MJLorton Solar Power and Electronic Measurement Equipment Forum
General Category => General Discussion => Topic started by: MJLorton on March 23, 2015, 10:59:25 AM
-
This is an excerpt from an email I received that folks might find interesting:
From Chris of http://www.AnalysIR.com/
*Hi Martin*
I thought you might like to know that I am sitting here in Dublin,
Ireland - developing a new range of IR products for makers & pros alike.
I was just about to ship off a 'pre-production' unit to a user in North
America today & decided to do some more tests before posting.
To cut a long story short, I noticed that the MCU was resetting at
certain times and I wasn't looking forward to a drawn out software
debugging session, particularly as I was confident the firmware was well
tested. (= it would have been very hard to find)
Then I remembered watching your video about power drop over USB cables
and realised I had put in a very thin extension USB cable on the module
and was outputting higher power than normal.
The interesting thing was that the error was occurring only on low
frequency IR signals @ higher duty cycle. (Essentially longer pulses,
because on the shorter pulses the on-board caps were kicking in and able
to cope)
This resulted in the MCU resetting due to voltage drop.
Taking out the 'cheap' cable solved it and we are back on track again -
all thanks to your video.
Small world & thanks again......
*Rgds*
PS: If you are ever looking at IR remote control, let me know & I will
send you a copy of AnalysIR to play with. (One piece of test equipment
you definitely don't have in your lab :-) )
Also, have a look at our blog where there are some interesting articles
related to IR & feel free to reuse any of the content on your vblog if
you ever plan an IR remote control series. Similarly if you ever have
any questions on IR just email me.
Further detail:
PS: The impact of a voltage drop in powering an AVR or any MCU would/should be obvious to many/most people. I suspect most would understand that a reset would occur (i.e. brow-out protection would kick in). However, the not so obvious bit is the voltage drop across a USB cable which is where your video comes in. Again, this only occurred when we were pushing the current (IR transmission power) to the limits. Did another batch of testing today with the shorter USB cable and everything was still fine.
Ironically, that 'thin' USB extension cable could well become a valuable part of the test harness in future.
The rating for the AVR @ 16MHz is 4.5-5.5V, so things start getting hairy at 4.5V.
In our case brown out detection was configured for 4.3V +/-0.2V, which automatically issues an internal reset once that level is reached - thus reducing the risk of corruption/damage via rogue instruction execution etc.