Plasma achieved!

Current images of fusor efforts, components, etc. Try to continuously update from your name, a current photo using edit function. Title post with your name once only. Change image and text as needed. See first posting for details.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

__UPDATE__

Since my last update, significant changes and improvements have been made to my fusor project.
Unfortunately, the TCP 121 I purchased on eBay for my TPH 062 proved persistently non-functional, even after a visit to Pfeiffer. Despite efforts to diagnose and repair it, the pump remains silent. Because of this, to avoid further delays and complications, I invested in a much simpler air-cooled diffusion pump from eBay. To efficiently control it electronically, I'm implementing a TRIAC dimmer with specific modifications. Additionally, I've ordered an NPT to KF adapter for the foreline port to ensure compatibility and a reliable seal. Once these components arrive, I anticipate a substantial improvement in the vacuum conditions, thus advancing the progress of my fusion experiments.
However, I have recently noticed one problem. The seller indicated that the diffusion pump's flange was an ISO63-K. Since, from the images, I could clearly see bored holes instead of an edge to be clamped, I thought it was a typo. But when I placed my throttling valve on the pump's flange, the holes did not align. Has anyone ever seen this type of flange, and how does it work?

ISO63 - F Throttling Valve
ISO63 - F Throttling Valve
Diffusion Pump
Diffusion Pump

Furthermore, I've encountered persistent challenges with the Edwards AGC. Initially, it experienced issues indicated by loud bangs and internal sparking. Investigation revealed a faulty connection on one of the 440V capacitors, leading to recurrent short circuits and arcing. I replaced the capacitor, ensuring better isolation and resolving the problem. Upon inspection, three components—an NTC Thermistor, a fuse, and a MOSFET—were evidently damaged due to a current surge. After replacing them, I still had no luck. So, I have decided to take another route and make my own pressure gauge controller. I have found 2 videos on YouTube that describe the general concept and how to approach the construction of one. I will use an RJ45 port connector to connect directly to the gauge and a potential divider paired with an ADC to convert the signal and send it to an Arduino. Everything will be powered by an LED Driver that outputs 16W, and I will use some step-down transformers to supply the correct voltage to the gauges. I will keep you updated on that adventure once all the components arrive.

PCB of the AGC near faulty capacitor
PCB of the AGC near faulty capacitor
Result of arcing
Result of arcing

Lastly (and probably the most exciting), I am enhancing the fusor's mobility by constructing a purpose-built trolley. I repurposed a robust metal garden trolley sourced from the hardware store for the main framework and wheels. The trolley's sturdy chassis and air-filled wheels form the foundation. I assembled a strong base using wood slats and planks to ensure stability and durability. The next phase involves constructing and assembling the top section, a critical component dedicated to securely holding and stabilizing the fusor during transportation.

Screenshot 2024-03-07 235649.png
Screenshot 2024-03-07 235641.png
Screenshot 2024-03-07 235615.png

While I've faced setbacks in my fusor project that are a bit disappointing, I do find solace in tackling the constant problem-solving challenges that arise. Despite these issues, I believe I'm getting close to achieving fusion. The next steps involve using a PEM cell to produce deuterium and testing for fusion using a neutron detection tube and the silver activation test. However, I'm still uncertain about the power supply and hope to find a suitable one at a nearby university or company.

I encourage everyone to share their opinions, as I'm genuinely interested in hearing different perspectives. Whether it's insights on dealing with faulty equipment or suggestions for other aspects of the project, I'm open to any input. Shortly, I'll be posting another update to document the DAQ system and the GUI I've developed for it.

Cesare M.
User avatar
Dennis P Brown
Posts: 3279
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Plasma achieved!

Post by Dennis P Brown »

Activation requires about 30 kV or so power supply. As for measuring neutrons, any good detector will do fine down to 20 kV. I have dealt with a flange like your DP but I made my own adapter from an aluminum circular plate. With a lathe it is a rather straight forward process. But I guess that isn't an option.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

__UPDATE__

Since the last update was posted, there have been some additions and changes to the setup.
First of all, thanks to Liam David's help, we successfully configured the TPH 062 to operate using the TCP 040 controller. Additionally, we acquired another controller, the TCP 121. However, despite our best efforts, it seems this unit is experiencing issues—potentially a faulty relay or IC—preventing the turbo pump from spinning up as expected.
Drawing on Dennis's recommendation, we contacted a nearby university workshop, where we were able to custom-machine an adapter for the diffusion pump. Seeing the potential benefits, I've decided to keep the diffusion pump handy as a backup or substitute for the TPH 062, ensuring our setup remains resilient despite potential failures down the line.

