Rigol DS1054Z rotary encoder mod *works!*

[tv84]

Hmm, my (old) DS1054Z is still at 04.04.04.02. Unless Rigol changed the way their firmware update works, you can't go back to an older firmware version unfortunately.

Sure you can!  That is very old news -> Here

[klausES]

It seems to be a problem that the Alps encoder E12 "E" series stops with its rest position on the flank.
If you carefully outsmart the detent by hand, i.e. stop at the tip of a detent, then it would work.

I'm also not satisfied with the quality of this Alps type.
The detent is too hard, in particular for a small button diameter.
The feeling when turning is inferior, as is the feeling when pressing the button.
No comparison to the encoders that Rigol has now installed (at least to mine in the newer generation of the keyboard / encoder board).
They run smoothly with little play and a good pressure point.

Looking for a higher quality encoder. Preferably one of the two who rigol uses on timebase and volt / div with detent. 
So far I have not found the original ones.
Rigol will certainly not build it itself, but buy it somewhere in China...

[ebastler]

It seems to be a problem that the Alps encoder E12 "E" series stops with its rest position on the flank.
If you carefully outsmart the detent by hand, i.e. stop at the tip of a detent, then it would work.

I'm also not satisfied with the quality of this Alps type. [..]
Looking for a higher quality encoder. [..]

Regarding the functionality problem you experience, it remains a fact that the same type of encoder has been used successfully by many others. (Including myself; mine is a "STEC08" from Reichelt.) You could go back to the .02 firmware to verify whether a change in .03 has caused the problem; see tv84's hints on "downgrading" the firmware. Or you could try swapping the quadrant signals to see whether there was a change in the board layout which now makes the wrong signal critical in the detent position.

Regarding the different feel, I doubt that you will be able to identify the exact encoder type used by Rigol. (And if you did, you would have to be quite lucky to find a source for single quantities.) Alps and Bourns are the two major Western brands for encoders, I believe, and are widely available from the usual distributors. You can get types with a metal shaft instead of plastic, which may (or may not) be better quality. 

But don't make this an endless search for perfection; it's about using the scope after all!

[klausES]

Tried another Bourns encoder today (PEC12R-4220F-S0024).
Its latching is not on an edge (datasheet and measured in real terms).

The same behavior can be seen as with the Alps.
There are 2 steps in the menus etc. between two screenings. 
As with the Alps, speed also plays no role in this behavior.

The Bourns can be handled better because it does not have the pointy grid as the Alps, but none of this was planned.

What else did they change with the newer board besides smd?
Actually impossible illogical because these two encoders (apart from standing on the flank at the Alps, the Bourns doesn't do this),
if you think away the purely mechanical screening, do nothing else than the original encoder.

At the moment I can't think of anything else.
Why is this happening?

... except converting the original encoder to a grid.
I've already started ...

Another warning:
When unsoldering the Alps (i.e. the 2nd unsoldering process on this board at the point), four of the five soldering pads said goodbye.
And i am experienced for smd soldering ...
The board looks really nice, but it seems very sensitive here.

[TurboTom]

Could the Actel ProAsic that's sitting on the front panel board and that's probably doing the low-level interfacing of the individuals switches / encoders to the SoC, be configured differently on the new models? This means, the configuration of this "small FPGA" isn't changed during a firmware update. That would be quite a bummer since now, the "easy" encoder replacement isn't an option anymore (even more so due to the new SMD installation of the encoders).
But it should be possible to find a suitable, detented encoder with a single phase change between two detents, after all.

This encoder in the 9pulse / 18detent version may do the trick -- difficulty will be obtaining it in small quantities.

[klausES]

It does not mean that this FPGA could not be changed by firmware,
but what good does it do to a user if Rigol does not do this because they do not need to do it.

If this is the case (it seems to be so), would the original encoder with its 20 regular changes
(A and B counted together measured) cause 40 changes there, or am I wrong?
But then how does it recognize the direction of rotation?

Then it would still be possible to try an encoder with 30 detents with 15 pulses.
There are a few. Also from Alps.

Such a crap ... all over again ... 

[klausES]

Had already started building the original encoder 20 mechanical detents that takes up
so little space that it still fits in the front (it's almost done), but that would be useless. 

