DIY Wavemaker

[Jimbo]

I finally have some free time and am going to use a microcontroller that I have laying around, actually have several.

The device will be based on a dontronics dt106 board and utilize a pic 16f876 processor. I will use the pic simulator IDE basic compiler to develope the program. Circuit board is available for just a few bucks

The first version will incorporate a couple of standard relays, no soft start feature. This wavemaker will control two devices

on/off cycling modes-
Alternating, one pump on, one off
Surge, both pumps on/off
Oscillating (several pulses per side) (this will not be programmable)

Timer on will be adjustable from 10-300 seconds (5 minutes)
Timer off will be adjustable from 10-300 seconds (5 minutes)
15 minute feed button
Reset button (reset to 10 second and alternating mode)
The unit will store settings in the processors eprom

LED's 3 for to display mode select
Blinking LED for timer on (faster means less and slower means faster)
Blinking LED for timer off
2 LED's to distingish pump on/off
Total of 7 LED's

Push switches.
1 for mode select
2 for updown on timer on select
2 for updown on timer off select
1 for reset
Total of 6 push buttons

Phase II controller will incorporate soft start features and speed adjustment thourgh pulse width use and a couple of solid state relays. This type of unit should be able to control tunze stream pumps.

Basically all the same features of above but will have speed select, additional LED and two more switches.

I could use a lcd but don't want to overcomplicate the unit before I even get started.

What features would you want in a wavemaker. I could put on a photocell to slow down the wavemaking at night. Do you think it should be a 2 or 4 channel device. The PWM unit will only control 2 devices because of limitations of the processor's PWM output features, although this can be overcome it would take a great deal of additional programming. But you could always plug in two powerheads. to one control outlet.

[Geoff]

maybe a random setting? or maybe a 12 hour tidal setting?

G~

[Jimbo]

Been doing some work on this thing all day and giving a bunch of thought to what I want, what others might want.

I've dumped the idea of just a relay driven unit, you can get them anywhere.

So it's the solid state relay unit that does PWM. Now the cool thing about this is that the powerheads will never completly shutdown, this should alleviate wear and tear and should function good with turbelle streams and Rio Seio pumps. Now the low output or effective off will need to be adjusted depending on the pump. For instance a pump may still spin with a 50% duty cycle, that means the the pulse is suppling 120volts to the pump 50% of the time effectivly running it half power, even so it may barely be turning, some pumps may require more like 60-70% duty cycle for the low setting. This will need to be settable via switches, user input. So the pumps connected to the unit need to be the same type or at least compatible with the low duty cycle setting.

I've decided to go ahead and use a lcd, it can display alot more information. In the long run it's just easier to display info rather than pulsing led's and allows easier integration of more functions.

Geoff can you tell me what you mean by tidal, I was not going to put a realtime clock on the unit but the circuit board has spot for one. The code to set the clock can be a bugger with the current basic compiler I'm using, don't think I want to add that code on the first prototype although I have a clock chip and can be wired in, for later integration.

Definitly a random function for pump on/off.

[Geoff]

not necessarily a real time tidal simulator. just something that would change occilate between banks of pumps with a short overlap time inbetween. maybe every six hours with hour or two overlap between banks.

i think this would be usefull for those with really long tanks, maybe not so usefull it the tank is less than 4'

Just curious- are you also on the diyspeakers list?

G~

[Jimbo]

Geoff I don't think that will be to much of a problem, I've decided to use a different microcontroller that I have, it has a much more sophisticated basic and issues with the basic pic IDE won't be an issue with this other unit (setting up the realtime clock). But of course cost goes up, but for me I already have the thing. But they are about $50 bucks with a realtime clock and a million io lines Well it's a 40 pin chip so more than we would ever need for such a project.

www.kronosrobotics.com (Dios controller)

So the hardware is basically all here and tested, I still need to order another solid state relay but have one to play with.

So here's what has been crossing my mind.

1. How about 12 programmable events that can be assigned to do things at certain times over a 24 hour period.
2. 12 programmable settings for the above events from the following
a Turbulence (both power heads turn on full blast for defined duration)
b Push-Pull (power heads alternating for define duration)
c Wave Turbulence, both power heads ramp up and down together over
Specified duration.
d Wave alternating, powerheads ramp up and down opposite of each
other
e all settings above should have max limits set for duty cycle, so if you don't want full power during a wave alternating session at night.

