Minimalist computing, PLC with an STM32.

Finding a PLC that is relatively simple to use and above all, affordable and also easily expandable is not that easy.

You can use hardware based on Arduino, which costs about a third of what a standard industrial type PLC costs, or use an industrial PLC. The prices range according to the solutions adopted from €300 to €1,200 in Europe, at the end of 2022.

More troublesome, programming software is more or less proprietary and more or less flexible. An Arduino solution will always be more flexible than an OEM solution because you can access CPU resources. What industrial solutions prohibit. but Controllino type boxes are not easy to obtain, especially where I live. It is therefore difficult to offer a finished product in this format.

In fact, and this is my case, if I want to develop specific solutions, neither of the two previously mentioned worlds, Arduino or Industrial, allow me to shape a finished product. I am forced to be dependent on higher layers. From an economic point of view, these upper layers capture most of the added value by producing the least effort because none of these upper layers is in contact with the ground.

To be able to offer specific solutions and at the same time capture the maximum revenue, there are not many solutions. You have to master everything, the hardware, the software and the knowledge that will allow you to do something different but relatively profitable.

Let's go back to the late 80s. PLC standards were not yet established as they are today. So there was some design freedom. Most of the small companies then worked on the VME bus. The advantage was that it was enough to make a printed circuit board and a small front panel to be able to offer a processor or input/output card adapted to the customer's needs.

In fact my basic training allowed me at the time to build complete systems with 68000 processors, based on OS9. It was working very well.

Since that time, generalist manufacturers have increasingly 'closed' hardware and software, so that any personal development is made impossible. Of course, even with proprietary solutions, everything is still possible, but at what cost and with what constraints!

So I decided to create a small processor board olmost VME standard. The bus connector is not VME standard, it is useless. Why? Because an STM32 type processor has absolutely all the internal resources necessary for the development of a PLC. It is just necessary to establish high-speed serial communication links to create a high-performance system at low cost.

Compared to the resources required to create the same type of board based on MC68000, this has absolutely nothing to do in terms of complexity or processing capacity. An STM32 commonly operates at 80MHz, which is 10 times faster than an MC68000@8MHz, especially for PLCs where complex CICS instructions are not used.

So, here is the fruit of my reflections :

This board is only the processor board. It has no inputs/outputs (or almost). It is intended to be supplemented by other cards with various types of interfaces. In the spirit of VME, this allows maximum agility in the configuration of an automatic processing machine.

The result is a very simple design. The software to be used for the development of the application program will be that supplied as standard by STmicro.

In fact, anyone with some knowledge of microcontrollers and C language will be able to program this board. When I have tested it, I will publish the project files thus allowing to start writing code directly.

And in addition, this solution will therefore be open source ;-)

Addicted to GoWin fpga's!

It must be recognized that when one has tasted these GoWin circuits and the simplicity of the development software, it becomes very difficult to resist the urge to use them for a whole host of projects.

I had the opportunity, a few months ago, to replace an Intel FPGA-based module with a GoWin module that I made myself, in a project to rebuild the Micro-Professor MPF-1b:

Multitech - MPF-1 / MPF-1B

The result was that : 

I also use a GoWin FPGA in the final version of my MIDI switch. It works very well.
So why not continue with the recreation of an SP12? In fact at first I wanted to recreate a Drumulator. But given the work done on the subject and the little difference in fact with an SP12, I decided to go with an SP12. 

So I gathered the majority of the digital part as well as the sampling in both directions on a small FPGA-based board. 

It should look like this:

I received the circuit boards a few weeks ago and have just mounted the digital parts:

This time the FPGA system is a bit more complex since the circuit is used with some external memory circuits. In fact, the FPGA does not have enough internal memory to accommodate all of the memory associated with the SP12 processor. No room for sample memory either. Also, some of the memory must be non-volatile and this is not possible with the internal memory of the FPGA.

For the moment, I have not powered this board. I will do that in the next few days.

The next objective will consist in starting the program of an SP12 in this board.
Nice challenge!!!

Korg Kronos main board.

Regarding electronic music devices, I'm more of a 'hold style'. I like beautiful, well-constructed main boards. While consulting on Internet which deal with this subject, I have just discovered that the Kronos range from Korg, works on an Intel motherboard base, and more particularly the D510MO version. 

The first generation of Kronos:

The Intel mother board :

In the idea, why not. The machine works with a customized version of Linux, and some additional boards made by Korg.

But hey, given the price of the machine, we could have expected a professional motherboard, to say the least. But no, this is a low-end motherboard whose long-term reliability is absolutely not the objective.

Personally, I don't really like this way of doing things. We will say that Korg produces good material... but much too expensive for my taste given the way it is developed. You do the same with a low-end PC at 500€, a good sound card and a T1 used for its keyboard and bought for $125 (as me), all for less than $1000!

Note that this probably allows convenient management of the touch panel display. Because for the rest, still very small buttons resembling those of your FM tuner from the 80s, designed to be used 10 times a month like those of the T1. Unacceptable for equipment wanting to be professional!

I may not make friends here, but really, is it serious?!!!

A small audio amplifier based on NE5532.

A small idea, by the way. Why not build a small, compact linear audio amplifier?

In fact, I had the opportunity to test a small amplifier based on an operational amplifier of the TL074 type during the development of the audio part of a Drumulator.

This concept of amplifier exists, obviously. I have only taken examples that can be found on the Internet.

32 circuits equivalent to 64 AOPs are placed in parallel. Push-pull type topology therefore 128 AOPs in service. This should provide enough power for more than adequate rendering. In addition, this circuit does not require any adjustment.

An example of how the circuit can be rendered:

I have some finalizations of track sizes to do and I think I will make this little prototype.

Attempt to clone an SP12 (3)

I just received the SP12 PCB.

Another circuit to build and test. However, this becomes quite difficult due to the lack of components.

Fortunately, I will be able to test the majority of the functions of the board thanks to a few tips. Some of the components will only be available from April or May next year: terrible situation!

MIDI SWITCH(2)

The final production version of the PCB has arrived.

Finally, the printed circuit is very small. I have gathered the MIDI ports in packs of four, which will make it easier to cut out the front panel of the case. I brought back the micro-usb power socket on the front panel.

It is planned to be able to connect a PC equipped with a small software to facilitate the configuration of the switch. I don't have a MIDI ID so I decided not to use the MIDI interface for this.

This possibility, like that of also serving as a USB/MIDI interface, is planned for the next version. 

Attempt to clone an SP12 (2)

Construction of the SP12 core continues. I routed the board and sent it to manufacturing.

Preview of routed faces. 

Both sides:

Front face:

And the back :

Finally, a 3D representation:

I just have to wait for the reception of the printed circuits to begin its assembly and the first tests...