You just made my point, I should not have to look for it.rpdom wrote:It seems you buy cheap micro USB cables.qwertinsky wrote:Obviously this problem is due to voltage drop in the USB cables . I am sure you would be lucky to find anything larger than 28 gauge wire in a micro USB cable.
Anything up to 20AWG is available if you look for it.
If they did that then we'd have 1,000,000 dead Raspberry Pis from folks using the wrong polarity and/or the wrong voltage. Barrel connectors are a monster PITA.qwertinsky wrote:
If they put a "real" power connector on it to begin with, this would never have been an issue.
remind me again what the standard power connector for 5v looks like?qwertinsky wrote: If they put a "real" power connector on it to begin with, this would never have been an issue.
power connector with no standards whatsoever? That'll be great. I'll just plug my 20VAC wall wart with a barrel connector into your Pi, shall I?
I got your point. You can't be bothered to look for it and are quite happy to use cheap knock-offs]You just made my point, I should not have to look for it.
A diode that cost .001 cent and a 5v regulator .02 cent at 1,000,000 quantities would prevent that.DougieLawson wrote:If they did that then we'd have 1,000,000 dead Raspberry Pis from folks using the wrong polarity and/or the wrong voltage. Barrel connectors are a monster PITA.qwertinsky wrote:
If they put a "real" power connector on it to begin with, this would never have been an issue.
No it would not! Considering a user could easily plug in a 17V supply, your .02 cent 5V regulator (i'm assuming you mean a bog standard 7805 derivate) would either need a heatsink the size of a PI, or it would overheat immediately. Even an optimal 8V for a cheap 5V passive regulator would need a large and expensive heatsink. and you price estimates are ridiculous, and leave out a lot you are not mentioning, like decoupling caps, 1500uF buffer caps etc. also an 7805 cannot supply 2A, and low drop replacements are more expensive, only totally phased out china junk is priced at anything nearly as low as you claim.qwertinsky wrote:A diode that cost .001 cent and a 5v regulator .02 cent at 1,000,000 quantities would prevent that.DougieLawson wrote:If they did that then we'd have 1,000,000 dead Raspberry Pis from folks using the wrong polarity and/or the wrong voltage. Barrel connectors are a monster PITA.qwertinsky wrote:
If they put a "real" power connector on it to begin with, this would never have been an issue.
Code: Select all
#!/usr/bin/python
import RPi.GPIO as GPIO , time
redLED=35
GPIO.setmode(GPIO.BCM)
GPIO.setup(redLED, GPIO.IN)
powerlow=0
while True:
if(GPIO.input(redLED)==0):
print "POWER dipped below 4.63v"
powerlow += 1
else:
powerlow =0
if (powerlow > 3):
print "Low power for " + str(powerlow) + " seconds"
time.sleep(1)
Using Mike's code, I made a script which uses ANSI control codes to constantly display the LED status in the top right corner. Could someone test it out with the undervoltage message? I don't have the means to do that here.mikerr wrote:I wrote this quick python script to view the power led status over ssh,
may be useful as a template for others:
Code: Select all
#!/usr/bin/python
import RPi.GPIO as GPIO , time
from subprocess import call
import shlex
import sys
import signal
def handler(signum, frame):
# call(shlex.split("echo -ne '\e[u''\e[0m'"))
sys.exit()
signal.signal(signal.SIGINT, handler)
redLED=35
GPIO.setmode(GPIO.BCM)
GPIO.setup(redLED, GPIO.IN)
powerlow=0
#call(shlex.split("echo -ne '\e[s'"))
while True:
if(GPIO.input(redLED)==0):
call(shlex.split("echo -ne '\e[s''\e[1;56H''\e[1;31m'POWER dipped below 4.63v'\e[u''\e[0m'"))
powerlow += 1
else:
powerlow = 0
call(shlex.split("echo -ne '\e[s''\e[1;78H''\e[1;31m'ON'\e[u''\e[0m'"))
if (powerlow > 3):
break
time.sleep(1)
#call(shlex.split("echo -ne '\e[u''\e[0m'"))
print "Low power for " + str(powerlow) + " seconds"
It doesn't have to be 2m long to be "bad enough"!:electronicsguy wrote:cool. Does any 2m USB cable work, or a specific, low-cost one?
What does "vcgencmd version" report?mannkeithc wrote:Ig with some test mode configured that forces it to display the red square "over temperature" warning at start-up?
Sounds like the over-temperature warning. If your firmware/kernel is from a week ago, then that will have the fix.mannkeithc wrote:The small square is predominantly red with an orange tint on the RHS, so it is not pure red, and as you say could be be "under voltage" warning arising due to transient voltage levels when the device is first powered on. The small square does not include any blues or yellows - i.e. did not resemble a rainbow like the big square, hence I first thought it was a "temperature" warning.
