Alps rotary encoder gone crazy

[hgg]

Hi,

I have an Aiwa CX-NS74HR stereo that uses an ALPS rotary volume control and started doing
strange things.  When you turn the volume up, it refuses to do it or goes up just a level or
go down most of the times...  It generally wants to turn the volume down instead of up... 

So, I was wondering if the problem is more likely to be with the encoder or the circuit.
Is there a way to test the encoder out of circuit with a multimeter?
If it is the encoder how can I find a compatible replacement?
(Only the word ALPS is written on it...)

Any help is appreciated!
George.

[Skimask]

Pick one...

http://www.digikey.com/scripts/dksearch/dksus.dll?FV=fff4001e%2Cfff80033%2C3400006%2C408003c%2Ce000003&k=rotary+encoder&mnonly=0&newproducts=0&ColumnSort=0&page=1&stock=1&pbfree=0&rohs=0&quantity=&ptm=0&fid=0&pageSize=500

[SeanB]

Most likely to be the encoder having dirty wipers inside. You can open them ( carefully) and clean them, then relubricate them and reassemble. This is a known failure mode of them, when the one output or the other gets noisy from wear.

RS. Farnell or Element 14 probably have an equivalent, you just need the same pinout and the right mounting and shaft size.

[hgg]

Thanks!
I will try to open it first.

How can I choose from the list when I don't know the number of pulses?
Will it matter?

[FrankBuss]

I've used the ALPS encoder once. To test it, you need two pullup resistors, see the schematic on my project page:

http://www.frank-buss.de/attiny/index.html

I would then use an 2 channel oscilloscope to check the quadrature signal.

[SeanB]

ALPS EC11B vertical pot, 15 pulses per rotation. ALPS part  EC11B15244A7 or RS  623-4079 and costs about 5 Euro. The RS page has full data on it with dimensions.

[hgg]

Thank you for the suggestions!
FrankBuss I don't have an oscilloscope, yet.  :-)   The Siglent is in the mail right now, travelling  :-)

SeanB, I opened the encoder and its a 24 pulses per rotation.  You can actually see the small contacts.


I did not know how it worked but now its obvious.  All the contacts were very dirty and oxidized!
I cleaned the rotary base and the three contacts.  It has three copper contacts that are forced on
the rotary plate with just the tension of copper itself!  Very bad design.  This thing is destined to fail!

Anyway, I cleaned it, re-tensioned all the contacts and made sure that they had the right distance
from each other.  (I cut the width on one of them shorter because they were almost touching!)





I reassembled the parts together and made a small video to show some small test using two
multi meters in continuity mode.  (Sorry for the very annoying sounds...  )



I think that this is an easy way to quickly test a rotary encoder, although it does not tell anything
about the 90 degree phase.  Maybe if you can analyse the sound of the beeps using a steady
rotation speed, you can find out the phase between them.  I am not sure.

I will try now to solder it back on the board and reassemble the stereo.

[SeanB]

I am going to guess it is a Phillips branded stereo.............. They are rather well known to have this issue.

[ptricks]

I had a stereo once with a bad encoder. Every time you went to turn up the volume it would work but do it 20 points a turn, from near silence to full blast !

[hgg]

Its a very old Aiwa CX-NS74HR that my uncle gave me to repair...

That was it!  Indeed the fix worked like a charm!  I don't know though for how long.
I am sure it will fail again because of its terrible design.

    :-)

I will be buying a new one.  SeanB can you help me find one?  Its a 6mm, 24 pulses/sec.
Its something like this : http://media.digikey.com/photos/Panasonic%20Photos/EVE-GA1F2012B,%20EVE-GA1F2024B.jpg
but its slotted as well.  The height of the black dial is about 1.7cm from the square base.


Thank you very much for the help!   Great forum!

The question though remains to be answered by the forum experts.
Do you know if I would have been able to check the integrity of the encoder by analysing
the multimeter beeps to see if the 90 degree phase is correct?  No oscilloscopes...   

[SeanB]

Yes using the multimeter would have worked, provided the continuity test can show noisy contacts.

The RS website has a few encoders, just search the RS part number, and look up the similar items there, all are quite cheap. RS seems to now be charging a delivery fee ( at least here unless you are over a certain amount) so order some other stuff at the same time to amortise the delivery cost or to get over the floor limit.

[hgg]

I cannot find one with the top groove, but I might make it myself.

