Progress on recently acquired (broken) A500+

demolition · 18 May 2013, 23:09

Bought a 500+ not long ago on Amibay, and just starting a thread here with my progress in fixing it (or trying to).

So far I have verified what the seller says that the dodgy looking tracks all measure fine, however I've 'reinforced' them as well as several nearby vias. I've reseated all DIL chips (and all the pins in the sockets looked fine), trying booting without the CIA chips, and tried switching with another working Gary chip.

I've changed JP3 to switch high/low chipmem and still green screen, so I don't think one of the memory chips are at fault, however I have some sockets and new ram chips in the mail, so I can replace them if needed.

I checked all data and address bus lines and they also measure fine.

I have a PLCC puller in the mail and when I get it, I will take out Agnus (although seller already had it out and cleaned it). But I have another identical Agnus in my A600 I can test with.

If this also doesn't affect the problem, I'll bring the board to work and try to capture the chip select signals on the chipmem on a scope.

Any other ideas?

Loedown · 19 May 2013, 05:10

If changing the jumper still results in a green screen then logic dictates that it could be a stuck address / data line, the OE pin may not be changing state or CS. Do resistive checks to Vcc and GND yield the same results for each data and address pin? You may have one of the pullup resistor networks in backwards ( I've seen this before with Commodore's kwality )

Look also for stuck reset lines to various chips. Just my thoughts without looking at the schematic.

roy bates · 19 May 2013, 05:17

you could try to replace parts on the data path if its not the memory.
then check the sockets for corosion on gary and all other parts.
i would be looking at u12 first then u32,u34 u35 etc.(replacing one at a time)

demolition · 01 June 2013, 01:37

Finally got a little time to spend on this and have a few more conclusions.

I replaced all chip ram with some good quality sockets and put in 8 brand new ram chips. No change, so the old rams were probably good.

All address and data pins on the RAM chips are high impedance relative to Gnd/Vcc and connections towards U34/U35 are as they should be. _OE are connected to Gnd on all chips and _WE measure ~68 ohms against U35, so this is also good.

When I wrote the original post, I forgot that Agnus in the 600 was soldered in, but I managed to find another 2MB Agnus to test with, and that also didn't help.

RAS is pulled to Vcc with 4.7k on all chips and CAS measures 68 ohm against U35.

Seems to me like there isn't many places left for the problem to lie, and that would be the chips that roy bates mentions, U12, U32, U34 and U35. So that should probably be my next focus.

kipper2k · 01 June 2013, 01:41

Check area around Gary, remove the old socket, put in a new one as the original socket will prob have oxides on them that cant be properly cleaned and very important, clean all pins on Gary itself by sanding them as the oxides will cause bad connections. On the three A500+ i have fixed the problem has not been bad ram even though the A500+ would give a green screen on boot

demolition · 01 June 2013, 01:52

Ok, will do that. I did try with another known working Gary chip but if the socket is bad that won't do any good obviously, although I did try cleaning the socket with some contact spray. Not a magic product, I know.

Gary is located close to the battery, so it seems likely that the problem could be around there. When I have the socket out, it will also give me a chance to check out all the traces below the socket.

Edit: Just did a continuity check from the top of the Gary pins to the back side of the PCB and all pins checked out.

roy bates · 01 June 2013, 13:18

i would start replacing the 74 logic now as its the only parts left apart from u32(the 74ls244 under gary first then u32 and the rest of the data bus after if it still dont wont)
youll see why i said u32 after the 74ls244 when you see what it is.

demolition · 02 June 2013, 01:08

Did some more poking around in there, but still no resolution.

Don't have replacement 74xx ICs yet, so couldn't do much but look at the signals with the scope. Looks like all dram address and data pins are floating when not used, so no outputs seems to be hanging.
I checked U10 and U12 and they also look like they are working, buffering the input to the output pins when _OE was low. The only thing was that the voltage of some of the output pins were a little low when pulling high. The high voltage was on some pins only about 3.7V, but since that is well above the minimum specs of the 74LS244, I'm guessing that shouldn't be a problem.

I could also verify that CAS, RAS and WE were being asserted on the ram chips. I didn't look at all signals together, so couldn't verify that it was actually at the correct times.

Was thinking that it could be useful to create a custom kickstart ROM which would just do read/writes at a couple of specific dram addresses in a loop. This way, it would be much easier to see if what is happening on the various lines is actually correct.

GadgetUK · 02 June 2013, 20:00

Thanks for posting this, I am in exactly the same boat. I believe the RAM is OK on mine also.

Long story short...

