Author Topic: replacement rotary encoder  (Read 1954 times)

0 Members and 1 Guest are viewing this topic.

Offline con-f-useTopic starter

  • Supporter
  • ****
  • Posts: 807
  • Country: at
replacement rotary encoder
« on: December 10, 2021, 09:37:27 am »
So, I'm looking for a replacement of the pictured Diehl rotary encoder from a 1986 microwave, I know next to nothing about. It's 20mm by 20mm, and I'm not sure if it's gonna be compatible with just any standard modern one. Can one of you fine gentlemen reassure me, and point me to a couple of options?
 

Offline HB9EVI

  • Frequent Contributor
  • **
  • Posts: 722
  • Country: ch
Re: replacement rotary encoder
« Reply #1 on: December 10, 2021, 10:01:24 am »
I'm not so sure about the encoding scheme it is using; at a first glance it doesn't look like gray encoded.

In general ALPS is a good source for almost any kind of rotary encoders, also for older devices, but at least you should know what code the old one uses. Maybe somebody else can give you a hint.
If ALPS has nothing you can check for a Bourns encoders
 
The following users thanked this post: con-f-use

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 6526
  • Country: fi
    • My home page and email address
Re: replacement rotary encoder
« Reply #2 on: December 10, 2021, 12:07:11 pm »
So, I'm looking for a replacement of the pictured Diehl rotary encoder from a 1986 microwave, I know next to nothing about.
It looks like it has only three pins, with the outermost pins either not connected or connected to the center pin depending on the rotary position, with possibly 12 positions.

Could you use a DMM (even a cheap one!) in diode/beep mode to construct a table for the positions?  I realize it is a bit of work, but if you do it systematically, it should not take you more than 15 minutes or so.  Name the three pins A, B, and C.  In each orientation, check the continuity between A-B, A-C, and B-C.  If no continuity, mark 0 or Open; if continuity/beep, mark 1 or Closed:
$$\begin{array}{r|c|c|c}
\text{Position} & A-B & A-C & B-C \\
\hline
~ & ~ & ~ & ~ \\
~ & ~ & ~ & ~ \\
\end{array}$$
That table will tell exactly what the encoder is, and you should easily either find a replacement, or devise a replacement.
 
The following users thanked this post: con-f-use

Offline con-f-useTopic starter

  • Supporter
  • ****
  • Posts: 807
  • Country: at
Re: replacement rotary encoder
« Reply #3 on: December 11, 2021, 10:45:52 am »
So, thanks Nominal Animal and HB9EVI, it's rare to find so helpful and friendly people in a forum, when you ask relatively basic question like mine.

The problem with the encoder is, that it seems completely shot. I measured the connection between the pins and all of them are open, i.e. no connection. That is the case even if I turn the knob, for all positions of the knob. That was the case even why I hadn't opened the encode up and it was still soldered on the microwave's controller board. Seems like the contacts are complete abraded/corroded away.

So I attached a better picture of the contact pcb. There's three pins on the pcb which I have label R (red), C (common) and B (blue). R is connected to the left three pads, B is connected to right three, C is connect to the inner ring. The knob turns a bar which I have colored brown in the last picture. I think the way it used to be was, that the bar made contact with the inner, black ring (pin C) all the time. Then depending on the knob position, you get BC connected, then RCB, then RC, then none, when you turn clockwise, if that makes sense. There's 24 clicks for a full turn.

« Last Edit: December 11, 2021, 11:18:27 am by con-f-use »
 

Offline HB9EVI

  • Frequent Contributor
  • **
  • Posts: 722
  • Country: ch
Re: replacement rotary encoder
« Reply #4 on: December 11, 2021, 11:04:21 am »
to conclude from one of the pics in your first post, the contacts seem still quite ok, as far as that can be estimated from distance.
but on the contact pcb seems to be quite much residue from the contact movements; maybe there is some surface leakage which prevents proper evaluation of the encoders actual position. you could give it a try to remove some of the residue with a contact spray and wash it off with IPA afterwards.
 

Offline con-f-useTopic starter

  • Supporter
  • ****
  • Posts: 807
  • Country: at
Re: replacement rotary encoder
« Reply #5 on: December 11, 2021, 11:39:06 am »
Yeah, I'd rather replace the whole thing, because the pins do look worn and corroded on the tips, if you look up close in natural light and now that I've taken the thing apart and re-assembled a couple of times, they are all bend, too.
 

Offline HB9EVI

  • Frequent Contributor
  • **
  • Posts: 722
  • Country: ch
Re: replacement rotary encoder
« Reply #6 on: December 11, 2021, 11:44:16 am »
sure that you'll have to replace it; but you still don't know the encoding scheme; cleaning it up and getting it back to life at least for an evaluation of the encoding scheme could help you finding the right replacement.

in the meantime you can try to order 1-2 encoders on ali for a few cents; as quick drop-in replacement it could give you a clue.
 

