Tesla Coil Show Controller Project

Mads Barnkob

I am building a Tesla coil show controller that can control up to 6 Tesla coils at once. It will feature individual interrupters, touch screen controlled MIDI interrupter, safety features to ensure a safe method of working with the Tesla coils before a show, under a show and after a show. It will all be housed in a transportable 19" rack mount flight case for ease of transportation and to protect the electronics.

The 6 channel MIDI interrupter is the feature rich Syntherrupter made by Max as he shares all the plans, documentation, code and schematics.

The users and readers that know about my Tesla coils will know that I "only" have a small Tesla coil, a medium Tesla coil and a large Tesla coil. So how does this correspond with building a show controller that can take up to 6 coils at once? Well, I just have to build some more Tesla coils :)

Due to time constraints due to things like life, work, education and more or less sleep, this project is split into some smaller phases.

Part 1 - The parts
Part 2 - Programming Syntherrupter
Part 3 - 6 outputs build
Part 4 - Interrupters
Part 5 - Syntherrupter and MIDI
Part 6 - Audio input panel
Part 7 - Safety panel
Part 8 - Power supply

Part 1 - The parts
I bought 19" transportation rack, panels, power bar and some other parts for the previously released video about flight case construction. I bought it from the German company Adam Hall that is a supplier to many smaller retailing companies of disco, event and show business equipment like audio, light and transportation gear.


https://www.youtube.com/watch?v=Ad2129sZv6w

The plan
layout rev 0.2.png

The plan is not set in stone, so feel free to chime in with suggestions and questions about the possibilities :)
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Max

Holy shit this is biig  :o lots of nice stuff you got there!

One thing I haven't really mentionned so far is that you should - if possible - not power the Nextion screen through the Tiva board. In my case it barely works but I can see the 5V dropping because of the current consumption. In addition, when reducing the backlight brightness I get a nice ripple on the 5V bus.
Your screen is bigger and will need even more current so you should definitely power it separately. Nextion recommends a 5V/1A supply for your screen (500mA for mine).
I'll add that info to the wiki, too.

Can't wait for the upcoming videos!
Max

GrantV

WOW, that looks like a very exciting project! Nice one :-)

The only input I have for now would be to put all the mains inputs/outputs on the back panel and keep them away from the top of the unit to keep things seperated and tidy.

I see that you have 1x 230VAC in and 7x 230VAC out. Are these outputs to power the 6x tesla coils? (If so, I'd be using a 32Amp 3Phase input and connector, in which case, I'd personally rather build a seperate power box and just run the controls for the relays etc. from the main controller to the power box using a DMX cable or similar).

Anyways, I'll be watching your progress with interest!

Mads Barnkob

Quote from: GrantV on January 28, 2021, 07:37:40 PM
WOW, that looks like a very exciting project! Nice one :-)

The only input I have for now would be to put all the mains inputs/outputs on the back panel and keep them away from the top of the unit to keep things seperated and tidy.

I see that you have 1x 230VAC in and 7x 230VAC out. Are these outputs to power the 6x tesla coils? (If so, I'd be using a 32Amp 3Phase input and connector, in which case, I'd personally rather build a seperate power box and just run the controls for the relays etc. from the main controller to the power box using a DMX cable or similar).

Anyways, I'll be watching your progress with interest!

Thanks :)

These outputs are only meant for controllers, chargers and power supplies for what will be at a "control desk". All power supplies for the coils themselves is handled by a separate power supply splitter and I will cover that in the part with safety. It is paramount to me that I can operate this completely isolated from mains, but also be able to run on mains as things like 7" touch screen and safety relays can be quite power hungry as Max states from his experiences with the smaller 5" screens.
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Mads Barnkob

In the part 2 video I go through features of the syntherrupter, firmware flashing, microcontroller / display programming and test of the Tiva development board and Nextion display.

Having the intelligence version of the Nextion display did result in some issues. For some reason Nextion decided that a more power version of their display should have a whole range of new properties on screen elements and not interpret a project file the same way that the basic / enhanced versions does... Some of the missing graphics was fixed by copying links to graphics to an additional configuration field called pic1, but the way nightmode is made, it was not possible to get the right graphics for that.

Max has since this video was recorded released a new Nextion display software version that will make it into part 4 :)


https://www.youtube.com/watch?v=jQZj8DPNvJg
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Max

I take that video as a huge compliment, thank you! Looks like the wiki did its job so far? Obviously, if you have any suggestions, please tell me  :)
Interesting that you found the default password on the User Settings page; it's also included in the Getting Started article. Guess you were too quick with playing around for that text  ;D Anyways, you found it and didn't have to ask.

The issue you show at 19:48, where the signal doesn't match the on screen values anymore is really weird. Ontime seems to be fine but the frequency is not. What really surprises me is that it stays at 700-ish BPS even when you turn the slider to lower values. I haven't experienced that behavior myself and honestly no idea what could cause it. When you mentioned it in the Syntherrupter thread I had a different picture in mind (videos say more than a thousand words). This won't be solved by the modified firmware I posted there either.
I'm inclined to say it's some quirk of the Intelligent series but this is wild speculation. Please tell me if the issue remains.

What exactly was that 1-3us delay you were measuring at the end? Was that from the transistor switching the transmitter?

I have to say, the UI runs a lot smoother on those Intelligent displays. Shouldn't be a surprise though, considering that they have a 200MHz CPU instead of a 48MHz one. This btw doesn't help with faster firmware updates  ;D that is limited by the Tiva microcontroller which has to pass the data from one serial port to the other one. And this only works reliable up to 512kbaud/s - no matter what display it's connected to.

Btw TJC is the OEM of the displays. There's zero doubt about that even though Nextion itself had made for years the wrong claim that there was absolutely no connection with TJC. Fact is the display modules look the same, have the same parts and the TFT files for Nextion and TJC differ in only 1 value: the model name (plus some checksums depending on that).


Kind regards,
Max


Mads Barnkob

Quote from: Max on February 12, 2021, 10:08:36 PM
I take that video as a huge compliment, thank you! Looks like the wiki did its job so far? Obviously, if you have any suggestions, please tell me  :)
Interesting that you found the default password on the User Settings page; it's also included in the Getting Started article. Guess you were too quick with playing around for that text  ;D Anyways, you found it and didn't have to ask.

The issue you show at 19:48, where the signal doesn't match the on screen values anymore is really weird. Ontime seems to be fine but the frequency is not. What really surprises me is that it stays at 700-ish BPS even when you turn the slider to lower values. I haven't experienced that behavior myself and honestly no idea what could cause it. When you mentioned it in the Syntherrupter thread I had a different picture in mind (videos say more than a thousand words). This won't be solved by the modified firmware I posted there either.
I'm inclined to say it's some quirk of the Intelligent series but this is wild speculation. Please tell me if the issue remains.

What exactly was that 1-3us delay you were measuring at the end? Was that from the transistor switching the transmitter?

I have to say, the UI runs a lot smoother on those Intelligent displays. Shouldn't be a surprise though, considering that they have a 200MHz CPU instead of a 48MHz one. This btw doesn't help with faster firmware updates  ;D that is limited by the Tiva microcontroller which has to pass the data from one serial port to the other one. And this only works reliable up to 512kbaud/s - no matter what display it's connected to.

Btw TJC is the OEM of the displays. There's zero doubt about that even though Nextion itself had made for years the wrong claim that there was absolutely no connection with TJC. Fact is the display modules look the same, have the same parts and the TFT files for Nextion and TJC differ in only 1 value: the model name (plus some checksums depending on that).


Kind regards,
Max

The wiki is very nicely done, you went over the edge in details, help and information there! Hardly even 0.1% of all projects have that level of documentation :)

I still your idea of update rates being a problem could be it, that the faster clock of the Intelligence series is at play. Because it happens more often, the faster you move the sliders. But I will look further into this when working on part 4.

The falling edge delay was measured at the optical receiver on the DRSSTC driver board. So that delay is rise time of transmitter and fall time of receiving circuits added together.
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Max

Here's the thing that makes the faster moving theory so unplausible: you have the apply mode set to ,,On Release" which means that the display only sends a command to the Tiva when you lift your finger off the screen. Unless you somehow manage to just barely touch it, causing many press/release events, the Tiva just gets one single command per movement - no matter how long and fast that movement is. I thought you had the apply mode set to ,,Immediately", which means that it sends new values continuously, potentially causing a buffer overflow. Hence my original fix was to increase the buffer size.

To make absolutely sure it is not the movement speed you could set the apply mode to ,,Manually" (btw you can change the mode by pressing the bottom right button that says ,,On Release" by default). Now you decide when the command is sent to the Tiva. If you can still make it happen, there's something really wrong. I have heard f.ex. that some people had serious issues with serial communications on nextion, where the systems print function just did not work. If this is indeed the case on the Intelligent series, I can't help you unfortunately. Also, since i have no real use for an (expensive) Intelligent display, I'll not buy one myself, which of course limits my debugging abilities.


Kind regards,
Max

Mads Barnkob

In this video I go through building a 6 optical output panel with switches for selecting different interrupters, indication LEDs and dry test speakers.


https://www.youtube.com/watch?v=DCFt82_SrNI

The panel features for each of the 6 channels:
- Optical output transmitter IF-E96E, 1mm plastic fiber
- Output LED indicator
- Switch to turn on/off optical output transmitter
- Switch to select between two different interrupters
- Switch to turn on/off dry test speaker
- Dry test speaker

Schematics are made in https://www.expresspcb.com/pcb-cad-software/ which is just an old habit of mine for creating real fast schematics and PCB layouts, when I etch the PCBs myself.

Everything on one PCB
DSC_4032 (Custom).JPG

Switches and wires, luckily I still have a lot of color-coded multi conductor cable that I cut up to use the single wires. Makes it very easy to work with large wire bundles.
DSC_4134 (Custom).JPG

Test of the finished panel
DSC_4163 (Custom).JPG

Schematic (PCB + files in bottom of post for download)
opto_transmitters sch rev 1.1.gif

https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Mads Barnkob

6 regular DRSSTC interrupters are build from the free and open designs from Dan / Profdc9 on highvoltageforum.net: https://highvoltageforum.net/index.php?topic=353.msg2320#msg2320

Drawings are made with the free draw.io software, which can be used both online and a downloadble dekstop application.

panel_layout_3u_interrupters.png


https://www.youtube.com/watch?v=pOEmQ9KFFWY

GS6A1434.jpg

I built 7 interrupters, one for each channel and a master interrupter, that by a switch, can take over any channel. Each channel can also switch between normal and burst mode. The interrupter also has a "sound" input option, which can take any kind of regular audio output and transform it into a on-time limited square-wave signal. This feature is first implemented in another part, as this panel could not fit anymore control inputs.

DSC_4236.jpg
DSC_4722.jpg
DSC_4724.jpg

Using the color coded wires, it was easy to keep track of the hundreds of wires. Each interrupter repeats the individual colors I used for the output panel channels, but also for the interrupter itself it uses identical colored wires for each interrupters control inputs. This way its easy to distinguish each channel, but also rule out any faults as all channels looks identical.

DSC_4728_.jpg

https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Max

Once you're done you got to let us know how many meters of wire you ran through this thing!

Regards,
Max

Mads Barnkob

#11
In this video I mount, update and test the Syntherrupter with its MIDI capabilities and gives a walk-through on how I am using it. Some of the details from the video is written as a part of this post.


https://www.youtube.com/watch?v=JB4-n_mJty4

First I had to upgrade the firmware, since this was about half a year since it was first done, Max had put out 6 new releases in the meantime, which also included support for the 7" Intelligence version of the Nextion display. So no more custom editing and changes to the display, I can just down the image from Max to it!

The first tests after upgrading the Syntherrupter firmware from https://github.com/MMMZZZZ/Syntherrupter/tree/v4.1.0 v4.1.0 to https://github.com/MMMZZZZ/Syntherrupter/tree/v4.2.0-beta.3 v4.2.0-beta.3, was to test the simple mode with coil settings with oscilloscope. It was confirmed to be 100% accurate with an example of 280 Hz at 60 uS on-time.

DSC_4994.JPG DSC_4995.JPG

For the PC MIDI Setup, I ran through the suggested programs in the Syntherrupter documentation: https://github.com/MMMZZZZ/Syntherrupter#pc-midi-setup . The free SynthFont 1 program gave me a lot of problems with virtual COM port locking up after each played MIDI, so I had to close the program and open it again. I then tried https://projectgus.github.io/hairless-midiserial/ Hairless MIDI<->Serial and https://www.tobias-erichsen.de/software/loopmidi.html loopMIDI together with my regular https://www.midieditor.org/ MidiEditor program. That worked flawlessly.

To get into the coil settings, its good to remember that buttons with additional settings behind, are highlighted with a double frame in the Syntherrupter UI. You access these deep menus with a long press, as defined in ms in the overall settings menu.

The cutout for the display was not great, I have unfortunately misaligned something, so its shifted 2 mm both left and down. But overall you do not really notice it and the layout of the panel is satisfactory.

DSC_5003.JPG
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Mads Barnkob

In this video I install and test the analog audio interrupter and gives a walk-through on how I am using it.


https://www.youtube.com/watch?v=EYUts1-Pc40

The analog interrupter audio input is quite simple and build up around a single LM339 op-amp. It has a toggle switch between normal BPS interrupter output and the op-amp output. The potentiometer is a frequency bandwidth limiter setting somewhere between allowing a maximum frequency from 100 up to 500 Hz. Overall on-time is still limited by the normal potentiometer for on-time adjustment.

GS6A2705.jpgGS6A2706.jpg

The amount of wires is getting out of control. I am however very happy about the color-coded wires, as its still easy to follow single wires back to its origin.

GS6A2707.jpg
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Mads Barnkob

In this video I install and test the safety features.

A lot of the original plan to have the safety relay in the show controller has changed and instead turned into a separate project, with building a soft start safety box that receives a special emergency signal from the controller.


https://www.youtube.com/watch?v=U-NVW-oHbA4

A encoded signal (666 Hz) is sent over fiber optical link to a soft start safety box (bandpass filter from 400 to 1100 Hz) where a industrial grade safety class circuit is made to cut off mains supply to the Tesla coils in case of a malfunction in the monitored circuit or a pressed emergency stop.

Schematics
safety_circuits ver1.5.jpg

PCB files for ExpressPCB program are attached as zip file
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Mads Barnkob

The show controller can be powered from 230VAC or from a USB power bank. I install a input, power supply, USB meter, power bank, switch and USB outputs.


https://www.youtube.com/watch?v=xJr3eVNY2RM
https://kaizerpowerelectronics.dk - Tesla coils, high voltage, pulse power, audio and general electronics
https://www.youtube.com/KaizerPowerElectronicsDk60/join - Please consider supporting the forum, websites and youtube channel!

Cameron Prince

#15
Hi @Mads,

Thanks for sharing this awesome project. Mine is almost complete.

I ran into some issues getting the receiver for the safety controller working and was going to have trouble packing that many components onto the boards in my setup. I went a different route with an LM567 tone decoder. The attached schematic also contains an LM555 timer which acts as a ~0.5 second on delay. In both the 74123 and 567 circuits, I noticed that the output was going high for a brief moment just as power is applied, regardless of input signal. The timer prevents this initial on pulse.

The 567 circuit as shown in the attachement is configured to accept ~650-750Hz so it works with your original transmitter, as-is. In my testing with the transmitter circuit, I was getting closer to 700Hz rather than 667Hz. If need be, R4 can be adjusted to alter the center frequency.

A 2N3906 may be needed on the 555 output, depending on the current needs of relays, etc. which are to be triggered.

Thanks again,
Cameron

567.pdf