Just thought I'd share this info if it helps anyone.
Power supply used is the Pi 5v 2.5amp. No other peripherals were attached to the Pi 3.
I had an issue where the MightyHat would reset when attempting to power-up the Pi 3 using the onboard On/Off button. The fix was to add a filter cap to the input of the mosfet switch. The value I used was 1000uf 6.3V. I did try 100uF, but this wasn't large enough and I only had these two values to hand.
There's a market outline on the MightyHat board(R3) underside for a cap, but it won't accommodate a cap of this physical size. A lower value should suffice though if needed.
Outcome, Pi 3 and MightyHat working great!!
Hi probson - thanks for sharing your solution and findings!
The part I'm not really following is this:
Quote from: probson on November 27, 2016, 05:43:38 PM
I had an issue where the MightyHat would reset when attempting to power-up the Pi 3 using the onboard On/Off button.
Could you possibly give more detail, how would it reset, after what interval?
Hi Felix,
Hope this info is more useful.
On pressing the On/Off button on the Hat the LCD display and back-light would extinguish, as if you had removed power, and the then reinitialize as it does on first power-up. This happened every time I pressed the On/Off button.
The Hat worked okay when removed from the Pi 3.
With the Hat plugged back into the Pi 3, I measured the 5v line to the input side of the power mosfet and didn't observe any dip when pressing the On/Off button. This was using a DMM.
As a further test I shorted out the '5v on-link' on the Hat. When I applied power to the Hat, both the Hat and Pi 3 powered up okay.
I then decided to try adding a cap to the 5v line which resulted in a good result.
I probably should have said "appears" to reset as I didn't have a DSO to hand to be certain, nor a schematic of the Hat.
Okay, so everything is OFF when you press the power button (including the LCD etc) right?
Then when do they have a chance to turn ON then back off (reset)? I gather there must be some delay otherwise how do they reset before even being powered up?
That was my point of interest.
What is your power source? Just a battery (what is measaured voltage?)? Or is do you power from 5v?
To see dips you really need a scope and use a trigger. If the Hat tries to power a load that is too demanding, there would be a dip of a few ms to a few dozen ms, that's impossible to capture on a DDM except if it has a very fast sweep and a MIN-MAX mode that can capture that lowest voltage, still you could not see the actual waveform, just a number.
In the end whatever works will solve the issue and I did put the Capacitor pad just for this purpose, although generally it should not be needed. The cap is a reservoir that will charge before you power the Pi and will provide the needed boost till the booster can catch up with the demand.
So much easier when you can see something, so have a look at this,
https://dl.dropboxusercontent.com/u/119565528/MHat%20powerup%20issue%20comp.mp4
The power supply is the genuine Pi power supply rated at 5.1v 2.5A. I also tried other another supply.
Ok I fully get it now :)
Thanks for the video.
The power surge cannot be met without the capacitor, that makes some sense. I might start including a capacitor in the kit...
So you're good now, you've solved the problem right?
Yes, problem solved :)
To add, I really like the MightyHat. It's a nice tidy solution and well throughout. Great work Felix.
Quote from: probson on November 29, 2016, 03:07:43 AM
Yes, problem solved :)
To add, I really like the MightyHat. It's a nice tidy solution and well throughout. Great work Felix.
Awesome, really happy to hear that. Thanks for the feedback, much appreciated!
Felix - I recently received my MightyHat, and just made it through the assembly pages, to find myself with an unexpected component, not mentioned anywhere in the assembly instructions. A 1000uF capacitor. While I appreciate that the cap was included, and is a fix for Pi3 + MightyHat not starting up cleanly, it would help the more neophyte of us if the assembly instructions could reflect the new component - or at least make mention of it.
Obviously I found this thread, and that explains the part.
Fantastic products and love the site. -Mike
Hey Mike,
I think in most cases it's not even needed, maybe that's why I postponed or forgot to do it.
But I just added a little section in the assembly guide that shows this.
Thanks for the heads up,
-Felix