Maybe this is compatible:
http://gr.rsdelivers.com/product/alps/ec12e24204a7/17-5mm-vertical-encoder-24-pulses-24-det/7295791.aspx

I will order some other stuff as well.  Thanks SeanB!

[free_electron]

This is a common ailment for these alps encoders (especially that model...) the material used for the little blocks is too soft and over time you scrape off minute flakes of metal.... This creates a leakage path ( ive seen some that were rightout shorted. )

Do as you did ( if you don't have new ones) open it up , wash it down in an alcohol solution, dry , apply a little bit of nonconductive lubricant and close it again. Will work happily for years after ( and then you have to do it again. )

[hgg]

I didn't put the lubricant...  Did not have a dielectric grease.
What about silicon grease?

[SeanB]

A thin film of car oil will also work, do not use silicone grease, it goes hard with time.

[hgg]

Ok, thanks.

[Zapro]

Good job on the repair. I do the same repair from time to time on these encoders.

Just a pity that you removed a contact spring. There _IS_ a reason for the two contact springs - if one jumps for a split second, the other one will take over, resulting in a more stable readout.

They all oxidise over time, and a clean and re-lubrication with Electrolube or Super-Lube will make it as good as new. I always fill them to the brim with grease, so no air can get to the contacts. This makes the control a bit "harder" but haven't been a problem in the stuff that I repaired.

// Per.

[hgg]

Hi, you are right and I knew that removing it was not a good idea, but they were very close together,
almost touching and I think that they would have caused a problem in the future.

I did not though put any lubricant inside and it looks like that this is a must if you want to keep the
oxidation to a minimum.  I think you should use a very low viscosity lube because the friction drag might
bend and short the little contacts.  (Especially for the very low quality ALPS encoder that I have)

I have ordered two from RS and if they are not compatible, I will clean and lubricate the old one again.
Thank you!.

[1unc]

Hi.. I tried to watch your 2 VIDEOS, and get message "This is PRIVATE" and I can watch them.. why? What do I need to do to see videos? You can reply to me directly, at unc80@yahoo.com.  Greg

[FrankBuss]

Youtube did a really silly change recently. All unlisted videos before 2017 are set to private, if you don't fill out a form saying "I don't want Youtube to do this". Might be the reason for the 2 videos. I guess many videos are gone now, if the creators don't care.
https://support.google.com/youtube/answer/9230970

[1unc]

Hey George.. thanks for posting this topic.. you helped me repair my Rotel RSX-1057 receiver with the same exact issue!  When I'd turn volume knob up or down, the display #s would go crazy..up or down at random and the volume wouldn't change like it should. I've collected, restored and sold/kept vintage audio for 30+ yrs... and now retired from the corporate world.. that's all I do. One of my best regular clients just gave me the Rotel, and said.. "a tech told me it has a bad rotary encoder and a replacement part doesn't exist".. Well, I never even heard of a "rotary encoder" or understood what it did. So I started some web searches and found this/your post about the same issue.. and after reading the thread, thought..I have nothing to lose, and since I couldn't find any parts online... after removing a million screws to get the faceplate apart and get to the encoder.. I CAREFULLY took it apart...used some alcohol to clean the gooey/thick old grease that was gumming up the rotary part that moves over the super fragile contacts...I applied just a little PhonoLube grease , only to the shaft area but NOT to the contacts... reassembled the encoder.. got faceplate and everything back together, crossed my fingers, prayed, and VOILA !!!! The volume worked like it should... numbers on display slowly went up and down as I turned dial and output sounded great.  Thanks for creating this thread/post... you're a LIFE-SAVER man !!!  Greg in Greensboro, NC

[1unc]

Thanks Frank.. so you mean the original creator of the youtube video is the only person, who must do the "opt out" deal, to make video viewable by anyone else? Crap.. I really wanted to watch it, but by reading this whole thread.. I gambled and cleaned the rotary encoder in my Rotel RSX-1057, and now it works great! See my comments/post below...

[Doctorandus_P]

Sloppy design is a common failure mode of such encoders.

In lot's of example circuits you see a (ceramic) capacitor directly over the switch contacts to clean up the bouncing of the contacts a bit. This capacitor is charged "slowly" by the pull-up resistor when the switch opened, but completely shorted when the switch closes.

Even small ceramic capacitors can deliver very high peak currents. 10A is no exception, while such encoders are probably specified for a max current of somewhere between 25mA and 200mA. And such currents can easily damage the switch contacts over time.
If your circuit has such capacitors, you should also add series resistors to them to limit the maximum current to a safe value.