PEM Electrolyzer Dryer on Heltec ESP32 w/ Webpage

Discussions and images as necessary related to microcontroller or computer instrumentation and fusor control software and hardware.
Post Reply
User avatar
russssellcrow
Posts: 68
Joined: Fri Dec 27, 2019 12:40 am
Real name: Russell Crow
Contact:

PEM Electrolyzer Dryer on Heltec ESP32 w/ Webpage

Post by russssellcrow »

The Fusor Group at the Dallas Makerspace has repurposed a Heidelberg PrimeSetter air chiller dryer to become a PEM Electrolyzer Dryer for our Fusor.

We began by getting a PEM cell from a Horizon solar car toy. The instructions show you how the cell works, and claim it can really make the car go, we didn't test that part.
Horizon_PEM_Cell.jpg
We connected the PEM cell into the chiller dryer, added an extra stack of Peltiers between the Heat Sink and the Chiller block, and built a controller using Heltec's WiFi 32 Kit, (a 32 bit 200Mhz microcontroller). If you try this with an Arduino UNO, you might have a very hard time. The device looks like this, with the major components labelled, these are front and rear views.
PEM_Electrolyzer_Dryer_front.jpg
PEM_Electrolyzer_Dryer_rear.jpg
The Electronics Schematic is old school, because this is a one-off device. After debugging on a Protoboard, we made a working model which fits in a small Hamamatsu Laser container.
PEM_Electrolyzer_Dryer_Schematic.jpg
Heltec_PEM_breadboard.jpeg
After adding the WiFi "Network Name" and "password" to the code, compiling and uploading, the Heltec Wifi 32 Kit will reboot and attempt to attach to the Wifi Network. If it connects, you will see the IP Address which your WiFi Router has assigned to the Heltec. Then you can bring up the Web page which controlls the device in almost any browser (we use Chrome or Firefox) by entering the IP Address, i.e.
//192.168.137.159

Here is functional Block Diagram of PEM Electrolyzer Dryer, which shows the connections of WiFi to the Heltec, Heltec to the Power Electronics, Power Electronics to the Peltiers and PEM cell. Next we will include the Thermocouple for temperature feedback, and augment the Arduino code with a PID block, but the device you see is functioning as is.
PEM_Electronics_BlockDiagram.jpg
During operation, we proceed with the following steps:
1. Start the Fusor vacuum and let it pump down.
2. Fill the PEM with Heavy Water, allowing a little extra to come out of the H2 and O2 ports. These will become water walls to seperate the Deuterium Gas from the air in the lines, and to show us how fast the gasses are being made. Not to worry, the Filters will trap the few drops which make it through (which reminds me that I need to add a trap for the O2 line which vents in the Heat Sink fan!)
3. Turn on the Mass Flow Controller (another Heltec WiFi 32 Kit and controller we will post soon), and open the MFC to allow the vacuuming down of the line to the PEM Electrolyzer Dryer (there is a valve between the two to prevent the vacuum from emptying the PEM cell).
4. Close the MFC, and slowly open the PEM valve. You can watch that D2O bubble pick up speed if the valve is opened too fast.
5. Once the Fusor is <5 microns, we re-open the MFC to balance the Fusor chamber pressure to about 30 microns. The usual value is around 25% on the MFC Webpage.
6. Fire up the CO2 Laser Supply and strike a plasma.

The following file is the Arduino code, and an object file for the Heltec WiFi 32 Kit. If you have already converted your Arduino IDE to program the Heltec you won't need the .h file. If you haven't, then follow the procedure here:
https://heltec-automation-docs.readthed ... start.html
Fusor_Electrolyzer_Gas_Dryer_v51_shared.zip
(3.68 KiB) Downloaded 303 times
Attachments
PEM_Elec_Dry_webpage.jpeg
User avatar
Dennis P Brown
Posts: 3147
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: PEM Electrolyzer Dryer on Heltec ESP32 w/ Webpage

Post by Dennis P Brown »

Excellent post - very good explanations and photo's. I was curious how you deal with the air in those two large 'filter' system holders before adding the deuterium? Are they first vacuum evacuated? Also, where did you get that chiller cube or was that in the original device you salvaged?
User avatar
Rich Feldman
Posts: 1470
Joined: Mon Dec 21, 2009 6:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: PEM Electrolyzer Dryer on Heltec ESP32 w/ Webpage

Post by Rich Feldman »

Nice report.
I wonder about the effectiveness of drying gas by chilling/condensation.
Depends a lot on the amount of coldness.

Have never seen a quantitative measurement of moisture in H2 or D2 from PEM electrolyzers; we know there's a little.
Someone here reported neutron assays with and without some kind of dryer; was it conclusively different?

With traditional wet (bubbly) electrolysis, we can figure the gas comes out saturated with water vapor.
I did the exercise when filling balloons in summer of 2018.
At 20 degrees C, water is 17 g/m^3; about 10% by weight in deuterium.
At 0 degrees C, water is 5 g/m^3; 2.7% by weight in deuterium.

