EEZ Bench Box 3 (BB3)

[IanJ]

Ohhhhhh, so looking forward to getting my BB3.......I've even started clearing an area on the workbench for it's permanent stay........not to mention developing my own modules for it.

Roll on delivery day.

Ian.

[hydrabus]

Hello prasimix,

I'm CrowdSupply backer of EEZ-BB3 and it seems there is no any news since about 1 month, I hope all is fine on your side.
Do you have some update about the "EEZ Bench Box 3" ?

Best Regards,
Benjamin

[fuzzoli]

Hi Benjamin,

The updates are over on this thread: https://www.eevblog.com/forum/crowd-funded-projects/eez-bench-box-3-sequel-to-eez-h24005/msg3063776/#msg3063776

[prasimix]

Sorry hydrabus, time simply passes and the things promised to me do not happen. In short, I have everything except the DCP405 module for which I don't know what the status is (not to go into details now). I believe this will be known through the next week.

[prasimix]

Another achievement on BB3 + ULX3S integration track (previously mentioned in post #259): images from two sources can be combined with overlaying and displays on the same touchscreen. On one side we have content generated from STM32 running BB3 firmware, and from other side is ScopeIO on ECP5 FPGA:







In the next step, we will try to synchronize a higher resolution source (800x480 or 1024x600) and provide the combined image on the HDMI output (for presentation on an external monitor or projector).

[prasimix]

... and BB3 screen content on HDMI monitor! Now you can imagine to have e.g. control menus on the small touchscreen and live data (e.g. data logger, etc.) on the big screen: monitor or projector for presentation or for classroom work, etc.

[danielbriggs]

Wow! 
This project just keeps on developing!

[AlanS]

... and BB3 screen content on HDMI monitor! Now you can imagine to have e.g. control menus on the small touchscreen and live data (e.g. data logger, etc.) on the big screen: monitor or projector for presentation or for classroom work, etc.

Or for someone whose eyes are not so good!

[goran.mahovlic]

Or if you like higher resolution (1024x576x60) with fixed frame buffer

[prasimix]

I spent some time to check what is available in STM32H7 family which offers both single and dual core MCUs. It seems that migration to it sounds like no brainier since it gives more speed for less money. For example, STM32F769IIT6 costs € 11.08 on Mouser (qty. of 100), while STM32H743IIT6 costs € 9.58 and offers 480 MHz for system clock, and 1 MB of RAM (while F7 offers 216 MHz and 512 KB).

However, the question remains, can one easily switch from F7 to H7, i.e. are they pin compatible when selecting package with the same pin count that is LQFP-176 in our case. Migration is unfortunately not straightforward, but it is not impossible either. I was able to put all the necessary functionality on the H7 device, I even managed to improve the following things:

• SDRAM address bus is increased to 13-bit, that allows us to connect up to 32 MB
• SD card bus is expanded from 1-bit to 4-bit (I wonder how much this could speed up access to the SD card!)
• Touchscreen controller got a dedicated I2C bus

Something that may be missing is USB 2.0 which unfortunately cannot be realized without external PHY.

Of course it all fits together on a PCB of identical dimensions as the existing F7 MCU module. The only new thing I had to add was a 25 MHz external oscillator (XO1) and a 1:2 clock driver (IC5) because there were no free pins on the MCU to define the clock output for Ethernet PHY.

[krho]

Don't shoot the messenger but. Why don't you upgrade to LQFP 208... This would probably mean that you can do USB 2.0. and also add QSPI flash if there will be need to have memory mapped resources (fonts, images, etc.)
The CubeMX should do that upgrade pretty seamlessly.
Also there is(was) a mess with H series. Some registers were remapped and some frequency changes and no part number change was made  so you need to be careful.

[prasimix]

Yes, I believe that migration to LQFP-208 is also possible, and could be more challenging if we want to preserve same PCB dimensions, but even that is not so critical since there is a plenty of room to grow inside enclosure. Only width has to be preserved that we don't need to change front panel holes.
Thanks for warning regarding H7 mess. If you have more detail or link to discussion about it let us know, particularly if proposed STM32H743IIT6 is mentioned.