Battery leak.
Initially just black screen.
Removed battery and cleaned up.
Extensive testing done on all PCB tracks surrounding.
2 tracks repaired. (green screen 11 flashes then reboot from this point)
Gary socket replaced as precaution - no corrosion or dmg under it.
Address and data lines of RAM checked.
Piggy backed RAM to quickly rule out open lines interally.
CIA chips swapped around, and replaced with known working.
Just replaced the 74LS244N nearest battery, corroded bad, no damage under or around it. Still green screen.
Just removed 74LS373N (chip on order), no dmg under it and all nearby tracks OK.

I have checked the tracks around the battery until my eyes hurt. One thing i've just spotted is in the top left pin of the right hand CIA, there acid in that one pin of the socket, but nowhere else on the PCB inbetween?!?!?

I do know from experience that a faulty CIA can give green screen. I will resocket that CIA after ive fitted the 373 when it arrives.

Edit: also done the following btw. Checked PSU (working on 500 and 1200, voltages 100%
Swapped Agnus - wasnt exact model so It was running in NTSC mode, still green screen, put original back. Soldered (for good measure) agnus pins, CPU pins and ROM pins. Tried another kickstart, paula, super denise, CPU and Gary... Its becoming one of these 'OMG wtf is wrong with it 500's'... Last time I had one like this it was the RAM and I ended up changing all 16 (a500 not +) DRAM chips, I hope to god its the 373 or that CIA socket (hoping bad track under it). I could of course have missed a track next to the battery. Unfortunately my scope is in storage atm =\

GadgetUK · 02 June 2013, 21:04

I am measuring 120 ohms from 5v rail to ground, that seems low to me?

mc6809e · 02 June 2013, 21:32

Quote:

Originally Posted by GadgetUK

I am measuring 120 ohms from 5v rail to ground, that seems low to me?

5V^2/120ohms = 208 W

That does seem high. An A500 shouldn't draw more than 35W. It's possible a chip has an internal short.

Let the machine cool, then turn on the power and immediately being feeling chips. If one chip has an internal short, it will become hot very quickly -- much more quickly than the other chips.

GadgetUK · 02 June 2013, 21:41

Quote:

Originally Posted by mc6809e

5V^2/120ohms = 208 W

That does seem high. An A500 shouldn't draw more than 35W. It's possible a chip has an internal short.

Let the machine cool, then turn on the power and immediately being feeling chips. If one chip has an internal short, it will become hot very quickly -- much more quickly than the other chips.

Hmmm, more confused now. I am measuring across the VCC pin of the 74xx chips to ground (without power connected), and 120 ohms seemed low to me, not high.

demolition · 02 June 2013, 21:43

Can't conclude anything from that. My working A600 measures 90 ohm from Vcc to Gnd when off. That doesn't mean that the resistance from Vcc to Gnd stays at 90 ohms when powered on.

Edit: Also, 5^2 / 120 = 209mW.

GadgetUK · 02 June 2013, 21:45

I've ordered a 139, and replacement 74F244PC and 74F373PC as well, might not need them just want the spares anyway. I will post an update when the first 373 arrives and has been fitted.

Quote:

Originally Posted by demolition

Can't conclude anything from that. My working A600 measures 90 ohm from Vcc to Gnd when off. That doesn't mean that the resistance from Vcc to Gnd stays at 90 ohms when powered on.

Edit: Also, 5^2 / 120 = 210mW.

That's fine, thanks for info, that's reassuring if anything,

mc6809e · 02 June 2013, 22:54

Quote:

Originally Posted by GadgetUK

Hmmm, more confused now. I am measuring across the VCC pin of the 74xx chips to ground (without power connected), and 120 ohms seemed low to me, not high.

The lower the resistance, the higher the current and power.

mc6809e · 02 June 2013, 22:55

Quote:

Originally Posted by demolition

Can't conclude anything from that. My working A600 measures 90 ohm from Vcc to Gnd when off. That doesn't mean that the resistance from Vcc to Gnd stays at 90 ohms when powered on.

Edit: Also, 5^2 / 120 = 209mW.

That's true. I'd still hunt around for anomalously hot chips.

roy bates · 02 June 2013, 22:59

parts dont necessarily need to get hot if there faulty.

demolition · 02 June 2013, 23:16

No, but if one chip gets significantly hotter than the rest, then something is probably up with that one.

In my A500+, the hottest ICs are the CPU, Agnus and Denise which all seems approx the same temp. Slightly less hot is Gary and Paula. I can't feel the heat from the RAM chips, and also barely from the ROM.

GadgetUK · 03 June 2013, 11:47

Quote:

Originally Posted by demolition

No, but if one chip gets significantly hotter than the rest, then something is probably up with that one.

In my A500+, the hottest ICs are the CPU, Agnus and Denise which all seems approx the same temp. Slightly less hot is Gary and Paula. I can't feel the heat from the RAM chips, and also barely from the ROM.

