Building a small scale KCLO3/04 Cell

[MrB]

Lots of cheap kit will produce full power BUT cannot handle the internal heat build up over time, then components or solder joints start to heat, stress and fail.

 

Doesn't really seam to be the case, since it died in 5 minutes, with an extra fan blowing on it.

@ 12A that just shouldn't happen.

 

I've run mine at the full 20A since we started talking about it back and fourth in this thread, using a PSU tester, with resistance load, just to burn of what goes in as heat.

 

I guess it boils down to what sort of LED PSU we are talking about. There is a bunch, so who knows.

[AustralianPyromaniac]

I don't know exactly what the issue with it was but it did get to hot to touch in just a few minutes befor deing. Also the supply didn't have its own cooling fan it was one I installed onto the housing. The fan was external but still well placed and blowing directly at the components that were hottest.

[WSM]

I don't know exactly what the issue with it was but it did get to hot to touch in just a few minutes befor deing. Also the supply didn't have its own cooling fan it was one I installed onto the housing. The fan was external but still well placed and blowing directly at the components that were hottest.

 

I'm wondering if your input power matches the requirements of the power supply? What is the basic power used in Australia? Here in the US it's 120 Vac 60 Hz; is it different there? What does the power supply require or need (check the nameplate data)? Also, is there an input voltage switch and is it set to the correct voltage?

 

If you have a DMM, check the exact voltage at your wall outlet. Is it fixed or does it vary over time (do the lights go brighter or dim sometimes)?

 

This may not help. They may just be inferior power supplies, but who knows? I'm just throwing ideas out there.

 

Good luck.

 

WSM

Edited June 15, 2017 by WSM

[AustralianPyromaniac]

Thanks for the ideas but I think it was just a bad supply for the job. The voltage switch was definitely set to the 240V side and it was advatised as an AU power supply and shipped from within Australia so there shouldent be any problems there. This is the link to the supply if anyone wants to take a look at it.

 

https://www.ebay.com.au/itm/182395628016

 

Judging from the feedback, people are having success with this product but I assume they're not making much chlorate so I think it's just what isn't being used for that's making it fail.

 

Hopefully will post some photos soon as the cell is going very well. On my first run I produced around 450g in 4 days but this run (currently running) I'm expecting more. Thanks again for all the help getting this project up and running!

[insutama]

There is a thread on here called making potassium per/chlorate that has loads of info but it is about 200+ pages long.

User WSMs blog also has alot of info.

 

This is where I get my electrode from for my cells. He sells the Ti and MMO anode and cathode together for $25 which is decent and they work well:

 

http://store.theamateurchemist.com/electrodes/

 

I have a perc cell that is 2.7 litres and I use a 5 volt 20 amp power supply with it but it requires a water bath to keep the temp down after it has been running a while.

Electrodes are 1 inch apart.

 

Chlorate cells are easy as you can see how far done they are because the product precipitates on the bottom of the cell.

2.5 litre cell with a 5 volt 20 amp supply and MMO anode and a Ti cathode and keep temp around 70 C using a water bath if needed and you should get heaps of chlorate.

I will try find my specs from my first chlorate cell if you would like

Good luck.

So with the little 2" X 3" MMO and Titanium anode/cathode from http://store.theamateurchemist.com/electrodes/ which i just purchased, I need a 20 amp 5 volt power supply ? or can i got smaller ?

 

If just making kclo3 with MMO and Titanium anode/cathode would they ever wear out how often would they need to be replaced ?

Edited July 7, 2017 by insutama

[WSM]

So with the little 2" X 3" MMO and Titanium anode/cathode from http://store.theamateurchemist.com/electrodes/ which i just purchased, I need a 20 amp 5 volt power supply ? or can i got smaller ?

 

If just making kclo3 with MMO and Titanium anode/cathode would they ever wear out how often would they need to be replaced ?

If the cell conditions are ideal, the MMO and titanium electrodes could last a very long time (potentially years).

 

To be on the safe side, figure one year, but see for yourself how it goes. Take good notes and report your results here, later.

 

WSM

 

Edit; with one cathode, I anticipate a maximum current draw of about 12 Amps (double that with two cathode plates, surrounding the anode). If the supply has a larger capacity, the size of the electrodes will limit the current draw (a larger power supply can't "force feed" more current through smaller electrodes. the electrode size will limit the amount of current consumed).