[prasimix]

A simple example of how to use Node-RED to communicate with BB3:

[Corona_Spark]

I have been following this project, and I am also preparing to DIY an EEZ BB3. I am in China and I can’t see the pictures and videos when browsing the posts on the EEV BLOG website, which makes me very frustrated. In addition, my English is not good, so I can only read the posts through machine translation, and many of the content is not very understanding.
The work I have done recently is to analyze the schematic diagrams of each module, organize the BOM, and replace the components that are not easy to buy in China or the more expensive components with those that are relatively easy to buy and cheaper.(I also want to buy a kit directly to support this project, but the exchange rate of USD to RMB is too high, which is a big expense for me. So, I decided to DIY an EEZ BB3 by myself, but it did not help this project. I am very sorry.) The EEZ BB3 package is designed very well! I am slowly studying its essence. After finishing the BOM, I will conduct trial production of the first machine. Recently, I have seen the appearance of DCM224 module, and I am looking forward to its hardware system release.
On the https://www.envox.hr/eez/eez-bench-box-3/introduction.html website I saw a detailed explanation of the schematic diagrams of each module of EEZ BB3. This part is great! It is very worthy of my study, but it seems that the schematic diagram explained is an old version, which is different from the latest version of the hardware, which is a pity. I don’t know if there is the latest version of the document explaining the hardware principle? In this way, it is much easier to understand the design principle corresponding to the latest version of the schematic diagram and this document (I have poor analog circuit knowledge ).
In the choice of MCU, I think it is indeed inferior to using STM32H743IIT6. Now this chip has a low price and strong performance. The price of this chip used in our recent project is much cheaper than STM32F769IIT6, and it is better to buy. The version of the chip we use is older, the main frequency can run up to 400MHz, the new version of the chip can run up to 480MHz.

[prasimix]

Thanks for your inputs...

Recently, I have seen the appearance of DCM224 module, and I am looking forward to its hardware system release.

That's right, the DCM224 is a variant of DCM220 derived from modifications made to make it easier to pass EMI testing for CE. The main difference is the transition from a 2-layer to a 4-layer PCB and the reorganization of the power loop. It was also added that the MCU controls the switching frequency of two LT3763 controllers with counterphase signals. The option of frequency dithering has also been added, which can further enhance the EMI image.
For easier differentiation, I changed the output voltages and currents so that it goes up to 24 V and about 5 A (not calibrated), so the new name: DCM224. Finally, the option of PWM dimming power output has been added.

On the https://www.envox.hr/eez/eez-bench-box-3/introduction.html website I saw a detailed explanation of the schematic diagrams of each module of EEZ BB3. This part is great! It is very worthy of my study, but it seems that the schematic diagram explained is an old version, which is different from the latest version of the hardware, which is a pity. I don’t know if there is the latest version of the document explaining the hardware principle? In this way, it is much easier to understand the design principle corresponding to the latest version of the schematic diagram and this document (I have a poor analog circuit level ).

Yes, and I have no excuses for that. I have set relatively high standards and now I need to maintain them. I need to update the project pages as soon as possible as it will help other people who backed the "EEC BB3 Enclosure & Bare Boards" kit to more easily follow the design (they are welcome here and in the newly opened topic). I am currently finishing updating the user manuals (English and German) and I believe that I will soon be able to update the project pages as well.

In the choice of MCU, I think it is indeed inferior to using STM32H743IIT6. Now this chip has a low price and strong performance. The price of this chip used in our recent project is much cheaper than STM32F769IIT6, and it is better to buy. The version of the chip we use is older, the main frequency can run up to 400MHz, the new version of the chip can run up to 480MHz.

I'm aware of that, that's why I made the new MCU module version with H743, but the layout of the new PCB is just the beginning: we need to assemble the module and then the most important thing follows: adapt the firmware to work with two cores and work reliably. I think the new module will have to wait since the F769 does its job very well, albeit for a slightly higher price.
Priority will be given to new peripheral modules: we need to finish the three we started (MIO168, SMX46, PREL6), but also start working on some new ones. I am thinking primarily of the 2Q/4Q power module and I would love to see progress on high precision DMM, something based on a @Kleinstein design or @jaromir design.

[Corona_Spark]

Thank you for replying to me so early! Thank you very much for answering all my questions! It is the high standards of your project that attracted me, and your spirit of dedicated research is very worthy of our study! Many R&D engineers around me are too impetuous, and they are not careful in researching technology. This disappoints me very much, and I really want to change this situation. We have also established an interest group to study and study the EEZ BB3 project, hoping to learn more knowledge from it and develop good research and development habits. We will continue to pay attention to the EEZ BB3 project, and once again thank the EEZ BB3 team for their efforts!

[Corona_Spark]

That's right, the DCM224 is a variant of DCM220 derived from modifications made to make it easier to pass EMI testing for CE. The main difference is the transition from a 2-layer to a 4-layer PCB and the reorganization of the power loop. It was also added that the MCU controls the switching frequency of two LT3763 controllers with counterphase signals. The option of frequency dithering has also been added, which can further enhance the EMI image.
For easier differentiation, I changed the output voltages and currents so that it goes up to 24 V and about 5 A (not calibrated), so the new name: DCM224. Finally, the option of PWM dimming power output has been added.

For a more complex power system and EMI testing is required, it is difficult to design a two-layer board, the ground plane is not continuous, the ground impedance is large, and the wiring is more difficult. Four-layer board is indeed a better choice, and the cost will not increase too much.

