Page 1 of 1

KiCad

Posted: Mon Apr 20, 2015 7:53 am
by Slaxx
I'm trying to use most of the Components that are used on the Crazyflie 2.0 for further development. I remember Bitrcraze mentioning that the KiCad Project file won't be available until everything is cleaned up. How about the components library?

Could you in addition name the quartz that is used?

Re: KiCad

Posted: Tue Apr 21, 2015 9:22 am
by tobias
One of the major problems is with the components libraries as that is one of the things that has been changed and needs a lot of updating.

The quartz that is being used is the Epson: TSX-3225 16.0000MF18X-AC3

Re: KiCad

Posted: Tue Apr 21, 2015 2:50 pm
by Slaxx
Ok thank you for the answer. I wrote most of it by now anyway.
Since this is a 16 Mhz Quarz I guess its the one for the NRF? What about the 8Mhz for the STM?

During the creation of the schematics some further questions came up:
Whats the BALUN that is used?
Why do you use 2 different LDO (NCP702SN30 and LP29785-30) since they both deliver 3V with almost the same current?
VCOM and VFILT are they just debug variables?

Re: KiCad

Posted: Wed Apr 22, 2015 8:40 am
by tobias
What about the 8Mhz for the STM?
8Mhz is the Murata: CSTCE8M00G15L99-R0
Whats the BALUN that is used?
Johanson Tech. 2450BM14E0003T
Why do you use 2 different LDO (NCP702SN30 and LP29785-30) since they both deliver 3V with almost the same current?
LDOs are optimized for different things. The NCP702 has a low quasient current but still good SNR with is important for off mode and the radio. The LP29785-30 is a more general one and only on when system is on.
VCOM and VFILT are they just debug variables?
VFILT is there to optionally filter, with LC filter, the radio supply voltage to avoid interference from motor PWM noise. Wasn't needed.

Re: KiCad

Posted: Thu Apr 23, 2015 10:26 am
by Slaxx
Thank you.

In addition I think I found two discrepancies
1.) In the hardware specs (wiki and seedstudio) you write that a 8kb eeprom is used. However the product id is clearly a 64kb eeprom (else it would be 24AA08...)
2.) In the main schematic you write 10n and 4.7u on VDDA (over the caps C4-C9). Shouldn't this read 10n and 1u on VDDA according to your caps. However in the data sheet of the STM32F405 it is written 100n and 1u.

Cheers,
Fabian

Re: KiCad

Posted: Fri Apr 24, 2015 8:07 am
by tobias
In the hardware specs (wiki and seedstudio) you write that a 8kb eeprom is used. However the product id is clearly a 64kb eeprom (else it would be 24AA08...)
The eeprom is 64kbit and we write 8kbyte instead (bits are not that informative)
In the main schematic you write 10n and 4.7u on VDDA (over the caps C4-C9). Shouldn't this read 10n and 1u on VDDA according to your caps. However in the data sheet of the STM32F405 it is written 100n and 1u.
Yes the comment in the schematics is wrong. We use 10n and 1u for VDDA. The 10n is there to filter higher frequencies better not to add an additional 10n. The tolerance for the 1u is already 0.9 - 1.1u. For detailed decoupling e.g see this paper. 100n and 1u will most likely work well, probably only 1u as well. Testing will give you the answer.