Not really sure whether I should put this topic here or in Project Guidance but hey. So I am going to attempt to build a HHO Generator (Hydrogen, Hydrogen, Oxygen). It separates water molecules. Now a big issue with these is controlling current flow. Its easy enough to do with your electrolytes but if you live in a place where 20 degrees is a nice day you need to pump in a crap ton more electrolytes to keep your water from freezing. Rather than spend $80 on a PWM control unit I figure I can build one with an arduino. What would be a good mosfet to use for 11-15 volts at 0-30 amps. Preferably a mosfet that could handle more amps than that but up to 30 is a definite. On another note, is it possible to control the switching frequency on an arduino and if so, could the frequency be variably controlled using a POT? I haven't played with my arduino in some time so I have gotten rough around the edges with it.
You need to choose a mosfet that has a sufficiently low Rds(on) that the power dissipation at the maxuimum current (30A) is small enough for the mosfet to need more than a small heatsink. The Rds(on) figure you use must be quoted at no more than the gate drive voltage you will be using, and no less than the maximum current. The static power dissipation is given by:
P = I^2 * Rds(on)
Up to 1W you need no heatsink, and above about 5W you start needing a large heatsink. So you are looking for a mosfet with Rds(on) of 1 milliohm or less with no heatsink, or 5 milliohm or less and a modest heatsink.
You may find it easier to use two mosfets connected in parallel, each with its own gate drive series resistor. Mosfets of the same type and brand share current well, so each mosfet will carry close to 15A. To keep the power dissipation in each mosfet to no more than 1W when carrying 15A, the Rds(on) can then be as high as 4 millohms.
As well as the static power dissipation, there is also dynamic power dissipation due to the finite switching time of the mosfet. The dynamic power dissipation increases with frequency. When switching such high currents - and especially if you increase the PWM frequency - it is advisable to use a gate drive IC such as MCP1407, in order to ensure fast switching and low dynamic power dissipation. If you power the gate driver IC from the 11-15V supply (provided it will never exceed 18V), then this also gives you a wider choice of mosfets, because you can use mosfets that need 10V gate drive.
well if that mosfet can handle 120 amps, would only pushing it to 30amps, and rarely and for short bursts of time pushing it past that, require very big of a heat sink? Would attaching it to a car firewall under the hood make a sufficient heat sink? problem though, I need a 5 volt gate voltage to drive it with an arduino pin. Arduino runs a fairly high switching frequency and I am looking to probably lower that switching frequency which would also help with the mosfet heat management. Another thing, I am going to be wiring in a volt/ammeter on the negative side of my load so would i be better off with a p channel or n channel mosfet and if using the n channel, would i put that between my load and the ammeter or between the ammeter and negative ground? It would be more convenient for me to wire the mosfet to ground after the ammeter because then i would only need to run 1 wire through my firewall under my dash
wikitjuggla:
Would attaching it to a car firewall under the hood make a sufficient heat sink?
Just because it's metal, dosn't make it a good heatsink.
Different metals have different thermal gradients and I suspect that the metal used in the car's firewall will not make for a very efficient heatsink.
Having said that, I guess it's better than nothing.
The other thing you have to watch out for is what is the tab of the mosfet connected to or bolting it directly to the chassis will result in a very short lived fet/circuit or both.
Yet another thing to watch for (depending on your frequency of operation), is the extra capacitance you introduce between the tab of the fet and the chassis being used as a heatsink.
The tab of the fet will make up one plate of the capacitor, the heatsink the other and the insulator would be the dielectric.
wikitjuggla:
well if that mosfet can handle 120 amps, would only pushing it to 30amps, and rarely and for short bursts of time pushing it past that, require very big of a heat sink?
Ignore the 120A rating, focus on the power dissipation. For short bursts, you can rely on the thermal mass of the heatsink to some extent. You still need to make sure that you do not exceed the maximum junction temperature rating of the mosfet. If you want to use a single mosfet then you could use this http://uk.farnell.com/international-rectifier/irfb7434pbf/mosfet-n-ch-40v-195a-to-220ab/dp/2253785 but it is not logic-level so you would definitely need a gate driver chip.
wikitjuggla:
Would attaching it to a car firewall under the hood make a sufficient heat sink?
Steel does not make a very good heatsink. Also, I would expect the environment in the engine compartment to get quite hot anyway.
wikitjuggla:
problem though, I need a 5 volt gate voltage to drive it with an arduino pin.
I'm suggesting you use a gate driver chip, in view of the high crrent you will be switching. So not 5V.
wikitjuggla:
Arduino runs a fairly high switching frequency
I wouldn't call 490Hz high.
wikitjuggla:
and I am looking to probably lower that switching frequency which would also help with the mosfet heat management.
It will reduce the dynamic power dissipation.
wikitjuggla:
Another thing, I am going to be wiring in a volt/ammeter on the negative side of my load so would i be better off with a p channel or n channel mosfet and if using the n channel, would i put that between my load and the ammeter or between the ammeter and negative ground? It would be more convenient for me to wire the mosfet to ground after the ammeter because then i would only need to run 1 wire through my firewall under my dash
Wire the ammeter between the load and the mosfet drain. Preferably, use a Hall current sensor so that you only need bring thin wires through the firewall.
wikitjuggla:
.......So I am going to attempt to build a HHO Generator (Hydrogen, Hydrogen, Oxygen). It separates water molecules.
