1.6.5 fatal error

im having trouble accessing 24c32 on the ds3231 modules using the playground 24lc65 demo. fresh download of 1.6.5 gets fatal "ld.exe has encountered a problem" XP error whenever wire.h is included. the old 1.0.5 fails to write/read the eeprom even after changing 50 to 57. all that comes out is garbage (probably ff).

i checked the other threads on this but nothing worked. any suggestions what i might be doing wrong?

john1993: any suggestions what i might be doing wrong?

It looks like you are updating your Arduino IDE every couple of days, while at the same time you are using a Windows operating system that was released in the year 2001.

What about updating your Windows operating system to Windows 7 at least, which was released about 6 years ago?

Or if you want to use ancient Windows XP further on: Why not install Arduino 1.0.5 then? Please explain more about the problem with 1.0.5, I don't understand what "fails to write/read the eeprom even after changing 50 to 57" means.

When using the I2C scanner sketch, does the I2C scanner find your I2C EEPROM module?

updating os every month is not an option. or ever for that matter in my particular situation. and as mentioned in post #1 i do have 1.0.5 there. (and every other version from 0013 for that matter)

anyway, the 1st problem is solved. it took several tries but copying ld.exe from an old version fixed the error. strange developers prefer the faulty one to previous that worked... progress is a wonderful thing.

still no go for 24c32 using 1.6.5 or 1.0.5. again odd that send and receive need to be changed. was there any real reason for that? still curious if anyone got these modules to work with arduino? i can do it with asm and the rtc is ok. these are very cheap and popular products so should be some working sketches somewhere.

btw to answer your question 50 is the part id if pins 1-3 are gnd. 57 if high or unconnected which is module default. i did try both. new eproms are all ff and that is also what comes back when i2c fails. so not much help there.

john1993: btw to answer your question 50 is the part id if pins 1-3 are gnd. 57 if high or unconnected which is module default. i did try both.

You mean the "I2C address" of the module?

Normally, you never guess or "try" an address, but you use the address that you have to use.

If you don't know for sure which address to use, then use the "I2C scanner" sketch first to find out which addresses are active on the I2C bus!

What's the output of the "I2C scanner" sketch if you try that with your RTC/EEPROM module?

0x68 should be the adress of the RTC, so if you have an EEPROM, there should be a second device with another adress to be found on the bus. Which address is it that the I2C scanner reports to you?

thats good advice. i will try scanning.

when the "proper" address (i prefer the term device id to avoid confusion with ee memory address) dont work it makes sense to try others. the jumper circuit on that module is weird so not always easy to tell pin state. i did change jumpers to be sure but you can never be too sure with some of this stuff from The Evil Empire.

in any case thanks for the suggestion. either by scanning or backtracking from asm i will get this. the only thing known for sure is i have done something stupid.

thanks to your advice problem solved. i dont know why this didnt occur to me before. scanner bailed me out for mpu6050, hcl5883, baro, and many others. guess i was too busy dealing with that ld.exe issue.

turns out address/id comes up as 56. idk why because jumpers appear open and seller says 57. maybe resin bridge. anyway back on track, thanks.

well thats typical. for half price modules anyway. pin 1 on the chip is unconnected which caused 56 instead of the industry standard (and seller docs) 57. i do have a few others coming from different sources and will be interesting to see if they are all like that.

i have to decide whether to keep the pcb stock and tell everybody to change code or versa visa. maybe depends on state of the new ones. in any case well worth 45 cents for entire module with battery which is about 1/3 what the chips alone cost.

i would still be interested to hear what others experienced with the cheepee ebay ones.