Problem with rigol scope knob

[Bored@Work]

right so the bouncing issue is down to failure of the mechanics of the switch/knob

No, all mechanical switches bounce. And bouncing behavior can change over time, which is not a failure per-se. If there is no proper debouncing it is a design failure - typical mechanical switch behavior hasn't been taken into account.

[Feanor]

I've seen this too, brand new switches and knobs are great you could clock your PC off the nice clean edges that they produce, a few months later and the wave form they produce looks like a lunar landscape.

You could probably check this with your oscilloscope! I am laughing just picturing the scene now. Debugging a faulty scope with said faulty scope.

[kderh]

OK after my horizontal scale knob became totally random I decided to open the scope and clean the encoder, and now it works much better than before as it skips very rarely now.

If someone wants to try this here are some tips:

-Heat up the warranty seal with a regular hair blower, then you can peel it off without damaging it
-There are 2 screws under the carrying handle
-Do not touch the 2 screws next to the power socket for now, remove them later.
-To get the power switch out of the way, simply pull it out with pliers.
-Wiggle the back case around until it comes off
-Remove the other shielding elements (back, top)
-Remove all the front knob caps ( just pull them off)
-Remove the power supply as there is an additional screw behind it that holds the front panel
-Disconnect the input board by lifting the white flex cable off the main board
-Remove the other screws holding the front panel, then you can take it off
-Remove the input board (careful with the flex cable)
-Desolder the bad encoder and lift the 4 hinges on all sides to open it (or just replace it with a new one)
-Clean all contacts inside (I used a Q-tip)
-Reassemble the encoder, solder it in and reassemble the whole unit

This worked for me, hope it stays this way for some time... Mechanical encoders suck...

[saturation]

Great job, kderh, all you need are pictures and this is a perfect hack/fix.

OK after my horizontal scale knob became totally random I decided to open the scope and clean the encoder, and now it works much better than before as it skips very rarely now.

If someone wants to try this here are some tips:

-Heat up the warranty seal with a regular hair blower, then you can peel it off without damaging it
-There are 2 screws under the carrying handle
-Do not touch the 2 screws next to the power socket for now, remove them later.
-To get the power switch out of the way, simply pull it out with pliers.
-Wiggle the back case around until it comes off
-Remove the other shielding elements (back, top)
-Remove all the front knob caps ( just pull them off)
-Remove the power supply as there is an additional screw behind it that holds the front panel
-Disconnect the input board by lifting the white flex cable off the main board
-Remove the other screws holding the front panel, then you can take it off
-Remove the input board (careful with the flex cable)
-Desolder the bad encoder and lift the 4 hinges on all sides to open it (or just replace it with a new one)
-Clean all contacts inside (I used a Q-tip)
-Reassemble the encoder, solder it in and reassemble the whole unit

This worked for me, hope it stays this way for some time... Mechanical encoders suck...

[wd5gnr]

Well I got mine apart... encoder removed (vacuum really helps), cleaned, and resoldered. Now I'm about to put it back together. Wish me luck. Thanks for the steps ;-)


Bah! I got it back together enough to try it and it is no better (then again no worse). When I took the encoder apart, there was a membrane type switch in a green plastic frame and a white plastic ring on top with some position pins. The black part had a spinning disk of sorts. I used contact cleaner on both parts and scrubbed with a Q-tip. It didn't look like that I expected, but what do I know.

I'm debating if I want to take it apart again -- its a pain -- and try again. It appears to be an ALPS encoder -- does anyone have a source? I'm guessing its from this series: http://www.alps.com/WebObjects/catalog.woa/E/PDF/Switch/Encoder/EC11/EC11.PDF but would need a place that sells reasonable quantities.

Measurement wise the body is 12.5mm wide and 14mm tall. Shaft is 6mm in diameter. 14.4mm from the tip to the bottom of the body and 5mm shaft length.

Digikey has:
http://www.bitechnologies.com/pdfs/en12.pdf  - 17.5mm reach bit 5,, shaft length. Maybe trim the shaft a bit. There were 3 "12mm" encoders with switches so I think I'll order all 3 and see how they might fit. The are only a couple of bucks each.

