I'm at wits end, so I'm finally giving up and coming to the forums.
I have a theoretical problem that I need solved (my circuit is quite a bit more complex),
I have a uC that get 5v through a linear regulator connected to a 6V PSU.
The PSU also supplies logic voltage to another uC (servo controller)
I have a servo arrangement that draws nearly 10 amps @ 6v
I am on a time limit, I have a very large collection of parts, but EVERY SINGLE TO-XXX package I have is a F*!@#NG N-Channel MOSFET. (My pick was either an IRF520 or 3205)
I, again, have no access to a P-Channel (no not the internet), but since all of the grounds are tied for communication, there seems to be absolutely no way to achieve my goal. That's a big problem.
Logic Voltage on the Servo Controller is separated by a jumper to VSS (servos)
I need both uC's to power up, but allow switching of the MOSFET which opens and closes the VSS connection.
Everything was working with a house relay, except that in certain situations, the relay would cause a brownout on the main uC (I don't want to use a relay for other reasons, please don't point out relay solutions!). If a MOSFET is truly a switch, then why the hell can't I seem to use it as one?
I don't care if I have to use 15 MOSFETS, there is a way to do what I am asking and I know it.
The N-channel fets that you have listed need 10volt gate/source to switch.
So unless you have 16volt available, you can't use them to switch a 6volt rail.
What current are we taking about. 10Amp, or less.
Which regulator are you using for the micros. Most regulators need more than a volt difference.
6volt to make 5volt is cuting it close.
Switching an extra circuit to an already powered circuit can cause a voltage dip if the second circuit has a supply cap that has to be charged. That can cause a micro to reset.
Solution could be a bigger cap across the supply of the primary circuit.
Leo..
The IRF 3205 have a Vgs to Amp table of 11 amps @ 5v.
(I've tested this and it does indeed work as spec'd)
My 10AMP margin is very forgiving as the circuits draw 2amp at standby and 4 on move. I've seen fluctuations up until 5 amps. (but I need to futureproof the control system)
The 3205 works properly without the communication grounds tied, but I really don't want to have to isolate the boards to get the setup to work.
The entire system is run off of an AC to DC circuit measured with a .01V ripple and a measured maximum output of 6V@7A. I understand that my supply is lacking, but it's just a test circuit, the real robot supply will be spec'd later.
The Logic of both boards are using an LDO (I haven't found the part number) with a voltage drop of .3 to .5 volts. (12-5.5V in ---> 5v out) However the servo pins are all hardlined to the main supply (why I chose the LDO and a 6v supply)
I have modified the circuit and included 2 paralleled 3300uf 10v caps. These provided a noticeable boost in performance, but the controller seems to still act finicky. I'm going to have to troubleshoot that later.