A3010 with RAM Control Line Failure POST error

Arc/RPCs, peripherals, RISCOS operating system & ARM kit eg GP2x, BeagleBoard
Post Reply
philpem
Posts: 35
Joined: Fri Apr 04, 2014 5:42 pm
Contact:

A3010 with RAM Control Line Failure POST error

Post by philpem » Sun May 11, 2014 9:34 am

Hi there,

I've dragged an A3010 out of the garage with the intention of gifting it to a friend. Unfortunately in the years it's been out there, the CMOS battery has leaked and wrecked a considerable part of the motherboard -- almost the entire right side.

I've removed the old battery and repaired all the obviously bad PCB traces I could find, and now I'm in a position where the machine at least gets to the POST test. It turns out the only error bit set is "RAM Control Line Failure".

I've checked all the PCB tracks between the RAM and the CPU (via various resistor packs) and they seem good. I've also shotgunned the 74HC139 (RAM decoder) and installed new jumper caps -- still nada.

I'm left scratching my head here -- everything that handles RAM "control lines" looks good, but the POST still says the control lines are bad... Does anyone have any ideas what I could or should check next?

Ordinarily I wouldn't put much effort into repairing one of these machines, but it's a Learning Curve machine with a hard disk drive (mini Podule) and I'd quite like to see it run again...
And, to be quite honest, this has become something of a battle of wills 8^)

Cheers,
Phil.

steve3000
Posts: 1983
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by steve3000 » Sun May 11, 2014 2:30 pm

Hi, that sounds like a huge job, well done to get so far!

I've only had this error once, when resurrecting an A540 (and that's quite a different machine) - for me the problem turned out not to be battery damage, but a hairline scratch which had fractured a track near the MEMC (presumably caused by mishandling of the RAM card at some point in its previous life).

Back to your machine though - one thing you definitely should check is that you have removed any RAM upgrade and that you have reset the RAM size links correctly, back to 1Mb - as the A30x0s need to be told how much RAM is fitted, or they will fail POST.

Steve

philpem
Posts: 35
Joined: Fri Apr 04, 2014 5:42 pm
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by philpem » Sun May 11, 2014 2:56 pm

Hi Steve,

Yeah, it has been a big job -- lots of tracks patched with 38AWG wire-wrap wire and a magnifying glass. The Metcal has been paying for itself these past few weeks!

The machine never had a RAM upgrade; it's always been a "stock" 1MB(?) A3010.

I've checked the RAM sizing links though I now wonder if the corrosion might have damaged the link pins (it certainly made a mess of the jumper caps). They check out OK with a multimeter but you never know?

Cheers,
Phil.

User avatar
daveejhitchins
Posts: 4879
Joined: Wed Jun 13, 2012 5:23 pm
Location: Newton Aycliffe, County Durham
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by daveejhitchins » Sun May 11, 2014 6:24 pm

Hi, Phil:

With battery damage you'll need to check UNDER any ICs in the vicinity. If there are any tracks under the ICs, within the battery damaged area, they could have suffered corrosion too!

Hope that helps?

Dave H :D
Parts: UM6502CE, GAL22V10D, GAL16V8D, AS6C62256A, TC514400AZ, WD1772, R6522, TMS27C512, AT28C256
Products: ARA II, ARA III, ABR, ATI, AP6, MGC, AP5 . . .
For a price list, contact me at: Retro Hardware AT dave ej hitchins DOT plus DOT com

philpem
Posts: 35
Joined: Fri Apr 04, 2014 5:42 pm
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by philpem » Sun May 11, 2014 8:18 pm

I've just had another look at the machine...

After replacing the jumpers and caps, I've gotten no further. Tracks all test as good according to the schematic.

But there's a catch...

IC22 is absolutely red hot. So I think it's safe to say I need at least one new RAM chip, more likely two :(

The part number on the chips I have is HM514260JPB - the schematic lists them as "256Kx16SOJ". Does anyone have any of these kicking around, or know of an equivalent?