The 2Q/4Q power module is absolutely awesome! Looking forward to it!

[exe]

Question: can dcp405 generate waveforms? If so, what's the maximum frequency? I need a 100Hz signal (either sine or triangle). Of course it will be uni-polar, but that can be fixed with a DC-blocking capacitor.

[prasimix]

The easiest way to perform that is to apply external signal generator on Rprog inputs on the front panel and activate Rprog option (see Section 11.5.3). Here is an example of 200 Hz sinawave 2Vpp that should generate about 32 V on the DCP405 output (attached load is 4R7 + 4R7):



Another example is 120 Hz triangle (50% ramp):



... and 120 Hz square wave:

[exe]

Thank you very much, this is more than I hoped.

[prasimix]

Today we finalized support for SMX46. This is the first step in the direction of transforming the BB3 from a programmable power supply into a T&M chassis, which was the initial idea after the closure of the H24005 project. It is supported now in master firmware v1.3 and got its own firmware that can be easily downloaded without an additional debugger/programmer as is the case with the DCM module. The first use case for this module is to enable automation of module production testing, and we will start with the DCP405.
SMX46 got all the necessary SCPI support (which included adding a new ROUTe subsystem), and here are a few screenshots of what it looks like on the BB3 display:

3 module view (second slot):



Maximized view:



Settings page:



Analog output calibration:



Relays cycle counters:



Switch matrix routing and label settings:

[goleszek]

Hello,

I'm new to this thread and I'm really impressed with the capabilities of the firmware of Bench Box 3 tool.

A few years ago I've started a similar hobbyist project called Labtool, supporting the usage of the various measurement modules. My device consists of the backplane with slots for up to 10 add-on modules, housed by a laser-cut metal housing.

The main capabilities of a backplane are:
- 7 inch LCD, 800x480 (or 6.5'' 400x240 with touch capability)
- MCU STM32F429 2MB Flash with external 32MB SDRAM, 64MB SPI Flash and battery backup RTC
- bunch of embedded interfaces like ethernet, 2xUSB, RS232, RS485, LIN, 2xCAN, I2C, SPI, UART, PWM, DIO, AIO

The purpose to design this tool was the need to have a freely programmable test & measurement device, capable to run even complicated test sequences thanks to PLC-like firmware. The idea reflects the capability to use each of the existing I/O and it's set of capabilities (like the current, voltage, baud rate, duty cycle, frequency, phase, etc.) as a parameter or result storage of any logic or arithmetic sub-block.

Each I/O card, sized 60x122mm, communicates with a backplane through 32-pin connector, providing +5/+12 supply, 16MHz system clock, SPI, I2C, and a few dedicated bi-directional lines. On every card exists small I2C configuration memory, which provides card information (like name, type, calibration data, etc.) during the initialization phase or after hot-swap.

Up to now I've designed and tested six different modules: 6-channel relay board, 8-channel high side driver module, 2 and 4 channels DDS signal generator, spectrum analyzer DC-50MHz, dual programmable resistor decade, and a 4-quadrant power supply +-20V/+-2A . And of course, the port extension for debugging . Some photos are attached herein.

Unfortunately, this project consumed me a lot of time and due to other professional activities was not yet fully finalized, mostly its software part. So this is why I'm impressed with the complete functionality of your design.

In a case someone will find this design useful, I can share schematics and some build guidance.

[Andrew McNamara]

[...]
Up to now I've designed and tested six different modules: 6-channel relay board, 8-channel high side driver module, 2 and 4 channels DDS signal generator, spectrum analyzer DC-50MHz, dual programmable resistor decade, and a 4-quadrant power supply +-20V/+-2A . And of course, the port extension for debugging . Some photos are attached herein.

Nice work! Surprising how much like the BB3 it is, too - clearly there's a need for an open platform like this.

[prasimix]

Great work @goleszek! The project seems to include a lot of ideas that have been implemented or are planned for BB3. Some of your modules could be a great addition to the BB3, some like the 4-quadrant PSU have been under consideration for a long time.
If there is a good will to share the details, I suggest you open a GitHub account and put there everything you want to share and specify license. Of course there is no reason not to do that here on the forum, in this thread or a new one.

In theory, someone might be able to try and adapt the BB3 firmware part to your chassis when you get a richer GUI and solid SCPI support.

By the way, I'm also interested in more details about your PLC-like firmware i.e. PLC functionality/programing principle, how far you've come and if you have any plans about it.

[goleszek]

Hi @prasimix,

Yes, I thought about putting everything "as is" to Github, but probably as a next step this will need my prompt support on that (which, unfortunately , I can't provide due to lack of time).

Anyhow, I will prepare some details regarding 4-quadrant PSU and PLC-firmware basis, which I will post here soon - maybe someone will find it useful.

For now, I've added herein some pictures, which didn't fit in my previous post (sorry for low quality of some of them).