Hi again,
Sorry for bothering you again with my still existing optimissm to find a way pushing out more NOx percentage as already possible with the different
approaches I'v seen here so far. Pls. don't hit me too hard, but critics are wellcome!
Not so long ago I posted a flat design in the Main Birkland-Eyde Fred but in the mean time did a lot of reading in anchient papers of the 19xx and
came to another approach (suitable for small scale at home) which should allow the fastes possible cooling rate IMHO.
I attached an old school hand drawn picture which may amuse you (which is maybe a good start)
In that design I switched from the continous AC arc to an intermittently DC arc. Which is triggered always if the capacitor reaches the breakdown
voltage of the air gap. So I can set up the frequency by capacity and the power of my supply.
The idea is when the powerful cap-discharge takes place, the very high peak current will trigger the magentic valve and nearly at the same time a
short gas (Air + O2) burst blow out the hot gases in the pipe in the opposite of the arc chamber.
The preload gas pressure determines the amount of the gas burst of each action (also the spring load and nozzle diameter are critical for
dimensioning) I know that there are a lot of details needeg (electrode cooling e.t.c) which I did not draw at the moment. I only want to share it with
you and get you opinion if this way is worth to follow.
I hope that it is possibel to run this at least with 20Hz (or 20 cycles/s ignition blow out and ignition)
The stored energy in the cap is used for opening the valve and heating the air, so this has to be a fat one.
So for a personal max. design (fitting to my power supply) one discharge should deliver 50 Joule and if I have not miscalculated that means a cap. of
220nF charged at 10KV.
I have no idea how much the magnetiv valve will consume from this 50 Joule, but maybe one of you have a rough calculation.
I know that the moved masses should be as small as possible. I also do not want to use an iron core. (I hope I can drive it with eddy current effect)
So well now, feel free to attack
If I estimat an induction of about 30uH then the estimated flash duration will be roughly 100-200usec so -hm -ok the valve must be quite fast... I
have no idea how fast can a self tinkered magnetic gas valve like that in my ugly drawing be ? I guess 1-3 ms would be like hell, would it?
[Edited on 13-5-2020 by FranzAnton]
[Edited on 13-5-2020 by FranzAnton]Twospoons - 13-5-2020 at 14:52
As a reference point for your valve I use these in a product.
They take ~10ms to open. And they are fairly fast, as solenoid valves go.
Rather than using the discharge current to directly operate the valve, I would suggest you use it to trigger a separate driver circuit. Otherwise you
might face issues with arcing inside your valve coil, as when the arc fires a very large voltage (several kV by the look of it) will appear across
your valve coil.
[Edited on 13-5-2020 by Twospoons]FranzAnton - 14-5-2020 at 00:26
OK, thank you for that feedback. 10ms is a value which I may not be able to fall below with a self construction. I feared that the devil will reside
in detail.
The next thing I am not quite sure how to solve (in a simple way) is to clear the arc. Because the HV power supply is connected directly on the cap.
there will be too much energy delivered so that it is unlikely that after the cap has lost it's main charge the arc will stop. (Because the burning
voltage is much less that the ignition voltage) So it will keep burning.
To switch on/off the (self build) inverter on primary side is a lot of stress for my simple design and the (passive)snubber will eat a lot of
energy...
In best case the gas burst will blow out the arc, but if not I have no plan B at the moment.
Some detailed explanation about my valve (why I hoped that it may be fast)
Compared to yours the inductans is about 2 magnitudes less, because there is only one fat winding which will produce the magnetic field.
I estimate a peak current at least 1kA ore more. So the Voltage over this winding is not so high that I have to fear arcing. Because winding
resistance will be 1-2 mOhm.
I will also not use an iron core that has a great inertia. The pusher inside should be an aluminium tube (in which eddy currents are generated)
And these eddy currents generating a magentic field which push the tube against the magnetic fiels comming from my outside winding.
Sorry for that bad english explanation, but with that configuration I avoid the main parts making a real valve from a shop slower.
If you not completely against that "arguments" I think I should give it a try only to assemle this kind of valve to check if it will work that way.
The super madness would be if the gas burst can blow out the arc. I unfortunately have no idea and experiance how hard it is to blow out an arc like
that.... looking forward to further opinions...
The valve I'm using has a fixed core in the electromagnet, which drives a thin steel diaphragm that carries the valve seat. It also only moves about
0.01". So even though it is steel, it is still light and fast.
An aluminium diaphragm could probably be set up in a similar fashion, if you really want to go that way.
[Edited on 14-5-2020 by Twospoons]FranzAnton - 14-5-2020 at 02:23
ok right. I expressed myself missleading. the 30uH was a rough estimation for the whole assembly not only the inductance of the valve, because the
parallel LC resonance circuit sees also the wires from cap to valve from valve to arc and back to cap.
If it is much lower, I will be happy, so that would lead to a shorter arc time.
So you are right, my estimation of 30uH is really high. If it is less, my peak current will be also much higher.
But I am curious, if it is possible with such a short but high current burst to open a valve for a short time. I think I will spent some time in
workshop to figure out. I will post the outcomes for sure and some pics for amusement Twospoons - 14-5-2020 at 16:05
Just FYI, these pics are of the internals of the valve I use. Just thought it might provide inspiration.
WGTR - 14-5-2020 at 21:04
To add another idea, perhaps instead of controlling the valve electronically it would be possible to use a resonant brass reed to control the air
flow:
Even a rotary valve would be simpler - just a rotating shaft with a hole through it, driven by a motor. Synch the arc to the shaft with a hall sensor
or optical sensor.FranzAnton - 15-5-2020 at 00:01
Thank's a lot for your inspirations and fine pics! I will think over all that stuff and make a drawing of my first attempt which is aligned also to
the materials and machines I have on hand in my "workshop" chaos
As you all quickly realized the tricky thing is the syncronisation and that within the short time periods given in that design.
For now I do not investigate more about the arc clearance and hope that the gas burst is able to blow out the arc. wg48temp9 - 28-5-2020 at 11:50
Perhaps you need something like this 55cm arc from a 5kW half bridge inverter.
