DiagROM V1.0 Released

[CRP]

sorry I used the wrong words, here is what I did in order:
1) connected usb-to-ttl adapter to my pc, and there I connected tx and rx pins to make a loopback and verify that the adapter is working, which it is.
2) connected rs232-to-ttl adapter to the above, and repeated loopback test at db9 connector, again everything ok.
3) connected amiga serial port to rs232 adapter, pins tx->rx, rx->tx and gnd->gnd, but when I turn on amiga I get no consistent output, just occasionally random characters, but they look more like the result of noise to me than an incorrectly transferred output from the amiga; nonetheless, diagrom's status line shows serial 9600 bps, so it does recognize something at the serial port.
4) suspecting that the ami's serial port is not working, built a loopback connector for the amiga, connecting pins 2-3, 4-5-6, 8-20-22 but did not attach the leds to +12v and -12v pins, started diagrom's serial test and they all fail, although test passes are much quicker than when nothing is attached to serial port, so again I guess this means that something is recognized.

So my question is: should I first get leds to complete the loopback, or should I move straight to debugging the components in the amiga around the rs232 port? I already swapped the CIAs, and read I should also check caps and resistors, and finally U38 and U39.

thanks

Originally Posted by Chucky

you should not SHORT anything.. you should CROSS so TX on the Amiga is connected to the RX on the PC and the RX on the Amiga is connected to TX on the PC..

if you get an echo. you have shorted and done it wrong as when you send data. it is getting back to the PC again.. and thats wrong and a reason why you do not get any data.

[Chucky]

Diagrom always say 9600 bps unless you have a loopback adapter attached (then it says LOOP)

so. if you get random data and are using 9600 BPS on your PC.. paula is not getting all 8 bits needed to produce serial data.. so I would do a continuty-test on all datapins from paula.

[tnt23]

Originally Posted by CRP

I switched the two CIAs, but nothing changed. Can I assume the two rs232 chips are fried?

On an A500 I was resurrecting for a friend, either U38 or U39 was bad and would not let the computer start normally. Got extremely hot right from turning power on. (We just chopped both off, did not bother having working RS232). Could be worth checking.

[plasmab]

The CIA does not do serial on the amiga. It’s Paula

EDIT: the CIA’s are just used for the handshakes which DiagROM does not use

[Chucky]

I do not think the TTLs are dead. they are hard to kill. I think that you have issues with datalines to paula it seems that you can get picture on diagrom. do a audiotest and listen.. same signals are used there.

Paula handles serial.. NOT CIA.. this is a common misstake.

[crashserious]

Minor update. I desoldered a bit over half the memory chips today (I was finding inconsistencies I couldn't figure out and had new ones in hand.


I did find that CAS on the higher numbered 6 of 32 (Last?) chips was missing completely. Which lead me to a bad CAS trace between the 7th and 6th highest numbered chip. It does sort of appear to have been some form of corrosion, though there's no damage close to the battery so I can't account for it's source.



Getting there. A rev 4.5 schematic would come in so handy right now since it has more chips than a rev 6. grrrr.

Originally Posted by Chucky

this is most likely not a memorychip. but a trace that is bad.
but IF memory (it can draw thebit all high.. or all low) a easy method figuring out what memorychip it is. is to short the datapins of the memory ICs and look on screen (as it is updated in realtime) and see what chip you can short without 2 stripes gets changed. then you have found the chip.

but now I am quite surtain it is a trace or pullup/down resistor.

[crashserious]

Making progress!!! It's no longer displaying a green screen! Now it's YELLOW with teh Kickstart ROM in, and DIAGROM appears to work just fine!

I replaced all the RAM chips for good measure (and socketed them) and fixed a CAS trace between the 25th and 26th RAM chip causing CAS to not be present of the last 6 chips. I replaced all the 74244's, 74245's and 74737's.


I also had to fix a data line (D8 I believe) from the 73737 between the ROM and RAM. (I may have damaged this trace replacing the 74737)

The tests all work as far as I can tell. but the Fast Mem (slow or fast) reports: "Bad Block start at $FFFFFFFF and ends at $00000003 with a size of 0KB".

The serial logs prior to this I don't see issues.

(One final tidbit that I tried on a whim.... I cut J500 since I also have a 8Mb Zorro II IDE card to put in this thing. I get greeted with a kickstart screen!!!! booting with gotek floppy works, after enabling the RAM on the Alfa Data AT-BUS 2008 by un-jumpering JMOFF it sees the memory and the IDE CF card is formatting now. The A2090A doesn't seem to work, but that could be a bad HDD I bought off ebay, since that drive is very small, I may not care a lot about it.)