P12336-ND
987-1194-ND
987-1195-ND

1 of each is less than $5 total.

Sigh....

[wd5gnr]

Well it turns out I didn't have it completely apart last night which is why it looked like something was missing. Unfortunately, it still didn't clean so I'm going to try ONE MORE TIME.

The attached picture is the encoder apart (enhanced to show detail so a little grainy). The green part is the switch. What I had last night was that off but the black body of the encoder remained attached! Duh. You have to get the board lock legs spread apart to get the black part off.

I Q-tipped the encoder wheel (16 pads so I guess its 32 pulses per revolution?) an sprung the contacts a little with a screwdriver.

A few notes about taking the whole thing apart:

1) The white cable is in a ZIF socket. You have to release the two black tabs at either side of it and then it just comes out with no force. If you are pulling the cable you don't have it unlocked. It should just "float" out with no force.

2) The reason I mention this is that getting it back in is HARD. It is very hard for me at least to keep the ZIF lock open on both sides. The best method is to flip the scope upside down or at least tilting down in the right way so that the lock wants to fall down. Then once you get the cable in (should be none of the exposed metal showing) you can flip it back and push the locks down with a small screwdriver.

3) As far as I can tell the power supply bundle of wires has to come off the main board. I don't think the connector on the power supply comes off.

4) If you see an LED "broken" on the front panel board don't panic. Its for the logic analyzer and you don't have it (well, if you do, then panic). Looks like they make all the boards and just pop the unwanted LED off with a pair of dykes or something.

I'm going to try again. I'm getting pretty good at putting it back together.

Correction. I think it must be 24 pulses per rev. There are 3 contacts. So for 1/3 of the rotation any given contact is the common contact, the A contact, and the B contact. So per rev terminal A has 3 contacts brushing 8 pads for 24 pulses per rev. The good thing about this is I have figured out all the stuff I never knew about mechanical encoders.

[Mechatrommer]

be carefull not to damage the parts, as the problem might lies outside it, not in the encoder. the device seems easy to clean/repair.. clean the contact pad and push the 3 contacts lever a bit outward, but not too much to avoid misalignment. you must have a way to verify that the signal is something like the pic below (from your EC11.pdf) (ie 3 phase signal). if you have verified the signal is ok, then i think the problem is not the encoder, it must be something else. Happy hacking

[wd5gnr]

Yeah I thought the same thing. But I still think it is the encoder based on occam's razor and the fact that it got worse over time until now it is unusable. The last encoder I mentioned from Digikey looks like it should fit ok and the only real difference is it has detents. If it doesn't work than I will have to look at what's reading it. I did think of swapping two of the encoders but decided it wasn't worth the trouble since I have a new on en route. But if the swap caused the problem to stay it would have to be the circuit not the encoder.

[Mechatrommer]

if you cannot verify the 3 phase signal, then easier for connectivity test for each pin while rotating the knob.

[wd5gnr]

Well it turns out it was both.

When I started out the encoder was "flaky" but you could coax it. Then eventually it got to where it would just "jitter" and just flop around the current position when you turned the knob.

I replace the encoder with Digikey 987-1195-ND. The fit was exact and since I'd had the other one open I knew it had the right number of pulses per rev. However, it has detents so you get a little click as you turn it. Otherwise perfect.

Replacing the encoder made it "flaky" again. You could get it to move but mostly only in one direction and sometimes randomly the other way. I opened it back up and there are two diodes behind each encoder. One of them read open both ways. I don't keep SMD diodes around (I have started keeping a bunch of SMD resistors around) but I managed to remove one off something I am not using. It was a Schottky diode (the one in there was something like a 1N914 I'd guess from measurements of the other ones) but judging what the circuit was doing I didn't think the forward drop was going to matter. Now it works like brand new!

Rigol has offered to send me a new UI board under warranty which is SUPER of them. Haven't got it yet though. I did buy from a US distributor, by the way.

