Smoke, No Mirrors - I Blew Up My Electrolysis Project

Thread Starter

verticaLabrador

Joined Mar 4, 2019
5
Hi All,

I'm actually not looking for homework help, I'm actually 38, but I'm such an absolute n00b that I feel that is where this question belongs.

WP_20190303_17_18_23_Rich.jpg

I've built a large electrolysis tank to de-rust car parts. You basically use a DC supply and some washing soda crystals to blast away the rust from the cathode. In order to power it you need a 'decent' DC power supply for which I'm using an ATX power supply from an old Dell Optiplex 760. I have an unlimited supply of these from work (with managers permission of course) as we scrap them regularly.

However, the process is obviously very hard on the power supply in question as it's effectively running at 100% and they inevitably burn out. As a result, I asked an electronics friend to suggest a resistor to put in place to keep the PSU running at about 70% output. He recommended the below (10WRJ) however after running for only five or six hours it blew and the output dropped to 3V. It used to be a nice anodised copper colour now it's burnt silver! :) Essentially however it did it's job as the PSU is fine and still kicking out 12V.

Maybe someone here knows more about electronics and can suggest a better resistor or circuit or resistors running in parallel. Interestingly. While the ATX system is down, I've got a laptop PSU in instead (I'm sure it'll melt down soon) which is 18v/70w and it's having a way more aggressive impact on the de-rusting process. I get decent bubbles with the ATX PSU, but with the laptop supply it's like a deep fat frier! Obviously higher volts means more bubbles?

WP_20190303_22_58_46_Rich.jpg WP_20190303_22_59_02_Rich.jpg
 

Alec_t

Joined Sep 17, 2013
11,269
Welcome to AAC!
That 10W1RJ would probaly be still alive today if it had been mounted on a decent heatsink. That's what those two bolt-holes are for :).
 
Last edited:

jpanhalt

Joined Jan 18, 2008
9,354
Simple answer: Your current is too high for your equipment.

Your tank looks great. I am envious. Mine is much smaller. I use a 50W wirewound resistor to limit current. Like this, but oblong rather than round:
upload_2019-3-4_6-21-58.png

Got it at a surplus store for a reasonable price. One thing to consider is to use a lower concentration of electrolyte (in your case, sodium bicarbnonate). Removal takes longer, but I am not sure that is a disadvantage. Also, have you considered using phosphate? It leaves the steel somewhat phosphated as a protective coating.

As for heat sink, when I need it, I put the resistor in a non-conductive bucket of water.
 

Yaakov

Joined Jan 27, 2019
1,612
Welcome to AAC!
That 10WRJ would probaly be still alive today if it had been mounted on a decent heatsink. That's what those two bolt-holes are for :).
Yes, try again with a 20W mounted on a good heatsink with thermal paste. It will last much longer.

Why 20W? Because it won't cost much more, if any more, and it will provide more dissipation for a longer life.
 

Thread Starter

verticaLabrador

Joined Mar 4, 2019
5
Hi Alex!

Thanks for the response. I always wondered what that was for. Easy when you know how really. I wonder is there a heatsink I can find on the bazillions of Dell Optiplex's at my disposal... ;) :D Gotta laugh sometimes.

Few questions if I can:

1. I always thought that resistors were sort of 'passive', in that, they didn't heat up etc, but this one (especially! :D) obviously, eh, does quite nicely! :) Is there a difference between a resistor and a load generating resistor?

2. I'm wonder if effectively the resistor was using 10w and the de-rusting tank say 90w of a 100w output, is the PSU running at 100% anyway?

3. The poles running through the resistor seem a little on the light side for the amount of juice that's running through it. Would I be better off with a bunch of 'lighter' resistors running in parallel? Is there something beefier I can use?

4. Wonder why 18v 70w generates far more hydrogen/oxygen than 12v 180w?

Thanks by the way. Help is much appreciated. It's basically making a broken process work in a shed tucked away in Ireland belonging to a decent sys-admin who knows his IT stuff, and loves project cars, but really doesn't want to learn another discipline that seems more than witchcraft than technology (electronics! :D)
 

Thread Starter

verticaLabrador

Joined Mar 4, 2019
5
Yes, try again with a 20W mounted on a good heatsink with thermal paste. It will last much longer.

Why 20W? Because it won't cost much more, if any more, and it will provide more dissipation for a longer life.
Thank you! These things only cost cent's so no harm at all.
 

Yaakov

Joined Jan 27, 2019
1,612
Resistors are passive, literally all they do is heat up!

There is no such distinction.

You are reducing the voltage on the electrodes, not the load on the power supply.

Voltage is the key to electrolysis.
 

Thread Starter

verticaLabrador

Joined Mar 4, 2019
5
Simple answer: Your current is too high for your equipment.

Your tank looks great. I am envious. Mine is much smaller. I use a 50W wirewound resistor to limit current. Like this, but oblong rather than round:
View attachment 171535

Got it at a surplus store for a reasonable price. One thing to consider is to use a lower concentration of electrolyte (in your case, sodium bicarbnonate). Removal takes longer, but I am not sure that is a disadvantage. Also, have you considered using phosphate? It leaves the steel somewhat phosphated as a protective coating.

As for heat sink, when I need it, I put the resistor in a non-conductive bucket of water.
Very interesting. I used an old 1000L IBC tank I got for free from a friend, zipped the top off with an angle grinder, then created the cathode plates out of old scrapped roof sheeting. I'm going to switch to using central heating radiators soon for a more sci-fi/heavier gauge steel look.
 

Yaakov

Joined Jan 27, 2019
1,612
You are reducing the voltage on the electrodes, not the load on the power supply.
Let me restate this. The reduced voltage should result in reduced current since effectively reducing the load. The resistor is intended to reduce electrode voltage, though.

Sorry for the poor wording.
 

Thread Starter

verticaLabrador

Joined Mar 4, 2019
5
Resistors are passive, literally all they do is heat up!

There is no such distinction.

You are reducing the voltage on the electrodes, not the load on the power supply.

Voltage is the key to electrolysis.
Very true. But I'd love to reduce the load on the PSU. I tend to leave it running overnight or over a day or two gently bubbling away. Hence I hate when I blow a PSU as I don't know it's stopped until I need the component if you get me. I've heard that back in the day hotrodders swore by 24v battery chargers for this. Considering voltage seems to have such an impact I have considered bolting the ATX PSU to a DC/DC converter but maybe that's overkill and I just need to walk away from ATX and go more with 18v-24v transformers... but the ATX PSU's are free! :)
 

jpanhalt

Joined Jan 18, 2008
9,354
BTW, the watts your resistor needs to dissipate can be calculated as (I^2)*R. Anything more than about the resistor rating will get quite warm.
 

Yaakov

Joined Jan 27, 2019
1,612
You should make some actual current measurements to see what is happening.

You can get a cheap hall effect current meter that will also let you read voltage and monitor the setup, that probably has other diagnostic advantages as well.
 
Top