does someone currently have a non-working version, that works in 1 x0xb0x, but not another? xlarge, when you got the ones back that were faulty, did they work in your x0x?
the reset circuit most likely not the problem. the microcontroller will not start up until the reset is engaged, so the BOD, or watchdog timers, or anything else will not be in effect.
Presenting: x0xlarge the non-unobtanium cpuupgrade
Moderators: altitude, adafruit_support_bill, adafruit, phono, hamburgers
Please be positive and constructive with your questions and comments.
-
- Posts: 3155
- Joined: Fri Feb 17, 2006 5:35 am
-
- Posts: 151
- Joined: Wed Jul 11, 2007 2:02 pm
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
Hey guest.
I haven't gotten any back just yet - one is on it's way though.
The two people in this thread that have reported problems still have them - and they all worked in my x0x before shipment.
I was, yet again, reading the datasheet for the 2561 and a note about startup reset caught my eye:
p373
1. The Power-on Reset will not work unless the supply voltage has been below VPOT (falling).
But, that is the 5V rail and that never dips below 1.6V (POT) once it's stable. (my x0xb0xers 5V doesn't dip at all)
That is what got me thinking if it's not stable when.. But then again, if the 5v rail wasn't stable the gate wouldn't open up either.
Either way, that would be easily testable by reseting the chip manually i guess.
I've tried to contact you about the schematics, could you please drop me an email?
[email protected]
I haven't gotten any back just yet - one is on it's way though.
The two people in this thread that have reported problems still have them - and they all worked in my x0x before shipment.
I was, yet again, reading the datasheet for the 2561 and a note about startup reset caught my eye:
p373
1. The Power-on Reset will not work unless the supply voltage has been below VPOT (falling).
But, that is the 5V rail and that never dips below 1.6V (POT) once it's stable. (my x0xb0xers 5V doesn't dip at all)
That is what got me thinking if it's not stable when.. But then again, if the 5v rail wasn't stable the gate wouldn't open up either.
Either way, that would be easily testable by reseting the chip manually i guess.
I've tried to contact you about the schematics, could you please drop me an email?
[email protected]
-
- Posts: 3155
- Joined: Fri Feb 17, 2006 5:35 am
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
ok, that is good. if the one that comes back works in your x0x, then it has to be a powersupply or bad connection issue. were the broken ones able to talk over usb at all? bootload or acknowledge presence in any way?
- altitude
- Posts: 995
- Joined: Wed May 11, 2005 5:17 pm
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
Could current draw at start up be the issue? I had at one time used 100R resistors for some dim LEDs and the box would only boot intermittently, I measured the current draw at boot up and i was pulling 700+ mA (then dropping) on the 5V rail! Could a 162 be more tolerant of voltage drop at boot up?
-
- Posts: 3155
- Joined: Fri Feb 17, 2006 5:35 am
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
the original x0x firmware didnt have the brown detection enabled (at least if i recall correctly). the new firmware does, so the large current draw shouldnt be a problem, it would just reset the micro, and it would start over from the beginning. there might be some odd case where it continually resets itself, as the startup code draws a lot of power, but that is not as likely.
-
- Posts: 151
- Joined: Wed Jul 11, 2007 2:02 pm
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
I haven't received any back yet but one x0xlarge darffader has didn't work in one x0xb0x but it does work in another.
Unfortunately the one that didn't work had to go back to it's owner rapidly so we can't lab on that machine.
But it prooves the point, atleast on that x0xlarge, that it is something different in the x0xb0xes rather than the x0xlarges. Observe i did not use the term "wrong".
Unfortunately the one that didn't work had to go back to it's owner rapidly so we can't lab on that machine.
But it prooves the point, atleast on that x0xlarge, that it is something different in the x0xb0xes rather than the x0xlarges. Observe i did not use the term "wrong".
-
- Posts: 363
- Joined: Wed Jul 08, 2009 7:10 pm
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
I've got 2 x0xb0xes, but just 1 Xlarge CPU board for now. I can say that the CPU board works perfectly in both of them.
x0xb0x 1 is a Substyler/x0xsh0p kit from 2 or 3 years ago.
x0xb0x 2 was self sourced, based on PCB's from Adafruit (and using the BOM here). Built around 4 years ago.
Some mods on both but nothing too wild.
Tried a firmware update on it last night via Midi. Works great!!
Im looking forward to new firmware builds that really make the most out of it now
x0xb0x 1 is a Substyler/x0xsh0p kit from 2 or 3 years ago.
x0xb0x 2 was self sourced, based on PCB's from Adafruit (and using the BOM here). Built around 4 years ago.
Some mods on both but nothing too wild.
Tried a firmware update on it last night via Midi. Works great!!
Im looking forward to new firmware builds that really make the most out of it now
-
- Posts: 3155
- Joined: Fri Feb 17, 2006 5:35 am
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
a quick thing to try on a non-working x0x, is to manually reset the micro. this can be done after letting the x0x boot up for a minute, then briefly short out c105. you can do this multiple times and see if it ever comes up.
