Scope clock STM32

[paulvr]

This is my first post in this forum and I would like to introduce you to my version of the scope clock. Mainly inspired by the work of David Forbes and Grahame Marsh.
The work is not finished yet, but I have reached a milestone for myself in that all components are on one PCB and only one voltage source is required, nm. 12 volts DC.
The high voltage is made using a standard Würth WE-flex transformer, plus and minus 250 Volt. The filament voltage is provided by a buck and forward converter and a standard Würth push-pull transformer (6.3 Volt to 600mA).
The MCU is an STM32G491, mainly chosen because of the integrated 12-bit DACs. The image composition is completely DMA controlled and works independently of the CPU, which therefore has a lot of time for other tasks.
The characters and graphic shapes are encoded as SVG path.
An ESP32 has been added to achieve the correct time according to the NTP. Operation is entirely via a web server running on the ESP32. The software is not finished, and not all functions work (work in progress).
The next step is to try out other CRT tubes and make a housing.

Regards,

Paul

[leo oel]

This is great, will the project be available for repeating?

понедельник, 26 февраля 2024 г. в 11:12:05 UTC+2, paulvr:

[ken Henriksen]

Hi Poul.

 

That is a great job nice work, I have been working for this some time but not to extend you got. Would it be possible to see the schematics ect?

 

/Ken

Paul

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/4e60fa43-a9b5-40b5-8381-5a968b45d480n%40googlegroups.com.

[Paul Andrews]

Wow. As others have said, that is great work, and yes, would be very interested in trying to do this myself (to replace my frankensteined scope clock). I love the idea of using SVG to encode the shapes, leaves open a lot of scope (no pun intended!) for future mods.

[David Forbes]

Paul,

This is quite a nice project. It looks like a very clean display.
I have run out of the custom transformers I had made in 2006, and I'm quite interested in using off-the-shelf ones in the future. Would you be willing to share a schematic of what you made?
I also appreciate using SVG rather than rolling your own graphics. That's a nice standard solution that makes sense in this age.

[paulvr]

The flyback converter is built around the TI LM5156H. This IC was selected because it has an external mosfet, which means the voltage on the primary side can be increased considerably and the winding ratio does not have to be very high. An external mosfet generally also has better properties that improve efficiency. The chosen mosfet is very small and difficult to solder, as an alternative I can recommend the IPP17N25S3-100.
The flyback transformer is from Würth from the WE-FLEX series. The insulation voltage is not great considering the output voltage, but I have not been able to detect any problems. The achieved efficiency is between 70 and 80% depending on the input voltage and output power. The output power easily reaches 6 to 7 Watts depending on the input voltage. Minimum input voltage approximately 9 Volts, maximum 16 Volts limited by the LM5156H.
The filament power supply is built around TI's SN6507. That is a fairly new IC and one of the few forward converters I have found that can supply up to an ampere and more. Together with a Würth push-pull transformer it is a good combination to supply the 6.3 Volt up to 600 mA. The SN6507 has programmable current limiting, but that doesn't really work well with the filament which has a very low start-up resistance. The converter remains in current limiting mode when the programmed current is too low. Efficiency in combination with the buck converter approximately 80%.

Op maandag 26 februari 2024 om 17:30:59 UTC+1 schreef David Forbes:

[Richard Scales]

I too would be interested in a kit if you are considering making such a thing available.

- Richard

[Benoit Tourret]

So do I,

I have a 3LO38i waiting for a while in his box.

[Toby Thain]

On 2024-02-26 4:12 a.m., paulvr wrote:
> This is my first post in this forum and I would like to introduce you to
> my version of the scope clock. Mainly inspired by the work of David
> Forbes and Grahame Marsh.
> The work is not finished yet, but I have reached a milestone for myself
> in that all components are on one PCB and only one voltage source is
> required, nm. 12 volts DC.
> The high voltage is made using a standard Würth WE-flex transformer,
> plus and minus 250 Volt. The filament voltage is provided by a buck and
> forward converter and a standard Würth push-pull transformer (6.3 Volt
> to 600mA).
> The MCU is an STM32G491, mainly chosen because of the integrated 12-bit
> DACs. The image composition is completely DMA controlled and works
> independently of the CPU, which therefore has a lot of time for other tasks.
> The characters and graphic shapes are encoded as SVG path.

Looks great.

Does the signal generation use integrators or is it direct 2 channel DAC
output to amplifiers and plates?

--Toby

> An ESP32 has been added to achieve the correct time according to the
> NTP. Operation is entirely via a web server running on the ESP32. The
> software is not finished, and not all functions work (work in progress).
> The next step is to try out other CRT tubes and make a housing.
>
> Regards,

> Paulscope_clock_DG7_32_v0_5.jpg

>
> --
> You received this message because you are subscribed to the Google
> Groups "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to neonixie-l+...@googlegroups.com

> <mailto:neonixie-l+...@googlegroups.com>.

> To view this discussion on the web, visit

> https://groups.google.com/d/msgid/neonixie-l/4e60fa43-a9b5-40b5-8381-5a968b45d480n%40googlegroups.com <https://groups.google.com/d/msgid/neonixie-l/4e60fa43-a9b5-40b5-8381-5a968b45d480n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Grahame Marsh]

Hi Paul

Your clock is lovely. Glad that I able to help in some way.

I used (9 years ago!) DXF format for my graphics and now reading about SVG it looks to be a superb format for scope clock fonts and graphics. I look forward to seeing the evolution of your clock!

Grahame

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.

To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/4e60fa43-a9b5-40b5-8381-5a968b45d480n%40googlegroups.com.

[paulvr]

I'm not sure what you mean by integrators. The DAC outputs are configured as unbuffered, to use the full range 0 tot 3,3 Volt (internal STM buffers are not rail to rail). So, the DAC output is connected to an opamp buffer and from there to a amplifing and summing opamp, and further to a transistor high voltage class A amplifier. Very straight forward.

Op dinsdag 27 februari 2024 om 08:51:14 UTC+1 schreef Toby Thain:

[Christian Riise Wagner]

Amazing work! I've been working on porting a fork of David Forbes' code to an STM32G474. Got a starfield and 3D cube showing, but got somewhat stuck. Now looking at designing a power supply.

Are you planning on sharing the software?

[paulvr]

This is a follow-up to my first presentation of my implementation of the scope clock. In response to the comments I decided to change the name to SVG clock.

Most of the time was spent designing and making the housing. Inspired by the round housing that David Forbes once made, I continued with a transparent tube as a starting point. A prerequisite is that all potentiometers are accessible from the front. See the result in the image. Fairly compact and worked out well in my opinion.

The hardware has gone through several iterations, mainly to make it suitable for a wider variety of tubes.

Of course, the software is never finished and the possibilities for adding functions are endless.

Regards,

Paul

Op maandag 26 februari 2024 om 10:12:05 UTC+1 schreef paulvr:

[martin martin]

Excellent work!

I made a similar Nixie clock in an acrylic tube also.

~

mcve...@gmail.com

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.

To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/1f525095-eda8-46d3-b10c-50907de1b301n%40googlegroups.com.