[MrWeedster]

Hi.

Recently i wanted to fire up my origin 300 after months of collecting dust.
Sadly it doesn't start anymore. After giving it power the message "No L1 comm" comes up on the display.
I did not yet hook up a console cable, but w/o L1 there would not be much to see i think.

I already reseated RAM, VRMs and even exchanged the I/O board with a spare i got laying around, but no luck, still same message on the display.

What else could be wrong? Maybe the DALLAS chip is dead?

[jpstewart]

Maybe the DALLAS chip is dead?

Possibly. I have the same problem on my O300. ISTR back in the Nekochan days tracking it down to a dead Dallas chip (DS1742W-120) based on some search results. I haven't managed to find a replacement DS1742W (they're out of production now) nor have I tried to re-work one to use an external battery. But that's what my research indicated was the problem.

I'm surprised replacing the IO board didn't help, though. I thought the DS1742 was on it. So either you have two IO boards with dead batteries or it's not the Dallas chip that's at fault.

[MrWeedster]

I'm surprised replacing the IO board didn't help, though. I thought the DS1742 was on it. So either you have two IO boards with dead batteries or it's not the Dallas chip that's at fault.

I replaced the I/O board using the DALLAS from the replaced board, because i only have two left: One in the tezro, and one in the O300.

[MrWeedster]

It was indeed the DALLAS. I had a good one left which did the trick.
Now it works again.

[jpstewart]

I'm resurrecting this old thread to add another data point about the Dallas chips.  My original one was a DS1742W-120.  I couldn't find an exact replacement at a reasonable price, but I got a new old stock DS1742W-150 and it is working fine. 

I honestly don't know if the -120 vs -150 is significant in the case of the O300 or not.  Maybe I'm just lucky that the -150 part is working for me.  Maybe they'll always work.  But I'm putting the info out there for anyone else who is experiencing the "No L1 comm" error and needs to replace their Dallas chip.  Don't get too hung up on the -120 vs -150 part.

[Raion]

How's the job in terms of difficulty? Is there a lot of sensitive parts around it?

[jpstewart]

It's super easy.  The Dallas chip is socketed.  It's on the underside of the IO8 assembly, so the whole thing has to be removed but that's a few screws and a couple of ribbon cables.  There's really nothing to it.  Finding the chip was the hardest part.

(The only thing that threw me was that the screws holding the case shut were Philips head, and inside were Torx.  I spent longer than I'd care to admit trying to figure out why the screwdriver wasn't grabbing before I realized the difference.  I hate when equipment plays mix-and-match with screw head types.)

Since the Dallas chips are obsolete, I tried to modify my original one with an external battery.  But I just ended up destroying the chip instead.  That's a much harder job, but I believe people here have done it.  Replacement was a piece of cake.

[weblacky]

The problem is, every Dallas chip model is a little different (hence why I posted the instructions on drilling the Dallas chip from an Indigo2 successfully a while ago, I'd never seen instructions for it before...kind of wondered why no one really responded to that). Each model has to be destroyed by someone to find the battery leg locations, polarity, and relative positions inside the epoxy. Then you know where to focus correctly. You use the pin legs to anchor your location for each model.

You could help others if you posted your destroyed chip (as long as we can tell where things were and the pin layout of the legs is still in tact) up on this thread. If you can see the battery and it's metal connections, while seeing the chip's legs and orientation, that should help others to successfully drill their DS1742W.

Also, from memory, I think the numeral after the model number is the amount of battery-backed storage the in chip. Good to know putting in a larger storage chip isn't an issue. But I think that's all it is.

[jan-jaap]

Each model has to be destroyed by someone to find the battery leg locations, polarity, and relative positions inside the epoxy.  Then you know where to focus correctly. 

I had a whole bunch of these things X-rayed a couple years ago. The pictures went down with Nekochan I'm afraid.

[jpstewart]

You could help others if you posted your destroyed chip (as long as we can tell where things were and the pin layout of the legs is still in tact) up on this thread.

I never got inside the chip.  I was clumsy and broke several pins before getting very far into the process.  At that point I wasn't going to waste time on opening it up since it was never going to work with missing pins.

However, apparently the DS1742W is also used in some Tek scopes so there is some info in the forums at eevblog.com about replacing the DS1742W's battery or replacing the whole chip with a different one on an adapter board.  (There's a github project at https://github.com/james10952001/DS1744W...2W-adapter for that.)  That was my next option if the new chip didn't work out.

Also, from memory, I think the numeral after the model number is the amount of battery-backed storage the in chip.  Good to know putting in a larger storage chip isn't an issue. But I think that's all it is.

Interesting.  I had assumed that the -120 and -150 were access times:  i.e., 120ns and 150ns parts like so many RAM chips with similar suffixes.  But I don't know where that assumption came from.