitions at the MK2400 input to 18v, have become open circit - you should be getting 0-18v pulses at the MK2400. The zeners have been dissipating too much power. Normally you would include a series R to limit the zener current and provide some voltage drop. Get some new 18v zeners of the same type (500mW?) and fit a series resistor with each based on a combined zener and load current of about 20mA and a max volt drop across the resistor of 12v (30-18) - that gives the R at around 600R. Do some more tests on that basis and adjust the value of R if necessary - the 100nF caps would seem to be unecessary. R needs to be at least 0.5 watts using these numbers.

Rgds, Boxxy

 

[Digitalwoodshop]

Koch,

As a retired Navy Analog Fire Control Tech FCC(SW) I was looking for the result of the problem. What was the fix?

Thanks,

Al