[SOLVED] BigQuad: brownouts when punching throttle

[theseankelly]

Hi Forum

My bigquad build is suffering from brownouts (the CF2.0 board reboots) when I punch the throttle from zero to 100% very quickly.

I'm powering the CF2 with a wire coming from my PDB's 5v supply into one of the ESC pins as recommended. The PDB has a steady 4.85v output which doesn't fluctuate at all, even during the punchout. The voltage across VUSB on the CF2 is 4.52v, but I see it drop to 4.50v when I punch the throttle to 100%. I'm not using a very good meter so I don't know if the instantaneous drop is actually lower than that.

When I take the throttle from 0% to 100% a bit more slowly, I don't get brownouts, and VUSB stays at a steady 4.52, so there's definitely something instantaneous about this problem and not an overall load at higher throttles.

Any clues on where to look?

I'm also curious what causes the .33v drop between my supply and the VUSB power -- I'm not very knowledgeable about electronics but I thought I remember textbooks telling me the voltage drop across a diode was going to be more like 0.7v?

Thanks!
Sean

[tobias]

Hi Sean!

I can start with what I know. The 0.33V dropout is from the schottky diode mounted on the BQ board to prevent backwards current. Schottky have much lower dropout but can't handle high voltages.

The dropouts sounds really strange. Are you sure you don't get a voltage dip? have you measured with a scope? The power management chip (BQ24075) input should work down to 4.35V according to the datasheet but maybe there are tolerances.

Some possible solutions:

• Have a battery connected to the CF2. However the battery will charge with up to 500mA which might make the PDB overheat. Possible to change to 100mA though but that requires updating the nRF51 firmware. Also required to power off CF2 with button...
• Short the schottky diode to remove the dropout. Don't power the BQ board and USB at the same time though.
• Maybe it is possible to make the PDB output 5v-6v instead of 4.85V?

[theseankelly]

Hi Tobias -- thanks for the info.

I brought my copter in to the office today and put it on an oscilloscope: I don't see any measureable voltage dips across VUSB or the 5v regulator output. Any other electrical factors that might contribute? Current draw? Would it help or be useful in any way to hook the other three voltage inputs up to the 5v source?

I do have another 5v regulator sitting around somewhere -- I plan on hooking that up straight from the battery instead to see if it help too.

I also haven't looked in to software to see if the ESC code is crashing or anything. I suspect that's not the cause though, because this issue doesn't reproduce when I power the CF from USB in addition to the battery source.

[tobias]

Did a test on a CF2 here and the LEDs start to flicker at a VUSB of 4.55V and shuts down at 4.50V. I have no idea why it isn't 4.35V as the datasheet says... Anyway that is most likely the cause, your VUSB is too close to 4.50V. I would just short the schottky (D1) to get rid of that drop. I don't think anything will break even if with the PDB is powered and VUSB at the same time but I would try to avoid doing it .

[theseankelly]

Thanks Tobias. I think I might just run a separate pololu 5v regulator to power the cflie, since I've definitely found it useful to have USB plugged in for cflie debugging while also powering the esc's with a battery, and I'd rather not risk anything.

[theseankelly]

On a second look, there's an SS34 on the 5v supply coming off my PDB, so shorting the SS12 on the bigquad deck is a safe solution even if I end up plugging in USB, right?

[tobias]

Ahh, that explains why you get 4.85V. If the SS43 is only used as reverse current protection then it might be better to short that instead. It depends on the effect you want. If you short the SS12 on the BQ-deck the USB will power the BQ peripherals (GPS, OSD, receiver etc) which might be a good thing when you develop (you don't have to power the motors).