Exactly the same in mine. I would hazard a guess that the warmest chips are CPU and Agnus. Cant feel anything on RAM.

I might get my laser thermometer onto the job later just to see if one RAM chip is warmer but I don't think its a RAM fault.

Occams Razor - It's going to be a damaged track I reckon. Although whats going through my mind is that many years ago maybe this 500+ green screened (data path, RAM or multiplexer fault) and then went into storage until now.

demolition · 09 June 2013, 00:45

So I just replaced the Gary socket with a high quality one and no change. I removed the 6 logic chips, U10, U11, U12, U13, U34 and U35 and replaced them with sockets as well so I could try out various ICs. Turns out the problem was with one of the LS373 ICs (U11/U13). Replacing it gave me a nice purple screen with a magically floating floppy disk.

Now that I have a MB with quality sockets for the logic and RAM will also make it a lot easier in case I need to test ICs in the future.

18 May 2013, 23:09	#1
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	Progress on recently acquired (broken) A500+ Bought a 500+ not long ago on Amibay, and just starting a thread here with my progress in fixing it (or trying to). So far I have verified what the seller says that the dodgy looking tracks all measure fine, however I've 'reinforced' them as well as several nearby vias. I've reseated all DIL chips (and all the pins in the sockets looked fine), trying booting without the CIA chips, and tried switching with another working Gary chip. I've changed JP3 to switch high/low chipmem and still green screen, so I don't think one of the memory chips are at fault, however I have some sockets and new ram chips in the mail, so I can replace them if needed. I checked all data and address bus lines and they also measure fine. I have a PLCC puller in the mail and when I get it, I will take out Agnus (although seller already had it out and cleaned it). But I have another identical Agnus in my A600 I can test with. If this also doesn't affect the problem, I'll bring the board to work and try to capture the chip select signals on the chipmem on a scope. Any other ideas?

01 June 2013, 01:52	#6
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	Ok, will do that. I did try with another known working Gary chip but if the socket is bad that won't do any good obviously, although I did try cleaning the socket with some contact spray. Not a magic product, I know. Gary is located close to the battery, so it seems likely that the problem could be around there. When I have the socket out, it will also give me a chance to check out all the traces below the socket. Edit: Just did a continuity check from the top of the Gary pins to the back side of the PCB and all pins checked out. Last edited by demolition; 01 June 2013 at 02:12.

02 June 2013, 20:00	#9
GadgetUK Registered User Join Date: Mar 2013 Location: Manchester Posts: 189	Thanks for posting this, I am in exactly the same boat. I believe the RAM is OK on mine also. Long story short... Battery leak. Initially just black screen. Removed battery and cleaned up. Extensive testing done on all PCB tracks surrounding. 2 tracks repaired. (green screen 11 flashes then reboot from this point) Gary socket replaced as precaution - no corrosion or dmg under it. Address and data lines of RAM checked. Piggy backed RAM to quickly rule out open lines interally. CIA chips swapped around, and replaced with known working. Just replaced the 74LS244N nearest battery, corroded bad, no damage under or around it. Still green screen. Just removed 74LS373N (chip on order), no dmg under it and all nearby tracks OK. I have checked the tracks around the battery until my eyes hurt. One thing i've just spotted is in the top left pin of the right hand CIA, there acid in that one pin of the socket, but nowhere else on the PCB inbetween?!?!? I do know from experience that a faulty CIA can give green screen. I will resocket that CIA after ive fitted the 373 when it arrives. Edit: also done the following btw. Checked PSU (working on 500 and 1200, voltages 100% Swapped Agnus - wasnt exact model so It was running in NTSC mode, still green screen, put original back. Soldered (for good measure) agnus pins, CPU pins and ROM pins. Tried another kickstart, paula, super denise, CPU and Gary... Its becoming one of these 'OMG wtf is wrong with it 500's'... Last time I had one like this it was the RAM and I ended up changing all 16 (a500 not +) DRAM chips, I hope to god its the 373 or that CIA socket (hoping bad track under it). I could of course have missed a track next to the battery. Unfortunately my scope is in storage atm =\ Last edited by GadgetUK; 02 June 2013 at 20:08.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
My A500+ progress so far.	Mogsy	Hardware pics	16	02 March 2013 09:20
Broken A500	absence	support.Hardware	22	09 October 2011 23:13
Recently acquired A3000	AmigaJules	Hardware pics	9	07 October 2008 10:24
A500 VS IBM PS/1 - My project (Work in progress)	aquanaut	Hardware pics	21	17 May 2008 07:27

19 May 2013, 05:10	#2
Loedown Precious & fragile things Join Date: Feb 2009 Location: Victoria, Australia Posts: 1,946	If changing the jumper still results in a green screen then logic dictates that it could be a stuck address / data line, the OE pin may not be changing state or CS. Do resistive checks to Vcc and GND yield the same results for each data and address pin? You may have one of the pullup resistor networks in backwards ( I've seen this before with Commodore's kwality ) Look also for stuck reset lines to various chips. Just my thoughts without looking at the schematic.

