how to test an encoder

[Slim shady]

hi all
never tested an encoder before but i'm required to proove that the encoder i have is faulty and what may be its cause.
wasnt enough for sum person that when i replaced the encoder the machine worked fine and didnt over run.
i have a digisolver model no. e7n-g0256-5t0ir
i've split the housing and have a smaller motor inside with 6-8 wires connect. 2 are looped at the connect between the houseing and motor.

im lead to belive i need an oscilliscope but where to start is above me...
anyone can help???
slim:-?

 

[nakulak]

http://universalencoderchecker.com/

 

[Slim shady]

thank you

thank you

this is an option.
is there a way to use an oscilloscope?
and if so what will i nedd and how?

 

[a.bisnath]

can you contact the maufacturer?

can you contact the maufacturer?

they sometimes give a procedure or or test point checklist to verify the operation of the encoder.They have become quite complex now.recently saw one with smart electronics in side

 

[gar]

100624-1924 EST

Slim:

I will assume it is an incremental optical encoder.

Get a manual on the encoder. Determine the supply voltage and normal current. Connect to an adjustable voltage regulated supply with adjustable current limiting. If the encoder is 5 V or whatever set the power supply voltage to this value first before connecting the encoder. make sure you do not apply reverse polarity.

Monitor the current as you adjust the supply from -10% to +10% of the nominal voltage. +/- 0.5 V for a 5 V nominal value.

Make sure none of the output leads are shorted together or to anything else.

It would be useful to know what type of output was supplied.

Connect a Simpson 260 between one output that is not a once per revolution signal and the minus side of the power supply. The output should probably read toward the + supply voltage or toward the negative (probably called common or ground) without rotation. Rotate the encoder shaft and the output should be about 1/2 the supply voltage. Do this for each output lead. The once per rev output won't do much that you can see with the Simpson.

More later. Try this first.

.

 

[justdavemamm]

hi all
i've split the housing and have a smaller motor inside with 6-8 wires connect. 2 are looped at the connect between the houseing and motor.

That would lead me to suspect that it is not a digital encoder but rather a resolver. A resolver typically has 3 coils.

See http://en.wikipedia.org/wiki/Resolver_(electrical) for the basics.

See http://focus.ti.com/lit/an/sbaa144/sbaa144.pdf for waveforms on testing.

General you'll need a sine wave signal generator and a dual channel scope to test it.

Googling the mfg and part number turns up little about this leading me to suspect that either this company is defunct a long time ago or has been adsorbed by another company.

It might be from here: http://www.autotechcontrols.net/resolver/r_4.htm

 

[Slim shady]

thank you

thank you

thank you
i believe i have everything i need to test this resolver now except the diagram from the supplier
this is the correct supplier from autotech i've just fired an email to them about more information.
once i have this i'll be a smarter person.
thank u all let u know hoe it goes..

 

[gar]

100629-0714 EST

Slim shady:

If you have a resolver, then you have a two phase device. This would typically have 6 leads. Each pair is a coil. Coils are very reliable devices if their power dissipation is not exceeded.

Two of the coils are stators and mechanically offset by 90 degrees. Also there are opposite poles for these displaced by 180 degrees. These stator coils are excited by sine waves that are electrically 90 degrees apart.

The rotor has one coil coupled by slip rings or a rotary transformer. Slip rings can be a reliability problem.

If you have a resolver, then you should be able to do DC resistance checks on each coil. One coil of a resolver I have is 110 ohms.

When properly excited the output of the rotor will have a relatively constant output amplitude and an electrical phase angle relative to one of the stator coils that is proportional to the mechanical angle of rotation.

For testing without two phase excitation you could use a 5 V 5 kHz audio oscillator to excite one stator coil. There should be little coupling to the other stator. Output from the rotor should remain in phase with the excitation for half a rotation and then change phase by 180 degrees for the other half rotation. The amplitude will change with rotation.

Then change the excitation to the other stator coil and a similar result should occur, but the shaft angle will be shifted by 90 degrees relative to minimums and maximums.

.