 

Good luck.

Edited July 7, 2017 by WSM

[insutama]

Thanks this is gonna be a side project as i have enough kclo3 and kclo4 i just think it would be nice to be able to make my own kclo3.

 

With one anode and cathode how mucb chlorarte would i be able to produce per day of running electrolysis and how pure would the chlorate be that i get from MMO and titanium?

[AustralianPyromaniac]

Thanks this is gonna be a side project as i have enough kclo3 and kclo4 i just think it would be nice to be able to make my own kclo3.

 

With one anode and cathode how mucb chlorarte would i be able to produce per day of running electrolysis and how pure would the chlorate be that i get from MMO and titanium?

The yield completely depends on the size of the cell. My cell is a 2.5L that runs at about 2L and has 2" x 6" Ti and MMO electrodes. In one run it produces arrond 1kg of chlorate from 800g of chloride but the theoretical yeild is higher. I don't think I'm running my cell long enough.

 

Assuming 100% conversion from Cl- to ClO3- and accounting for loss in recrystalisation you should theoretically get 625g of chlorate for every liter of solution your cell can hold. This is assuming a 400g/L KCl solution.

 

As far as the purity, It should be high enough for pyro uses as long as you do a few things. Use a good source of KCl, boil the solution well to remove hypochlorite and perform a slow recrystallisation of the mother liquid to keep all the remaining KCl in solution.

 

As far as a source of KCl, if your in the USA using water softener is a good source but if you can't find it, or don't in the USA Muriate of Potash is perfectly acceptable. I have been using in with no problems in my cell.

[insutama]

I live in canada but i actually have 1lb of lab grade potassium chloride i will use this for my first attempt then try and source some locally after that.

 

Using my smaller electrodes 2"x3" would 1 liter of chloride solution be ideal or what would be the ideal cell size for these smaller electrodes ?

[AustralianPyromaniac]

I live in canada but i actually have 1lb of lab grade potassium chloride i will use this for my first attempt then try and source some locally after that.

 

Using my smaller electrodes 2"x3" would 1 liter of chloride solution be ideal or what would be the ideal cell size for these smaller electrodes ?

Well at least if anything fails on your first run you will no it's not the chloride! Expensive though.

 

I am by no means an expert on these things but I believe you should be able to use a large cell like a 2L and reduce your electrode spacing to something like 3-4 cm. Certanly if you wanted to be on the safe side using a 1L would work.

 

The main issue with running a larger cell is the fact that the electrodes my not be able to keep it all hot enough. I would experiment with a 2L at 3.5cm if I was you but again, 1L is a safe bet.

 

Also, if it did get too cold you could insulate it with something. Not a very elegant solution though.

[insutama]

I dont need to make much chlorate its more of just a science experiment i find it intresting so ill probally start with 1l and maybe shoot for 5amps

[pyrojig]

I have had great success with pc power supplys . Not saying they are the best but are priced right for a cheap sys

[insutama]

What about a laptop power supply

[AustralianPyromaniac]

What about a laptop power supply

For the size your making it should work but why not use an ATX computer supply? It will last longer and can be used for any cell in the future. There also easy to find as junk in old computers.

[Baldor]

 

 

Firstly a warning, I purchased a 20A LED power supply and tried to use it, but within 5 minutes it overheated and died even with a cooling fan blowing on it. So my advice to anyone building a cell is, do not use and LED power supply. Other than that I had no problems and ended up just using an ATX power supply which is working splendidly without any overheating.

 

 

 

Sorry I'm late advising against this ebay power supplies. They say they are protected against overcurrent, but it's false. I burnt more than one playing with induction heaters, the safety thermostat trips when the power transistor have already died, and usually dies with the rest of the supply. They work well if you are sure you will not short circuit them or exceed the rated current, but with applications with varying current draws, are a liability.

 

I'm now using an adjustable 60V 20A for playing with induction. It's overkill for electrolysis but a nice tool to have., An ATX power supply with fast fuses should survive a lot of beating.

