Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: agr on December 17, 2010, 07:52:46 PM
-
Hello,
I'm trying to install zip chips into my recently acquired A3000. The chips are TC514402AZ-70. I've now put 8 of those in U850 to U857 (bank 0). I've put them with pin 1 on the left, next to the capacitors (can read the numbers on the chips from the front of the A3000).
The jumper J852 is put toward the front of the A3000, where it says 1Mx4.
No error on boot, no smoke; the problem is that the new ram is not shown in the WB titlebar or in ShowConfig.
Can someone help me make it working ?
Thanks !
-
Hello,
I'm trying to install zip chips into my recently acquired A3000. The chips are TC514402AZ-70. I've now put 8 of those in U850 to U857 (bank 0). I've put them with pin 1 on the left, next to the capacitors (can read the numbers on the chips from the front of the A3000).
The jumper J852 is put toward the front of the A3000, where it says 1Mx4.
No error on boot, no smoke; the problem is that the new ram is not shown in the WB titlebar or in ShowConfig.
Can someone help me make it working ?
Thanks !
Those are static column rams.Should work assuming they are good ram,but i have run into oddities with some A3000's and toshiba ram before.What ramsey/Dmac revisions do you have? is this machine 030 or any accelerators installed?3.1 roms?
-
It's a standard 030, KS31, board rev 9, no accelerator.
Ramsey is 390541-04
DMAC is 390537-02
-
It's a standard 030, KS31, board rev 9, no accelerator.
Ramsey is 390541-04
DMAC is 390537-02
EDIT: sorry for the double post but I having problems with chromium browser on this board and in amibay (also edit button does not work on chromium)
-
Amiga 3000 memory is installed in "banks" or groups of 8 chips per bank. For instance, if you installed just 4 megs (8 chips) the first zip would go into the first socked, then you would skip 3 sockets and plug in the next zip, then you would skip 3 more sockets and plug in the next zip, and so on. The zip ram is not plugged into the zip sockets in a row, like the DRAMs are.
-
I've mixed static and FP RAM, if that's what you're doing, make sure FP is in the first socket of the first bank.
-
All the chips are the same: SC.
I have the same problem with the 32 chips fitted. After system boot, 0 other ram :(
EDIT: chips are 70ns
-
All the chips are the same: SC.
I have the same problem with the 32 chips fitted. After system boot, 0 other ram :(
EDIT: chips are 70ns
Dude, I've been there. If you have money, then don't waste your time and invest in ZorRAM (or ZIP to SIMM adapter if you will stumble upon one). Of course ZorRAM is a bit slower than on-board RAM, but it will save you a lot of headaches related to ZIP RAM chips...
My A3000 came with 4MB of Fast. I've tried to install more by adding ZIPs. Theoretically chips I've acquired were compatible with those installed in my A3000 - same type, same speed, installed in correct banks, etc. But it would never not work. In my opinion A3000 onboard RAM is extremely difficult do expand. Some people did this and had no trouble at all, but I've wasted many hours and got nothing! Finally I gave up and bought ZorRAM, which is an excellent RAM expansion.
-
Are the ROMs A3000 type? (Is SCSI working?)
On my 3k several pins of the ZIP sockets had no solder, you might want to inspect them. Have you tried different ZIPs in the first bank? If there's a defect chip, none will be detected.
-
Yes, the ROMs have a sticker saying A3000 on them, branded Escom !? Don'know if they are the right one but the A3000 works fine and boot from the 512MB SCSI HDD. I did not inspected the soldering of the ZIP sockets as it require to completely remove the motherboard. I have also tried a set of 256k chips without success.
I also tried an A2058 Z2 memory board and it works fine.
Do you know any way to diagnose Ramsey ?
There is one strange modification on this motherboard, a 42MHz oscillator add-on wired next to the FPU. Any info on that is welcome ?
-
I've found something intersting: I moved 1MB of chip ram to the fast ram DIP sockets and it works !
So, I will check the connections between the DIPs and ZIPs banks 0, it can be the soldering issue...
-
Forget messing with ZIP chips, contact DJBase on this site and inquire about his ZIP to SIMM converter for the A3000 :)
-
Any damage to the motherboard in the battery area?
