For some time now I've been working on an approach to invention and problem solving that, in my house, has become dramatically known as "The Invention Engine". It's a formalized structure to problem solving, and so far it's been working very well. It's all in my head at the moment, but at some stage I'll put it out to the world in some form. I'm testing the process again by finding a solution to triggering a siphon with an ultra slow flow.
In some recent trials, it looks like there may be some advantages to a constant flood grow beds, rather than flood and drain. This relies on having highly oxygenated water, but you should have that anyway because that's what your fish like. Some plants prefer flood and drain, but a lot seem to enjoy constant flood.
I can't help but think that one potential downside to constant flood, is that there may be areas that could become stagnant. Flood and drain, on the other hand, pulls all the water out on each cycle, so there are no areas that miss out on turnover with new water. My fear is that it might be possible that constant flood could leave some areas to could go rotten over time.
I thought I might make a grow bed that was constant flood for most of an hour, drained once quickly, then have it repeat that pattern for ever.
Tricky.
Constant flood works by having a standpipe sticking up through the bottom of your grow bed at the height you desire the water to sit. You pump water in constantly, and it overflows into the fish tank via the standpipe.
So my solution would be to put a bell over the standpipe, but with an oversized standpipe that would be too big for the flow, so that it didn't trigger.
Now all we need to do is introduce a new, quick flow of water once every hour to trigger the siphon.
The Invention Engine first casually mentioned I should just put a second pump on a timer, and have it supply the hourly additional dump of water. I told it that a solution like that was too easy and expensive, and really needed a rethink.
I added the following parameters...
1. that it should require no additional pump
2. that it should require no electronics or timers
The Invention Engine suggested I divert a flow of water to a bucket sitting over the grow bed, put a second bell siphon in it, and have the water dump quickly to trigger the main siphon.
"OK" I said. But how do you make a siphon that takes an hour to trigger, but still delivers a rush of water to the grow bed. A siphon with a very slow inflow can only trigger if it had a very small standpipe. Any flow that took an hour to fill a bucket, would only trigger a tiny siphon, and only deliver a tiny amount of additional water to the grow bed. A tiny additional flow would mean some very accurate calibrations to make the grow bed's main siphon trigger. In fact it might not be possible, and would definitely not be reliable.
"Make it huge" The Invention Engine replied.
"I want it small" I insisted.
"Oh" said The Invention Engine, "A small bucket that takes an hour to fill, but then dumps quickly? That's a different question altogether".
No comments:
Post a Comment