help with reverse engineering a couple of found servos's

[nick mulder]

Hi,

I recently took a punt and bought (sight unseen) the remnants of a tour buses 'satellite finder' hoping to find some stepper motors -

Instead I got two DC brush motors with encoders - I dont know too much about them tho, the encoders seem to be around the 1000 ppr range and have five pins coming out of them but only 4 of them are wired - do I have an absolute encoder ?

1 wire is red, 1 brown and 2 black - what is a reasonable guess on the pin-outs ?

the disk is thin metal with the encoding bits are very (very) finely cut out, the sensor wraps around the disk with about 1mm either side - is it a hall-effect sensor or IR LED/phototransistor set up ?

The motor itself I'm assuming is 24v and it runs with two 9v batts in series - but what interests me is that it has 4 brushes but only two are wired (two of them 90deg/next to each other), the other two are still intact but the connecting tabs have been snipped off so they cant be accessed without unscrewing the brush cap/holder thingy's ...

Why snip the other two poles off ? can I use all four brushes and how ? or is it fine with just the two ?

more info: If I pull out a brush and rotate the shaft 360deg I see there are 18 commutator contacts per/rev ...

A few questions there, but for now the most important are the encoder pin-outs and whats the dealio with the 4 brushes
any help appreciated

cheers,
nick

[NC Cams]

Dumb but important question: Any ID tags on motors or or encoders???

[molitovv]

Originally Posted by NC Cams Dumb but important question: Any ID tags on motors or or encoders???

Nope - else I would have probably got some more info on them from a good web search - nothing so far, and I cant get any info on the satellite finder they are from either...

will try to get some pic's but I think I have done a good job describing them in any case

[NC Cams]

Sorry, had to ask.

Seen cases where the info was there but not provided.

No offense meant, hope none taken....

[Al_The_Man]

Originally Posted by nick mulder - but what interests me is that it has 4 brushes but only two are wired (two of them 90deg/next to each other), the other two are still intact but the connecting tabs have been snipped off so they cant be accessed without unscrewing the brush cap/holder thingy's ... Why snip the other two poles off ? can I use all four brushes and how ? or is it fine with just the two ? Whats the dealio with the 4 brushes

Are you sure the pairs are not connected internally, this is the way it is normally done. If you can measure continuity between any two, there should be zero resitance between two coupled brushes.
Al.

[molitovv]

Originally Posted by NC Cams No offense meant, hope none taken....

nah none taken ! cheers (;

[molitovv]

Originally Posted by Al_The_Man Are you sure the pairs are not connected internally, this is the way it is normally done. If you can measure continuity between any two, there should be zero resitance between two coupled brushes. Al.

Pretty sure they aint from what I could see, but I am at work right now and cant double check - I'll get the meter on them tonight

If they are connected internally what would be the use of the tabs that were snipped off ? was it just a redundancy that made it easier in terms of production cost or ?

anyways will check first

cheers,
Nick

[unterhaus]

I wouldn't mess with the brushes. I can't imagine that they weren't connected in service, so unless someone messed with them after that, it should be ok.

On the encoder, best case is that there is a 5v line, ground, and two phase lines. The fifth line would probably be the index line. It probably isn't too hard to figure out what lines are power and ground, then the other lines can be monitored with a voltmeter to sort them out. Ground is usually very easy to figure out, since it goes to everything. You should be able to figure out power since it is feeding one side of the led's, positive side of electrolytic caps, and if you are lucky, the power pin of an ic that you can look up online.

Halls need a magnet, not a slotted disk. What you are seeing is a led emitter/phototransistor.

[nick mulder]

Originally Posted by unterhaus On the encoder, best case is that there is a 5v line, ground, and two phase lines. The fifth line would probably be the index line. It probably isn't too hard to figure out what lines are power and ground, then the other lines can be monitored with a voltmeter to sort them out. Ground is usually very easy to figure out, since it goes to everything. You should be able to figure out power since it is feeding one side of the led's, positive side of electrolytic caps, and if you are lucky, the power pin of an ic that you can look up online. Halls need a magnet, not a slotted disk. What you are seeing is a led emitter/phototransistor.

I tried gnd on the brown line and +ve on the red and the LED lit up nicely ...



The rest of the pins are still puzzling tho - there are three pins left for the phase/index - but only two are wired up - the 3rd and the 5th both of these are black.. gnd and +ve being 1 and 2 and the 4th lead is left unconnected.

Upon scoping them I get a solid square wave from pin 3 running at relative frequencies to my turning the motor shaft at ~3.5v peak to peak (with a hint of duty cycle corresponding to I'm guessing the encoder disk slots)

pin 4 I get the same relative frequencies but the wave is much smoother with rounded 'corners' at the peaks and the verticals are visible in the scope (ie. they aint as vertical as the output of pin 3) - it is also only ~1.2volts peak to peak... aaaaaand : not as much duty cycle apparent either

pin 5 which is wired up from the old satellite finder gives me a static ~1.5volts with no response to shaft movement (in either direction)

I am new to these, so dont know what to look for - keen to learn tho, any help appreciated

cheers all,

Nick

[unterhaus]

the index pin is only going to change once per cycle, so you might miss it, or it's just not there.

I'd find it very strange, but what you are describing almost sounds like an analog encoder on one channel, and a regular ttl encoder on the other channel. I'm having trouble imagining how you use such a thing, unless you condition the analog signal into ttl, and it happens to be the other phase from the existing ttl signal.