I posted pictures of the tear down.
https://www.eevblog.com/forum/index.php?topic=2092.0

I don't remember if I mentioned in the notes but the only thing holding the back shield on is the jack nuts for the DB9. Be very careful as the metal work is not so good. I have a top shield screw that will no longer tighten, and one of the bottom case screws chewed some plastic and won't go in well either. But the case is a tight fit so you really can't tell. In fact, I've been leaving out the screws under the handles and can't tell the difference. I suppose if you had it bouncing in the back of a truck, maybe. But mine sits quietly on a bench.

[Zyvek]

Rigol has offered to send me a new UI board under warranty which is SUPER of them. Haven't got it yet though. I did buy from a US distributor, by the way.

You are quite the master, going beyond breaking the warranty seal, and still getting warranty service!  Awesome, plus kudos to Rigol for the service as well!
 

[Mechatrommer]

Well it turns out it was both.

both what?

[wd5gnr]

Both the encoder and something on the board. Originally the motion was jerky in one direction and then it became almost nonexistent. Replacing the encoder got me back to jerky and replacing the diode made it work 100%. My guess is the diode was bad from the start. Then the constant pulling and twisting on the encoder to try to get it to unstick took the encoder out.

[Mechatrommer]

so i guess it was the diode only, not the encoder. from what i understand of your post #35.

[wd5gnr]

No, I think it was both. Let me put it in time order.

1) New scope. I noticed that sometimes the trigger level was "flaky"

2) Soon I can only adjust the level up noisily. So you can reset to the baseline (push the button) and move up, so it was usable. At this point, I think the diode had completely failed.

3) Assuming it was the encoder I am twisting the encoder, trying to pull it out while turning, push it in, etc. Contact cleaner, compressed air, etc. Nothing helps.

4) Over a few months the problem becomes worse. Now twisting the encoder knob just causes the trigger level line to jump up and down maybe one step in either direction. The only way to get useful position is to use the 50% key. At this point, I think the encoder has failed.

5) I take the thing apart and clean and respring the encoder several times. No change.

6) I replace the encoder. Now I am back to the behavior in #2, so the new encoder did change the behavior. It just still didn't work.

7) Replacing the diode under the encoder restores normal function.

So I think the encoder was OK to start with, but eventually failed prior to me opening the thing up.

[Mechatrommer]

ok, thanx for the report. will be helpful if some of us got this problem.

[oliver602]

Searched google about this and got this forum!

I have a month old DS1052E and have noticed its horizontal scale knob is jittery when trying to zoom in or out fairly quick. All the rest are fine. Still under warranty so pretty reluctant to open it.

[crypton]

Ehh, apologies for ressurrecting a dinosaur (maybe it can be changed to a new topic IF necessary but I would keep it here under one place / same issue). 

I've found an unit that wasn't in use due to those flacky encoders.

Very well, followed the instructions and I have now used these encoders:

858-EN11-HSM1AF15 - for channel timebase - works flawlessly.

For the other 4 encoders there's been suggested:
858-EN12-HS22AF20
652-PEC11R4315FS0012

Some claim that these "work" but for both of them I see them skip many values. For the 'menu' encoder they skips about 3 values. For instance intensity increases/decreases in increments of 3.
Don't remember this happening on the original encoders.

Position is also incerementing in bigger steps. When I ever so slightly move the encoder I could get smaller increments but with dentents the steps seem quite uneven - not good.

So I've spent quite some bucks(euros ) on testing different encoders for the menu and hor., ver. positions (ch and trig) but these 4 havent worked as expected. 

Curiously enough I haven't seen any reference to what encoder one should use for that to work on Rigol DS1052E. Maybe someone can point me (and for future reference) to the 'correct' ones ?

[Chicopy]

Have anybody successfully replaced the Horizontal position knob (or any of the 4 that look the same) on DS1052E ?
I saw several discussions about the encoders used, but I still miss clear answer what type is 100% perfect, either original type or a one that works exactly as original.
Some types mentioned became obsolete, some unable to find..

[MF-jockey]

I was able to solve the problem with additional capacitors on my OWON SDS7102. Should also help here, as the technology in the Rigol is often the same.
You can find the solution to the problem here:
https://www.eevblog.com/forum/testgear/owon-sds7102v-problem-with-encoders/msg3479200/#msg3479200