Author Topic: Open Source Scope- A Case In Point (Read 10481 times)

Old Printer · « **on:** May 23, 2018, 06:35:24 pm »

To put out an open source scope, a company with deep pockets will have to roll the dice. While thinking about this and wondering if it might ever happen, a somewhat similar situation struck me. I am a life-long model airplane hobbyist, mostly radio control, and that is how I became interested in electronics. Microprocessors had a huge impact on radio control equipment, particularly transmitters. Complicated models with many parts to control, and craft like helicopters and multi-rotors (drones) saw a huge advancement. There were only a handful of manufactures sharing the market, Futaba, JR Propo, Multiplex and Horizon are a few. Six or seven years ago a small startup Chinese company, FrSky Electronics, decided to market a transmitter with open source software called Taranis. Model builders are a perfect target, that as a group love to tinker. FrSky did it right and used excellent quality components and put out a well engineered product. It was a huge hit and word spread through the online model forum's. It was easily the biggest moddeling product rollout I have ever seen. There was already a group, kind of like a small Sigrok, of software type enthusiasts with an open source project underway. FrSky piggybacked on that and have stayed with the original plan. I don't know their bottom line, but they continue to innovate and offer new products, and there is always a waiting list for the first run.
So if any of the companies with the where-with-all to make a scope like this happen, take a good long look at the FrSky Electronics company, and how they have done since their Taranis transmitter was introduced (2013 I think). Like Yogi says, you could look it up

mikeselectricstuff · « **Reply #1 on:** May 23, 2018, 07:06:45 pm »

Most RC gear is used by hobbyists who like to tinker. Most scopes are used by professionals & educators.
I'm pretty sure the scope market is way smaller that RC modelling.
Most scopes do everything reasonably well, so few users will feel the need to tweak - it's just a tool.

We will never see an open source scope from a big manufacturer.
We may see open source firmware for an existing scope, but only a small number of people will use it.

ebastler · « **Reply #2 on:** May 23, 2018, 07:21:37 pm »

Quote from: Old Printer on May 23, 2018, 06:35:24 pm

Six or seven years ago a small startup Chinese company, FrSky Electronics, decided to market a transmitter with open source software called Taranis.
[...]
There was already a group, kind of like a small Sigrok, of software type enthusiasts with an open source project underway. FrSky piggybacked on that and have stayed with the original plan.
[...]
So if any of the companies with the where-with-all to make a scope like this happen, take a good long look at the FrSky Electronics company, and how they have done since their Taranis transmitter was introduced (2013 I think). Like Yogi says, you could look it up

So, what was there first -- the FrSky company and their transmitter project, or the OpenTX open-source software? This sounds less like a company doing the right thing and open-sourcing a software package they have developed internally, and more like a startup taking the cheap-and-easy route by feeding on an existing open source project?

So, feel free to go ahead and develop your own scope firmware. Then stand by and wait for a Chinese startup to pick it up and make it part of their new product. Not sure whether this is a desirable model?

t1d · « **Reply #3 on:** September 12, 2018, 02:42:38 am »

Open Source oscilloscopes are already out there... Just a Google away...

mikeselectricstuff · « **Reply #4 on:** September 12, 2018, 07:59:06 am »

Quote from: t1d on September 12, 2018, 02:42:38 am

Open Source oscilloscopes are already out there... Just a Google away...

Show me an Open Source Scope that is anywhere near as functional as even the cheapest commercial one.
It won't happen, because nobody needs it enough.
Openness alone is not a good enough reason - it has to offer something new.

ebastler · « **Reply #5 on:** September 12, 2018, 11:07:53 am »

Quote from: mikeselectricstuff on September 12, 2018, 07:59:06 am

Show me an Open Source Scope that is anywhere near as function as even the cheapest commercial one.
It won't happen, because nobody needs it enough.
Openness alone is not a good enough reason - it has to offer something new.

Agree. Or at least it has to be on par hardware-wise with commercial scopes, to enable users to add something new via the open-source platform.

Most open source "scopes" seem to be toys with 10 MHz sampling rate or less. https://www.scopefun.com/ offers two channels with 100 MHz sampling rate (each?). Has anybody found something better than that?

Jeroen3 · « **Reply #6 on:** September 12, 2018, 12:35:39 pm »

I don't think there is a business model in Open Source Scopes.

You need a large investment to build the hardware. And you need to build the software.
Open Source is not free. And I believe the return value from the open source community is not very high.

I mean, Qt is a very large "open source" project, but it cannot exist without Qt Group selling licences for non-open source features. It is such a massive project that for anyone capable to work on Qt and push changes would be spending his time working for a different employer. Only if that employer values open source, this is viable.