Other features as mentioned in previous post would be low setting of duty cycle for each powerhead, you will need to calibrate the unit for your system. A concern here is that if you set it too low you may have a problem with the power head quitting, from be clogged or such. If this happens it's possible it could overheat, but probally not an issue, but something I will be testing on some older units floating around.
I ordered two Seio 820 super flow pumps for my tank

I have not done much programming on the DIOS so it will take me several days to become familiar with it and another 3-6 weeks to get the above stuff working.

For those saying why not get a aquacontroller 2 is that you don't need a computer to program it and you can set speed limits, not just on/off. Cost of this could possibly be around $75 with all parts.

[Jimbo]

Downloaded and printed the updated Dios manual. Will be refreshing myself over the weekend.

I've been thinking of wave action in different parts of a reef
First we need to define areas of the reef.
1. Fore reef from coral exposure or down several meters, very turbulent at times.
2. Outer reef coral allways under water, low turbalance gental water movement to strong water movement, but tidal action
3. Lower reef and lagoonal areas probally same as two, but there might be some type of upwelling of the water in deep water areas. Need some divers to chime in here and tell us

What type of system do you have?
Sps tanks probally have type 1 or 2 most likely a mixture of the two.
Lps tanks are probally types 2 and 3
Softy's I'm not very educated in this area.

Although we might want to duplicate mostly a certain biotope, often we may want some time of very turbulent action to suspend deitris.

Something that has crossed my mind is that on the upper reef often from wave action corals are subjected to air bubbles. It would be easy to add a couple of relays to operated a electric valve to add air to the pump. What might the benefits of adding some air to the wave action at times. Well the air / water interface causes more friction and may aid in cleaning of the rock and corals. This may also benefit the skimmer in some ways. Does anyone know if direct contact of air bubbles to sps's some how benefits them in other ways? The downfall to this is spray that would be generated, but we could easily introduce air just a couple of times a day. Or maybe there's no reason to do this at all

We can also include in the programming wave alternating that we have stronger incoming wave generated by the program than outgoing. Incoming tide and opposite to simulate outgoing. It's interesting about incoming and outgoing, incoming would be more turbelent and outgoing is just a push of water in one direction, during a outgoing tide we still often have some wave action though and it would be good to quantify this.

Let's also take a look at turbulence at different times of incoming and outgoing tidal cycles. For instance is there a typical cycle of wave action, say 5 incoming rolling waves and then out going movement influx of water during incoming, is there a general rule of thumb to how much outgoing vs incoming water during tidal movement in either direction, how might this change at different times of tidal movement? Should we include a luner cycle to our tide movement? If I remember right at low tide and high tide there is hardly any water movement, how long does this period normally last. If someone can supply a chart of water movement and definitions that would help greatly in defining a water movement system that would duplicate nature. You might ask do we really need all that, probally not, but it would still be cool.

If your a diver give us your take on water movement on the reefs.

In closing..
I would like enough information to quantumize tides and water movement so we have the ability to really duplicate nature. Hopefully someone has already done this and can point us to a website that can give us charts as to water movement cycles during incoming and outgoing tide cycles. A chart of monthly and yearly tide movements might be beneficial also.

Thanks

[Jimbo]

I'm going to put the question of water movement over in the thinktank. We really need to break down types of water movements and quantify them and best ways we can implement them with the types of technology that we have at our disposal

[Jimbo]

Well hooked up the dios to my solid state relay and plugged a maxijet 900 into it. Unfortunatly I quickly found out that my relay is only rated for a on/off switching time of about 10hz, 10 times a second. This is not fast enough, can make the PH unit buzz and whir, but can't really control the speed.


So I started hunting on the net for AC motor speed controllers. Most of the stuff just confused me LOL. I'm pretty good with digital electronics, but analog and expecially inductive circuits give me a headache. I did find the circuit attached that might be useful.
You can see it's discription here
http://home.maine.rr.com/randylinscott/aug99.htm

The site shows the circuit but does not go into it's operation much. Looking at the circuit the bridge rectifier converts ac120 to dc120. I believe the the scr is getting turned on/off by the rest of the circuit, but am confused from there on. So if you can explain the rest of the circuit and supply a wave form of the output (is it a square wave) that would be great

Although I couldn't control the speed I did write a simple program to run the PH through a simple cycle.