Now I am thinking of building a mechanism that prevents the shaft from rotating when the encoder is pressed.
That is the real reason for the whole misery ...

All a bullshit.

[klausES]

The "shaft brake when pressing the encoder" works perfectly with the original encoder.
Just installed and now tested for an hour.

One would have to make an effort (intentionally with force) to jump to an adjacent menu item when pressing.
And the feeling (turn and button) is simply better with the originals than with the Alps (sorry, I like the Alps otherwise) and Bourns.

If I had known all of this beforehand ... it can be that easy. after that you're always smarter ...

You didn't have to remove / unsolder the encoder,
With the last brake variant now running, you would not even have to open the housing. 
It stays that way now.

[Adrian_Arg.]

so these encoders don't work either "Alpes ec12e2424407"

[klausES]

Since this is technically the same as the Alps I tried, probably not.

Remember, this only affects the new DS1000Z with the new control PCB (with the SMD encoder) such as with mine.
The older versions seem to be running (that would be the Alps type that many used).

[ebastler]

The "shaft brake when pressing the encoder" works perfectly with the original encoder.

While I am happy with the detented encoder in my older scope, this makes me curious. I understand you have made a mechanical mod?
Could you share a photo or a design drawing please?

[Fungus]

Now I am thinking of building a mechanism that prevents the shaft from rotating when the encoder is pressed.
That is the real reason for the whole misery ...

Add a push button at the side of the encoder, avoid the problem entirely.

Can the PCB be hacked to exchange the "clear" button with the encoder's push button? That would be a huge improvement in the UI and still leave the 'scope pretty on the outside.

[klausES]

...I understand you have made a mechanical mod?
Could you share a photo or a design drawing please?

Yes, it is mechanical.
The principle is very simple, it depends on the properties of the material and its exact mass.
It is complicated to explain with words alone. I'll make a drawing the days.

...Can the PCB be hacked to exchange the "clear" button with the encoder's push button?...

As a mechanical micro button (short stroke) or like the original combs in the layout with conductive rubber
(the free surfaces are coated with solid copper underneath the soldering resist, you could mill an area on it).
That would be tedious but possible and you would have to work very precisely for the timing to the front panel.

On the other hand, one would have to consider whether changing the grip is really worth striving for with every operation ?!?

As simple as it is (almost idiotically simple) and to praise yourself (stinks as they say here  ),
I played with it for quite a while yesterday because it is really great to use.

[tooki]

I am trying to find out whether the encoder ordered does not exactly match the one mentioned here.

The ALPS EC12E2424407 often mentioned here differs from my ordered, if at all, only a little bit from the mechanical properties
and dimensions but it should be the same from the electrical side (I hope ...) 
It is an Alps STEC12E08 rotary encoder, 24 pulses / 24 detents, vertical.
Find only a very bad datasheet of this STEC12E08 ...

And I already had doubts because the STEC12E08 is funny to buy only from a single dealer across the country
and this type does not even appear in the manufacturer list on the Alps website ... 
Not even Mouser or Farnell think of this type.

Reichelt is using an extremely old part number.

It looks like the part was originally STEC12E08. Sometime in the 2000s, Alps renumbered its encoders, and it became the EC12E2424407. Later, it was discontinued, which is why it's not found on any modern Alps datasheets for the EC12 series.

I've attached below an older EC12 series datasheet that includes the EC12E2424407, as well as a 20 year old Alps catalog that lists the STEC12E08.

I am trying to find out whether the encoder ordered does not exactly match the one mentioned here.

The ALPS EC12E2424407 often mentioned here differs from my ordered, if at all, only a little bit from the mechanical properties
and dimensions but it should be the same from the electrical side (I hope ...) 
It is an Alps STEC12E08 rotary encoder, 24 pulses / 24 detents, vertical.
Find only a very bad datasheet of this STEC12E08 ...

At the moment it is not quite sure whether I might have caught a defective encoder when ordering. 

As it is now, I cannot leave it. 50% of menu items can only be selected between two stop points ... 

No, it's not broken, it's just the wrong part for your scope.

But why do I have the problem that two menu items are jumped per click? 
Unfortunately I had not tried the encoder with the older firmware.
Since the update to 04.04.04.03 I have not come back to the previous 00.04.04.04.02 to test the encoder.