-
- Posts: 151
- Joined: Wed Jul 11, 2007 2:02 pm
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
I've gotten a response from one who has trouble: It stays "dead".guest wrote:a quick thing to try on a non-working x0x, is to manually reset the micro. this can be done after letting the x0x boot up for a minute, then briefly short out c105. you can do this multiple times and see if it ever comes up.
He tried it in different modes, several times... no go.
-
- Posts: 151
- Joined: Wed Jul 11, 2007 2:02 pm
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
More response; darffader lowered the BOD and that resulted in a working cpu - in one troublesome x0x.
We will test with more and see if that really is the key.
We will test with more and see if that really is the key.
-
- Posts: 3155
- Joined: Fri Feb 17, 2006 5:35 am
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
that does seem to imply a powersupply problem. one where the BOD gets repeatedly triggered at startup. how many milliamps is darffader's powersupply? the 7805 regulator should be able to do an amp easily. i wonder if there is a momentary shorting of power to ground via the microcontroller.
1. has any new code been added that modifies an output port in startup or bootloader?
2. it might be good to add an "all LEDs off" function to the bootloader. that way all the LEDs dont turn on at once during startup.
the LEDs are the only thing capable of drawing significant power when used correctly. so its either that, or some pin being assigned incorrectly, or shorted.
1. has any new code been added that modifies an output port in startup or bootloader?
2. it might be good to add an "all LEDs off" function to the bootloader. that way all the LEDs dont turn on at once during startup.
the LEDs are the only thing capable of drawing significant power when used correctly. so its either that, or some pin being assigned incorrectly, or shorted.
- antto
- Posts: 1636
- Joined: Thu Apr 15, 2010 3:21 pm
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
guest: darffader's adapter is 300mA, xlarge's is almost 800mA
yes, i've added more code into the bootloader on top of your modifications: changing the interrupt vector (to enable interrupt-driven MIDI) and then enabling the "spi" in order to have the same LED functionality as in the firmware
then i set all LEDs to a specific "pattern" which can't be mistaken for a "random" one, thus it is very easy to tell if the bootloader is running or not
here's how the bootloader looks like in action: http://www.youtube.com/watch?v=IEW8H299tjg
i was also suspecting possible trouble with the LEDs because i've seen x0xes with all kinds of crazy combinations of LEDs, and who knows if they have wisely-chosen resistors to them.. so the LEDs potentially vary a lot
however, we just tested with xlarge's cpumod, set to BOD=4.3V (which again doesn't work):
- i added a 1s delay before any LED functions in my bootloader (right after the pin initialization) == still doesn't work
- i moved the 1s delay even earlier, before the pin initialization, before the interrupt vector, basically at the very top of main() == still didn't work
- if that wasn't enough, we tested the original x0xb00t2.hex (right out of your zip, without any modifications) == nope
then, changed just the BOD to 2.7V ==> your bootloader works (mine too)
so uhm.. i'm clueless
yes, i've added more code into the bootloader on top of your modifications: changing the interrupt vector (to enable interrupt-driven MIDI) and then enabling the "spi" in order to have the same LED functionality as in the firmware
then i set all LEDs to a specific "pattern" which can't be mistaken for a "random" one, thus it is very easy to tell if the bootloader is running or not
here's how the bootloader looks like in action: http://www.youtube.com/watch?v=IEW8H299tjg
i was also suspecting possible trouble with the LEDs because i've seen x0xes with all kinds of crazy combinations of LEDs, and who knows if they have wisely-chosen resistors to them.. so the LEDs potentially vary a lot
however, we just tested with xlarge's cpumod, set to BOD=4.3V (which again doesn't work):
- i added a 1s delay before any LED functions in my bootloader (right after the pin initialization) == still doesn't work
- i moved the 1s delay even earlier, before the pin initialization, before the interrupt vector, basically at the very top of main() == still didn't work
- if that wasn't enough, we tested the original x0xb00t2.hex (right out of your zip, without any modifications) == nope
then, changed just the BOD to 2.7V ==> your bootloader works (mine too)
so uhm.. i'm clueless
-
- Posts: 3155
- Joined: Fri Feb 17, 2006 5:35 am
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
well, that definitely confirms its a powersupply issue. im suprised its drooping that low. a 300mA powersupply might be a bit low, but should still function.
antto: do you have an oscilloscope? it would be interesting to watch the 5V line at startup. another thing that can be checked is the steady state level of the 5V line. if its a bit low, around 4.6V, then it wouldnt take much to get it to spike down to 4.3V
i will take a look through the code and see what i can come up with
antto: do you have an oscilloscope? it would be interesting to watch the 5V line at startup. another thing that can be checked is the steady state level of the 5V line. if its a bit low, around 4.6V, then it wouldnt take much to get it to spike down to 4.3V
i will take a look through the code and see what i can come up with
-
- Posts: 3155
- Joined: Fri Feb 17, 2006 5:35 am
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
also, xlarge, what size is the powersupply bypass cap on the daughterboard?
-
- Posts: 151
- Joined: Wed Jul 11, 2007 2:02 pm
Re: Presenting: x0xlarge the non-unobtanium cpuupgrade
100nFguest wrote:also, xlarge, what size is the powersupply bypass cap on the daughterboard?
Please be positive and constructive with your questions and comments.