Mau1wurf1977, what version 8433 is that motherboard? Probably a v1 or v2 if the RTC is soldered. Some v3's had RTC sockets, some did not. I think all v3.1's had RTC sockets.

feipoa wrote:

Mau1wurf1977, what version 8433 is that motherboard? Probably a v1 or v2 if the RTC is soldered. Some v3's had RTC sockets, some did not. I think all v3.1's had RTC sockets.

Wow you must know these boards well...

It's a version 2!

I really hope the RTC will fix the cache issue I have (see my other thread). At least I will be able to disable it in the BIOS.

Yup, I know this board pretty well. I've even modified it to accept 1024 KB of cache. But this particular board already has the TAG bit set correctly. There is still a good chance the RTC will fix your issues. I noticed this board is fairly particular with having a RTC with a working battery.

EDIT: Let's not hijack this thread. Redirect questions concerning this motherboard to 486 stuck at Updating ESCD... or the World's Fastest 486 thread.

feipoa wrote:

Yup, I know this board pretty well. I've even modified it to accept 1024 KB of cache. But this particular board already has the TAG bit set correctly. There is still a good chance the RTC will fix your issues. I noticed this board is fairly particular with having a RTC with a working battery.

That's fantastic that you know so much about this board.

I will take images of the BIOS chip and version numbers. Maybe it got flashed?

Mau1wurf1977 wrote:

Using a ruler I measured 9mm!

I think a top loaded option would be a good idea.

feipoa wrote:

You can get all this information from the datasheet. See attachment.

I hope you find a way to keep the battery internal to the unit.

Thanks!

Feipoa, I was asking because it wasn't actually a Dallas-branded chip but a 3rd party clone. The package may have been slightly different.

In other news, I'll probably be traveling over the holidays so production/shipping might be delayed until early February. Takes about a month to get the boards from the pcb manufacturer so they'd arrive while I'm away.

Maeslin wrote:

Feipoa, I was asking because it wasn't actually a Dallas-branded chip but a 3rd party clone. The package may have been slightly different.

I have an ODIN OEC12C887A and a Dallas DS12887 sitting right in front of me, out of the motherboard, and can assure you that they have nearly identical dimensions. The ODIN has slightly more tapering towards the top, which equates to about 400 microns less width at the top surface. So you get slightly more real-estate using the Dallas cap.

Please note that you cannot replace every board you want with another battery module. There are many boards that only work when you replace the module with one of the exact same type.
For me it is not a wise descision to replace a battery module with a "new" one because the problem with the included battery still remains. You just buy yourself a bit of time (1-20 years, depending on the battery state - it beginns to discharge as soon as the module is produced , not when plugged into the motherboard). IMO it is much wiser to solder a CR2032 battery holder onto the module. That might not look very nice but gives to full control over the battery and CMOS buffer.

FGB,

I believe his design will either be open-faced with removable battery, or have a removable cover and a removable battery.

Open-faced with removable 1220/1225 battery. Modified the layout to accommodate a top-loading battery holder instead of the side-loading one. It'll be a bit thicker since I have to fit the RTC on the underside of the board. Here's the new version.

Here's the battery holder that will be used:
Harwin S8411-45R

Maeslin, how does the overall height compare with the Dallas 12887, not including the pins?

feipoa wrote:

Maeslin, how does the overall height compare with the Dallas 12887, not including the pins?

Pretty much the same height, 0.330" give or take 30thou. The frame holding the pins has the same height as the ds12885 itself.

New update. I've managed to lay out a 3rd version of the board, this time with all components on the same side. It uses a smaller DS12885 package and a smaller crystal.

Main benefit of having all the components on the same side is that the entire package may end up up to 0.1" lower.

Due to OSHPark production constraints, I'm going to start with a batch of 3. If anyone is interested in trying one out (and if the 1st one works in my system) I can them out.

Pending an answer as to my out-of-pocket expense, I will try one out.

Expense would be zlich for the beta. I'll swallow the costs. Later versions, if I stick with the TSOP package, would likely be between $15 and $20. Unfortunately the DS12885 in TSOP is a few dollars more expensive than the SOIC package. 🙁

Actually, all things considered, I'll produce V2 instead of V3. Yes it's a bit thicker, but the chip is far more readily available in SOIC package.

Zero cost? How can I resist. Put me down for 1.

I think this whole idea is great, having an easily replaceable battery instead of the Dallas module.

I'll probably end up buying a few if they work out, my first computer has one and my Cousin's 486 has had the mod done to it. I stuck it in the wrong way the first time I inserted it so it semi broke I think. It turns on, but eats a whole CR2032 in a week >.>

It's taking longer than expected, but boards have been ordered and shipped. Shipping is being sluggish.

I'll test the first batch when I get it, (if they work) send the ones I won't use out to feipoa and Tr3vor (one each) then share the design on oshpark if anyone wants to make theirs themselves.

I ordered 12 boards myself, so if the beta is successful and past the 2 volunteers who will get a free board each (again, if they work), I'll sell the rest to any interested.

A V2 might follow to try and fit everything in a lower profile, but the difference in height would be negligible.

Glad to hear things are moving, albient slowly.

Finally got the PCBs! Just waiting for a parts order that should arrive within a day or two and I should be able to assemble some of the first batch.

Also found a much cheaper chinese supplier for the DS12885s, so I'll be testing a few of those alongside the more "legit source" ones to make sure they're not fakes.

Should have them assembled and tested before the end of next week.

What is the added cost to have had the PCB print shop arrange for component soldering?