DLO-138 OScope

What are you developing?
Post Reply
willfly1
Posts: 17
Joined: Wed Jan 04, 2017 5:13 pm
Location: GTA, Canada

DLO-138 OScope

Post by willfly1 » Sun Feb 26, 2017 5:38 pm

DSO-138 Oscope firmware

DLO-138 is alternate firmware for DSO-138 and adds second analog channel, two logic channels, more sample depth and serial port data output.

Code: https://github.com/ardyesp/DLO-138
pic3.png
pic3.png (153.54 KiB) Viewed 2847 times
pic2.png
pic2.png (164.74 KiB) Viewed 2847 times
pic1.png
pic1.png (173.51 KiB) Viewed 2847 times

User avatar
ahull
Posts: 1662
Joined: Mon Apr 27, 2015 11:04 pm
Location: Sunny Scotland
Contact:

Re: DLO-138 OScope

Post by ahull » Mon Feb 27, 2017 8:55 am

Nice :D

I see stm32duino.com gets a mention on the github main page, and Stevestrong's fast TFT lib also gets a mention here...
https://github.com/ardyesp/DLO-138/tree ... src/TFTLib
- Andy Hull -

User avatar
ThunderOwl
Posts: 135
Joined: Fri Oct 28, 2016 6:04 am
Location: Riga, Latvia
Contact:

Re: DLO-138 OScope

Post by ThunderOwl » Mon Feb 27, 2017 10:20 am

This is sure interesting take on things! :) That being said, I wonder, does anyone moves things to F4xx already? Because russian-speaking osci-making community are doing that, they are taking the lead! :D
F407, for example. There are chinese boards with that (I have one). Yes, those are not almost-free as BluePills, but, I understand that hardware is much much more capable and non-restricting
In every man's childhood, first 40 years are the hardest

stevestrong
Posts: 1838
Joined: Mon Oct 19, 2015 12:06 am
Location: Munich, Germany

Re: DLO-138 OScope

Post by stevestrong » Mon Feb 27, 2017 10:31 am

The question is what brings the F407 more than F103 can offer.
If the display update rate is "acceptable" for the user, I see not reason why to go for F407 (although I am aware of all advantages that it would bring).

Btw, @willfly1, congrats, simple and therefore cool look! I like very much the design!
Last edited by stevestrong on Mon Feb 27, 2017 1:17 pm, edited 1 time in total.

User avatar
ThunderOwl
Posts: 135
Joined: Fri Oct 28, 2016 6:04 am
Location: Riga, Latvia
Contact:

Re: DLO-138 OScope

Post by ThunderOwl » Mon Feb 27, 2017 10:52 am

My "not-so-educated guess" would be usage of better internal ADCs, working in tandem (can be done on beefy versions of F303 too, four ADCs. Example: UCE-DSO-210 by uCore). Probably amount of RAM. Moving display away from SPI.
In every man's childhood, first 40 years are the hardest

User avatar
rexnanet
Posts: 209
Joined: Wed Mar 16, 2016 10:34 am

Re: DLO-138 OScope

Post by rexnanet » Mon Feb 27, 2017 11:43 am

The F4 versions have much faster clock speeds so FFT and display issues will be a lesser problem.
Also have a loooot more RAM = bigger sample buffer.
And It enables the use of an external faster ADC, maybe on the 50MSPS range...

I'm assembling some PCB adapter boards with a F429 and F746 to do some experiments...

It would be great to replace the manual gain switches (and gain pins) with menus and MCP6S92's :)
Might have a go on this :)


P.S. BTW is there a video with this setup/code? :) how many FPS?

stevestrong
Posts: 1838
Joined: Mon Oct 19, 2015 12:06 am
Location: Munich, Germany

Re: DLO-138 OScope

Post by stevestrong » Mon Feb 27, 2017 1:14 pm

Dual 1Msps ADc readings is no problem for F103, the question is whether there is need for more. If yes, then it is F407 a clear choice.
Otherwise the TFT update, I can tell you, is very fast with the current lib (8 bit parallel, not SPI). Check it out here.
As I told, if there is no speciality requested (like FFT or higher sps), I am convinced that F103 is still a good choice.

But this does not mean that I will not test it with my F407 board :)

User avatar
ThunderOwl
Posts: 135
Joined: Fri Oct 28, 2016 6:04 am
Location: Riga, Latvia
Contact:

Re: DLO-138 OScope

Post by ThunderOwl » Mon Feb 27, 2017 1:34 pm

As far as I know, thing that "I don't need higher bandwidth, cause I look at low frequency signals" is common misconception, for digital oscilloscopes. More samples per second means more real information taken about signal (and not, say, constructing signal from 4 points). So, if we want to see some bad spikes in not so high frequency main signal - we need as much points digitized as possible. More professionals than me - please correct if I said not completely correct.
P.S.: every DSO's dream is to became like analog one :) (but with all that superb servicing features that digitals has)
In every man's childhood, first 40 years are the hardest

User avatar
rexnanet
Posts: 209
Joined: Wed Mar 16, 2016 10:34 am

Re: DLO-138 OScope

Post by rexnanet » Mon Feb 27, 2017 1:52 pm

stevestrong wrote:Dual 1Msps ADc readings is no problem for F103, the question is whether there is need for more. If yes, then it is F407 a clear choice.
Otherwise the TFT update, I can tell you, is very fast with the current lib (8 bit parallel, not SPI). Check it out here.
As I told, if there is no speciality requested (like FFT or higher sps), I am convinced that F103 is still a good choice.

But this does not mean that I will not test it with my F407 board :)
The link you sent doesn't have a video... it's the github page..

I was referring to video on the SPI display version :) That would be nice to see.

I also agree with you ThunderOwl. More MSPS is always better :)

willfly1
Posts: 17
Joined: Wed Jan 04, 2017 5:13 pm
Location: GTA, Canada

Re: DLO-138 OScope

Post by willfly1 » Mon Feb 27, 2017 2:37 pm

rexnanet wrote: P.S. BTW is there a video with this setup/code? :) how many FPS?
Sorry, don't have a video and didn't calculate FPS either. Its fast enough, that I didn't bother with it. The minimal flicker comes from 8 bit parallel display and erase/drawing of waveform in a single pass.

Post Reply