Moving on to the electronics aspect of the fusor, we've made significant strides. As previously mentioned, the pressure gauge controller met its demise some time back, prompting me to seek a replacement solution. After exploring various options, I ultimately constructed my own circuit for reading voltage and translating it into pressure readings. To proceed with this idea, I procured a matrix board from Amazon and sketched the circuit's layout (detailed schematics will be attached at the end of this paragraph).
The circuit begins with an LED driver to convert mains voltage to a stable 24V DC output. This output then enters a DC-DC buck converter to refine the voltage supplied, safeguard against surges, and offer flexibility in regulating the voltage supplied to the gauge. While the Edwards APG NW16 gauge currently in use accepts a voltage input ranging from 20V to 36V, the inclusion of the buck converter serves to future-proof the controller. It ensures compatibility with potential upgrades to the supply, accommodating higher voltage outputs or facilitating the use of different gauges that may require lower voltages.
Next, the voltage is channelled into the appropriate pins of an RJ45 port, with a voltage divider employed to lower the voltage from the output signal pins (3 for + and 5 for -) for compatibility with a 16-bit ADC. This setup allows the signal reading to be relayed to an Arduino, which then transmits the data to the graphical user interface (GUI). The voltage is converted to pressure in the Python file using a polynomial regression model. While sourcing an equation for this conversion proved challenging, I leveraged a table from one of the gauge's datasheets containing a modest number of data points correlating pressure to voltage. Employing the Numpy Python library, I conducted a polynomial regression with 80 terms, ensuring the model's accuracy.
The GUI utilizes this model to translate voltage readings into pressure values, displaying them on the interface. Further along in this post, I'll provide images and videos showcasing the GUI in action. As depicted in one of the visuals, I can remotely control the reactor using my tablet, thanks to a VNC service running on the Raspberry Pi and screen mirroring functionality.

Relay_PCB.pdf
Mosfet for Relay Control Schematic.
(125.4 KiB) Downloaded 75 times
FusorPressure.pdf
Voltage to Pressure Converter Schematic.
(166.21 KiB) Downloaded 76 times


Image of the Voltage Pressure Converter Board.
Image of the Voltage Pressure Converter Board.


Current state of the box containing all the electronics while showcasing the immaculate cable management.
Current state of the box containing all the electronics while showcasing the immaculate cable management.


An image showcasing how I can control the reactor from my tablet by mirroring the GUI.
An image showcasing how I can control the reactor from my tablet by mirroring the GUI.



This is an image taken during a run to test the pressure gauge. The meter and the plot both relay the pressure.
This is an image taken during a run to test the pressure gauge. The meter and the plot both relay the pressure.



Here's a snippet of the Graphic User Interface that controls the reactor (for now, once the data is recorded into the system, all the meters, except for the pressure indication, will be filled by placeholder values).
Here's a snippet of the Graphic User Interface that controls the reactor (for now, once the data is recorded into the system, all the meters, except for the pressure indication, will be filled by placeholder values).



The second electronics update is that I have implemented relays on the turbo and roughing pumps. These SSRs are, in turn, toggled using Mosfets powered to a 12V power supply and triggered by the Arduino. Each breadboard in the control box is connected to its own 12V supply using a simple adapter from Amazon. This ensures that all the components don't draw excessive currents from the adapters or the Arduino's 5V output (I also provided schematics for the Mosfet circuit). The respective switches on the GUI can turn the pumps on.



SSR mounted on top of the TCP 042 to toggle the turbo pump.
SSR mounted on top of the TCP 042 to toggle the turbo pump.

SSR mounted on the casing of the roughing pump.
SSR mounted on the casing of the roughing pump.



Lastly, a particularly intriguing development involves the automation of the needle valve's operation, achieved through a combination of a 3D-printed mount and the previously installed stepper motors. Using the slide bars at the bottom of the GUI and the "Set Valves" button, I can precisely control the stepper motor's rotation to adjust the needle valve's position, opening or closing it as needed. The mount securely holds the needle valve's knob in place, facilitated by screws.

