Replacing a volume control digital encoder.

[indyld]

I'm looking to work out the spec of a volume/switch encoder of type the Bourne/ALPS make with 3+2 pins (2 pins being on/off). It's actually for a car radio but I figure those that work with digitial mixers may have experience of the spec of encoders in these kinds of applications. Or at least some ideas on how I might try to reverse engineer the situation.

There aren't any markings on the faulty one and the exact look of the shaft etc. isn't something I can find searching online plus obviously what it looks like isn't the key thing. The exact form factor is not the issue, the problem is working out what might replace it. Is there a common range of specs used for this kind of thing to start seeking an alternative?

I have a o'scope etc. and imagine that the number of clicks plus however the pulses are arranged are the main differentiators. I may be able to power the circuit for testing, maybe not.

Cheers.

Edited February 21 by indyld

[themadhippy]

a rotary encoder is a rotary encoder, about the olny thing different between makes apart from the physical size is the number of pulses per rotation and whether its dented or not.There not much more than  a fancy spdt switch. 

[indyld]

29 minutes ago, themadhippy said:

a rotary encoder is a rotary encoder, about the olny thing different between makes apart from the physical size is the number of pulses per rotation and whether its dented or not.There not much more than  a fancy spdt switch. 

I've looked at different types and their applications and it seems likely that this may just be a mechanical one with what are effectively a bunch of rotating contacts, so when am next testing I will first check this. Would make sense as this is a consumer product and the encoder just needs to tell a controller to turn the volume up or down with relatively few rotations over its lifetime.

[sunray]

There are too many variations of rotary encoders to give a simple answer, My limited experience of 4 units were all completely different:

1)  3 connexions from a pair of proximity sensors and a magnet repeatedly sent one way or the other,

2)  a series of 4 contacts repeated right round which sequenced either 1234 or 4321 on 5 wires but the 5 wires had voltages on them so the output was a saw tooth rising or falling,

3)  5 connexions but I did't open it to see the mechanics 2 connexions simply saw the pulses and would park closed or open, other 3 formed a common and left/right movement

4)  contained what looked like a PROM on 5 wires.

[KevinE]

it will be a mechanical quadrature encoder. The thihgs to watch out for are as follows:

 

1. number of clicks (detents). Some encoders are 'smooth' and have no detents at all. Get one with too many and the circuit may see the pulses too quickly and refuse to move the figures, due to the way the anti-bounce circuitry works.

2. number of pulses. The pulses are a pair with a phase angle in between them, which is how the thing knows which way you're turning it. You can get a pair per detent, a pair per 1/2 detent, or 2 pulse pairs per detent. Get this wrong and the parameter you're changing will alter out of sync with the detents, not a huge issue but it can be annoying to use if you're altering something that must only go up or down one, for example.

3. Phasing. The pulse can change either on arrival at a detent, or half way between two. Far more common is to change half way between detents, but if it's the wrong type, the setting might randomly change up or down one even when you're not touching the control. (this scenario is rare).

4. Pulse direction (A leads or follows B). This is a common variation. This determines which way the setting will go depending on clock/anticlock rotation. Get the wrong type and turning the knob clockwise, for example, may result in the volume going down, and vice-versa.

They're cheap enough - order one, and if it works first time, great! If not, try another 😃

 

 

 

Edited February 21 by KevinE

[indyld]

Cheers. I've gone for the 'buy something that seems similar and see what it does' option (although I think I count 28 detents and can only find 24s). To gain more information, I figured I can either put the encoder back in circuit if possible for testing, or connect it to a small test rig with a few components, and use 2 channels of the 'scope to see what's what in terms of leading pulse etc.

Am pretty sure they are the mechanical type though as they don't seem complex enough to have hidden magic inside and a continuity test between A / B and GND/Com/Whatever reads as on/off/on/off....

At least the mechanical are the less complex to match up.

[KevinE]

you can break the thing open to count the stripes on the 'pot track'. It will tell you the number of pulses, (one stripe per pair) and of course the detents you can count by feeling them.

The other thing which I suppose is obvious is getting the right shaft. Usually they're a d-type with a shoulder. Get the shoulder in the wrong place and the knob will either stick out,  or sit down low and eventually after a few knob presses, will ground itself out at the bottom and you'll lose the switch action. It's all great fun isn't it.