Cheers,
Phil.

pkersey
Posts: 205
Joined: Wed Feb 06, 2013 11:22 pm
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by pkersey » Mon May 12, 2014 12:31 pm

philpem wrote: The part number on the chips I have is HM514260JPB - the schematic lists them as "256Kx16SOJ".
Phil.
"HM" looks like a "Hitachi" prefix. Checked eBay and found a Chinese seller dealing chips which are actually Hitachi, so they could be that, but they are going for about £13, quite a kick...

http://www.ebay.co.uk/itm/HM514260DJ7-S ... 58269609f

I don't know whether or not those are actual pictures of the product in there, but the prefix in the one they are selling is "DJ7" so there's a possibility the component packaging could be different (if that's the same component at all).

steve3000
Posts: 1983
Joined: Sun Nov 25, 2012 12:43 am
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by steve3000 » Mon May 12, 2014 1:17 pm

If the RAM chips prove tough to source, keep an eye on eBay/etc. for 4Mb upgrade cards for the A3010 - as these bypass the original 1Mb.

Alternatively I think the Acorn 1 to 2Mb upgrade for the A3010 was simple plug-in chips? These may well be the same/compatible chips as the internal 1Mb, which you could extract and use?

Might be worth a post in the 'wanted' section.

pkersey
Posts: 205
Joined: Wed Feb 06, 2013 11:22 pm
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by pkersey » Mon May 12, 2014 2:27 pm

philpem wrote:The machine never had a RAM upgrade; it's always been a "stock" 1MB(?) A3010.
steve3000 wrote:If the RAM chips prove tough to source, keep an eye on eBay/etc.
Hi again! The RAM upgrade chips are HM514260AZ8. There's another ebay seller dealing these way cheap:

http://www.ebay.com/itm/2-PCS-HM514260A ... 257400d416

Are these the same as used for IC22? Can't check from here now. Worth take a look, this seller put an actual picture.

User avatar
danielj
Posts: 7160
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by danielj » Mon May 12, 2014 2:36 pm

No, the ones on the PCB are SOICs, that's not to say you can't wire the ZIPs in in their place though!

d.

pkersey
Posts: 205
Joined: Wed Feb 06, 2013 11:22 pm
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by pkersey » Mon May 12, 2014 3:59 pm

danielj wrote:No, the ones on the PCB are SOICs, that's not to say you can't wire the ZIPs in in their place though!

d.
Aye, I thought that that could work! Anyway, I would rather go to the purist approach and fit RAMs exactly as the originals unless there was no other way.

philpem
Posts: 35
Joined: Fri Apr 04, 2014 5:42 pm
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by philpem » Sun Jun 08, 2014 12:02 pm

Cheers for the info, guys. It looks like in my sleep-deprived state, I misread the part number on the chips.

It's actually HM514260JP-8.

I've posted a want-ad for either a 4MB RAM upgrade or a bunch of the SOJ-package chips...

My backup plan is to make a PCB which fits in place of the two ZIP RAMs and allows four to be fitted... I'll need to decide whether that's a viable plan, or merely a fool's errand.

Thanks,
Phil.

philpem
Posts: 35
Joined: Fri Apr 04, 2014 5:42 pm
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by philpem » Thu May 16, 2019 2:25 am

Holy Necrothreads, Batman!

I've just finished repairing this A3010. The RAM section of the PCB was beyond help, none of my many attempts to patch the tracks (or replacement RAM chips) gave anything but a RAM Control Line failure at POST. Thanks to the kindness of danielj and Andy_S, I've been able to piece together the parts for an A3010 4MB upgrade, and assembled the upgrade board this evening.