For whatever reason, it looks like Alps used to make lots of 1:1 detent-to-pulse ratio encoders, but little by little moved entirely to 2:1.

It appears that Rigol changed from 1:1 THT encoders in older boards to modern, probably 2:1 SMD encoders in newer boards. If you put a 1:1 encoder in a circuit expecting a 2:1, you get twice as many pulses per detent, causing the software to move twice as fast as you wanted.



For the original THT board: I'm a bit puzzled by the fact everyone is using EC12 series encoders in a board that clearly uses EC11 series originally. For a given shaft length, the EC12 is 0.5mm taller than the EC11, which explains why the knobs sit proud when using an EC12.

---- Edit ---
From what I can tell, for everyone with a board using the 1:1 THT encoders, the optimal Alps part number is STEC11B13 (EC11, 20mm shaft, with button, 20 detents, 20 pulses; AKA EC11B20244), and the equivalent Bourns (which is an active part) is PEC11L-4120F-S0020 (PEC11L, 20mm shaft, with switch, 20 detents, 20 pulses).

For those with a 2:1 SMD encoder board, it'd be one of the following Alps models:
EC11J0924411
EC11J1524413 <-- probably this one
EC11J0925403
EC11J1525402

Those differ in number of detents/pulses (# of pulses in red) and button travel (in green, 4=0.5mm, 5=1.5mm travel). My guess is that 0.5mm travel is probably the right one, but I don't know whether 9 or 15 pulses is better. AFAIK, either will work, just at different speeds obviously. Given that Bourns only makes this in a 30-detent/15-pulse version, my hunch is that the second one (in bold) is likely what Rigol is using, insofar as the Chinese clones will also probably be of the most widely available type. The equivalent Bourns part is PEC11S-9220F-S0015.

As detailed in a reply below, it turns out that they measure the shaft length differently on the plastic-shaft encoders vs. metal-shaft ones.  The plastic ones include the body, the metal ones do not, adding 5mm length. Bourns carries both the THT and SMD metal ones in the needed 15mm height, but the SMD one doesn't seem to be in stock anywhere. Alps doesn't even have part numbers for them.

So here's the updated info.

From what I can tell, for everyone with a board using the 1:1 THT encoders, the optimal encoder is the Bourns PEC11L-4115F-S0020 (PEC11L, 15mm shaft, with switch, 20 detents, 20 pulses).

For those with a 2:1 SMD encoder board, it'd be Bourns part PEC11S-9215F-S0015, which nobody seems to have in stock.

--- /Edit ---

I've attached the current Alps and Bourns datasheets, too.

[tooki]

Last datasheet, since it pushes the total attachment size over the limit.

[tooki]

Tried another Bourns encoder today (PEC12R-4220F-S0024).
Its latching is not on an edge (datasheet and measured in real terms).

The same behavior can be seen as with the Alps.
There are 2 steps in the menus etc. between two screenings. 
As with the Alps, speed also plays no role in this behavior.

That, too, is a 1:1 ratio encoder. I think it's highly likely you need a 2:1 encoder. Besides, isn't that a through-hole encoder? Why would you buy that instead of an SMD one? Try one I list in the post above.

The Bourns can be handled better because it does not have the pointy grid as the Alps, but none of this was planned.

What do you mean by a grid? And what do you mean by a pointy grid??

What else did they change with the newer board besides smd?
Actually impossible illogical because these two encoders (apart from standing on the flank at the Alps, the Bourns doesn't do this),
if you think away the purely mechanical screening, do nothing else than the original encoder.

What?

... except converting the original encoder to a grid.

Again, grid??

Another warning:
When unsoldering the Alps (i.e. the 2nd unsoldering process on this board at the point), four of the five soldering pads said goodbye.
And i am experienced for smd soldering ...
The board looks really nice, but it seems very sensitive here.

That sucks.

I do find SMD pads to be much, much easier to lift than THT ones, generally speaking.

[klausES]

I still wanted to try a 30 detend 15 impulse encoder,
luckily I don't need that anymore.

The Bourns can be handled better because it does not have the pointy grid as the Alps, but none of this was planned.

What do you mean by a grid? And what do you mean by a pointy grid??...

I meant detent and whether the detent runs fine and soft
or mechanically hard and difficult to tackle at the tips.

Thanks for your detailed answers.

[Adrian_Arg.]