'This program will operate the solid state relay from
'Port 4 that will supply power to the Powerhead
'Simple program to pulse the powerhead 100 times for a duration 
'1 second on and 1 off, then turn the PH on for 1 minute.
'Then off for one minute and start all over again
func main()
dim i as integer
output 4	 'set port 4 for output
low 4		 ' turn off port 4
mainloop:
for i=1 to 100
   high 4		'turn on powerhead
   pause 1000 ' wait one second
   low 4		 'turn off powerhead
   pause 1000 ' wait one second
next i		 ' do this 100 times
high 4	 
pause 60000	'wait one minute
low 4 
pause 60000
goto mainloop
endfunc

Thanks

[Jimbo]

Been studing AC motors all day. The 60hz sinewave generated by AC voltage determines the rotational speed of the unit. I could not find much data on rotional speed of most motors but the ones I did find run from 1700 - 3000 rpm a minute. This makes sense since it the 60hz sinewave that drives them.

So this now known, by slowing the sign wave we can slow the motor down.

But to build a programmable signwave generator is more of a hastle. The circuit above should do the job, but the 10k Potentiometer needs manual adjustment. Fortunetly there are microcontroller interfacable digital Potentiometers.

I contacted www.microchip.com and they where kind enough to send me a sample of there MCP42010, it has 2 programmable 10k MCP42010 devices in it. If it could not get any easier the Dios has a library function for the programmable potentiometer.

So tomorrow it's off to norvac electronics to pick up the rest of the components..

[1979camaro]

wow...ambitious project...not my bag, but i wish you tons of luck!

klingon eh?

[Jimbo]

My cat's name is klingon also, been working on the motor driver today, I'm hoping the motor will work off a square AC wave. I knew that old rio pump would come in handy.

[Jimbo]

I was doing some load testing and that old rio pump is drawing several hundred watts when it is supposed to draw 55 watts 1/2 amp. Measured the resistance, it's rio2500 @ 60 ohms, that would be about 2 amps and 240 watts, but when they get spinning the alternating magnetic field increases the resistance so amps should go down. If you have a rio2500 please check it's resistance.

Something I wonder is that if these cheap pumps where primarily designed for 50cycle power and they just relable them 60 cycle when they sell them in the states. If they where designed for 50 cycle they may not function as efficiently at 60 cycles.

I've decided not to use the above schematic, in place I will be converting 120ac to 120dc and using 4 power transisters to create an alternating current, but it will be a square wave, not a sine wave, but don't see this as much of a problem, but have not achieved that level of testing yet. The nice thing here is that we can use PWM on the 120dc, by adjusting the frequency and duty cycle of pwm I'd be we can get more efficiency out of the existing pumps. I think the square wave will utilize more power in the pump so some compensation with the pwm duty cycle may be necessary to bring power consumtion to the correct level. Hay we may find that 70hz and 80 duty cycle with this setup yealds greater efficiency. Anyway it will be fun to play with.

[Geoff]

ahh, yea, what Jimbo said. i feel i am pretty good at DIY, but jeesh, i am so lost. i never was really good at the electronics side of all of this.

very impressive so far !!

G~

[Jimbo]

Having some trouble with isolating the 120vdc that will drive the motor via the use of normal transisters. So I'm going to get some opto/isolaters to drive the power transisters, this will seperate the low voltage section from the high voltage section. It only took be 3 days to figure this out LOL. But have learned alot about transistors

[Jimbo]

Finally have the motor controller functioning, I only burned up 4 transisters in the process Guess that's not bad for a learning experience. One issue I have is the transisters I'm using are only rated for 80 volts, really need some rated for 120 or better. I can make any induction motor spin slow now

Some thing's I've learned, there's a difference between a sign wave and square wave in relation to the power the magnetic coil gets. You can probally run the motor at 60/80 volts with a a/c square wave. Although it would be hard to get 80 volts regulated, have looked around and you would need a transformer rated for 80 volts, they are not cheap about $20. So 120 it is... Once I get some more materials will post the schematic and some sample code for controlling the motors speed.

Triacs are really cool devices, they can switch just about any voltage, 120 upto 600 volts ac/dc, thier cheaper than relays and won't wear out. You could even run the power from your flourescent ballast and have the circuit turn on each light individually. I will have at least a couple of channels the controller that will use these devices.

Velleman makes a 4 channel triac board
http://www.gibsonteched.com/vk2634.html look here, probally be getting one of these to add to the conroller, just a wavemaker?, well don't see any reason not to use the extra i/o ports for something