Jimbo's Reef Controller Part 2

[Jimbo]

Formally wavemaker controller, but it is evolving

Just about have the prototype 6 channel triac board, each triac can switch on/off a 480w device. Easy enough to have 2 six channel boards for 12 channels. Issues with a triac board is that it almost requires a second power supply to isolate it from the microcontroller. But two small 5v wall transformers should not be a problem, very cheap. The triacs themselves are only a buck a piece plus opto isolators about $30 each. A manufactured circuit board should only run $10-$20 dollars. Am going to play with some of the timing regarding running pumps at different speeds this week.

Lighting is a interesting issue, some have multiple 250-400 watt bulbs. So the software will need be able to hook lights to multiple outlets, also this would support other on/off time for actinics. But even two or three outlets should be plenty, but a 3rd channel board is possible, there's enough ports. Not to mention refugium lights, but those are usally under 400 watts. I should have a circuit board designed by next week, probally a month before I have a sample, I'm out of money this month anyway LOL

I think more important stuff like PH, Ph, ORP and Temp control would take presidence over lighting that you can just hook to a timer. PH control after I played with the PH control and temp sensor.

I have a couple of dallas semiconductor ds1822 temp sensors on the way, should be easy to interface. Checked out radioshack today and they have project boxes at the cheapest price that have seen yet. $6-7 for large box.

[Jimbo]

I've been looking at reactance and it's association with "My Description of Soft Start" To me soft start is limiting the current that the pump gets when initially starting. Once started the alternating magnetic field creates reactance (resistance to the current) Let's take a closer look at reactance in a pump.

I have Zoo Med PS-40 pump and it is rated at 23w. So it's reactance (resistance) would be about 628 ohms when running. But if you measure the coil resistance it only measures 23ohms 3%. So during the initial start up on the first cycle the pump gets a max of 27x the current or 621 watts if we start the pump at zero phase 1/120 WOW!!!. Now it's easy to see why starting and stopping a pump often can depolarize it's magnet. In reality we don't have any control over what part of the phase we start the pump on a wavemaker so it's probally on getting 200 watts on average if my calculations are correct .

So how long do we leave the signal on under a sine wave from zero crossing to get 62.1 watts or 10% total current during startup? I'll show the calculations and the result in a couple of days, an interesting engineering problem.

[Jimbo]

Been playing with the circuit today. Although I was unable to get a variable speed out of the power head I can reduce startup chatter and current during initial startup. Hopefully this will prolong the life of the powerhead.

Although my zero ac crossover circuit is not neccessarily starting at zero crossover it does give a consistant measuring point. I've found that I can reduce almost all chatter with a delay of 6ms with a MJ 900 and Seio pumps. I tried a zoomed pump and it chatters no matter what setting I use, but less around 5ms .

Now it may be still possible to run the pumps at variable speeds but not using triacs. I think you need a ac cycle stretcher using a SCR and programmable pot may work. I get magnetic feedback into the circuit that can cause some problems, some clamping diodes maybe necessary on ports that the pumps are connected to. Only have a problem if I unplug the pump while the the microcontroller is running. Seems to kill the processor until reset.

I have the pumps turning on and off as slow as 1/4 second with decent results. If nothing else this should facilitate some lower water flow out of the pump.

If you would like a pump tested you can send it to me. I would like to get a turbelle stream for testing (the cheap model). I think the more expensive ones run off dc voltages, so it would probally be dangerious to play with those.

Anyway it's time to put this portion of the project away and work on interfacing the temp probe (ds1822) ic.