Bitcraze Forums

Hey guys,

Ive got two flies and both are showing the same problem. Not sure what happened.

I was charging one of the flies and it got pretty hot, that was the first indicator that something was fishy. I dont remember that happening to the second one, but either way they both have the same symptoms.
All batteries are newish.

- Voltage reported when starting to charge a depleted battery: ~3250mV
- Charging times are normal and they do not get hot (just as warm as usual)
- Voltage reported when green light stops flashing: ~3530mV.( I remember this usually going to ~4.1)
- After 40s of steady hover flight, the red light stays on (usually this would occur after 5ish minutes)
- I can fly another 3-4 minutes with the red light staying up (usually I would land the flie and it would disconnect 30s later)
- The blue light only blinks once upon power on (used to be blue all the time while on)

So something fishy is up.

With a fully charged battery:
I recorded these voltages while having it plugged in without battery:
Name | V | Pin # +- | Measured voltage
VCOM | 5.0 | +16 -19 | 5.0
VCC | 3.3 | +15 -17,18 | 1.93
VCCA | 2.8 | +20 -19 | 2.82


I recorded these voltages while having it plugged in with battery and just battery (same results):
Name | V | Pin # +- | Measured voltage
VCOM | 5.0 | +16 -19 | 3.99
VCC | 3.3 | +15 -17,18 | 1.88
VCCA | 2.8 | +20 -19 | 2.8

VCC should be 2.8 or 3.3 right? 1.93 seems low.

I should note that one of these quads has had a 5.8ghz fpv system attached, the other the 2.4ghz system. I think problems started occuring after using the 2.4ghz system..

The 2.4ghz fpv system now behaives differently. Usually I would turn on the drone, connect, turn on the fpv system, and fly away.
Now, the moment I turn on the drone the signal quality of the crazyflie connect drops and results in a disconnect, but with the drone still switiched on and transmitting a nice fpv image.

Any ideas?

Cheers

Hi Omw!

Sadly I think you ran into the same problem as me, where you have shorted a part out - the VCC giving 1.8v is a dead giveaway and the rest of the symptons will match up with me too.

superroach wrote: You may be on to something there then.
Charges:
VCCA: 2.78v
VCC: 1.86v

I'll try connect it to the laptop and see what it reports.

Battery life wise I may get 30 seconds tops flight time.

Check out this thread:
http://forum.bitcraze.se/viewtopic.php? ... lator#p336

For what is needed for the repairs.

I'm going to have a go at doing it myself as well too.

In that post it shows you what you need to get, the one giving you the vcc reading like that and lack of voltage is this one:
http://www.digikey.com/product-detail/e ... -ND/949689 . There will be better and hopefully more local places for you, just put in the codes mentioned into google or better still into http://www.findchips.com

Look on your crazyflie, its in the rear corner between M1 and M2 (underside).

Thanks!
Ill check it out this week. Id love to know what causes this
I might have shorted one flie, but two back to back? Hmmm..

Oh, and if anyone is interested, I ordered the parts from here https://hbe-shop.de. One of the few places I didnt have to buy in bulk / as a company / order from outside the EU.

This is a bit worrying and doesn't seem to be caused just by a short any more. It seems to be related to putting on an extra load. Do you connect the FPV systems to the VCOM voltage? If I recall correctly you both use a step-up for the FPV? My first guess is that when the system is tuned on or off the step-up generate a voltage spike that damages the VCC reg. The question is why the VCCA reg isn't damaged though as it is almost connected the same way, and almost is the difference of a RC filter which might protect it from a spike... hmmm, I might be on to something. Unfortunately I don't have the stuff to test it my self.

tobias wrote:This is a bit worrying and doesn't seem to be caused just by a short any more. It seems to be related to putting on an extra load. Do you connect the FPV systems to the VCOM voltage? If I recall correctly you both use a step-up for the FPV? My first guess is that when the system is tuned on or off the step-up generate a voltage spike that damages the VCC reg. The question is why the VCCA reg isn't damaged though as it is almost connected the same way, and almost is the difference of a RC filter which might protect it from a spike... hmmm, I might be on to something. Unfortunately I don't have the stuff to test it my self.

