I've got an electric car for the kids that has only an On-Off switch for speed control. I was hoping to add a better speed controller than that. I would need to PWM 12V, about 10-15A. I was hoping to use a
555 to generate the pulses to the mosfet gates.
I saw this:
formatting link
I can appreciate the beefy freewheel diode and mosfets, but would I really need a FET driver? At what point do I need a FET driver? Would the output from a 555 suffice?
More current. FETs have capacitive gates; bigger FETs (or FETs in parallel) act like bigger capacitors. If you want them to switch fast (on *or* off), you have to charge them fast - this takes a HUGE inrush current, which a 555 normally can't provide.
More voltage swing. FETs need to be driven to their Vss levels to shut them off quickly, which for a P-FET might be higher than the voltage a 555 can put out.
Is it one of the small 'power wheels'? (that is the fisher-price name though there are other manufacturers.) With two small gel cel lead acid batteries?
If so you can add a simple speed control by putting 'about' a 1/2 ohm resistor is series with the battery. But I found that after a few trips the kids want to drive it 'full on' all the time.
Nope, got it at a flea market. Originally runs on 6V, but that didn't give a whole lot of power (step on the throttle, and nada). Put in a higher A-h 6V battery, and works a little better, but still not much power. Put in a 12V battery from a different project, and whoa, yee- haw! A bit too much power. Wires got hot, too. So I'm planning on re-doing the wiring, and adding a PWM control. Maybe a fan, but if I do that, would that just fan the flames...? :D
I'm still annoyed that the original 6V battery charger that came with it gave 8+ volts output across the battery terminals after charging overnight. I shouldn't have trusted them, but hey... I gave them the benefit of the doubt. Now I use a dedicated 6/12V adjustable Pb-acid battery charger.
Right now it's just sitting in the room, no one riding it, because the
6V battery (even the bigger one) doesn't give enough juice. A pity.
Nope, got it at a flea market. Originally runs on 6V, but that didn't give a whole lot of power (step on the throttle, and nada). Put in a higher A-h 6V battery, and works a little better, but still not much power. Put in a 12V battery from a different project, and whoa, yee- haw! A bit too much power. Wires got hot, too. So I'm planning on re-doing the wiring, and adding a PWM control. Maybe a fan, but if I do that, would that just fan the flames...? :D
I'm still annoyed that the original 6V battery charger that came with it gave 8+ volts output across the battery terminals after charging overnight. I shouldn't have trusted them, but hey... I gave them the benefit of the doubt. Now I use a dedicated 6/12V adjustable Pb-acid battery charger.
Right now it's just sitting in the room, no one riding it, because the
6V battery (even the bigger one) doesn't give enough juice. A pity.
Michael
Many years ago, before modern electronics, in a galaxy a lot like our own I set up a throttle that supplied voltage in 3 volt increments. That's when it was easy to tap a six volt battery. The kids liked it, it was like shifting.
Hmm... I wonder what will happen if I put the weak 6V and the stronger
6V batteries in series. Maybe that could work... otherwise all I have are just those two 6V batteries and two more 12V batteries, none of which are center-tappable.
Hmm... I wonder what will happen if I put the weak 6V and the stronger
6V batteries in series. Maybe that could work... otherwise all I have are just those two 6V batteries and two more 12V batteries, none of which are center-tappable.
Thanks,
Michael
Your setup would be like adding series resistance, well not like it, you would be. The old motorcycle batteries had the links between the cell exposed as lead bars. No problem to drill into the bar and add a contact point. Not so with modern sealed LA batteries.
A 555 can source or sink at 200ma. This is more than enough for driving the gate of a MOSFET at any reasonable frequency for this task. However, I would use the ebay link if you value the time used on this.
Hence the "in general". The driver the OP referred to could source or sink 14 amps, which may be needed in really big motor controls, or controls with really high PWM frequencies.
Assuming my math and assumptions aren't horribly wrong...
I would guess that you'd take the gate capacitance, voltage range, and switching time, and figure out how much current gives you that time factor.