-
The battery is starting to leak... but no damage yet.
@adz
If the zip chips are not working, I give little chance for a ZIP to SIMM converter to work on top of those zip sockets...
-
I've found something intersting: I moved 1MB of chip ram to the fast ram DIP sockets and it works !
With or without moving the RAM size jumpers (J852/853)?
If DIP 256kx4 work and ZIP don't it's clearly a problem with the sockets.
-
With or without moving the RAM size jumpers (J852/853)?
If DIP 256kx4 work and ZIP don't it's clearly a problem with the sockets.
Or a problem with one of each 1M (2 chips = 1MB) of ZIP chips! :)
Do you have a warranty for your ZIP Ram memory?
For the FPU, read this: http://www.amiga-hardware.com/downloads/a3000fpuupgrade.jpg
Check with SYSINFO or AIBB if your FPU is at 42MHz ?!?
-
Yes, with moving J852 to the 256k4 position. This board (rev 9) does not have J853.
I tested the connections between DIP and ZIP sockets bank 0 and they seem ok.
I re-tried now to put the DIP chips and they aren't recognized anymore :( :( (they are still ok in the chip ram section).
-
You should try testing the RAM by absolute address even if it's not added to the system pool, it should be there all the same.
I'm sure I had some tool for that but can't find it atm...
-
@agr: or simply electrostatic discharge (ESD) have destroyed all these chips! :-(
-
@VingtTrois I'm equiped with ESD condoms :) (antistatic wrist strap and desk mat).
I'm interested by the absolute memory testing solution, this is what the system may do to detect the ram chips...
-
@VingtTrois I'm equiped with ESD condoms :) (antistatic wrist strap and desk mat).
I'm interested by the absolute memory testing solution, this is what the system may do to detect the ram chips...
I can test for you each chips (for free), on A3000/030@25MHz Rev9.3 + 2x OKTAGON 2008.
-
@VingtTrois
This is very nice of you :) Thank you, but I just bought one of those Oktagon 2008 as you are using to test the chips. So if they work on the Oktagon, I will have 8Mb usable for sure. I guess this ram will be slower... but I just need a good working Amiga as a reference machine for my hobby development of AROS.
I still have to:
- replace the leaking battery
- install the network Xsurf card + AmiTCP or something
-
There is a clock speed jumper for Ramsey (J851) and a clock 90 jumper (J100) both need to be set for 16 or 25 Mhz. Most A3000's are 25 Mhz. But it's a good idea to check jumpers in any case. Your problem is most likely an open connection to any one of the ZIP chips. Could be a bent pin, poor solder joint or broken trace. If the OS can't find memory in the first bank it won't bother to check the others. It's worth the effort to fix the problem. ZorRam will only run at 2/3x speed of motherboard fast memory and supposedly only works with SuperBuster rev 11. As far as the 42 Mhz FPU mod, someone went to alot of trouble to do this. There's a simple cut and jumper mod on Aminet/hardware/hacks to run the FPU from the motherboard 50 Mhz oscillator. You won't see a big performance boost since the FPU is seldom used by the OS. Looks nice on Benchmark programs though.
-
There is a clock speed jumper for Ramsey (J851) and a clock 90 jumper (J100) both need to be set for 16 or 25 Mhz.
All these jumpers are set to the 25MHz position, like on:
http://amiga.resource.cx/photos/photo2.pl?id=a3000&pg=7&res=hi&lang=en
same board, same chips config, same jumpers
Your problem is most likely an open connection to any one of the ZIP chips. Could be a bent pin, poor solder joint or broken trace.
I checked with a multi-meter, the DIP and ZIP sockets are connected on bank 0; seems ok to me and I just don't like to remove the motherboard completely...
It's worth the effort to fix the problem.
Yes, I'm trying to :hammer:
Can someone post a ShowConfig output of his A3000 with FastMEM on the motherboard ? Thanks
I'm trying with a little prog to write then read a value back from the FastMEM, according to this memory map: http://gega.homelinux.net/AmigaDevDocs/hard_d.html#d-2. I tried $0700 0000 without success :furious:
-
Or you can use one of the memory testers on Aminet. :) There's a few of them, I've used one which was originally designed for A1000 expansions I think....