Mario87 · « **Reply #7 on:** September 12, 2018, 12:38:07 pm »

Probably as close as you'll get to an open source scope from a major manufacturer.... http://www.ti.com/tool/TIDA-00826

Mechatrommer · « **Reply #8 on:** September 12, 2018, 01:07:22 pm »

Quote from: ebastler on May 23, 2018, 07:21:37 pm

Quote from: Old Printer on May 23, 2018, 06:35:24 pm
Six or seven years ago a small startup Chinese company, FrSky Electronics, decided to market a transmitter with open source software called Taranis.
[...]
There was already a group, kind of like a small Sigrok, of software type enthusiasts with an open source project underway. FrSky piggybacked on that and have stayed with the original plan.
[...]
So if any of the companies with the where-with-all to make a scope like this happen, take a good long look at the FrSky Electronics company, and how they have done since their Taranis transmitter was introduced (2013 I think). Like Yogi says, you could look it up

So, what was there first -- the FrSky company and their transmitter project, or the OpenTX open-source software? This sounds less like a company doing the right thing and open-sourcing a software package they have developed internally, and more like a startup taking the cheap-and-easy route by feeding on an existing open source project?

So, feel free to go ahead and develop your own scope firmware. Then stand by and wait for a Chinese startup to pick it up and make it part of their new product. Not sure whether this is a desirable model?

this whole drama can be both a curse or a bless. a curse if you the original developer want to make profit out of it, better keep it closed source otherwise you'll lose in competition (sale), chinese will usually profit just a dime, or even negative if we try to calculate it. so this is also a bless, because the original OSHW is like a publicly available blueprint, but only the china can manufacture it the cheapest, hence benefiting users. comparing china with non-china company/manufacturers that copycatted/originated from OSHW, non-china usually priced more expensive compared to china version. recent example is prusa i3 mk3 that keep popping and annoy me in youtube ads, the price is more than double than clones. granted it has bells and whistles but its not like i will not be able to print with china version. ymmv.

back to OSHW DSO that imho wont go because:
1) DSO circuit is not simple like FrSky transmitter or rc controller, not everybody is willing to assemble and populate all components by hand, its not easy task.
2) no need OSHW, china have done it, Rigol, Siglent, Arm USB scope etc, you just cant get any cheaper. even if you can, it wont worth the trouble.

The End Thank you.

cstratton · « **Reply #9 on:** October 04, 2018, 05:43:21 am »

RC Transmitters really don't challenge technology at all. Electronically, it's just an MCU, an SPI packet radio intended for wireless mice,a display, and some input transducers. You could build your own, and many people have. It's just cheaper and easier to buy.

In contrast, affordable scopes with decent performance are all about pushing the parts beyond their alleged specifications. It's tricky design, and probably quite a bit of behind-the-scenes sourcing negotiation.

Replacing the oft-buggy or at least sub-optimal firmware of a popular scope model would be a worthy effort, however it's a big project (even just documenting the board is a bunch of high speed parallel routing into an FPGA or similar, not a half dozen few-MHz SPI lines to a radio) and not clear that the targeted model will remain available long enough to justify the effort.

xaxaxa · « **Reply #10 on:** October 04, 2018, 06:39:00 pm »

I don't know why people are focusing on open source oscilloscopes, arguably one of the worst candidate markets for introducing a new product. There are open source test equipment out there just as advanced if not more (spectrum analyzers, network analyzers) that are available off the shelf.

I don't think it is very difficult to develop an open source scope with performance comparable to commercial scopes, but I don't see much point. Existing scopes do what I need them to do, are good price-performance wise, and there aren't really many "secrets" to the design left to discover, so I think I'd be wasting valuable engineering effort if my company were to develop an "open source scope" product line.

TK · « **Reply #11 on:** October 04, 2018, 07:03:21 pm »

Digilent promoted OpenScope MZ to be open source. The design files seem to be public, not sure about the software

https://store.digilentinc.com/openscope-mz-open-source-all-in-one-instrumentation/

SiliconWizard · « **Reply #12 on:** October 08, 2018, 03:42:57 pm »

Quote from: TK on October 04, 2018, 07:03:21 pm

Digilent promoted OpenScope MZ to be open source. The design files seem to be public, not sure about the software

https://store.digilentinc.com/openscope-mz-open-source-all-in-one-instrumentation/

Yeah. It's very low-spec and this kind of design is accessible to hobbyists. Now even a basic 100 MHz/1 GSps scope that you can buy for 300 bucks is a whole other matter.