WhatsApp Video 2024-04-17 at 18.56.33_0adfec3e.mp4
A video showcasing the recently installed mount on the needle valve with a stepper motor. This ensures complete remote capability of the reactor. (The awful background noise is the cooling fan of the electronics box)
(5.23 MiB) Downloaded 68 times


Unfortunately, a recent setback occurred when the 5kV power supply overheated and malfunctioned, rendering me unable to conduct tests on the system with the plasma in the reactor. Despite this setback, I remain optimistic about the progress made thus far and would love to hear your feedback. If anyone is interested in delving deeper into a specific aspect of the setup, I am more than willing to expand on additional details and insights.
User avatar
Dennis P Brown
Posts: 3279
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Plasma achieved!

Post by Dennis P Brown »

Impressive electronics build and overall progress. Professional build. Little doubt you will achieve success in the near future.
User avatar
Taylor Romain
Posts: 28
Joined: Mon Oct 03, 2022 5:20 pm
Real name: Taylor Romain

Re: Plasma achieved!

Post by Taylor Romain »

Good idea with the polynomial regression for determining the pressure for the APG-M gauge, I will try this out for my gauge in LabVIEW. Are you using back suppression diodes with the magnetron gauge?
User avatar
Richard Hull
Moderator
Posts: 15120
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Plasma achieved!

Post by Richard Hull »

I noticed the two Arduino Unos. They are so cheap that even for the simplest of functions, hand wired in yesteryear, why bother, just slap on a slave Uno and command it to do your bidding which can be updated or changed as needed in code.

Great work and good luck getting back to work with a repaired HV supply.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

Thank you for the compliment, Dennis! Really means a lot.

Taylor, unfortunately the Magnetron doesn't work. I tried connecting it with my pressure controller, but I got no reading nor a voltage. The same was also the case with the controller, which is now broken. However, I did install a suppression diode for the pirani gauge. Its the D4 diode on the FusorPressure schematic. I found the specific properties on one of the datasheets: 1A min. surge rating and 100V reverse voltage rating.
For anybody using an edwards magnetron gauge is it normal for it not to show a reading, even when the system is in a vacuum? Am I missing something?

I agree with you Richard, they are very easy to use. I had some issues however, with the serial communications and making a system to distinguish between the two ports. To solve this I implemented a "Check Connections" button on the GUI that not only tells you wether both Arduinos are responding correctly, but also makes the green LEDs blink in sequence for 5 seconds each, so then you know wether Arduino 1 is actually recognized in the python program as Arduino 1 and not 2. But apart from that it was a piece of cake.
User avatar
Liam David
Posts: 550
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: PPPL

Re: Plasma achieved!

Post by Liam David »

Nice work!

It's unfortunate that the gauge doesn't output a simple logarithmic or linear signal. I'm sure you've double checked the regression, but do be careful of over-fitting data with a polynomial that has too high of a degree. You want a degree that is less than the number of data points, and ideally one that is much smaller.

Is that gate valve a two-position or continuously-variable valve?
User avatar
Taylor Romain
Posts: 28
Joined: Mon Oct 03, 2022 5:20 pm
Real name: Taylor Romain

Re: Plasma achieved!

Post by Taylor Romain »

Cesare_Mencarini wrote: Fri Apr 19, 2024 8:01 am Taylor, unfortunately the Magnetron doesn't work. I tried connecting it with my pressure controller, but I got no reading nor a voltage.
I've had some issues with this gauge in the past. I'm using the Edwards AIM-S gauge. Once, when it was powered with 24V and working properly, the next use I found myself having to adjust my connections, specifically the enable pins even after it had been functioning as intended before. Without a suppression diode, the current can quickly surge to 1A when powering off the sensor causing damage to the gauge. Only power on when the pressure is below 10^-2 mbar. Luckily, no polynomial regression is needed as the manual provides an equation for calculating the pressure. I hope you can get your gauge working again. It's a good one, especially when measuring at deeper pressures beyond what the Pirani gauge can measure.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

