Custom Korlia 12V controller

Hello all,
I think most folks agree that the 12V Korlias are great little pumps, but $200 for the contrller is a bit steep. Well I thought I might try to make a controller for them, and have have some initial success to share. Some of this is technical, but I'm hoping there's a few electron pushers out there who could give me some advice.

First, I started with the nano pump but most "features" should be consistent, except frequncies maybe.

The controller is a very simple and cheap microcontroller (picaxe) which drives the motor by switching two pins on and off (whenever the first in is the second if off) with a varying frequency. The signals from these two pins are sent to a standard IC for controlling motors (L293D) which then switches full power +/-12V square waves to the motor. All you have to do is feed it 12V from a good wall-wart, the whol2 thing costs under $20.

1) They do indeed run at 12V but this is not DC, nor traditional AC... using an h-bridge (+/- 12V square waves) the correct frequency is 500Hz.

2) The motors are reversible, you simply have to change whtehr you start with + or - 12V acroos the two wires leading to the motor.

3) They do not like hard starts/stops... best to come on slow and build up to full speed, but they don't seem run at less than 500Hz with the 12V supply. I hope the solution here is to vary the supply voltage with the drive frequency (probably through PWM at the enable pin of the driver IC).

4) Seem to require > 1amp to dirve at 12V.... my first driver IC (L293D) died on me due to surpassing it's current rating. Upgrade is planned with either using both channels on this chip side-by-side or will switch to L6205 IC. The picture attached is a vero board layout for this next prototype to show just how simple it is.

5) Here's how it works... I suspect the motor internals are a permanent magnet to bring the rotor to the correct stop position. Then only a single internal coil pair (a second is possible using a capacitor for phase-delay, but this would make varying frequency difficult). As the rotor turns from the stop possition to align with the field on the first half of the square wave, this sets direction, and inertia simply cares it through to the inverted second half of the wave (so long as it comes before friction kills velcoity and the internal permamnent magnet brings it back to the stop position)... which is why it only works very close to 500 Hz.

I'm now upgrading some components, and had some thoughts. Normally, if driven by a sign wave the voltage of an AC motor is usually expressed in RMS so a 12V rating would translate to ~17V peak-to-peak. Is that true here or would that fry my motors? Also knowing the relation between peak-to-peak voltage output and resonant frequency for the motor would be a HUGE help for figuring out a soft-start routine.

If anyone has experience with the factory controller, I'd love to hear about it, if only I could hook it to my occiloscope for 5 minutes all mysteries would be revealed.

The next steps once that is figured out, is a couple of knobs for setting average wave power/duration, and adding some code for a random number generator to vary things a bit around the average to keep it interesting for everyone in the tank. Hopefully when done, people will be able to use this info to make a cheap <$50 controller to run as many channels as they like, and they can program it however they want.

Anyother ideas/recommendations would appreciated.

 

Sorry if my spelling is horrible... I wrote this half asleep

 

I love that you are thinking of making your own controller for these. K+ for thinking outside the box!

I am trying to come up with a way to have controllable wavemakers in may tank as well.

The closest is the seio controller, that ramos both pumps up and down from 30% to 100%, but at the same time. if they could alternate, one at full, one at 30%, that would be perfect. The price is too steep at over 150 for that controller as well.

So, is there a way to make the controller not shut off all the way and ramp down to say 30% or so, then ramp up to 100% like the seio, but unlike the seio controller, alternate sides???


See the korallia controllable powerheads are not DC as some assume. They differ from the non controllable ones by simply having a different plug at the end of the cord, as far as I can tell.

I have a buddy that is a reefer as well that is an EE for rockwell automation/ allen bradley and he programs PLCs for a living. Maybe he can chime in if you get snagged, but your lingo has me convinced you can do it.


Heck, this EE I know even owns a seio controller and runs it on his maximodded maxijets. What if we can get him to take the seio controller apart and see if he can re engineer one and make his own?

 

Horkn - I'm glad someone is interested. When this is all figured out I'll publish schematics so that everyone can do this.

I'm hoping that 30% output is achievable... This qould be a key part of "soft-start". It would go something like reducing the drive frequency (how fast you switch between + and - ouput) to around 130Hz, and then turning the driver output on and off very quickly at say 10-20kHz. If the output were on only 30% of the time, the motors would "see" a lower voltage (~4V). The tricky bit here is that they're not likely to scale so conveniently, so I'll probably have to different duty cycles at each frequency untill I find the ones that work... Still do-able though.

Your buddy sounds like he would know exactly what to do.

I'll have to try the Seio model... I'm not familiar with those.

 

Well, we have the Koralia 4's 12v along with the controller and it works marvelously!!!!8)

 

Congratulations on your DIY controller. Thanks for sharing the steps.

 

InvertiGreat, here is the seio controller.

http://www.riopump.net/new_products_folder/ElecController.html
Note that it is made to work on their seio super flow pumps only, and not their seio prop pumps but I am sure it will work great on those as well sicne it works on MJ modded maxijets. I have a feeling the seio controller will work on any pump, but the prop pumps would be the only ones worth using a controller on.

Unless we can build a controller for under 50 bucks that is

 

The seio controller looks like a standard variable frequncy drive (much like what I'm trying to make for the Hydor pumps, but it looks like it works at 120V instead of 12V, which is why it's good for so many models... This kind of device is quite common, but the switching elements to work for 120V are more expensive, and as far as know, no easy solution exisits to add multiple channels expect stacking them, and linking the controls together... it would certainly be doable but I doubt it would be cheap. That's probably why they turn an on off together... it's easy to run many pumps of the same output becuase at 120V, currents are usually very small, hence-power strips.

The nice things about the hydor pumps is the fact that you can go all the way to stop, and even reverse, rather than just 30%... plus the fact that it's cheap and easy to add more channels since parts that work at 12V are ubiquitous. Also, since the current at 12V is 10x higher, this becomes your limiting factor, and it's actually better to make a seperate, independant drive elements for each channel. My bet is that since the one controller can be used from nano up to the 4's that your paying for a very over-engineered system if your running smaller pumps. which keeps the price high.

 

InvertiGreat, i have bben searching the web for ages and you seem to be the only one that has got anywhere with controlling these powerheads, have you got any further with this controller? also, do you think the koralia controller could be controlled via a pin out straight from a ardunio controller (5v PWM signal)?

 

ive been looking at the hydor wavemake 4 version (for christmas)....

I think its pretty damn cool that your willing to take on a challenging task on making your own....