19 May 2013, 05:17	#3
roy bates Registered User Join Date: Apr 2011 Location: birmingham Age: 55 Posts: 2,827	you could try to replace parts on the data path if its not the memory. then check the sockets for corosion on gary and all other parts. i would be looking at u12 first then u32,u34 u35 etc.(replacing one at a time)

01 June 2013, 01:37	#4
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	Finally got a little time to spend on this and have a few more conclusions. I replaced all chip ram with some good quality sockets and put in 8 brand new ram chips. No change, so the old rams were probably good. All address and data pins on the RAM chips are high impedance relative to Gnd/Vcc and connections towards U34/U35 are as they should be. _OE are connected to Gnd on all chips and _WE measure ~68 ohms against U35, so this is also good. When I wrote the original post, I forgot that Agnus in the 600 was soldered in, but I managed to find another 2MB Agnus to test with, and that also didn't help. RAS is pulled to Vcc with 4.7k on all chips and CAS measures 68 ohm against U35. Seems to me like there isn't many places left for the problem to lie, and that would be the chips that roy bates mentions, U12, U32, U34 and U35. So that should probably be my next focus.

01 June 2013, 01:41	#5
kipper2k Registered User Join Date: Sep 2006 Location: Thunder Bay, Canada Posts: 4,323	Check area around Gary, remove the old socket, put in a new one as the original socket will prob have oxides on them that cant be properly cleaned and very important, clean all pins on Gary itself by sanding them as the oxides will cause bad connections. On the three A500+ i have fixed the problem has not been bad ram even though the A500+ would give a green screen on boot

01 June 2013, 13:18	#7
roy bates Registered User Join Date: Apr 2011 Location: birmingham Age: 55 Posts: 2,827	i would start replacing the 74 logic now as its the only parts left apart from u32(the 74ls244 under gary first then u32 and the rest of the data bus after if it still dont wont) youll see why i said u32 after the 74ls244 when you see what it is.

02 June 2013, 01:08	#8
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	Did some more poking around in there, but still no resolution. Don't have replacement 74xx ICs yet, so couldn't do much but look at the signals with the scope. Looks like all dram address and data pins are floating when not used, so no outputs seems to be hanging. I checked U10 and U12 and they also look like they are working, buffering the input to the output pins when _OE was low. The only thing was that the voltage of some of the output pins were a little low when pulling high. The high voltage was on some pins only about 3.7V, but since that is well above the minimum specs of the 74LS244, I'm guessing that shouldn't be a problem. I could also verify that CAS, RAS and WE were being asserted on the ram chips. I didn't look at all signals together, so couldn't verify that it was actually at the correct times. Was thinking that it could be useful to create a custom kickstart ROM which would just do read/writes at a couple of specific dram addresses in a loop. This way, it would be much easier to see if what is happening on the various lines is actually correct.

02 June 2013, 21:04	#10
GadgetUK Registered User Join Date: Mar 2013 Location: Manchester Posts: 189	I am measuring 120 ohms from 5v rail to ground, that seems low to me?

02 June 2013, 21:43	#13
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	Can't conclude anything from that. My working A600 measures 90 ohm from Vcc to Gnd when off. That doesn't mean that the resistance from Vcc to Gnd stays at 90 ohms when powered on. Edit: Also, 5^2 / 120 = 209mW.

02 June 2013, 22:59	#17
roy bates Registered User Join Date: Apr 2011 Location: birmingham Age: 55 Posts: 2,827	parts dont necessarily need to get hot if there faulty.

02 June 2013, 23:16	#18
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	No, but if one chip gets significantly hotter than the rest, then something is probably up with that one. In my A500+, the hottest ICs are the CPU, Agnus and Denise which all seems approx the same temp. Slightly less hot is Gary and Paula. I can't feel the heat from the RAM chips, and also barely from the ROM.

09 June 2013, 00:45	#20
demolition Unregistered User Join Date: Sep 2012 Location: Copenhagen / DK Age: 43 Posts: 4,190	So I just replaced the Gary socket with a high quality one and no change. I removed the 6 logic chips, U10, U11, U12, U13, U34 and U35 and replaced them with sockets as well so I could try out various ICs. Turns out the problem was with one of the LS373 ICs (U11/U13). Replacing it gave me a nice purple screen with a magically floating floppy disk. Now that I have a MB with quality sockets for the logic and RAM will also make it a lot easier in case I need to test ICs in the future.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)