Other DSO manufacturers besides GW-Instek that use Xilinx Zynq-7000 series

[ebastler]

Thanks Pascal; I am really glad that we already have the model names for those modified Zynq scopes sorted out...

Seriously though: Who do you think should have an interest in developing and selling such a product? The scope manufacturers, instead of developing and selling a new scope model at bigger margins and revenues?  An independent aftermarket manufacturer, who will have to reverse-engineer multiple scopes, and deal with really interesting challenges of design ownership, EMC and electrical safety responsibility, product liability, warranty... all for measly revenues from a niche product??

No offense intended, but I would be curious about your professional background. Have you been involved with hardware development in a commercial environment?

[pascal_sweden]

I haven't stated anywhere that the goal is to make this a mass product.

Neither have I stated that it should be implemented on all of the current available bench oscilloscopes.
One from the list that I provided is enough. The list just gave an overview of the possible candidates, that is, cheap bench oscilloscopes that have an area for improvement in terms of processor performance.

It would be a proof of concept project, to validate if it can be done, and what the implications are.

Eventually this project and all the related activities behind it, would result in an open source bench oscilloscope based on a Xilinx Zynq-7000 architecture. It is the latter one, that I envision to become a mass product!

In this open source bench oscilloscope, all parts, including the analog front-end, the controls and the knobs, the display, the power supply, the hardware and processor architecture, would be owned by a legal company or an open source organisation.

Possibly there could be a partnership with an existing Chinese oscilloscope manufacturer, as they are extremely good at the build mechanics, the molding of the plastics, and the overall physical design.

The product would be released on the market after all necessary certifications and conformity tests would have been passed.

The open source bench oscilloscope would be an excellent product for educational purposes in an FPGA and DSP class room, and for students and thinkers at home that want to learn different aspects related to electronic systems:

Embedded Software:
Software development (C, C++, Java) on various hardware platforms
MCU and FPGA development (Atmel, ARM, Zynq, Xilinx ISE/Vivado)
Android front-end design and development of Linux device drivers

Chip Design:
Electronic Design Automation, HDL languages (VHDL, SystemC)
High-Level Synthesis, “All-in-C” based design, Virtual Prototyping
FPGA Based ASIC Prototyping for Hardware Software Co-Design

Regarding my background: I like to think out of the box, and I see opportunities where other people only see weaknesses. I see myself blessed with that mind set. I am not that technical anymore, given my current role, but I hold a Master degree in Electrical Engineering, and I am the personal holder of several patents related to electronic systems.

[nctnico]

Do you have any idea how many hours it will take to develop (design, write and test) the oscilloscope firmware? It doesn't seem like much but it is actually an incredibly complex piece of software so count on at least a couple of thousand hours.

You better just bite the bullet and buy a GW Instek GDS-2204E. If you want to add your own features to it then work out how the applications (which can be installed seperately) on these oscilloscopes work.

[ebastler]

Sorry Pascal; got you wrong regarding the commercial intent then. Probably the discussion of model numbers and stickers made me assume that you were suggesting a product.

As a hobby project -- if this tickles you fancy, feel free to give it a go!  It's not for me, though: For me, such a significant time investment should either result in something that is truly unique (not necessarily useful, but unique  ), or should be about using and/or conserving an old device that has some significance for me. While I like using my Rigol as an off-the-shelf tool, those low-end scopes are too much of a "throw-away" product to be the basis for a rewarding hobby project.  But that's just my personal taste.

[pascal_sweden]

No worries. I appreciate your remarks and concerns.

Regarding effort and business priorities: I do realize this is a big task, but one can achieve great things in the open source software community, if there is enough traction behind the project
This is definitely a joint-effort project, and not a one-man show!

If the underlying Linux device drivers are in place, and a high-level software API library is available to control and monitor the various functional areas in the oscilloscope and the various features related to signal processing, it could form a basis for new features and new software applications over time.

The time you would invest in building this product, would be rewarded with new knowledge and skills in various areas. So you would be rewarded with additional education and on the job training =)

I do agree that this project is probably an utopy! One where we would no longer have to wait months for firmware updates from the vendor!

However, I believe it is sometimes nice to play with ideas, envision something bigger, throw it in the ring for discussion, and get feedback about feasibility That discussion on its own already brings some value!

[ebastler]

If you want to get this off the ground, I think the way to go is to actually start working on it. That will get people excited, prove to them that you are serious, and will provide a nucleus from which you can try to build a larger team. Christer Weinigel's work on the Owon scope (in the SDS7102 thread) is a great example.

At least for my hobby endeavors, the problem is not that I'm short on ideas, but short on time to actually develop something. Assuming that most hobbyists are in the same boat, we are not that impressed by someone who just suggests: "Hey, you should be doing this!" 

[pascal_sweden]

Same here. Plenty of ideas, but also short on time unfortunately.

My ambition for this summer is, to become good on the Zynq-7000 platform and develop my own Linux device driver for the Linux distribution that runs on one of the ARM A9 processors inside the Zynq-7000.

Once I master that, which will probably take more time than expected, I want to get a better insight on how an Android software stack is mapped on top of a Linux software stack, and how Android device drivers differ from Linux device drivers.

[langwadt]

Same here. Plenty of ideas, but also short on time unfortunately.