-
RAM grows from $07FF FFFF downwards, so Bank 0 is 07C0 0000-07FF FFFF.
This was apparently done to get a contiguous chunk with accelerator RAM.
The RM is wrong as RAM starts at $0700 0000 for 16 MB, 64 MB were reserved for 'future' mainboard RAM.
-
I tried also 07F0 0000 as I have now 1MB: no success.
I've found a diagnostic rom for Amiga Forever (I have a license): logica-dialoga-200.rom. This may test better my A3000; it looks fine in WinUAE. How can I load this ROM (512k file) into my A3000 ?
-
I checked with a multi-meter, the DIP and ZIP sockets are connected on bank 0; seems ok to me and I just don't like to remove the motherboard completely... Yes, I'm trying to
The 1 meg Zips use one more multiplexed address line (ADDR 9) than the 256K Dips use. But if your sure you have continuity from the Dip to Zip sockets on all pins then the open connection could be from Ramsey to the Dip sockets. Also, J852 is long way away from Ramsey so I would check for possible open connection there. One side has +5V, the other Ground and the center to Ramsey pin 12. A3000 Schematics can be downloaded from many websites. A good small program from Aminet is Checkmem. Use the "Force" option to check areas outside of the memory list and use with Cpu Nodatacache.
Can someone post a ShowConfig output of his A3000 with FastMEM on the motherboard ? Thanks
RAM: Node type $A, Attributes $505 (FAST), at $7000000-$7FFFFFF (16.0 meg)
-
I bought an Oktagon 2008 on eBay, just for testing my zip ram chips and there is a problem at boot. The autoconfig screen shows this board 2092/8 as DEFECTIVE.
What do you think ?
a) my chips are defective
b) the Oktagon is defective
@VingtTrois I will have to send you my chips for testing. If you agree, please send me your address.
-
Take the chips away and look again.
-
Take the chips away and look again.
I did it and googled for: Amiga Oktagon 2008 zip orientation
and found this: http://brain-plasticity.safe-install.com/ (http://brain-plasticity.safe-install.com/) :whack:
and then I RTFM of the Oktagon in german :rtfm:
and found that I mounted the chips in the wrong orientation :madashell:
I fliped the chips and everything is ok.
> cpu nodatacache
> checkmem times 1000
... still running without error on the first half 8Mb.
Those chips are very resistant (also the Oktagon 2008...) !
Now for the A3k mb zip sockets, I did put the chips with pin 1 toward the caps, on the left when looking at sockets from the from of the A3k. Is that ok ?
-
Now for the A3k mb zip sockets, I did put the chips with pin 1 toward the caps, on the left when looking at sockets from the from of the A3k. Is that ok ?
Yes, thats correct. Bank 0 is from U850-U857. Now you need to find out why they don't work in the Zip sockets. If you have a PLCC extractor tool you could try removing Ramsey and reinstalling it. Also, it would be easy to do a continuity test from the Zip sockets to the Ramsey socket after the chip was pulled.
-
Hello Amiga fans,
After 15 years, I post where did this go and what is the problem now.
I don't remember when, but my A3000 was fine with 2M chip + 16M fast zip RAM. I think the problem was solved by setting some jumper in the correct position.
And now, recently I flushed some Zorro cards around. At some point everything worked including +8M zip on my Oktagon 2008 and a SCSIknife on it. Then I removed the Oktagon to attach the SCSIknife to it, reinserted the card and the 16M zip disappeared.
I've seen another topic here on this forum: 'Amiga 3000 fastram suddenly disappeared' https://forum.amiga.org/index.php?topic=74971.0
So as suggested I used ziptest. It shows faults on U850, U858 all the way up to U875. If I swap those zip with U854 .. U879, the result is the same.
Can anyone help me to solve this issue ? Should I ask in the other topic ?
Thank you
-
Does the fast memory work normally again if you remove all of the Zorro expansions?
I gather you've got 32 x 1Mbit ZIPs fitted to the A3000. Are you saying that ZIPtest indicates that the first three 4MB blocks of fast memory is not working, but the last 4MB is showing as good? Please post the entire test output from ZIPtest here as it's not obvious what's possibly good or bad, or what aspect of the test fails.