Since this is technically the same as the Alps I tried, probably not.

Remember, this only affects the new DS1000Z with the new control PCB (with the SMD encoder) such as with mine.
The older versions seem to be running (that would be the Alps type that many used).

What version of PCB does your rigol ds1000z have?

[klausES]

You mean specifically the encoder board ?

Mine probably has the latest with the encoders as SMD.
"Keyboard A V01.02"

Mainboard is "V01.04"

[ebastler]

One thing that baffles me in this context: As Rigol did change the PCB design and the BOM of the front panel board anyway -- why didn't they introduce a detented encoder for the menu selection knob?! They must have someone listening to customer feedback, including the discussions and advice on this forum? 

[tooki]

  If only... how many years did it take for them to fix the “pluses” typo?

It was clearly a conscious (if stupid) decision to use a non-detented encoder to begin with, not a bug, so I don’t see why they’d change it. Rigol definitely does not have a reputation of being responsive to customers.

[klausES]

The right choice would be to take one like the two new ones at the bottom left.

Certainly the same people as before thought it was right to take one without detent at this point,
(or you didn't want to annoy any of the thousands of older buyers now) ?!? 

[klausES]

There are two options (outside, without opening the housing).
The more effective of the two is the silicone ring on the inside of the shaft.
See image.

Version inside:
It must have an exact mass, then it is put under "pressure" early on, at the very beginning of the pushing, and its deformation
(since it cannot leave the V-shaped space) increases the friction between the shaft and the thickening of the shaft above
and the fixed neck of the encoder to increase the friction from turning while pressing.

Version outside:
(There it is easier to try different diameter and tensions to test it).
The ring may only have a slight tension without pressing (it should dampen the rotation "then" if only slightly)
and when pressed it is forced to increase its overall diameter.
This more tension also creates more friction and the shaft can be turned brake while pressing.

Everything depends on the right dimensions and materials of these sealing rings.
If these are selected correctly, braking works perfectly without the normal turning and pressing of the button being impaired.

Inside it has to be soft silicone (easily deformable), on the outside there is also a rubber or better a Viton ring.

I hope the translator doesn't build too much nonsense here... 

PS.
Silicone Ring inside = yellow
Rubber or Viton ring outside = red

[klausES]

This would have been my conversion of the original encoder (20 pulses) to 20 detent (pictures still in the early stages).
Unfortunately, it is useless if the software of the new smd encoder board now evaluates the encoder "twice" per step.

So work for free ... 

Before I throw in the scrap unfinished, I wanted to show you.

[tooki]

From what I can tell, for everyone with a board using the 1:1 THT encoders, the optimal Alps part number is STEC11B13 (EC11, 20mm shaft, with button, 20 detents, 20 pulses; AKA EC11B20244), and the equivalent Bourns (which is an active part) is PEC11L-4120F-S0020 (PEC11L, 20mm shaft, with switch, 20 detents, 20 pulses).

For those with a 2:1 SMD encoder board, it'd be one of the following Alps models:
EC11J0924411
EC11J1524413 <-- probably this one
EC11J0925403
EC11J1525402

Those differ in number of detents/pulses (# of pulses in red) and button travel (in green, 4=0.5mm, 5=1.5mm travel). My guess is that 0.5mm travel is probably the right one, but I don't know whether 9 or 15 pulses is better. AFAIK, either will work, just at different speeds obviously. Given that Bourns only makes this in a 30-detent/15-pulse version, my hunch is that the second one (in bold) is likely what Rigol is using, insofar as the Chinese clones will also probably be of the most widely available type. The equivalent Bourns part is PEC11S-9220F-S0015.

So I ordered a PEC11L-4120F-S0020 for my scope, and when it arrived, I was surprised to discover it's 5mm longer than the original. Turns out that on the plastic-shaft encoders, they include the encoder body in the shaft length, while on the metal-shaft ones, they don't!    So when I went by the 20mm shaft spec on the 12mm plastic encoder people recommended earlier, it actually required a 15mm shaft in metal.

I now have a PEC11L-4115F-S0020 on the way. But for the SMD models, I don't know what to suggest. Nobody seems to stock the 15mm shaft SMD version of the Bourns (PEC11S-9215F-S0015), and Alps doesn't even have a part number for a 15mm SMD version!

I'll update the post accordingly.