[WSM]

Sorry I'm late advising against this ebay power supplies. They say they are protected against overcurrent, but it's false. I burnt more than one playing with induction heaters, the safety thermostat trips when the power transistor have already died, and usually dies with the rest of the supply. They work well if you are sure you will not short circuit them or exceed the rated current, but with applications with varying current draws, are a liability.

I'm now using an adjustable 60V 20A for playing with induction. It's overkill for electrolysis but a nice tool to have., An ATX power supply with fast fuses should survive a lot of beating.

 

My best power supplies are those rated 0-? Volts and 0-? Amps, which usually have constant current capability (which is very convenient for tracking Amp-Hours and figuring the current efficiency of your cell). That said, I've used power supplies with less control successfully in most of my electrochemical experiments.

 

I also track the Voltage, Amperage and Temperature of my cells with inexpensive digital panel meters, purchased on eBay. If nothing else, tracking the current is the most useful measurement, I believe. Test the voltage where it enters the cell for the most accurate reading (avoiding the voltage drop through the leads).

 

WSM

Edited July 9, 2017 by WSM

[Baldor]

In V and A adjustables power supplies, you set the V, and limit the max A. The device will draw what current is needed at the fixed V. If max A is reached, the supply turns off, or V drops. It's perfect to protect both the supply and the device connected to it.

 

If I'm not mistaken, in electrolysis the important value to control is V (Correct me if I'm wrong), and current at a given V will be a function of type, number and size of electrodes, type of electrolyte, and distance between electrodes. (The configuration of the cell, or the equivalent resistance of the cell). It will change as the reaction progress. So if you know that you need 5V for a given reaction, you need to characterise the cell before choosing a power supply. For this, an adjustable power supply, or a beefy fixed one and an amp meter are the best choice. Once you have characterise it, you can use a cheaper and more adjusted to the cell needs one.

 

You may also need to limit the current to avoid overheating or damage the electrodes, but then Ohm law is your enemy. The cell have a fixed equivalent resistance, and to lower the current, you will need to drop the voltage.

[insutama]

how hard is it to wire up a ATX power supply also which would be a good one to buy ?

[MrB]

how hard is it to wire up a ATX power supply also which would be a good one to buy ?

 

Get a PSU with permanently attached cables. A modern PSU focuses most of its power on 12v, since "everything" in the PC is powered of this. Make sure your PSU actually has enough 5v Amps for what you want it to do.

Cut the 24 pin ATX connector of. Grab the green wire, and a "on/off" sticky switch,and connect it to one of the black wires, ground, this will be your powerswitch. (Green is power good, used as a sense connection, and when the motherboard connects it to ground, it sets the PSU to full power on mode.)

 

Grab all the red cables. Including those from "all the other" connectors. These are your 5v leeds. Generally speaking, it is ok to simply strip them all, and combine them to form a 5v connection, for the cell. There isn't many PSU's with split 5v rails, so that will be fine.

 

Grab all the yellow cables. These are your 12v rails. If your PSU is a single 12v rail design, it will have only one 12v watt rating on the label, and you can do as you did with the 5v leeds, grab any yellow cable in sight, strip the ends, and combine them for a 12v rail. If your PSU has a label showing a split watt / A rating for a number of 12v rails, you have to figure out which cable is connected to what rail. A PSU with 2 rails, and only a 4pin CPU powerconnector generally has rail one in the 24pin connector, and rail two in the 4 pin connector. Verify this with the diode tester on a multi-meter, and then test the yellow wires on all the other connectors, and pair them up. Two rail PSU's with a 8pin PSU connector typically has rail one on the first 2 pins, and rail two on the next two. Three and four rail PSU's are more expensive since these are KW units or more, but generally speaking the rails are split over the PCI-E power connectors, so multi-meter at hand, diode test, and pair all the cables together, and you should be fine.

Some, a lot, of PSU's state that they have mutliple 12v rails, but the diode test say they are all connected to the same rail. That is fine. (If you want to make really sure, measure the resistance between the cables, it should be in the 0-2ohm range.) For all intents and purposes we care about, these are actually single rail PSU's, they just have separate fuses limiting the max current, to keep the cables from melting. Combine them to a single 12v point. Try to keep the wires about the same length.

Do not, wire the 12v rails together, on a PSU that actually has multiple rails. (As in, the diod test says they are not connected.) They may very well be a touch out of sync in current, which may cause very weird issues, and fire hazards.