......On another note, is it possible to control the switching frequency on an arduino and if so, could the frequency be variably controlled using a POT? ......
Thank You
Hmmmmm, as soon as someone says "HHO", the pseudoscience alarm goes off.
The correct name is "an electrolysis cell", and yes, you can split water with it.
Sorry to rain on your parade, but none of those marvelous " perpetual motion" machines has ever delivered anything but economical loss in a controlled experiment.
You want to burn gasoline in an inefficient internal combustion engine (75%) loss, use some of the energy to generate electricity (10% loss) then put that electricity to work splitting water (50% loss), before you feed the electrolysis gas back into the inefficient combustion engine, expecting to gain energy.
It is a loosing game, and the only winners are the "visionaries" selling the expensive plans.
The OP said it would be used rarely and for short bursts of time. I was wondering whether he intended to use it only during braking, to recover some of the energy that would otherwise be dissipated as heat in the brake pads.
dc42:
The OP said it would be used rarely and for short bursts of time. I was wondering whether he intended to use it only during braking, to recover some of the energy that would otherwise be dissipated as heat in the brake pads.
That could be an interesting possibility, but I don't believe in that either for the following reasons:
braking power. The braking effect of the normal generator in the car is far too little
safety. If you conserved just a quarter of a cars kinetic energy as chemical energy in this way, you would have something really nasty! When people talk "HHO generators" they normally mix the hydrogen and oxygen right away, and that mixture is quite wonderfully explosive.
Safety II. Changes on the braking system is frowned upon in most of the world.
Peter_I:
Safety II. Changes on the braking system is frowned upon in most of the world.
You wouldn't need to change the braking system, it would be sufficient to detect current in the brake light circuit (which could be done non-invasively) and use that to bring the electrolysis cell on-line, providing a small amount of [additional] braking force via the alternator. But I too find the idea of storing a hydrogen/oxygen mix scary!
I've done a bit more reading about this, and it seems that (a) the purpose of adding the hydrogen/oxygen mix is to try to improve the combustion efficiency of the fuel/air mixture rather than add a significant amount of energy in its own right, and (b) to make it work efficiently you have to run the engine on a leaner mixture than normal, which involves hacking the signal from the oxygen sensor. I've no idea whether these claims can be justified.
I don't know anything about gate driver chips. And when I say reducing the frequency, I mean down to more like, say we are switching half on, half off, I want it on for something like a quarter of a second on, a quarter of a second off. And to stick to what i know, how about using several mosfets in parallel to allow me to use a 5v gate drive. Plus I already have an arduino plus several homemade arduinos. I don't have a gate driver. And as far as heat sink is concerned, How about just slapping an old school Pentium 2 processor heat sink on them. Those things are huge.
On another note, Can we please keep this discussion to mosfets and arduino. I have done my research on the HHO generators and have read a many forum about them. I don't mean to be rude, just trying to stay on topic and to each their own about the generators. And none of this has anything to do with my brakes. and i meant mosfet will be driving up to 30 amps continuous. I meant I will rarely (and when i do only in short, few minute bursts) drive it up and over 30 amps. And a mosfet with very little voltage drop would be best as well but idk if i will be able to have my cake and eat it too on that one
as far as the 10v gate voltage, what if i used a small step up transformer to bring the arduino pin up from 5 to 10 volts, but i'd prefer not doing it this way
wikitjuggla:
as far as the 10v gate voltage, what if i used a small step up transformer to bring the arduino pin up from 5 to 10 volts, but i'd prefer not doing it this way
Transformers only work using AC voltage, it cannot boost a 5vdc signal to 10vdc directly.
Yup, and that is a perfectly sound way to save some energy. It neither conflicts with thermodynamics, nor requires you to break any other well established principles regarding how the world is put together.
The efficiency is probably quite a bit higher too.
An even higher efficiency has been demonstrated using flywheels in city busses. They do at lot of stopping and starting, ideally suited for the flywheel solution.
wikitjuggla:
.............
On another note, Can we please keep this discussion to mosfets and arduino. I have done my research on the HHO generators and have read a many forum about them. I don't mean to be rude, just trying to stay on topic and to each their own about the generators. And none of this has anything to do with my brakes. and i meant mosfet will be driving up to 30 amps continuous. I meant I will rarely (and when i do only in short, few minute bursts) drive it up and over 30 amps. And a mosfet with very little voltage drop would be best as well but idk if i will be able to have my cake and eat it too on that one
On that note.
Yes, of course we can ignore that you believe in a fringe* science, and think you can beat thermodynamics.
At low voltage, MOSFETs can handle an impressive current!
*Edit: "Fringe" does not really apply to "HHO" in the sense of Yul Brown, Stan Meyer and Ruggero Santilli, "totally pseudoscientific fraud" is more correct.
