diy NVRAM

[Holden202T]

that should be ok.

i have a nvram with a very keen leg after an accident so i glued it to a eprom socket, it just fits in the case, so i'd say your setup will have a few mm clearance

[accelrate]

Did this project end up getting any further? I would like more information on how to build this. Thanks

[festy]

I'm getting a few boards made soon, using them as panel fillers to keep costs down, so I have to wait a bit.
If I ordered a run of just these boards by themselves, I'd have to get a whole lot made or the unit price would be uneconomical. There's not much point in these if they cost more to build than a DS1245Y

[The1]

you would have to look at oncosts though, atleast with yours you can just put a new battery on it. Can get them for nothin just abouts, so in the long term.

[accelrate]

I was worried it didn't end up happening. If I am up to the RT stage and you haven't finished, what are the actual components I'll need. I might just home job it, like you did.

[festy]

This by itself won't give you RT - it just goes in place of a more expensive DS1245Y NVSRAM IC, you'll still need an RT memcal board to plug it in to.
SRAM has a different pinout to eeprom, so you can't just plug it into a standard memcal unfortunately.
I've already made a small batch of basic RT memcal PCBs for the 808 (no knock filter support etc, I didn't need it) and have a few left over.
Rough cost would be something like $30 for the NVRAM module, or $60-70 with the RT board.
If you wanted an RT board with all the bells and whistles then VL400's would be the way to go.

[festy]

I should have a couple of modules ready for testing in a few weeks for the 808 and vs+ ecus, just waiting for the pcbs and parts to arive now.

[The1]

Here's a photo inside a DS1245Y, my fake one died, the battery was dead, unfortunatly the memory chip is stuck in the plastic, but you can see the rest of it, seems the top cap just pops off these and they have areas so you can measure the battery voltage, i tried to see if i could get the battery out but no luck.

[The1]

well i sat there for quite a while scraping the plastic off, so here's what's in it.

Henli Max CR1632 3volt battery

batt.jpg (125.22 KiB) Viewed 12850 times

Hyundai HY628100B 70ns 128K x8 bit 5.0V Low Power CMOS slow SRAM (Perhaps why they have corruption issues rated at 5v?)

nvram ram.JPG (248.56 KiB) Viewed 12850 times

[festy]

So are you saying this one was a fake?
That would explain why it had a build date of 2009 but the SRAM IC was made in 2000
Also, that battery is more than double the capacity of the one in mine - probably because the SRAM in yours is the cheaper "high" low power version, which needs twice the current to maintain data retention. Even so, that battery/sram combination is only good for ~6 years data retention, well below Dallas's 9-10 year spec.

And that battery in yours is a CR, not BR - the BR chemistry makes them extremely safe (but expensive), they're approved for use in everything from guided missiles to the space station - unlike the cheaper CR range.

Also, I'd be very surprised to see Dallas/Maxim still using an 8 pin DIP package in 2009 - and a non-RoHS 13 year old IC at that!

Haha, I just found the "datasheet" for your fake - http://chaiper.com/vossel/VS1245a.pdf so answered my own question about it's fake-ness.
They've just copied the Dallas datasheet and changed DS to VS everywhere, including very low resolution copies of the original tables and diagrams