My ambition for this summer is, to become good on the Zynq-7000 platform and develop my own Linux device driver for the Linux distribution that runs on one of the ARM A9 processors inside the Zynq-7000.

unless you have something that absolutely cannot run out side the kernel you usually don't need to write a kernel driver

With the standard UIO driver you can mmap() registers in the PL and access interrupts from the PL with select()
from user space

[pascal_sweden]

Siglent released a new oscilloscope based on the Zynq-7000 SoC architecture.

More details can be found in the separate thread:
https://www.eevblog.com/forum/testgear/new-siglent-sds1000x-e-oscilloscope-based-on-xilinx-zynq-7000-soc-architecture/msg1172684/#msg1172684

[tautech]

Siglent released are soon to release a new oscilloscope based on the Zynq-7000 SoC architecture.

More details can be found in the separate thread:
https://www.eevblog.com/forum/testgear/new-siglent-sds1000x-e-oscilloscope-based-on-xilinx-zynq-7000-soc-architecture/msg1172684/#msg1172684

Error corrected ^^^ 

Mid April is what I'm led to understand.

[kcbrown]

Siglent released are soon to release a new oscilloscope based on the Zynq-7000 SoC architecture.

More details can be found in the separate thread:
https://www.eevblog.com/forum/testgear/new-siglent-sds1000x-e-oscilloscope-based-on-xilinx-zynq-7000-soc-architecture/msg1172684/#msg1172684

Error corrected ^^^ 

Mid April is what I'm led to understand.

I think it'll be very interesting to see how the new scope does.  From Dave's videos, it looks like it's fast and reasonably responsive (doesn't look to have quite the instant response that Keysight's scopes have, but does any scope match that?).

[wpwrak]

Are there any new members in the Zynq-based scope class ?

So far, I know of (with date of introduction):
Instek GDS-2000E series (2016)
Siglent SDS1000X-E series (2017)
Rigol 7000 series (2018)

Siglent's seems to be the most "pure" Zynq-centered system, while Instek use at least one additional FPGA, and Rigol seem to have two custom ASICs (*). At least Siglent also seems to OEM for other brands.

(*) According to https://www.designworldonline.com/rigol-announces-new-7000-series-digital-oscilloscope/

- Werner

[nctnico]

There are more models from GW Instek that use the Zync. What is interesting about GW Instek is that they also seem to use Lua. This would allow a very easy way to extend the functionality without needing to rewrite firmware. However I never looked into this deeper.

[wpwrak]

There are more models from GW Instek that use the Zync.

Thanks ! So that would be the GDS-1000B series, according to the thread at https://www.eevblog.com/forum/testgear/a-review-of-the-gwinstek-1054b/ ? Appears to be a streamlined (and somewhat downgraded) iteration of the GDS-2000E platform. I wonder if they've gotten rid of the second FPGA.

What is interesting about GW Instek is that they also seem to use Lua. This would allow a very easy way to extend the functionality without needing to rewrite firmware. However I never looked into this deeper.

Interesting indeed. Sounds like something they might be able turn into an attractive and fairly unique feature without too much effort.

- Werner

[Markus@RohdeScopes]

Hi

I didn't recognize this thread until yet, but the RTH uses a Zynq too.

Markus

[wpwrak]

I didn't recognize this thread until yet, but the RTH uses a Zynq too.

Thanks ! Is it also one from the 7000 series ? At 5 GSa/s, I guess it's accompanied by some custom ASIC(s) ?

Background: I'm compiling a list of platforms that might be suitable for an attempt to replace the vendor firmware with something Open Source, in this context: https://www.eevblog.com/forum/projects/open-source-instrument-firmware-project-104048/

- Werner

[Markus@RohdeScopes]

It's from the 7000 series. The design is more than 2 years old (launched Jan 16). The whole DSP is done in the FPGA no additional ASIC. But other analog components are own ASICs (like ADC) and therefore it's not possible to replace our firmware by an Open Source FW. The open Source FW is an interesting concept but not very realistic. In details it's not so easy to build an instruments FW, which measure the values with the right accuracy.

[wpwrak]

The whole DSP is done in the FPGA no additional ASIC.

Nice !

But other analog components are own ASICs (like ADC) and therefore it's not possible to replace our firmware by an Open Source FW.

Yes, undocumented chips make everything a lot more difficult. Not impossible, but harder. Sometimes, manufacturers provide this kind of information when they realize that 3rd party firmware is more likely to increase their revenue than harm them, but whether and when depends on circumstances.

The open Source FW is an interesting concept but not very realistic. In details it's not so easy to build an instruments FW, which measure the values with the right accuracy.

Oh, I wouldn't expect it to be very good at the beginning, but Rome wasn't built in a day either My main concerns regarding accuracy are calibration, and to a lesser extent characterizing the analog path. It's likely to be difficult to make full use of manufacturer-provided calibration data - assuming that we can retrieve it, without a clear definition how it is to be used.

Hmm, just saw in the FOSS thread that someone already beat me to getting the basics going on the Siglent. Excellent

- Werner

[pascal_sweden]

Let´s reconvene on this topic in the scope of the Rigol MSO5000 series and higher!

[oliv3r]

Hmm, just saw in the FOSS thread that someone already beat me to getting the basics going on the Siglent. Excellent

- Werner

Two years later, I'm curious, anybody have a link to that thread for me to read? As I don't have any siglent hardware, I don't frequent those threads :p