If there's less water than that to begin with, none will condense on the "dryer" surfaces.
All models are wrong; some models are useful. -- George Box
User avatar
russssellcrow
Posts: 68
Joined: Fri Dec 27, 2019 12:40 am
Real name: Russell Crow
Contact:

Re: PEM Electrolyzer Dryer on Heltec ESP32 w/ Webpage

Post by russssellcrow »

Thanks for the kudos, Dennis.

Yes, you are correct, we vacuum down the filters and Chiller cube (these parts are rated at 10 Bar, so vacuum doesn't bother them). I neglected to put the valve in the Functional Block diagram which isolates the PEM cell, good catch. After evacuation comes the tricky part, we close the Mass Flow Controller, then open the isolation valve to allow the PEM cell to bring the high side up to atmospheric pressure with generated Deuterium Gas. If the valve is opened too fast (which is very easy), it sucks Heavy Water down the line into the Trap Filter (25um). I've toyed with the idea of using an earlier Needle Valve in this position. We originally used this Needle Valve to regulate the chamber pressure, but upgraded to the Mass Flow Controller when Josh Melnick loaned it (and a Bayard Alpert gauge) to our project.

The Chiller Cube was original equipment in the Chiller/Dryer Unit which came out of a Heidelberg PrimeSetter. This was a Laser filmaker for the printing industry, and the Chiller/Dryer kept their air bearing Laser spinner from self-destructing (at 54,000 RPM) from corrosion, by cleaning the compressed air of dust and water. They also used magnetic levitating mirror spinners, like the Pfeiffer turbopumps, but they were never as fast or stable.

Here is a Heidelberg Laser spinner.
ProSetter_Spinner.jpg

Under the aluminum safety cap (right side) is a Wollaston Prism, which reflects the incoming Laser beam by 90' onto the surrounding drum where photosensitive material (film or plate) is exposed. Inside the all brass housing is an air bearing axle, with a mechanical clearance of about 0.00015 inches. If the axle or bearing surface corrodes, it's curtains for the Laser spinner (this explains why the air chiller/filter was so ellaborate). On the side is a Laser Diode, which reflects off the shaft and sends timing clock pulses to the imaging computer which feeds the picture data to the Laser at about 125MHz. This was pretty cool tech in the 1990s.
User avatar
Rich Feldman
Posts: 1470
Joined: Mon Dec 21, 2009 6:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: PEM Electrolyzer Dryer on Heltec ESP32 w/ Webpage

Post by Rich Feldman »

The rotating-mirror scanner unit is also very cool; thanks for sharing.
Reminds me of Gerber flatbed photoplotters used in the printed circuit industry when tooling involved big sheets of film. The Gerber file format for art files is still standard today.

My words about air-drying effectiveness were poorly chosen; sorry about that.
Of course the unit can keep dew point below 0 degrees, to protect metal parts against corrosion in an industrial setting that might generally be humid, without ever needing to replace or regenerate desiccant material.

I just meant, it might not make a difference if D2 from the PEM cell is already dryer than the Chiller Cube's limit.

[edit] Reading more of your posts, I see where you added another stage of Peltier cooling.
And reported recovery of 3% of your heavy water. Yay! I was surprised that so much water comes through the PEM.

[\edit]

You still get the benefit of filtration. If and when my own PEM cells get into service, I want to investigate the moisture by trying to frost up a cold finger that's progressively chilled by, say, ice - boiling freon or propane - dry ice - LN2.
Resized_20211207_125230.jpeg
All models are wrong; some models are useful. -- George Box
User avatar
russssellcrow
Posts: 68
Joined: Fri Dec 27, 2019 12:40 am
Real name: Russell Crow
Contact:

Re: PEM Electrolyzer Dryer on Heltec ESP32 w/ Webpage

Post by russssellcrow »

Hey Rich,

Yes PEM cells come with some surprises, but we probably need to get used to this technology. While Tesla, Ford, and GM rush headlong into Lithium powered cars, those clever Japanese at Toyota aren't letting go of the Hydrogen power cycle. They have over 6,000 H2 powered cars on the road, and if they develop safe Hydrogen sequestering in Graphitic Carbon Nitride cells (another subject for another thread), that number could balloon easily.

Back to PEM...the next step is to upgrade the PEM cell we are using. The D2 gas is coming out with impurities, we think because this is such a cheap cell with a bargain basement membrane. I see there are Plug Power parts for sale on Ebay, but their cells are generally outrageously priced.

As to how the PEM membrane has evolved, this looks like a pretty good wrap up in Wiki:

https://en.wikipedia.org/wiki/Proton-exchange_membrane
User avatar
Dennis P Brown
Posts: 3147
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: PEM Electrolyzer Dryer on Heltec ESP32 w/ Webpage

Post by Dennis P Brown »

This is really good work - keep up the posts!
Post Reply

Return to “Control and Instrumentation Hardware and Software”