 

[Slim shady]

Resolver testing

Resolver testing

thank you gar this is exactly what i'm dealing with and its a heap of help thank you

If you have a resolver, then you should be able to do DC resistance checks on each coil. One coil of a resolver I have is 110 ohms.

dc resistance check?? to get a sine wave i need to apply ac current true??

i've found my coils rotor 17.4 ohms
and both stators 180 ohms each

voltage on rotor is 1.8 v
output volts 2.63 v

but no indication of ac or dc volts
please clarify??

pretty much i'll excite the rotor with a sinewave generator and need to the sinewave and cosine wave one the outputs of the stators, ??

this is best done with the osciliscope right or can an applied voltage on the rotor provide voltage references on the stator coils?

 

[gar]

100629-1836 EST

Slim:

My resolver rotor has a resistance of about 58 ohms.

The fact that you are getting ballpark reasonable DC resistance readings is good.

The following reference may be useful.
http://www.optoresolver.com/help/tutorials/resolver_v_encoder.htm

The following is even a more important reference.
http://www.ddc-web.com/FileLibrary/GenericDocs.aspx
Then download "Synchro/Resolver Conversion Handbook"

I was wrong on my description of the normal operation of a resolver. The above tutorial indicates the resolver is excited on the rotor rather than two phases on the stators. Instead both stator signals are in phase with the rotor excitation and the stator output amplitudes vary as the sine and cosine of the shaft angle.

If your resolver is a brush type, then the brushes could be a problem. If it uses a rotary transformer, then there is not much to go wrong except for burned out windings, won't normally happen, bad solder joints, or bearing problems.

You should be able to apply a couple volts at 5 kHz to the rotor and observed the output voltage on each stator. If you can lock the rotor to the housing, then you could tap the housing with a small hammer and see if there is a perturbation in the output voltages.

You need a good mechanical fixture and some DDC components to evaluate mechanical position to electrical output.

.

 

[Slim shady]

resolver testing

resolver testing

alright i understand the operation of the resolver
and just got information from the suppliers from
irwin research
the housing contains 26 pins ABCDEFGHIJKLMNOPQRSTUVWabcXYZ

pin letter function
A G0
B N/C
C G1
D N/C
E G2
F N/C
G G3
H N/C
J G4
K N/C
L DIRECTION CONTROL
M G5
N N/C
P G6
R N/C
S G7
T N/C
U G8
V N/C
W G9
a N/C
b +5VDC
c GND
X CASE GROUND
Y SPARE
Z SPARE


THE RESOLVER INSIDE IS FINE all three winders on this brushless resolver are showing good ohms.
i think i have a problem in the housing wiring which is unexcessable
my problem now is to understand what the housing is doing .
starting to get alittle confused

 

[gar]

100630-0659 EST

Slim:

My guess. Your housing contains a resolver-to-digital converter.

I would relabel GND as 5 V COMMON. There may or may not be any connection between pin "c" and "X". You can check this with an ohmmeter.

Pins labeled G0 thru G9, 10 bits, are likely TTL parallel data outputs.

"Direction control" seems obvious, but is unlikely an input and probably controls nothing, but rather is probably an output that indicates direction of rotation of the resolver.

The only two inputs are +5 and GND. The 5 V common (GND) will also be the reference for the outputs.

If the resolver is good, which is most likely, and the whole assembly is bad, then either the oscillator supplying excitation has failed, or the resolver-to-digital converter has failed or output drivers if these are external to the converter.

.

 

[Slim shady]

spares

spares

my problem was that the motor was over running
lead me to the resolver
which i replaced
and no change
until i replaced with another spare on the shelf and then came right
that made me believe i have 2 faulty resolvers and one good
but the maintence supervisor believes that he brought two fairly new in which he wanted proof that they were faulty

it is a digital output TTL with multiplexing
number of counts per turn 256 greycode

pin L is tied to =5 volts for counterclockwise
and 0 v (GND) for clockwise rotation

based on whats been discussed i'm feeling pretty cofident that there is a fault inside the houseing in the resolver-to-digital converter . which is not accessable

just reading the type of output

am i only needing three wires
+5
sig ref 0 volt
output

i'm guessing output to be direction?

 

[gar]

100701-0736 EST


See the following for a discussion on Gray code, and slightly less than half way thru is a table of Decimal, Gray, and Binary.
http://en.wikipedia.org/wiki/Gray_code

From your information it appears you have absolute encoder, this derives from the use of the resolver, with a parallel output in Gray code. From the meager information on the pins there is no indication of a multiplexing function.

Further on in the above reference is a description on constructing an n-bit Gray code.

.