> Hey all,
> I have a couple of questions for power switches which I believe Tim and I
> kind of touched on last meeting, but I'm still a little unclear about.
> The controller chips allow you to set a limit to the power dissipated in the
> N-channel FET, and also provides a couple equations to use in calculating
> this power limit. Tim, as we were calculating last meeting, the power
> dissipated under "normal" operating conditions is only about 0.17W, based on
> an Rds_on calculated to 6.75 mOhm, max current of 5A, and max junction temp
> of around 115C. However the equations provided to calculate power limit
> take into consideration max desired ambient, junction, and case temperatures
> whose values I am unsure of. In the old APS design notes, power limit to
> the FETs is set to 1/3 W which seems reasonable enough to my inexperienced
> mind, but may be wrong now that there are 8 switches instead of 4. Does
> anyone have any suggestions, or ideas as to a reasonable power limit?
The number of switches does have some bearing, but for now let's
do all the calculations independently of the number of switches.
The transistor we picked is pretty beefy compared to the job, so i would
suggest picking a moderate temperature limit for the transistor
junction, something in the 100-120 C range maybe, and see if you meet
the inrush requirement. At one point we mentioned an inrush requirement,
but i can't find it on the requirements page. I will fix that.
A reasonable inrush requirement is 100% for 100 ms.
There is an operating temperature requirement on the page, but it's not
clearly associated specifically with the aps. The stated requirement is
38 C. I'd like to bump that to 40 C just to get a rounder number. (I
will update the requirements page.)
I have something more to say about the requirements, but first i'll try
to address your specific question.
You don't need to know the case temperature ahead of time, you just need
to know the thermal model of the transistor from junction to ambient.
I'm looking at "STEP 3" in the tps2490 datasheet, The numerator in the
1st equation represents how many degrees to spare the transistor has at
a given ambient temperature and DC current.
Reading between the lines, they're saying that the case will eventually
come up to an equilibrium temperature determined by the quantity in
square brackets in the numerator. Now, what will happen if there is a
sudden transient overload? In the first few instants the case
temperature won't change even though the junction temperature will
increase considerably. This is because the "thermal mass" of the case is
large compared to the junction. So what they're ultimately calculating
here is how much extra dissipation the junction can safely take on after
the case has heated up under the continuous maximum load. They then take
70% of that and set there power limiting threshold to that.
Remember that (R_thetaJC) is the thermal resistance from junction to
case, typically in units of degrees Celsius per Watt. If you plug in
5 A, 40 C, and the thermal parameters from the transistor datasheet that
should be sufficient to calculate the power limit.
If you have more questions about this, feel free to keep asking :)
> I think I need a little clarification as to which chips need to be selected
> by tomorrow. My assumption is that all the chips we will need for the
> remainder of the project should be chosen, however this being the first time
> I've really done any chip shopping, finding the power switch chips has
> pretty much been my focus thus far. I think that finding the ideal diode
> thing for the umbilical should be pretty easy based on the relative
> simplicity of that particular circuit, but as for the pyro switches and
> igniters, I haven't had a chance to search at all. Should I be hustling to
> get these chosen as well at this point?
Ideally now we can be selecting chips at a fast pace since we've gotten
over the initial hump of selecting the first few.
Realistically, let's get the battery charger stuff selected, then think
about the pyro after Friday. I think Andrew will have more to say about
this at the Friday meeting.
David has looked at the battery charger issues some, and has some chip
ideas i think, so perhaps see what he's come up with and coordinate your
psas-avionics mailing list