Offline con-f-useTopic starter

  • Supporter
  • ****
  • Posts: 807
  • Country: at
Re: replacement rotary encoder
« Reply #7 on: December 11, 2021, 12:14:23 pm »
So as I said, I can't measure it, because there is no contact in any position - even after cleaning. From the design, I'd say the encoding turning clockwise is supposed to be:

Code: [Select]
Pos: RC RB BC
01: 001
02: 111
03: 100
04: 000
... from there it should repeat - as said, there is 24 positions.
« Last Edit: December 11, 2021, 12:19:44 pm by con-f-use »
 

Offline HB9EVI

  • Frequent Contributor
  • **
  • Posts: 722
  • Country: ch
Re: replacement rotary encoder
« Reply #8 on: December 11, 2021, 01:38:19 pm »
makes sense; then watch once the encoders from ALPS with 24 indents

https://www.mouser.ch/c/electromechanical/encoders/?m=Alps%20Alpine&resolution=24%20PPR
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 6526
  • Country: fi
    • My home page and email address
Re: replacement rotary encoder
« Reply #9 on: December 11, 2021, 02:29:54 pm »
I'd say it looks like 6 PPR, 24-position, ¼-cycle per detent contacting encoder.  (Any two of these is sufficient; typically they're described as 24-position, 6-pulses-per-revolution contacting encoders.)

Good news is these are easily available.  Bad news is that I can't find an exact drop-in replacement.  The closest thing I can find is the Bourns ECW series, say ECW0J-J24-BC0006L, ECW0J-B24-AC0006L, ECW0J-B24-BC0006L, or ECW0J-B24-HC0006L.  At least TME has some in stock.  Note that these are slightly larger than the Diehl encoder; see the PDF for the exact measurements.

The most problematic thing is the feet pattern and the asymmetric positioning of the pins.  To solve that, you could make a precise drawing (we're talking 0.1mm resolution or better) of the encoder, and make an adapter PCB in say EasyEDA so you can order a set of 5 for a few euros ($2 + shipping) at JLCPCB.  (It's really easy to do, and I certainly could do the PCB for you, if you took a picture with the board flat next to a ruler at the exact same distance from the camera.)  This will only add 2mm to the thickness.

It may be that the Bourns ES series (shaftless!), say ESD0D-S00-GC0006L (link to TME/AT) might be a better match: visually, the footprint and the dimensions seem to match very well (PDF) except for the pin spacing; but, if we had a precise image of the Diehl PCB (flat, with a precise ruler next to it), we could check.  The shaft itself can be made from 6mm plastic rod (PA66/nylon) with a lot of fettling with very few tools, to fit the existing knob.

So, if I may suggest, could you take another picture like the raw.png above, but with something behind the PCB so that it is absolutely flat; and with a precise ruler (at the exact same level as the PCB) right next to the PCB?  Making sure the camera is exactly centered on top of the board, so the image of the board is square and not a trapezoid? With 2× to 3× the resolution, so I can make a dimensionally accurate SVG to represent it?  Also, you should measure the thickness of the Diehl encoder, and precise dimensions of the stem and feet.
« Last Edit: December 11, 2021, 02:32:27 pm by Nominal Animal »
 
The following users thanked this post: con-f-use

Offline con-f-useTopic starter

  • Supporter
  • ****
  • Posts: 807
  • Country: at
Re: replacement rotary encoder
« Reply #10 on: December 11, 2021, 03:24:26 pm »
That is everything, I could have hoped for and more and so are you ;D. Thanks.

Since it's a one-off, I'm fine with a bit of epoxy, dremeling the knob a bit and botch wires, so I'll probably not bother with exact measurements and an adapter board. But it's nice to have the option now.

The body is 8mm thick, the metal part of the shaft is 11mm, the plastic shaft protrusion is 3.5mm. So the whole thing (8mm housing + 3.5mm protrusion + 11mm metal shaft sticking out) has a height of 22.5mm. The metal shaft has a diameter of 3mm and the four plastic pins on the underside are 14mm apart.
« Last Edit: December 11, 2021, 03:56:19 pm by con-f-use »
 

Offline con-f-useTopic starter

  • Supporter
  • ****
  • Posts: 807
  • Country: at
Re: replacement rotary encoder
« Reply #11 on: December 11, 2021, 03:47:20 pm »
Since you put so much effort into helping me, here's the best I can do dimension-wise. Assume everything that even looks remotely like it, are right-angles and the whole thing is symmetric and square save the three pins and where they stick out  :-DD
« Last Edit: December 11, 2021, 03:57:26 pm by con-f-use »
 

Offline Nominal Animal

  • Super Contributor
  • ***
  • Posts: 6526
  • Country: fi
    • My home page and email address
Re: replacement rotary encoder
« Reply #12 on: December 11, 2021, 05:57:06 pm »
Some dimensions were missing, like the width of the section containing the pins, but I bet it is 15.24mm.

Here is my best guess, at 1270 DPI (the image is 1270×1270, and exactly 1"×1" or 25.4mm×25.4mm).


To get the dimensionally correct SVG image, change the file name from mechanical.inkscape.svg.txt to mechanical.inkscape.svg; this forum doesn't support SVG file uploads, you see.

Anyway, this means that the Bourns ESD0D-S00-GC0006L (available from e.g TME) is a close match.  The legs are thicker, so you'd have to modify the plastic legs to fit into the existing holes.  The Bourns pins are so close to the body, that you might be able to desolder the three pins, and replace them with three U-shaped pins going above the board, and to the microwave PCB.  (I'd use solid-core wire, actually.)  And you'll need to make your own shaft to match the existing encoder; see the Bourns ES Series datasheet for exact dimensions.  If the microwave PCB does not have a hole in the right place, you might have to superglue the shaft to the encoder.
« Last Edit: December 11, 2021, 06:12:30 pm by Nominal Animal »
 
The following users thanked this post: con-f-use


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf