Multi-channel electroluminescent driver

Description

This circuit uses one or two LT3468 ICs to generate a high voltage (eg 200V) from a 6-12VDC input. The high voltage is distributed by either an HV507 or HV513 to 8 or 64 channels, respectively. When driving small EL panels or sections of EL wire, I'd recommend using the output channels in pairs, so the HV513 can drive four EL segments, and the HV507 can drive 32. There are two firmware Arduino sketchs -- one for linear control, and the other for matrix displays. The circuit is controlled by a SAMD21 M0 processor that has the Adafruit bootloader. I'd recommend buying Trinket M0's and removing the chip from the board. The LEDs, current-limiting resistors, USB port, decoupling caps, and reset button are also from the Trinket.

The high voltage is controllable from firmware. One of the microcontroller's timers is configured as a one-shot, which triggers an interrupt, which measures the high voltage level via a divider, and decides to turn on or off the LT3468.

Details

Main benefits of this circuit over existing EL drivers:

1. No loud 3KHz audible whistle noise. The transformer(s) in this circuit run at >1MHz, and do not produce audible noise. There is a small amount of noise from the EL wire/panel itself, but it's much less noticeable than typical EL drivers.

2. Smooth 7-bit or 8-bit grayscale control of each EL segment at >60 frame/second.

3. When powering 32 EL segments from the 64 channels (or 4 segments from 8 channels), the waveform from each pair should be balanced (no DC bias). The firmware doesn't take special care to always produce a balanced pair, but it's close, so there could be some improvement, and the hardware fully supports it.

4. All SMD. Total circuit height is 2.5mm.

Discussions

thekarlmiller wrote 09/23/2020 at 16:11

@Ben Krasnow - I hope you get this to know that your videos are very helpful. My career passion has been software for the last 40 years, but I have been a maker my whole life and your videos bring complicated materials science almost within the grasp of the everyday maker. I don't know why you do your projects and videos, but you should know that they are very inspiring and make a difference in other's lives.

Thank you.

Are you sure? yes | no

Ben Krasnow wrote 10/11/2020 at 18:23

Thanks! Let me know if you ever have a suggestion for a video topic.

Are you sure? yes | no

hudak wrote 11/15/2019 at 22:17

Anyone else have issues finding a 6.3uF SMD capacitor in a 0603 package?

Are you sure? yes | no

hudak wrote 10/03/2019 at 18:50

Hi Ben, great work! I wanted to ask if you could clarify the display approach. Did you use a passive matrix utilizing a first layer of horizontal bars in the back plane material, covered by dielectric and lumicolor and then use vertical bars of PEDOT to complete the capacitor? Did you incorporate any additional backplane material to maintain contrast via a small bus bar?

Are you sure? yes | no

ad wrote 05/09/2019 at 08:46

Hi Ben,

I'm very interested in your project. Is it possible to order a complete assembled print? And if so, can you tell me the price and how to order?

Thanks in advance,

Ad van den Bergh

Are you sure? yes | no

rcalaf.cot wrote 03/11/2019 at 03:04

Hi Ben,

Thank you again, I finally got a board working and hooked into a lumilor panel but nothing happened. I was using your Array sketch but no panel would light up. They where not bigger than the ones in your video. i was using the pin 32 and 33 with pixelvalues[0] as panel one then moved to pin 9 and 10 pixelvalues[5].

Then we used a 12V power supplied and it might draw too much Amp that fried the board.

Any recommendations to power up the board? i was using 9V batt but must likely not powerful enough. Would be cool to understand the match between the pixelvalues and the pins to solder the panels.

Thank you so much for your inspiration!

RCC

Are you sure? yes | no

Ben Krasnow wrote 03/12/2019 at 00:12

6 AA batteries (9V) is fine. A single 9V battery is probably not enough. If it's not working, check the +3.3, +5, and high voltage rails to start. If those are all fine, check the high voltage outputs from the IC.

Are you sure? yes | no

rcalaf.cot wrote 03/12/2019 at 16:29

Thank you very much for your answer :) any hints about the ports and the pixel values? Did you test what is the largest area that you lit up with Lumilor panels? Any recommendation?

Thanks again!

Are you sure? yes | no

rcalaf.cot wrote 02/12/2019 at 07:12

Good morning Ben,

One question, about the tiny transformers i guess is a 1:1 conversion. so the ATB322515-0110 should work. Or should we use a different configuration?

Great work by the way!

RCC

Are you sure? yes | no

Ben Krasnow wrote 02/13/2019 at 04:38

Both ATB3225 models have a 1:10 turns ratio. The larger of the two can handle slightly more current. https://product.tdk.com/info/en/catalog/datasheets/e6813_atb3225.pdf

Are you sure? yes | no

Jeremy wrote 12/14/2018 at 22:35

This is awesome. Are you planning to sell these drivers?

Are you sure? yes | no

Bharbour wrote 12/03/2018 at 23:29

You mention using two drivers for each wire or panel. Are you running them in bridge mode with one driver for the center and one for the "drain" wire, or just paralleling them for more current?

Also, what frequency are you running the drivers at?

Thanks,

Bob

Are you sure? yes | no

Ben Krasnow wrote 12/07/2018 at 07:25

The two high voltage converters run in parallel, and charge the same storage capacitor. The switch frequency is determined by the transformer impedance, and is about 1MHz in this case. The output driver is configured to have a max square wave output of about 8 KHz, but is determined by the firmware settings.

Are you sure? yes | no

ashwani.saurabh52 wrote 11/23/2018 at 14:58

this is really nice i want to know that are two outputs connected in parallel given by the LT3468. how much current can be delivered by the particular output pin of shift register. How large panel can be connected to this output.

Are you sure? yes | no

Ben Krasnow wrote 12/07/2018 at 07:29

Yes, the two LT3468 circuits are wired in parallel for higher total power capability. I haven't tested the shift registers for max panel size (or EL wire length). The 8-channel driver has much higher output per channel than the 64-channel driver.

Are you sure? yes | no

Zoé wrote 11/22/2018 at 03:11

Ben, amazing work, and you're very inspiring! Thank you.

How can I find the solder paste dispenser you showed in your video?

Are you sure? yes | no

Ben Krasnow wrote 12/07/2018 at 07:23

I bought it used on eBay, Metcal model DX-250. You can get the consumables here: http://www.testequipmentdepot.com/metcal/fluid-dispensing-systems/consumables/index.htm

Are you sure? yes | no

Mike Szczys wrote 11/09/2018 at 15:52

Your talk at Supercon on electroless deposition of copper was excellent. You've done amazing work on there, and I had no idea how your EL glove (and this HV driver) might link into that. It's an understatement to say you left the audience slack-jawed. Kudos on the apple-keynote-like "One more thing" that took everyone by surprise. Thanks Ben for being a huge source of inspiration for what can be done with determination and study.

Video of the talk for anyone interested:

https://www.youtube.com/watch?v=Tx2B5hI4w1U

Are you sure? yes | no

Ben Krasnow wrote 12/07/2018 at 07:29

Thanks, Mike! Supercon is the best!

Are you sure? yes | no

Multi-channel electroluminescent driver

Description

Details

Discussions

Similar Projects

Making the Ultimate Class-E Musical Tesla Coil

Piernas's Fast ZVS Mazzilli Driver

Magic Eye Color Organ

Solid state DC speaker protection.

Multi-channel electroluminescent driver

Become a Hackaday.io member

Just one more thing

Description

Details

Enjoy this project?

Discussions

Become a Hackaday.io Member

Similar Projects

Making the Ultimate Class-E Musical Tesla Coil

Piernas's Fast ZVS Mazzilli Driver

Magic Eye Color Organ

Solid state DC speaker protection.

Does this project spark your interest?

Report project as inappropriate

Send message

Remove Member