Programmable Solenoid Controller for Rectification Column

My university's chemical engineering department recently realized that a solenoid used to attract magnetic pins inside of its ancient glass rectification columns had shorted, and that it was re-installed improperly. The last person to attempt repair set it up to be a continuous-duty solenoid; but, to maximize the column and make it useful it has to be intermittent.

The specs for this solenoid are as follows:

Wiring: AWG 12 (Copper)
Amps: 18 A
Voltage: 12 V DC transformed from 120 V AC

Obviously, these configurations are too potent for the small Arduino board. We can setup a special frame from ABS fabrications, but we need to know how to program an Arduino to accept a timing sequence from the user to leave the solenoid on and then how long to leave it off and then to be able to change those inputs.

We know we'll need a relay, but we don't know what else we need. We don't have much programming knowledge for Arduino. We would probably have two transformers. One for the solenoid, which we would hook onto the relay board. The other would power the Arduino and knock the current and voltage down, so it wouldn't damage the chip.

Where do we begin? How can we make the input mechanism student/user friendly. Would I microprocessor yield better advantages here?

cjmccorm3:
Where do we begin? How can we make the input mechanism student/user friendly. Would I microprocessor yield better advantages here?

Start by fully explaining (and some photos will help, too) what it is you're talking about because based on your description I have no idea what it's supposed to be and do.

12V 18A is probably easiest to switch using a MOSFET or SSR. There are mechanical relays that can handle such current in DC but you'll have to look at automotive relays and other heavy duty ones. Switching an inductive load (solenoid) only makes it harder on the relay so make sure that thing has at the very least a good flyback diode installed.

Picture will take a bit, but I can better explain it.

Avoiding too much chemistry, the point of having this mechanism is that when you distill something and you want to keep the mixture coming out constant--you want to intermittently collect the chemicals coming out of the column. If you keep collecting continuously, then the mixture changes over time. Pulsing allows some liquid to drip back down into the still and then return more concentrated.

In order to accomplish this--the columns were designed with a small glass joint on the inside, which has at the end of the small glass arm a magnetic pin. We can attract these with neodynium magnets held in place, but it remains constant and non-intermittent. Prior to this a solenoid was in place which you could program with a box the manufacturer sold (no longer in business) that would pulse the arm pins as desired with the solenoid.

The solenoid was designed to pull about 15 - 18 A at 12 V DC current which was converted with a transformer from 120 VAC at 60 Hz. The wire gauge used when designing the last solenoid in place was too small causing over-heating, and since then the controller went missing. For many years the columns--quite good condition--were not touched. The last person wired the solenoid to a DC power adaptor for a laptop thinking it would be sufficient, but it did not pulse the magnet and it did not work properly with that circuit.

This system needs a controller. After more research, I believe an Arduino with a telenumeric keypad for entries and an LCD screen of 4-8 rows would be sufficient inputs and outputs to guide a script that would tell the relay how long to be switched on and how long to stay off while in use--especially since we can mount a GFCI plug in the box we would print to house everything. We can keep the Arduino on a separate adapter and place the transformer on the other one. However, the part we lack understanding of is finding the ideal relay and breadboard setup to successfully operate the circuit and to not kill the Arduino with back emf from the coil.

I've seen some forum posts on a solenoid valve that spoke of a bypass they used, but they were working with lower currents in DC. THe largest relay I've seen manufactured for Arduino is a 10 amp 5 V DC controlled relay for 120 - 240 AC power.

Lets get rid of the confusing solenoid term and use electromagnet. Do you still have the electromagnet? If so, can you measure the resistance of the coil?

Paul

cjmccorm3:
My university's chemical engineering department recently realized that a solenoid used to attract magnetic pins inside of its ancient glass rectification columns had shorted, and that it was re-installed improperly.

Do you have an Electrical and an Electronic Engineering department, even an IT department.
They should know about this bit of engineering, or how to service/fix it.
Amongst the staff of your University you must have some Arduino programmers.
Tom... :slight_smile:

However, the part we lack understanding of is finding the ideal relay

You appear to have skipped over wvmarle's suggestion to use a MOSFET. You said the coil is 18A at 12vdc. Logic level (controlled by 5v) power MOSFETS for 60v, 30 amps are quite common. For example

What are the on/off pulse cycles times? A "blink without delay" routine with different on/off periods should be good for the coil cycling.

The solenoid was disassembled and no one knows the resistance specs on the magnet. The metal felt like a cast iron rod and had a diameter of about 1.25 inches. They used 14 gauge or 16 gauge wire on something meant for more 12 gauge (where the heating came from we think); and, since then I've found several manufacturing manuals for the type of controller.

Most of the products they're selling now use a u-shaped electromagnet in small plastic box, and then they sell a programmable controller for the timing. These are a bit pricey though, and as it happens we do have an electrical engineering professor in our building. Since summer classes start back next week. He may be able to offer some insights.

However, since then we've seen about three ways to do it. One being to use a solid-state relay that activates on 5V and is meant for up to 60 V DC and rated up 25 Amps.

MOSFET pleases me too. I've reviewed a bit on those, but I think the problem there is that they require sharing ground. Even if we reverse back emf with a diode prior to coming back to the ground on the circuit...I think the higher current could fry the Arduino chip if they're sharing ground and it's 12 V instead of 5 V.

A resistor might step V and Amps down, but it still might not reduce it enough...and then it might cause a total reduction--which would limit the magnet field strength. That's a big question for me since I've never built such a thing, but there is also the option of using the relay to turn the power transformer on and off from the AC side. Then it's all isolated from the Arduino.

I suppose there's a few options. It seems we could build a magnet and use less current by bending two solenoid-type coils into a u-pattern and placing a metal plate across the two ends to direct the field.

cjmccorm3:
The solenoid was disassembled and no one knows the resistance specs on the magnet. The metal felt like a cast iron rod and had a diameter of about 1.25 inches. They used 14 gauge or 16 gauge wire on something meant for more 12 gauge (where the heating came from we think); and, since then I've found several manufacturing manuals for the type of controller.

So you mean the original wire has been replaced by a new one? That sounds like a problem to me, especially if a different type was used than original.

A way to reduce heating may be to start with a pulse, then reduce power using PWM. The moment the object has been pulled closer to the magnet it doesn't need as much strength to hold it close, than it needs to pull it from a distance.

MOSFET pleases me too. I've reviewed a bit on those, but I think the problem there is that they require sharing ground. Even if we reverse back emf with a diode prior to coming back to the ground on the circuit...I think the higher current could fry the Arduino chip if they're sharing ground and it's 12 V instead of 5 V.

Use an optocoupler and you can completely isolate the two. SSRs probably have it built in. Be aware that if you want to use PWM you have to keep the reaction time of the optocoupler in mind. It adds a delay.

Also you want to make sure you switch your MOSFET really really fast, as with this kind of current (especially during PWM) heat can build up really quickly if you don't switch fast enough: a partly open MOSFET has a high resistance. The specs you're looking at mean you'll have a high gate capacitance, which slows down the switching process. So you will need a gate driver as well.