If you briefly pull the _FAIL line low, does the fast memory then appear normally afterwards? You can do this by briefly linking U190 (real time clock IC) pin 2 to ground (the metal chassis) with the system running, which will force a hardware reset and initialise the fast memory controller (U890) to the correct power-up state.
Failing that, it may be worthwhile collecting more clues using DiagROM and using the 'Slow scan of 16MB fastmem-areas' to see what's going on. Again, post the test result here.
-
Hello and thank you for your reply.
Does the fast memory work normally again if you remove all of the Zorro expansions?
not anymore
I gather you've got 32 x 1Mbit ZIPs fitted to the A3000. Are you saying that ZIPtest indicates that the first three 4MB blocks of fast memory is not working, but the last 4MB is showing as good? Please post the entire test output from ZIPtest here as it's not obvious what's possibly good or bad, or what aspect of the test fails.
Here is just the result of the first part of ziptest:
ZIPTest 1.1 (2024-01-17) by Chris Hooper
CPU: 68030 with Burst
Memory controller: Ramsey-04 $d $38 (25.05 MHz)
Memory config: 1Mx4 (4MB per bank)
Memory refresh: 240 clocks (9.60 usec)
Data line test
Socket ADDR IO1 IO2 IO3 IO4 Socket ADDR IO1 IO2 IO3 IO4
-------- ------- ---- ---- ---- ---- -------- ------- ---- ---- ---- ----
U881 3.7 71ffffb Good Good Good Good U879 3.5 71ffffa Good Good Good Good
U873 2.7 75ffffb Good Good Good Good U871 2.5 75ffffa Good Good Good Good
U865 1.7 79ffffb Good Good Good Good U863 1.5 79ffffa Good Good Good Good
U857 0.7 7dffffb Good Good Good Good U855 0.5 7dffffa Good Good Good Good
U880 3.6 71ffffb Good Good Good Good U878 3.4 71ffffa Good Good Good Good
U872 2.6 75ffffb Good Good Good Good U870 2.4 75ffffa Good Good Good Good
U864 1.6 79ffffb Good Good Good Good U862 1.4 79ffffa Good Good Good Good
U856 0.6 7dffffb Good Good Good Good U854 0.4 7dffffa Good Good Good Good
U877 3.3 71ffff9 Good Good Good Good U875 3.1 71ffff8 1 1 1 1
U869 2.3 75ffff9 Good Good Good Good U867 2.1 75ffff8 1 1 1 1
U861 1.3 79ffff9 Good Good Good Good U859 1.1 79ffff8 1 1 1 1
U853 0.3 7dffff9 Good Good Good Good U851 0.1 7dffff8 1 1 1 1
U876 3.2 71ffff9 Good Good Good Good U874 3.0 71ffff8 1 1 1 1
U868 2.2 75ffff9 Good Good Good Good U866 2.0 75ffff8 1 1 1 1
U860 1.2 79ffff9 Good Good Good Good U858 1.0 79ffff8 1 1 1 1
U852 0.2 7dffff9 Good Good Good Good U850 0.0 7dffff8 1 1 1 1
I can post more logs, starting with Bank0 only. Note that nibble 0 and 1 are faulty, anyhow Bank0 only or all populated.
If you briefly pull the _FAIL line low, does the fast memory then appear normally afterwards? You can do this by briefly linking U190 (real time clock IC) pin 2 to ground (the metal chassis) with the system running, which will force a hardware reset and initialise the fast memory controller (U890) to the correct power-up state.
... will do that, I'm printing the schematics big enough for my eyes.
Failing that, it may be worthwhile collecting more clues using DiagROM and using the 'Slow scan of 16MB fastmem-areas' to see what's going on. Again, post the test result here.
I don't have DiagROM now. I have just one pair of EPROMs and a programmer. Please tell me which files I need flash so I do it the right way.
I appreciate your help :)
PS:
1. I have an Indivision ECS in the mix
2. found DiagROM.zip and DiagROMV2.zip via archive.org (diagrom.com is down for me)
3. have a pair of TC574200D and an UV eraser