I'm using MOSFETS to control a number of 12V automotive outputs ranging from 0.5 to 5A. They need to be high side switched. I'm looking for confirmation that I'm not doing anything silly in my component selection or circuit design..
The MOSFET i'm planning on is a IPD85P04P407ATMA2
U8 is an optoisolator.
Thank you.
That looks okey. Didn't check the MOSFET up...
It looks fine, but that's a lot of components for something that could be replaced by a single high-side switch IC.
You design a voltage divider for the gate voltage. Check that the ratio is appropriate for the required load current or desired Rdson.
For the linked MOSFET the effective Vgs of 5V looks typically sufficient for 80A, according to Diag. 7, but is not enough for the minimal Rdson at Vgs of 10V.
Consider the Rdson of 120-240 mOhm.
Did I miss a power dissipation spec for the chip? OP said they wanted to switch a max of 5A.
At 0.24 Ohm that sums up to about 6W heat per channel.
Sorry, I grabbed the wrong link for the datasheet, I corrected it in the original post.
I had originally been designing the voltage divider around a different MOSFET that had a lower (+/-8V) spec for Vgs.
The current MOSFET is +/- 20Vgs. If I understand correctly, I don't really need to worry about bringing the gate too low, correct? Even If I pulled the gate all the way to ground (like connecting to the opto with just a pullup resistor), and the vehicle was at full alternator output of the high 14V range, I wouldn't exceed the -20V spec.
Any reason not to just use a pullup instead of a divider?
Thanks, I was not aware of that product. That may be perfect as it appears to have all of the automotive protections (load dump, etc.) built in.
Are you saying that 6W is good or bad?
Also, in the datasheet is this table:
Not much, but the additional voltage swing would make it switch slower.
Also, if I'm understanding correctly, with the BTS716GB I can get rid of the flyback diode as that protection is built into the IC, correct?
Two of the channels power the 12V coil of a contactor with a 24 ohm coil (500mA)
Diagram post#1.
A voltage spike of 45volt on the supply, which is not uncommon in a car, could kill the mosfet.
If you don't need opto isolation, then circuit#4 on this page could be safer.
It uses 13mA current drive, which limits gate voltage to 13volt across the 1k resistor.
Gate drive is 7.5volt with a 12volt supply, which is more than your current circuit.
Needed, because the fet is not logic level.
The zener is there for extra security.
Leo..
I posted that after looking at the summary on Mouser. Digging a bit deeper, I see that the 5.3A is a max with all channels operating at maximum junction temperature. Individual channels are rated 2.6A unless you parallel multiple channels.
The takeaway here is (a) you should be able to find a single device that encapsulates your entire circuit (except the optical isolation) and (b) this series of devices probably has one that meets your needs, but this specific one might not be it.
It's a lot of heat to get rid of. At least a heatsink is required.
Right. The 6W mentioned before are for pulsed operation only, reduced to an average of 1.9W. The allowed 1.9W will mean a total permanent current of about 2.8A on 0.24 Ohm.
P=I²R, I=sqrt(P/R)
I've redesigned some load components, now the largest load is fuel pump at 3A startup, 1.5A continuous. The next largest is a air valve at 1.2A for 2 seconds, the remainder are less than 800mA.
I'm going to use this high side switch module.
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.
