« first day (2018 days earlier)      last day (2932 days later) » 

Nick Alexeev
Nick Alexeev
4:52 AM
@trilolil He is enjoying it too much. He can't get enough. He doesn't like the fact that the week is over.
 
 
6 hours later…
Asmyldof
Asmyldof
10:35 AM
@NickAlexeev Mainly that we're hiring FWdevs and it sucks, because there are none that fit
And I'm the asshole who needs to make that decision
 
 
1 hour later…
abdullah kahraman
abdullah kahraman
11:42 AM
Hey, remember my SD card was not working and we thought the reason was excessive ringing?
 
Asmyldof
Asmyldof
12:15 PM
I remember something to that effect.... but I also troubleshoot a bazillion SD problems, because for some reason not a lot of designs use them properly
Just a bad SD card?
Also happens
 
 
1 hour later…
abdullah kahraman
abdullah kahraman
1:41 PM
@Asmyldof Actually we thought the problem was the ringing.. But there is a WIZNET W5500 ethernet to SPI chip already that is working with no problems whatsoever..
So, because of my shoes, when I was working with the chip, I burned the SPI port somehow, due to electrostatic zapping :D
Now my order from Mouser of the new chip is here.. Whenever I soldered it, the problem was there, continuing its behavior from where I left it.
Then, I said, if the W5500 chip is working, NRF24L01 module is working, this should work, too..
I thought to myself that this is a problem with the SD card connector and hot air gunned the connector..
It worked!
But after some time, like 2-4 hours, it is back to non-working state again and I heat it and it works for about 2-4 more hours..
Changed the connector with new one, the problem persists :|
 
Marla
Marla
@abdullahkahraman That is an extremely long thermal time constant (2-4 hours).
Also sounds like something other than connector heated up, but you probably already thought of that.
 
abdullah kahraman
abdullah kahraman
2:00 PM
@Marla Yeah, but there is not much on the back and I was being very local with the hot air gun, low air flow..
 
Marla
Marla
@abdullahkahraman ok, agreed on only connector heating. But I still would think the thermal time constant (return to room temperature from being hot) would still be on the order of minutes rather than hours.
I just can't think of what the heat would change that would last for hours of operation
 
abdullah kahraman
abdullah kahraman
@Marla Yes, let me more carefully test it..
I will try to make physical stress like small bending etc.
 
Asmyldof
Asmyldof
3:05 PM
@abdullahkahraman SD cards are very noise and timing sensitive, since they contain no data management or validity engines like other mass storage systems
Combine that with a serial interface, rather than a relatively sturdy parallel bus
... riskycakes
 
abdullah kahraman
abdullah kahraman
3:17 PM
@Asmyldof Yeah, if I knew how to interface a serial flash that would be great!
 
Asmyldof
Asmyldof
It's even possible heating the PCB (if it's standard FR4) causes the signal propagation and reflection to change due to expansion
Try keeping it at 80degrees C for a day, see if it keeps working
 
 
2 hours later…
abdullah kahraman
abdullah kahraman
5:24 PM
@Asmyldof That came to my mind but I thought to myself that I am starting to overthink this issue and creating imaginary stuff :)
@Asmyldof Strangely, it is still working..
 
 
1 hour later…
The Photon
The Photon
6:46 PM
@Asmyldof That's really unlikely for the data rates involved in an sd card. I've gotta think a crack in a solder joint or in a trace is way more likel
Or even a cracked die, broken wirebond, or something like that.
 
Asmyldof
Asmyldof
@ThePhoton Unlikely doth not equalses impossible
I have seen less rugged SD cards choke on small dips and rings on the clock or VCC
More importantly: To start smoking the Mackerels, or the Chicken?
 
The Photon
The Photon
7:01 PM
@Asmyldof Smoke the marijuana. Grill the mackerels.
 
Asmyldof
Asmyldof
@ThePhoton Mackerel be one PHAAAAT doobey
Drippin'!
Litrully
 
 
5 hours later…
Nick Alexeev
Nick Alexeev
11:51 PM
I'm looking for an A/D converter with DMA. Somewhere between 2 Msps and 10 Msmps (in bursts). 8-bit resolution would be enough. Single channel.
It could be an A/D inside of an MCU. It could be a standalone A/D (perhaps with external memory). I'm hoping to find a solution that doesn't involve FPGA, though.
What are the odds? Does this sound plausible?
 

« first day (2018 days earlier)      last day (2932 days later) »