Liam David wrote: Fri Apr 19, 2024 2:31 pm It's unfortunate that the gauge doesn't output a simple logarithmic or linear signal. I'm sure you've double checked the regression, but do be careful of over-fitting data with a polynomial that has too high of a degree. You want a degree that is less than the number of data points, and ideally one that is much smaller.
Yeah, it is quite a pain, especially since the raw data itself produces a nasty graph. I was able to produce a regression after taking ln of both the pressure and the voltage to get a nice curve.
Thanks for the suggestion about the degrees—I turned it down to 12 terms, and it seems to work just fine. I had it set to such a high degree because near the maximum (atmospheric) and minimum pressure, the model was somewhat deviating, and I couldn't get a good atmospheric pressure value. Now that I have refined the model, as soon as I get back, I'll adjust the atmosphere potentiometer on the gauge.
Just out of curiosity, why am I worried about overfitting? Shouldn't I aim to get the best line for interpolation, which is closest to the data points?
By the way, the gate valve is continuously variable and only locks when it's fully open.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

Taylor_Romain wrote: Fri Apr 19, 2024 5:34 pm I'm using the Edwards AIM-S gauge. Once, when it was powered with 24V and working properly, the next use I found myself having to adjust my connections, specifically the enable pins even after it had been functioning as intended before. Without a suppression diode, the current can quickly surge to 1A when powering off the sensor causing damage to the gauge. Only power on when the pressure is below 10^-2 mbar.
If I remember correctly, I'm using the same one. Do you know if there is any test to check whether it is working or not?
When I'm back after the weekend, I'll test the gauge again. Now that I have a working turbo, 10^-2 mbar should be easy to attain. I'm going to use the same board I use for the Pirani and by looking at the datasheets for both, they seem to have the same wiring.
What do you mean by 'adjust my connections, specifically the enable pins'? That might fix my issue.
User avatar
Liam David
Posts: 550
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: PPPL

Re: Plasma achieved!

Post by Liam David »

Over-fitting can occasionally cause local maxima and minima between the data points and it's just a general thing to look out for. I don't know exactly what data you're working with and doubted it was a problem in your case, but I figured I'd point it out anyway. I suppose that in your case, if 80 terms works better, then stick with that.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

Thank you for the input and explanation, Liam! I understand now. I'll keep an eye out for it when developing future models.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

Taylor, I tried running some tests using the AIM-X-NW25 but was unsuccessful. When I connected the magnetron to my board, the indicated pressure on the Pirani was 1.6 10e-3.
Another concern is that the Pirani's datasheet says that its voltage ranges from 2 to 10V; however, the lowest I have ever seen it go was 2.3V (about 1.6 10e-3). This worries me about a possible turbo-related issue since it is rated for final pressures much lower than that, and it's running at 66% speed (also tested at full speed, but it made little difference) without pushing the Pirani to its limits. I'll never know for sure until I get the magnetron to work.
On a positive note, I collected pressure data from the fusor over time and got very good results. I'll post some pictures of the plots. I'll be happy to hear any suggestions or comments.

Note that the x-axis represents the index of a particular data point, not the time.
First dataset: The top graph represents a linear regression of the pressure data after all the pumps were turned off and the gate valve closed. From what I can see, there don't appear to be many leaks in the system beyond the gate valve.<br />The bottom graph illustrates the natural logarithm of pressure over time.
First dataset: The top graph represents a linear regression of the pressure data after all the pumps were turned off and the gate valve closed. From what I can see, there don't appear to be many leaks in the system beyond the gate valve.
The bottom graph illustrates the natural logarithm of pressure over time.



Second dataset: Both graphs illustrate the pressure over time during the Pirani calibration process and the pumps' activations later on. It seems quite nominal; however, when the gauge has reached a reading at about 760 Torr, it keeps increasing in small increments, even after 10 minutes (the waiting time recommended in the manual).
Second dataset: Both graphs illustrate the pressure over time during the Pirani calibration process and the pumps' activations later on. It seems quite nominal; however, when the gauge has reached a reading at about 760 Torr, it keeps increasing in small increments, even after 10 minutes (the waiting time recommended in the manual).
User avatar
Taylor Romain
Posts: 28
Joined: Mon Oct 03, 2022 5:20 pm
Real name: Taylor Romain

Re: Plasma achieved!