0xdeadbeef · « **Reply #13 on:** October 08, 2018, 03:57:47 pm »

Actually there was this at least one Open Source SW project for a (somewhat) commercial scope:
https://sourceforge.net/projects/welecw2000a/

ebastler · « **Reply #14 on:** October 08, 2018, 10:19:25 pm »

Quote from: 0xdeadbeef on October 08, 2018, 03:57:47 pm

Actually there was this at least one Open Source SW project for a (somewhat) commercial scope:
https://sourceforge.net/projects/welecw2000a/

Apparently that scope shipped complete with a strong motivation to rewrite its firmware.

"Welec W2000A - Worst oscilloscope ever??"
https://www.eevblog.com/forum/testgear/welec-w2000a-another-weird-oscilloscope/

0xdeadbeef · « **Reply #15 on:** October 09, 2018, 08:18:33 am »

I didn't say it was a good scope, just "somewhat" commercial (i.e. not a kit or so)

SiliconWizard · « **Reply #16 on:** October 09, 2018, 04:27:41 pm »

Indeed.

Anyway, a lot of commercial scopes seem to be using Linux as an OS (not sure what the "big" brands use though, such as Tektronix, but there's a lot of cheaper scopes that use Linux), so technically, it should be possible to get most of the source code from the manufacturer. I've not seen any releasing any source code so far though.

0xdeadbeef · « **Reply #17 on:** October 09, 2018, 05:31:23 pm »

I would think the big brands (Keysight, Lecroy, Tektronix) still use WinCE or something like that in the lower/mid range scopes. Like I'm pretty sure that the Lecroy Waverunner 3000 series uses WinCE. The upper range scopes typically use Windows anyway (partly because people expect Matlab-API support and stuff like that).

LapTop006 · « **Reply #18 on:** October 10, 2018, 06:24:00 am »

The Tek 5k series is Linux or Windows, I think all the Keysight's are embedded Windows.

The kernel modifications (and thus the parts you'd have an argument for under GPL) are likely pretty minimal, and would be unlikely to really help you build a new UI for existing hardware.

SiliconWizard · « **Reply #19 on:** October 12, 2018, 01:35:05 am »

Probably, although they may use various libraries that would make most of their software GPL.

As for Windows, Lecroy and Tek actually used regular versions of Windows in some of their scopes (not WinCE) in the 2000's. I just don't know what they're up to at the moment.

mikeselectricstuff · « **Reply #20 on:** October 12, 2018, 10:05:39 am »

Keysight use WinCE but they are looking at moving away from it as MS's stupid licensing means they can't buy licenses to ship scopes when MS discontinue it - this was mentioned in Dave's interview with the HP/Agilent/Keysight guy a while ago.

SiliconWizard · « **Reply #21 on:** October 12, 2018, 05:43:15 pm »

Quote from: mikeselectricstuff on October 12, 2018, 10:05:39 am

Keysight use WinCE but they are looking at moving away from it as MS's stupid licensing means they can't buy licenses to ship scopes when MS discontinue it - this was mentioned in Dave's interview with the HP/Agilent/Keysight guy a while ago.

I've heard that. That's incredible. They are really following a weird path at MS overall - walking all over their customers as though they didn't even matter. Anyway.

I think QNX would be a good match as an OS for high-end scopes. Very good stability and support. I've heard it's definitely not cheap though.

mikeselectricstuff · « **Reply #22 on:** October 12, 2018, 07:07:51 pm »

For a company the size of Keysight, rollling their own customised Linux would seem like a total no-brainer as it can be used across so many products.
Will be interesting to see if they ever totally replace the existing OS on existing products rather than superceding perfectly good hardware once MS pulls the plug.

bd139 · « **Reply #23 on:** October 12, 2018, 09:42:59 pm »

You don’t even need to do that. There’s a ton of supported commercial linuxes out there:

https://www.mentor.com/embedded-software/linux/

branadic · « **Reply #24 on:** October 13, 2018, 06:59:01 pm »

Quote

Actually there was this at least one Open Source SW project for a (somewhat) commercial scope:
https://sourceforge.net/projects/welecw2000a/

I've been one of the members in the project. We, Walter and me, also designed a piggyback board for the input stage, based on the LMH6518. All these files are available, but also the complete reverse engineering. Based on this OpenSource activities it's pretty simple to design the hardware.

By the way, after the piggyback board was available to the community a lot of new scopes such as Rigol popped up using the LMH6518, too. But we have been the first.

-branadic-


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Open Source Scope- A Case In Point (Read 10481 times)

Share me