Chris McSweeny gmail.com> writes: > By "another" are you implying that RPi is also a really badly designed > product?=20 I did not imply anything about RPi excepting for the fact that it is unfortunately possible to power it via the mini USB connector which, requiring 2A + startup surge, will cause most usual standards *compliant* (see below) such USB sources, and laptops, and motherboards, to collapse voltage or trip a PTC device placed there to prevent such misuse. Imo a fet switch put there specifically to prevent this would have been a good idea, especially if rated 0.5A and, thus, dirt cheap and tiny SOT323. Rpi is also capable to unfortunately power a master usb device with 3.3V from said USB connector when the master tries to go to sleep, which leaves internal buses and more powered inside laptops and the like, with 2A available to fry things in the other direction for a change. NOT a good design idea imho. USB 'standards' have been abused all the time and I also have an external USB hdd which has the criminal Y cable for power, following the theory that each USB connector should only be loaded with 0.5A at most. Of course it works with just one end plugged in, drawing 1A, since the makers of equipment also come to help the USB 'standards' 'following' equipment-makin= g and using crowd out there. Similar USB related discussions have occurred on piclist in the past, they can be found in the archives. The general rules for equipment which is meant for non end user lab use, such as modules which can be arbitrarily connected, which RPi and Arduino and others are, which may and *will* be abused by accident or misunderstanding, as I learned them many years ago in suitable courses, wer= e and *are*: - Fused at each power input to protect traces and wiring against fire. PTCs cost pennies now. - Reverse polarity protection if DC, optional overvoltage protection and possibly optional undervoltage protection (reset chip etc. keeping mcu from doing unsheduled things when undervolted). - Clearly marked supply type requirement, amps, and volts range for each connector meant to be used as a supply, possibly ON the board for devices supplied without a leaflet, such as duinos, i.e. on the silk screen. - Prevention of use of pins not meant as a power supply input as such, if possible, and in any case clearly marked as such (i.e. (!) DO NOT USE TO POWER BOARD legend on board near pin). Thin traces with tortuous routing near noise generating or sensitive circuit parts to auxiliary pin headers 'supplying' 3.3V and 5V are an excellent reason to avoid using them as powe= r inputs, especially for 1A and more. - Clearly stated current and voltage limits for all externally accessible pins, if possible marked on board silk if no leaflet is provided. I think that all this blurb would fit on the duno boards in tiny but visibl= e lettering about where the huge infinity logo is printed on the solder mask, at no extra cost... Just how hard is it it to print about this text there: Supply 9-15Vdc max 150mA or 5V 150mA.=20 Do not use both supplies at the same time. IO: -0.3 .. +5.3V max 5mA. It is ironic that duinos which use AVRs omit all this data (esp. current an= d voltage limits on all pins), when said pins are simply connected to a standard avr mcu on a tiny board, and the relevant data is in the AVR datasheet for all to see. This results in endless discussions by duino 'experts' who 'know' what did to not fry their 2nd or 3rd board and results in anecdotic 'knowledge' passed on as wisdom to those about to fry their 1s= t or 2nd, accompanied by an attitude problem indicating a mental age of 7 or below towards those trying to point at the 'real' datasheets which have all this info on maximum pin current limits (including ground and Vcc), decoupling (we don't need decoupling-it works fine without 'I tried' ...) and so on. I am so fed up with this crowd. --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .