If one SPI port is allocated from port P, that leaves 4 remaining pins for PWM use.Fred wrote:If there are spare FETs included they should be attached to the port P pins for PWM use.
- Jim
If one SPI port is allocated from port P, that leaves 4 remaining pins for PWM use.Fred wrote:If there are spare FETs included they should be attached to the port P pins for PWM use.
Even the correct LDO will get really hot unless it is heatsinked appropriately. I think I'd rather have it radiating heat into a sink that potentially radiating switching noise into the CPU wouldn't you?davebmw wrote:Firstly the LDO regulator gets really hot does anyone have any objections to ether getting rid of it totally of the CPU card and or using a high efficiency switching regulator circuit?
One CAN and no to 485.question is do we need RS 485? CAN if so how many?
As Jean said :-)Do we need BDM port?
Currently it's used as a "I'm alive" LED for when you first load your flash new code. I think for noobs this is a good thing to have and although it's across the port, the port is still free to be used for something else, the LED will just mimic the device. In summary, do as you please with it :-)I have noticed that there is an LED on PP7 a PWM port, why is this there can we get rid of and free up a PWM port?
I'm going to post a picture proving something you said was very right in a few minutes.Using as input will result in random behaviour, output goes low during reset
That a nice part but be careful about the pinout because it's not the same as the more common LM2937/LM2940.davebmw wrote:OK i have found this nice very capable LDO linear reg:
http://uk.rs-online.com/web/search/sear ... &R=5339404
LM1085IT-5.0, 5.0V, 3A, max 20V in, £2.15 from RS i'll get discount on that.
OK Cool thanks guys if there are any other features you want in the CPU area let me know and i'll squeeze them in.Fred wrote: Even the correct LDO will get really hot unless it is heatsinked appropriately. Agreed! dissipation is dissipation the laws of thermodynamics are pretty clear on that.
I think I'd rather have it radiating heat into a sink that potentially radiating switching noise into the CPU wouldn't you?
switching reg circuits only do that when designed by a complete muppet! correct use of reactives on the OP will kill all HF Ripple, but a linear reg is much simpler if not a bit less efficient. i'll keep it simple
Fred.
Yeah, put a couple of pins for the bootloader and reset jumpers/switches/headers etc. I like a bootloader switch on the outside of the case and a momentary reset switch is invaluable while testing and messing around. The bootloader needs something but the other one is optional and maybe even a bad idea.davebmw wrote:OK Cool thanks guys if there are any other features you want in the CPU area let me know and i'll squeeze them in.