 

Grab the orange leeds next. 3.3v will probably not be that useful, but just in case, grab them, join them up, and save them for some future purpose. They are always on a single rail, so they will be fine.

 

Grab the black leeds. Since this is ground, there are no "rails" to speak of, bunch them up, and you are done.

 

(Purple, blue white and gray, is, for all intents and purposes here, unused. Purple is the 5v feed for standby, blue is a -12v line. Gray is "Power OK" which is a signal line between the motherboard and PSU to let them agree on that voltage is stabilized, and everything is peachy. White, is currently unused, 20/24pin PSU's may, or may not, have -5v on this.)

 

Next up. Actually using it. Connect the 5v line to a test load. A 50-100w 12v light-bulb, or a couple, if you have to, to get in to the ballpark of the load where you are going to be using the PSU. Measuring V and A, of course. Take note of what your voltage actually IS. It is not uncommon for PC PSU's to control the 12v rail, or rails, and just step down the other voltages X amount, so with no load on the 12v rail, you might not be getting the voltages you expect. If that is the case, a dummy load will be required. Use the aforementioned light-bulbs. If the voltage regulation isn't working propperly your PSU will burn it self out in short order, so test it, make sure the voltage is right, if not, put a dummy load on the 12v rail, and try again. Multi rail designs, "usually" feed the 5v subsystem from the first rail, which is always the one present in the 24 pin motherboard connector.

 

Now, i cant tell you how hard this is. That comes down to personal experience, and how much one has been tinkering before hand, also, how one intends to do the actual connections, and so on.

Read it, twice if you have to, and decide if it's to your abilities, and go for it from there.

 

Um... Good luck.

[Baldor]

Or you can buy a board like this. and forget cutting cables.

 

I have one made like MrB, it is easy, but the board is cheaper and have all the necessary electronics to start and stop the PSU. And you can use the remaining molex connectors if you like.

[MrB]

Or you can buy a board like this. and forget cutting cables.

 

Just don't go and try to pull 50A over something like that. You need "all" the cables of the PSU, if you are going to use "all" the current it can provide for a given voltage.

Sure, exaggerating slightly, but still. The motherboard barely uses any 5v juice at all on a modern PC, trying to feed anything but a tiny clorate cell from only the 24pin ATX connector might work, but i wouldn't go there. Nice to see that it is at least fused of externally, but it doesn't say with what rating.

Also, i had to check, but, they only use 1 pin for every voltage, other then 5v, which got 3 pins. Well, at least they DID put 3 pins on 5v, so they load a few of the cables.

Edited July 9, 2017 by MrB

[Baldor]

This is why I say you have all the molex connectors available. Use the board for controlling the PSU, connecting other accessories that draw less current, and draw the power for the cell from one or more of the molex. Don't forget to use fuses here also.

 

There are other boards, the original one, if I'm not mistaken, from Dangerous Prototypes.

[MrB]

Running the cell of just one molex, is just as bad, if not worse,then running it of the board. I mean, drawing 5A over one cable is fine. Drawing 50, and the insulation will act like hot-melt glue.

Depending on exactly what you are drawing, you need a different amount of the cables, but "al of them" makes sure you can draw what ever the PSU actually can output.

 

I've never fused them tho. The PSU is fused internally, most the time with automatic, resetting fuses, sometimes with burners. Ideally one should fuse them still, but...

Edited July 9, 2017 by MrB

[insutama]

would this power source work

http://www.ebay.com/itm/30V-10A-Adjustable-Switching-Regulated-DC-Power-Supply-LCD-Digital-Display-/201872678744?hash=item2f008c9f58:g:wloAAOSwXYtY3J1F

how does it compare to the ATX supply ?

 

sorry im not very electrical inclined

[insutama]

how about this one? http://www.ebay.ca/itm/ASTRON-RS-7A-7-AMP-POWER-SUPPLY-Works/292148306938?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D222007%26algo%3DSIM.MBE%26ao%3D1%26asc%3D20140620091118%26meid%3Dbfb4619fcf1d434c8da7b02a5229b4d7%26pid%3D100005%26rk%3D3%26rkt%3D6%26sd%3D182652210182

or does it have to be adjustable ?