hi folks,
right. i've moved on to re-printing the 3d casework of the laptop, most of that will be in higher-temperature ABS. i'm about 50% the way through that, so another 2-3 days and i'll have an assembled prototype. some bamboo 1.5mm plywood panels are arriving from a U.S. supplier (the only one in the world that sells online 1.5mm bamboo!) are arriving shortly.
the jz4775 cpu card, it turns out that digikey sent the wrong PMIC. bizarre story: the ACT8600-QJ162-T became popular in the past few months because of a smartwatch that was developed in china using the ingenic jz4775 reference design. so the formerly popular ACT8600-QJ601-T suddenly became UNpopular, leaving the world's last remaining supplier as being digikey. so, that's a china-made PMIC, which got shipped to the U.S. (customs-charge), got shipped back to china (customs-charges), then provided by the agent to the factory, then the factory put it on the PCB, then they sent that to me (customs charges)... and then it turns out to be the wrong f*****g IC.
naturally it took a whole lot of shouting at digikey to get them to do a full audit of the stock of ACT8600-QJ601-T PMICs, but i've placed an order for replacements - to be sent directly to me - and will be putting those on when they arrive. also, only 1 out of the 6 PCBs happened to _not_ have solder-bridges under the DDR3 RAM ICs, so i've ordered some replacement DDR3 RAM and will be having a go at putting on 78-pin BGAs with the IRDA solder station. it'll be fun to try. ok.... correction: it'll be fun to succeed.
also even though it's short notice i'm looking to go to fosdem 2016 because i happen to be in den haag at the moment, it's only a train ride to brussels after all. i'll be bringing along the completed laptop prototype if anybody would like to see it.
l.
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Friday 15. January 2016 14.26.41 Luke Kenneth Casson Leighton wrote:
naturally it took a whole lot of shouting at digikey to get them to do a full audit of the stock of ACT8600-QJ601-T PMICs, but i've placed an order for replacements - to be sent directly to me - and will be putting those on when they arrive. also, only 1 out of the 6 PCBs happened to _not_ have solder-bridges under the DDR3 RAM ICs, so i've ordered some replacement DDR3 RAM and will be having a go at putting on 78-pin BGAs with the IRDA solder station. it'll be fun to try. ok.... correction: it'll be fun to succeed.
So, was it Digikey that mixed up the parts or the factory that put the wrong IC on the board?
As for the RAM, I guess that it made for the "bring up" process, as described in that bare-metal Ingenic work, to be slightly more exciting than it should have been. When you're probably only somewhat confident that the code will do the trick, I guess it's frustrating to have defective hardware involved as well.
Anyway, good luck with the printing and soldering!
Paul
--- crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
On Fri, Jan 15, 2016 at 2:02 PM, Paul Boddie paul@boddie.org.uk wrote:
On Friday 15. January 2016 14.26.41 Luke Kenneth Casson Leighton wrote:
naturally it took a whole lot of shouting at digikey to get them to do a full audit of the stock of ACT8600-QJ601-T PMICs, but i've placed an order for replacements - to be sent directly to me - and will be putting those on when they arrive. also, only 1 out of the 6 PCBs happened to _not_ have solder-bridges under the DDR3 RAM ICs, so i've ordered some replacement DDR3 RAM and will be having a go at putting on 78-pin BGAs with the IRDA solder station. it'll be fun to try. ok.... correction: it'll be fun to succeed.
So, was it Digikey that mixed up the parts or the factory that put the wrong IC on the board?
digikey. they just sent - a second time - the exact same wrong part, after i specifically advised them on the process by which they could verify what the part was.
As for the RAM, I guess that it made for the "bring up" process, as described in that bare-metal Ingenic work, to be slightly more exciting than it should have been.
yyyyeah.... but luckily there's debug information printed in the u-boot SPL section that allowed a diagnosis.
When you're probably only somewhat confident that the code will do the trick, I guess it's frustrating to have defective hardware involved as well.
it's why the "reduce it to one and only one change" is so absolutely critical.
Anyway, good luck with the printing and soldering!
thanks... :)
l.
arm-netbook@lists.phcomp.co.uk
-
Luke Kenneth Casson Leighton -
Paul Boddie