I started by removing the two RAM chips on the A3010's motherboard with a preheater and hot air. The last couple of battery-eaten tracks came away with them. I spent an hour continuity-testing all the ZIP-socket RAM address, data and control lines, revealing that aside from the ZIP sockets themselves being corroded by battery goop, the machine was fine. I installed new sockets (two rows of Harwin breakaway turned pin headers for each ZIP socket) and fitted the matching pins to the RAM board.

In went the RAM board, the jumpers were set for 4MB... and the machine booted to Supervisor for the first time in years! A delete-reset later and it beeped and booted to the desktop! :D =D>

Sadly the mouse port wasn't working - that turned out to be a couple of tracks that'd seen the sharp end of a screwdriver or similar implement at some point in the dim and distant past. After fixing that, I found that the middle mouse button didn't work (argh!) - this turned out to be a dead buffer on the 74LS241.
I didn't have an LS241 to hand so I stuck a 74HC244 in its place with pin 19 lifted and tied to pin 1. That and some contact cleaner in the socket got the mouse working. I'll fix it "properly" later...

I've given up on the rechargeable CMOS batteries both here and in the A3000 -- this is actually the second time this machine has had a battery leak (the battery was replaced previously) and the A3000's had the same. So I've installed a CR2032 in a coin cell holder with a 1N4148 diode in series with the positive leg. I'm not sure how long it'll last before it needs a new cell, but it'll be easy to swap when it goes.

After that debacle I found a few more open tracks on the parallel port which had been eaten by battery goo, so I patched those while I was in there. All on the parallel port, as it happens (unless I've missed some elsewhere). I've had a lot of practice with the Roadrunner pen!

Looks like my A3010 is good to go! Thanks everyone for all your help! :D

I'll be hunting down a copy of the Dealer Test Disc and testing the ports this evening, then it's on to finding my stash of software and installing some of it on the CF card..!
Last edited by philpem on Thu May 16, 2019 7:56 am, edited 1 time in total.

User avatar
danielj
Posts: 7160
Joined: Thu Oct 02, 2008 4:51 pm
Location: Manchester
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by danielj » Thu May 16, 2019 4:48 am

Well done! Epic resurrection!:D

philpem
Posts: 35
Joined: Fri Apr 04, 2014 5:42 pm
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by philpem » Thu May 16, 2019 7:57 am

danielj wrote:
Thu May 16, 2019 4:48 am
Well done! Epic resurrection!:D
Wasn't half. I did an IPC course on PCB rework and repair (IPC 7711/21B) through work a few years ago... I've certainly put those skills to good use.

On the plus side, at least I have practical experience fixing cataclysmic battery-induced damage...
Last edited by philpem on Thu May 16, 2019 7:59 am, edited 1 time in total.

Andy_S
Posts: 32
Joined: Sun Feb 17, 2019 6:02 pm
Location: Aberdeen, UK
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by Andy_S » Thu May 16, 2019 8:10 am

Take a bow Phil, great result. Be gentle with her as she comes out of her coma, probably thinks it's 1992.

Seriously though congratulations, phenomenal effort.

philpem
Posts: 35
Joined: Fri Apr 04, 2014 5:42 pm
Contact:

Re: A3010 with RAM Control Line Failure POST error

Post by philpem » Thu May 16, 2019 8:18 am

Andy_S wrote:
Thu May 16, 2019 8:10 am
Take a bow Phil, great result. Be gentle with her as she comes out of her coma, probably thinks it's 1992.

Seriously though congratulations, phenomenal effort.
She probably does, and the A3000 too... I haven't set the clock yet!

Thanks for the kind words - as you say, those RAM chips are a devil to solder. Drag soldering just creates bridges and false-opens.

I found the way to go is to get some 0.5mm or thinner tin-lead solder (you can make-do with 0.7mm) and a chisel tip about 3/4s the width of the pads. Cover the leads and pads in gel flux, then solder each pin individually. Inspect under a good microscope or magnifier and remove any bridges.

It took ages but it worked first time!
Last edited by philpem on Thu May 16, 2019 8:19 am, edited 1 time in total.

Post Reply