Post by Taylor Romain »

Have you tried adjusting the potentiometer? Since this gauge is used for deep vacuums, it could be set to turn on at a certain pressure (voltage). Try measuring the set-point trip level and see if any information can be extracted to further the diagnostic.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

Thank you for the suggestion. I will try that next time I get the chance and tell you how it goes.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

I "fixed" the AIM gauge. I am ashamed to admit that the solution was right under my nose. I never noticed until now that there is an additional detail in the schematic regarding the gauge's wiring. In fact, probably to further protect the sensor from high pressure, a switch is placed between pins 7 and 2 labelled "enable gauge switch." I wonder what that does.
I attached a photo of the gauge's output (reading about 10e-5 torr). The data looks reasonable, and it responds to the valves being operated.
To test the gauge, I modified my current board by adding a toggle switch between the two pins. Eventually, I'll replace this with a MOSFET so that when the pressure is adequate, I can turn on the gauge automatically.


WhatsApp Image 2024-04-27 at 00.54.48_60ca3cc8.jpg
User avatar
Taylor Romain
Posts: 28
Joined: Mon Oct 03, 2022 5:20 pm
Real name: Taylor Romain

Re: Plasma achieved!

Post by Taylor Romain »

Aye! I'm happy for you. The AIM-X gauge is great beyond the fore line gauge (Pirani) for accurate measurements at deeper vacuums. The gauge also has a 'set point' trip voltage which in essence, the gauge would begin outputting data after reaching a certain voltage threshold. According to the manual... "The AIM Gauge has a set-point facility; the set-point signal is an open drain transistor (FET) output, which is on when the pressure measured by the Gauge is below a preset pressure."

Still, you may want to turn the gauge ON/OFF automatically through LabVIEW, in your case.

Manual in reference: https://www.idealvac.com/files/brochure ... _Gauge.pdf
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

Thanks for all the support, Taylor.
Since you mentioned the set-point trip, do you have an example of why you would use it? I understand how it works, but I'm curious about the rationale behind triggering the gauge to output a signal only after a certain pressure threshold is reached, as opposed to keeping the potentiometer set to maximum voltage and activating the sensor whenever the pressure falls below a certain level to avoid damage.
Please correct me if I'm wrong, but both your explanation and the manual suggest that the gauge is on regardless of the set point; it just outputs the signal below a specified pressure.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

I was hunting for power supplies on eBay when I stumbled across a cheap listing for a Shimadzu 30kv power supply. The only issue is that the seller does not know any of the specifications or have the datasheets. He could send me some additional images of the supply (attached below).
I already wrote a couple of emails to the manufacturer, but I have had no luck so far.
The part number is 211-63325-01, and the serial number is 630555-3. I've tried searching the manuals, but they are nowhere to be found. The company's website also does not mention any power supplies.
Has anybody ever seen something like this or would be able to determine what different ports do?
Attachments
s-l200.png
s-l200.png (42.98 KiB) Viewed 1547 times
2.png
1.png
User avatar
Dennis P Brown
Posts: 3279
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Plasma achieved!

Post by Dennis P Brown »

Can't help any on what those connectors do but I notice a red "+" sign in a circle near the HV output. You would do well to check the Chinese characters out for that symbol and see if that indicates if this is a "positive" output power supply. If so, it is useless as a fusor supply.

That 2 amp fuse for the unit (120 v?) does not seem like a lot of power for 30 kV. Say 1.8 amp max, (a bit over 200 watts) then after the x-former likely even less power, maybe? Seems marginal at best assuming its a negative output supply..
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

The '+' symbol is the company's logo. As far as I know, the only information given is that the power supply has negative polarity (mentioned on the listing).
User avatar
Dennis P Brown
Posts: 3279
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Plasma achieved!

Post by Dennis P Brown »

Well, the unit will do detectable fusion with a very good detector system. So if its returnable, and they claim it works, might be worth a try.
User avatar
Cesare_Mencarini
Posts: 30
Joined: Fri Nov 25, 2022 7:23 am
Real name: Cesare Mencarini

Re: Plasma achieved!

Post by Cesare_Mencarini »

That's good to hear. I'll let you know more once the supply arrives and I have a chance to test it out.
Post Reply

Return to “Images du Jour”