tht / bpnode-bb Goto Github PK

Needs an additional pin, maybe some pull-up/down - unclear...

BP is a nice size for AA battery operation.
The existing regulator on the BP is frowned upon as being inefficient.

Just a thought, could we bring A3 through B1 to the BPE using blind pads as used on the I2C. We need a few pins to the left and right, A0 though A2 and B10/B11 through to GND to support the BPE. There might then be space for an RF95 radio. The differences in the radio pads is radical so I doubt we could dual purpose the radio area.

If extender was mounted under the BP we have the option to collect the 4 pin group carrying GND/DCLK/DIO/3v3 in order to use DCLK & DIO similarly to HyTiny extender - see below:
https://github.com/jeelabs/embello/issues/35
This approach supports all the programming options via the FTDI port.

Extender on top would mean light pipes or holes to the see the LEDs - possible.

BT1 has significance in the STM32duino world as a mechanism to trigger perpetual bootloader. The general case of a high impedance pulldown on BT1 is flipped to 3v3 to enable the perpetual bootloader feature.

Currently we have a hard pull down to GND which precludes pulling to 3v3.

It appears impossible to bring B6 & B7 into use as an i2c interface without major upheaval. Perhaps two surface mount pads on the underside of the extender above B6 & B7 with SMD pins?

The alternate i2c pins B8 and B9 looks possible by bringing though B8.

I wonder if there is space for one of jcw's µPlugs.

The board is off-center – looks like 1.27mm. Should be fixed in rev2.

Ran into a problem whereby STLink, Black Magic Probe and Ser+/folie could not take control while running the STM32duino bootloader and this sketch. Suspect code problems when running on STM32.

I needed to use BT0 to 3v3 to get back in control of the node. Used a jumper wire to the BT0 hole assigned to Red-Pill.

Just to make sure no-one tries to use it as a ser+ :)

Similar to HyTiny extender. Other STM32 boards I have use a Winbond W25X16 which is 2M byte.

The two image files are named Rev 1 when they appear to be Rev 2.

E1 is unterminated and E2 is grounded. Since unterminated defaults to low the chip enable address bits are both 0.

Great work Thomas. I have only noticed one thing, based on a sample of one I think that the BT0/BT1 block could move half a millimetre closer to your reset switch to make the fit perfect. The same effect can be had by slightly bending the centre pair of BT0/BT1 pins a couple of degrees towards the processor. I don't have any Red Pill units and therefore can't comment on the implications there.

The footprint is not compatible with the buttons I have. Review the footprint (or get other buttons)

Looks like it's working perfectly fine but there should be a diode in the reset line to make sure the BP is always able to do a clean reset.