Upgraded SDG1000X Plus AWG's coming

[Martin72]

What exactly is wrong with it?
I usually set the frequency I need via the keypad, for example.

[tautech]

What exactly is wrong with it?
I usually set the frequency I need via the keypad, for example.

Likewise but not always.

I commonly use < and > to select the exact digit to change then use the encoder. < Quick and accurate.

[Martin72]

This makes me so curious that I'm going to use the SDG1062X at work tomorrow.
Maybe I'll make a clip of it.
I don't remember any problems with the encoder wheel.
Maybe I've forgotten about it, we'll see.

[tautech]

This makes me so curious that I'm going to use the SDG1062X at work tomorrow.
Maybe I'll make a clip of it.
I don't remember any problems with the encoder wheel.

With considered and deliberate use the encoder works fine.

[pope]

What exactly is wrong with it?
I usually set the frequency I need via the keypad, for example.

Some times (quite often actually) I need to manually sweep through a frequency range.

[Martin72]

I also use it to find a resonance point, for example.
But in my case we're talking about maybe +/-50Hz and slow "start-up", which has never been a problem with the SDG1062X.
But there are supposed to be people who turn the encoder as if it wasn't smart and then complain that not every turn is acknowledged immediately.
Nobody can tell me that they are doing a serious task at that very moment.
I can't take that seriously.

[pope]

I can't take that seriously.

You don't have to. I'm just sharing my personal experience. It's not a secret that the encoder on the SDG1000X sucks 

[ebastler]

But there are supposed to be people who turn the encoder as if it wasn't smart and then complain that not every turn is acknowledged immediately.

By "smart" you mean that the encoder response accelerates when you turn it faster? I only have an SDG2000X -- on that one, there is no dynamic acceleration, and I think I like it that way: One encoder detent equals one audible click equals one step up or down on the numerical digit. Is the SDG1000X encoder implemented differently?

I do notice that the SDG2000X encoder starts to miss steps when I turn it a bit faster. Which is not an entirely irrelevant scenario in my use: Say I expect the need to "tune up" slightly, e.g. to find a resonance, and then find that I am still further off than I thought. Rather than changing the cursor to a more significant digit, I just give the encoder a turn or two, which I can do blindly (while looking at the scope). In that case the encoder response becomes spotty, quite noticeable from the irregular audible clicks.

It seems that Siglent took a very conservative approach with the encoder debouncing. Still better than too little debouncing, with the encoder making extra steps or moving backwards! But based on the experience with encoders in other products or my own designs, I think that they could have found a better balance.

[joeqsmith]

... Still better than too little debouncing, with the encoder making extra steps or moving backwards! ...

At least with the one we bought for a trial, it will randomly change directions.

[ebastler]

... Still better than too little debouncing, with the encoder making extra steps or moving backwards! ...

At least with the one we bought for a trial, it will randomly change directions.

Ouch... So they should probably copy the conservative debouncing approach from the SDG2000X.

The SDG2000X had issues with bouncing buttons in its earlier firmware versions, which Siglent fixed with an update a few years back. I don't recall any changes to the encoder behaviour, but maybe that was improved at the same time?

[joeqsmith]

Guessing that it can't be corrected with firmware or they would have addressed it.   Personally, I really like having an encoder for a UI and will frequently put it to use.  I'm not a fan of the detent but I can't overlook the fact the settings are not smooth.   

In this case, the product didn't cost much.  I have asked about other higher cost Siglent products having this potential problem but this thread reflects the typical response I receive. 

[Martin72]

By "smart" you mean that the encoder response accelerates when you turn it faster?

Deepl translation error...

It seems that Siglent took a very conservative approach with the encoder debouncing. Still better than too little debouncing, with the encoder making extra steps or moving backwards! But based on the experience with encoders in other products or my own designs, I think that they could have found a better balance.

Yes, "debouncing" is a matter in itself, how far you want to go.
If the "dead time" until a pending signal is regarded as a "real" one is too short, you get the notorious jumps.
If you choose it too long, the user has the feeling that he can turn "forever" until something happens.
Although I personally have no problem with the latter.
Now you can't expect high-quality encoders with the cheap SDG1000X and SDG2000X classes, although that would be an interesting separate topic in itself, how high-quality is it, in terms of being as bounce-free as possible.

[joeqsmith]

... Now you can't expect high-quality encoders with the cheap SDG1000X and SDG2000X classes, ...

Being an integral part of the UI, I certainly expect them to work, regardless of the cost.   I has asked if the higher end products were any better or not but so far, I have not got a clear answer.     

I picked up a cheap hand held multi-band radio a few months back (Tecsun PL-990).   It uses encoders for tuning.   I was expecting it to have problems out of the box but it has worked flawless.  I can turn the tuning as fast or slow as I want.   

[TurboTom]

...
Now you can't expect high-quality encoders with the cheap SDG1000X and SDG2000X classes, although that would be an interesting separate topic in itself, how high-quality is it, in terms of being as bounce-free as possible.

As if the encoder of the SDG6000X is any better...   All these small AWGs (form-factor wise) share the same "mechanical" hardware and a very similar, almost identical user interface. And IMO, this leaves quite some margin for improvement.

[Martin72]

As if the encoder of the SDG6000X is any better...

I didn't have the 6000s in mind at the time, but yes, they are also part of it because the "X" at the end indicates the cost effective series, be it for generators, oscilloscopes, analyzers...etc.

[slavoy]

I own a 1032X and now 2042X. The encoders works exactly the same. They do not change in direction, but are... sluggish. If you pass a certain rotation speed (which is low in this case) there is no more response improvement. Another issue - instant quick rotation of the knob gives almost no response. It will sometimes jump one value when I go fast half a turn 
There is nothing wrong with the encoder itself, its the software handling issue.