Making My Own SWG

[crokett]

I’m wondering if it is realistic to make my own SWG. The basic process is electricity is sent across the cell to split the chlorine out of the salt. I’m wondering about buying a replacement cell and building my own electronics to run it. Aside from
It would be fun, I could power it with a Raspberry Pi and make it a smart SWG. Is this realistic? I’d need to know what voltage/amps the cell operates at, etc.

 

[Brad_C]

I can't see why not. I've been reverse engineering my Australian Monarch ESC16 unit recently as I accidentally bought an ESC24 cell a couple of years ago and I wanted to know the difference so I can upgrade the unit from a 16 to the 24 (same control board, different voltage transformer and a couple of minor component value changes).

The 16 drives the 24 cell, but now it's summer and the mains voltage dips during peak periods I can see the output level dropping as the transformer can't supply enough voltage to keep the cell at the target current level. It's taken me 2 years to figure out my mistake and why the low salt light keeps coming on when we have a hot summers day. It wasn't until I was checking the mains voltage for another reason that it clicked. I always wondered why my pool seemed to require so little chlorine and I had to back off the SWG control. It is pumping out 24g/hr rather than 16g/hr I based the calculations on.

On this unit the 16 uses an 8 part cell (9 plates) and the 24/36/48 use a 12 part cell (13 plates). The units put a full wave rectified DC onto the cell and controls the average cell current and polarity using phase control of the rectifiers (target current 4A for both the 16 & 24). If they are running wide open and can't get a high enough current they light up the "need more salt" light. In "winter mode" they reduce the cell current down to 85% of target.

As they use a centre tapped transformer and a single triac on each side, they can control the average current and cell polarity by altering where they fire on the mains phase. Very clever actually. I've not measured the voltage reversal time, but it's every couple of hours of run time.
Sanitiser output is controlled by % of run time, so 50% has the cell on for 30 minutes in the hour in small bursts.

Not rocket science by any means, but quite a clever implementation. So yeah, with the right values you can drive the cell from a controlled DC source.

Frankly I can't see how much smarter you can make it though. More complicated? Certainly.

 

[crokett]

By smarter I mean able to control with my smart phone, Alexa, etc. I have an Intex SWG. The old model if you cut the power to it, when you turned power back on it would simply start up and then run the programmed amount of time. So I could control it with a mechanical timer, or possibly a smart outlet. The new model doesn't work like that. if it loses power, when power comes back it won't start working until you press a button, after that it will run the programmed time, then remember the time you started it.

You'll have to explain some of your post to me. I'm a computer geek, not an electrical engineer.

Also, I have no idea if this would cost less than just buying one, but it would be more fun and perhaps an upgrade over the intex.

 

[ajw22]

You'll have to explain some of your post to me. I'm a computer geek, not an electrical engineer.

Building your own SWG will give you an education in electrical engineering which is outside the scope of this site. It is all about voltage and current control.

Just buy a higher quality SWG then the Intex that comes on automatically when power is restored. SWG's are set it and forget it. About the only adjustment is % runtime and that changes infrequently once you get your pool dialed in. There is little reason to control a SWG with a phone.

 

[Brad_C]

Building your own SWG will give you an education in electrical engineering which is outside the scope of this site.

It'll probably give some excellent lessons in how to most efficiently extract the magic smoke from semiconductors also. Good lessons smell bad.

 

[Donldson]

Extracting magic smoke is really easy, I think the trick is keeping it in

Really doing this would only be an exercise in seeing if you can do it. You will be unlikely to replicate the performance or ease of use of a higher priced model and will likely spend nearly as much in the process (not even factoring in the value of your time).

Consider it like purchasing a used 7700k system but delidding the processor, adding liquid metal, then phase change cooling and overclocking it to the absolute limit. It's a fun project and you can take some pride in the results, but once you factor in the cost of all of that you could have had a 9700k with an AIO and gotten as good or better results.

 

[kairus00]

If you're using a pre-made cell, it's not really making your own SWCG is it? The cell is the hard part to make considering they're typically titanium plates that are then coated in another metal.

I don't see anything that would be difficult about putting together your own power supply for the cell as long as you can get a power supply that allows you to adjust the voltage or current per the cell's operating specs. You also need to factor in a flow switch into your design otherwise you'll have a pretty dangerous situation on your hand should your pump turn off while your cell is still live.

 

[chiefwej]

You are going to have to buy the cell, fittings and flow switch. That represents well over half of the cost of a system. You are very likely going to spend WAY MORE than the price of the system to build your own control unit. Unless there is a feature that you must have that no SWG system provides, this makes little sense. Even then you would be better off finding an existing controller and modifying it.