Note that the longer a FET is in the "in-between" state, the more power it wastes, so a faster switching time (for the same PWM frequency) means more efficiency.
So, let's say you wanted to switch 12v across a 2700pF gate in 10 nS. I = C * V / T, or 12 * 2700E-12 / 10E-9 = 3.24 amps.
Picking a 20amp fet at random... 2360pF, 5V, 0.2amps means a switching time of 60nS. If you ran the 555 at 12v, that'd be 140nS, for a maximum PWM frequency of 3.5 MHz (more likely, much less, if you want it to have any efficiency).
I'm a hobbyist and haven't any experience with motor drivers, so keep this in mind. I'm just going from a basic grasp, is all. But the basic equation for any capacitor is like this:
I = C * dV/dt
A mosfet's gate is roughly just such a capacitor. In the above case, the "C" used is the capacitance of the mosfet gate, usually referred to source but not always (I'll get there, shortly.) The gate-source capacitance is usually found by looking around for "Ciss" on the mosfet data sheet. There is also something called the equivalent gate capacitance (Cei, I think) and that is not always found on the datasheet, directly, but computed from the total gate charge (Qg) required (sum of a few things like the gate-to-source charge [Qgs] plus the gate-to-drain miller charge [Qgd] and something a little extra beyond that for overdrive.) Cei = Qg/Vgs and it is usually a lot bigger than Ciss. However you decide to get it, you figure out a C for the above equation. dV is easy. It's just Vgs -- whatever voltage you plan to use on the gate to turn the mosfet on (assuming your off state voltage is 0V.) dt is the time over which that voltage is achieved from the off state voltage. Okay, so that may be a problem.. figuring out dt. If you know how fast the drive works, you can plug that in. But that capacitance affects the speed, too. So that makes for a point of confusion.
Better may be to just realize:
Qg = C * Vgs
You get Vgs for free -- it's whatever you are driving with. Qg you get by extracting the Qgs and Qgd from the datasheet plus some estimate of the overdrive part. The effective C falls out. But you really don't care about it. You are faced with Qg and a designed Vgs and let the pieces fall where they may. If you know your voltage source has an eqivalent series resistance to it of 100 ohms and a Qg/Vgs of about 5nF, for example, then you know that about 63% of the voltage will be reached after R*C time, or 500ns. Figure three of these time constants to get to 95%, or 1.50us. Fast enough for some things. Not fast enough for others.
From the datasheet for that irfp2907z on the web site you posted earlier:
formatting link
There is Figure 6 on page 4. Qgs is the first ramp of that curve. It's about 46nC. The next flatish part is Qgd. Say that Vds is
12-13V for the car battery, so use the Vds=15V departure... Qgd =
34nC, which is the spread from 46nC to the reading of about 80nC where the curve rises back up along that Vds=15V line. Now, the overdrive part is the rest. If your Vgs is 10V, then this is the remainder from about 80nC to 150nC, or 70nC. But Qg, the total of the three, is just the sum or else just that final reading of 150nC for the typical gate charge. Assuming Vgs really is 10V for now, that means 150nC/10V or
15nF for Cei.
You could just use that value with what you feel you know about the effective series resistance of the driver output and get an idea of the time it takes to drive the mosfet on, as roughly three times R*C. Or, if you know what current it can handle well, you can reverse that and compute Qg/I, or 150nC/I, to get the rough time, too. For example, someone else posted that the 555 timer's output can handle
200mA. Let's say that's right. Then the required time to turn the mosfet on is roughly on the order of 750ns... about 1 microsecond, or so. If that is acceptably fast, then the 555 as granted would seem to be okay. If you really wanted 25ns instead, though, then the 555 simply wouldn't be fast enough for the job -- you'd need at least 30 times more current or about 6A to get that job done.
Hmm, I'm not much of an expert, but I would think that you might end up frying the motors if you push to much current through them. They're probably designed to run with the series resistance of the battery to limit the current. (What's the 6V battery voltage when the it's putting out 10 or 15 amps?) But if you don't mind wasting the power a resistor in series with your 12V battery may do the trick. Unless you are just wanting to learn about PWM circuits and all.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.