To be honest - I share the same concerns...
I have 3 flies, one with a 5.8ghz (http://www.fpvhobby.com/transmitter/21- ... itter.html) fpv setup, one with a 2.4ghz(http://www.fpvhobby.com/180-4-ghz-nano-size-vtx.html) setup and one exposed to a breadboard that I use for testing.

All fpv setups use the same voltage regulator http://www.pololu.com/catalog/product/2115

Ive only had problems since using the 2.4 system. I first thought I just shorted out a flie, so I re soldered everything to the spare flie. Im pretty sure nothing shorted out there, but it instantly stopped working too with the same symptoms. I can fly without the fpv system for a minute or so. If I enable it now the crazyflie disconnects but stays powered on - I guess there is enough juice left for the transmitter?

So yeah, Im a little scared to hook up the 2.4 fpv system to a working flie

I also have a switch attached to the fpv system, so we might need to distinguish between turning the flie on with the fpv system enabled, and turning the fpv system on after the flie has booted.

I was connecting to pins 16-VCOM - 17-DGND.

The pololu website does reference to LC spikes (http://www.pololu.com/docs/0J16), I guess I should read into it. Supposedly one can

suppress LC spikes by soldering a 33μF or larger electrolytic capacitor close to the regulator between VIN and GND.

Which tests would you do if you had the equipment?

If its any help omwdunkley, I'm using 19 and 16 for my power. Only real difference is the analogue ground.

SuperRoach wrote:If its any help omwdunkley, I'm using 19 and 16 for my power. Only real difference is the analogue ground.

But you also ended up with the same U9 voltage reg problems right? I think I tried different combinations and couldnt see a difference, so I just ended up using the pins that spatially made most sense in the moment. Ill prob end up using 16 & 18 to use a 1.27mm pitch connecter (eg as used for the battery plug/socket).

For general discussion, what could the pro and cons be of using DGND vs AGND?
-Oll

I guess there is enough juice left for the transmitter?

Yes, the system doesn't work that well on 1.8V. It is interesting that it does so at all.

Which tests would you do if you had the equipment?

I would take one of the damaged Crazyflies with the step-up connected and measure the VCOM voltage with a scope during turn on/off to look for any possible voltage spikes. I have attached a graph of the VCOM turn-on without anything connected. It turns on in about 4us. The TPS79301 regulators maximum input rating is 6V so if VCOM gets higher then that I guess it can be damaged.

For general discussion, what could the pro and cons be of using DGND vs AGND?

All the grounds are connected but have separate planes to improve noise within the voltage group. Analog/sensitive stuff should therefore be connected to VCCA and AGND. Digital stuff to VCC and DGND. The VCOM should be connected to GND or DGND. Remember that the analog supply is very limited due to the RC filter and should not draw more then an additional 10-20mA without changing the RC filter.

tobias wrote:

I guess there is enough juice left for the transmitter?

Yes, the system doesn't work that well on 1.8V. It is interesting that it does so at all.

Which tests would you do if you had the equipment?

I would take one of the damaged Crazyflies with the step-up connected and measure the VCOM voltage with a scope during turn on/off to look for any possible voltage spikes. I have attached a graph of the VCOM turn-on without anything connected. It turns on in about 4us. The TPS79301 regulators maximum input rating is 6V so if VCOM gets higher then that I guess it can be damaged.

Hi - I found an old oscillator and spent a day trying to get it to work!

Here are the results. No voltage spikes, not sure if I am happy or not.

Test setup:

+VCOM (pin 16)
-DGND (pin 17)

Measure voltage during startup without the 5v-step-up -> tx -> cam circuit attached and with the circuit attached.



Anything else I can test while I am at it?
Cheers!