[Chucky]

Either there is a bug.. or you do not have memory at that location..
I am in process of making a completley new memorytest..

[crashserious]

I've been thinking of getting a Programmer myself. I'd be willing to put the jumper back in and test it, if that helps.


At the risk of turning this into a request thread, I'd love to see some sort of indicator of which bar is which chip. I knew there were 32 of them-- but I didn't know what was where.

Originally Posted by Chucky

Either there is a bug.. or you do not have memory at that location..
I am in process of making a completley new memorytest..

[CRP]

I just got myself a cheap USB scope. The TTL serial signal comes out of Paula's pin 40 correctly. From there it goes into pin 9 of the MC 1488. The same signal, converted to RS232 levels, should come out of pin 8, but here all I get is a constant +12V signal, which should correspond to logical 0. So I am assuming that the MC1488 chip is dead. I will replace it and see what happens next.

UPDATE: replaced MC1488, now serial works fine

Originally Posted by Chucky

I do not think the TTLs are dead. they are hard to kill. I think that you have issues with datalines to paula it seems that you can get picture on diagrom. do a audiotest and listen.. same signals are used there.

Paula handles serial.. NOT CIA.. this is a common misstake.

I just received my Diagrom from GGLABS and I'm ready to fix several of my old A500 boards each with different issues. First system is a 6A with a chip mem error. I removed 1 chip where there seemed to be some liquid (coffee?) damage and replaced with new MEM chip but i think there is another or more that will need to be replaced. Can someone please help my figure out what this output is telling me. I looked all over the web and didn't find a "guide" for the Diagrom, if someone can please point me to one or good reference guides I would be very grateful. I REALLY want to fix these! See attached pic of Extended ChipMEM test.


CerebralCortex

[nolunchman]

my diagrom set works on a healthy 3000 motherboard and i am trying to use it on a damaged board (battery damage). i get no output from the serial port. although diagrom is not working with this board, that still may give clues where to look. what parts of the motherboard does diagrom need to still be functioning?

[Chucky]

Originally Posted by cerebralcortex

I just received my Diagrom from GGLABS and I'm ready to fix several of my old A500 boards each with different issues. First system is a 6A with a chip mem error. I removed 1 chip where there seemed to be some liquid (coffee?) damage and replaced with new MEM chip but i think there is another or more that will need to be replaced. Can someone please help my figure out what this output is telling me. I looked all over the web and didn't find a "guide" for the Diagrom, if someone can please point me to one or good reference guides I would be very grateful. I REALLY want to fix these! See attached pic of Extended ChipMEM test.


CerebralCortex

Addresserror can be much different things. this is a error exception so something is screwed.. my guess it is something handling memory.

also. it is not THAT often chips die. more traces/passives.

[Chucky]

Originally Posted by nolunchman

my diagrom set works on a healthy 3000 motherboard and i am trying to use it on a damaged board (battery damage). i get no output from the serial port. although diagrom is not working with this board, that still may give clues where to look. what parts of the motherboard does diagrom need to still be functioning?

Batterydamage on A3000 usually affects Paula, Denise etc. with datalines bad to paula serialoutput can be broken then it is harder. then you can check if powerled flahes, so you know that diagrom can run. then it is back to "oldfashioned" troubleshooting checking datalines etc to the chipset.

[nolunchman]

Originally Posted by Chucky

Batterydamage on A3000 usually affects Paula, Denise etc.

what about a 3000T motherboard? aside from the kick roms, the chips near the batt are the FPU and a few other unfamiliar ICs...
i've replaced the 40pin sockets and and jumped the traces but diagrom still doesn't output to serial. the paula and denise chips are on the opposite end of the motherboard.

here is a pic of the chips near the battery

https://1drv.ms/u/s!AnzbjMU9nUZpixIZhd8fols1iPke

[Chucky]

does powerled staert to flash?

and not sure but you maybe need to change some jumper is fpu is missing
cant remember now

[amigakit.com]

@Chucky

The ROMs are great work.

[nogginthenog]

@Amigakit - you should sell them (with a few € going back to Chucky of course)

[Chucky]

V1.1 out any day..

[Chucky]

http://www.diagrom.com/index.php/201...v1-1-released/