Troubleshooting an E1088 - Warning, long post

[ramegoom]

I'm trying to troubleshoot a friend's E1000 series machine. Here's what I have and maybe the experts will chime in.
The machine doesn't boot up, but instead, shows a "6" in the far right digit, all others blank. I believe when a 6 is shown, it indicated a memory problem. However, it doesn't move on to the next digit; it just sits at 6. Reset buttons on the hopper have no effect. The problem appears to be in the CPU board. Power supply checks OK, voltage on CPU checks OK.

Next, I pulled a CPU from a 2000 series machine, similar to the 1000 CPU - but different. I removed all four EPROMs from the 1000 board and installed them on the 2000 board, M1>M1, M2>M2, M3>M3, M7>P1 slots. Now, the machine fires up, although I have all zero's in the meter, well actually 00 (space) 000, would indicate 00 coins last played, 000 payed out. And, I get the "Insert Coin" to light. However, when I insert a coin, it immediately tilts. If I reset, it clears the tilt, and sets the Insert Coin lamp, and then I can sequence thru the diagnostics, i.e. 00004 pays 10 coins thru the hopper, 00005 prints out the six digit code of the M7 chip, and I can cycle thru the solenoids. But I noticed the reader light bulbs don't light when the Insert Coin is on (the reel readers have the large resistor, so I believe it has incandescent readers, and has the 50 volt fuse). Normal diagnostics dictate that if you get a 6, a second or two later, you get another number that is indicative of the actual chip that is at fault. But this one seems to hang up on the 6 and go no further. Oh, and all fuses check OK.

Next, I swapped the RAM chips and the microprocessor from the 2000 board to the 1000 board. Same 6 on the right digit. I swapped a few of the other socketed chips as well, still the 6 and no activity from the reset buttons. Pulled the RAM board off the CPU sockets, can't see anything glaring wrong. This machine is pretty clean, although the battery was dead-dead. I cut the battery out of the RAM board - it was a lithium battery and the board used the diode instead of the resistor so it wouldn't charge. I will replace the battery with a new Lithium battery.

I'm thinking the CPU is toast, can't find any burned components though. Nothing really heats up except the 5V regulators, which I believe is normal. The CPU gets slightly warm, but swapping the CPU with a known good one didn't change the problem.

I am open to any suggestions. Anyone see this hung-up-on-6 code before? Thanks for letting me bore you with a long post.

So, it appears that the 2000 CPU only partially works in an earlier 1000 machine.

[dhellis]

Typically a bbbbb6 (where b=blank) is an indication of a memory error but I have found that there are other things that can cause it. I would start with swapping
the ram chips on the E1000 ram board with known good chips. Next I would re-solder the connections that the ram board connects to. After soldering
take a small wire brush and clean the crap out of the pins and carefully inspect the pin sockets on the ram board for broken pins, I have seen more than just
a couple where these pins will break. I use a crappy USB microscope and it is still a bit difficult to tell.

Check continuity between Pin 7 of J6 on the MPU (Left side set of pins for the ram board) to pin 4
of U26 this is the Not Reset line (indicated by a bar over the word reset). Check the logic level on pin 3 of U26, that should be at a low and transition to high
when pressing the reset switch. If this is not happening then possibly U26 but could also be the small relay or even U5.

I have only seen 2 bad CPU chips and in both cases the error was 0bbbbb

[ramegoom]

Thanks Dan. The CPU chip is good, as evidenced by the swap with a known-good chip. I am thinking I will put soldered-in jumpers between the RAM board and the CPU board, just eliminate the connectors altogether. Also, I swapped the RAM chips, still sits at 6. It almost seems like the crystal isn't oscillating, where the CPU can't advance to the next instruction. I will try powering up the CPU and scope it out to see if the system clock is working.

Next step would be to replace the CPU board with a known-good one. Odd that the 2000 CPU 'sort of' works.

[dhellis]

Thanks Dan. The CPU chip is good, as evidenced by the swap with a known-good chip. I am thinking I will put soldered-in jumpers between the RAM board and the CPU board, just eliminate the connectors altogether. Also, I swapped the RAM chips, still sits at 6. It almost seems like the crystal isn't oscillating, where the CPU can't advance to the next instruction. I will try powering up the CPU and scope it out to see if the system clock is working.

Next step would be to replace the CPU board with a known-good one. Odd that the 2000 CPU 'sort of' works.

From what I have seen, if the clock is not running you will get b0bbbb or no display at all. Just the same, check at the crystal
that should be 4 Mhz, and pin 5 of U30 should be 2 Mhz. This problem can be a bear, I know because I have a board
that passes all of the tests on my tester but gets that same 6 error code when on the simulator. I have swapped known
good ram cards and swapped the M1 - M7 chips. I am about to the point of trying to locate some replacement pins
because all of the chips on the ram board have been replaced, the edge connectors have been replaced and the ram board
is good when swapped to another E1 board. The only thing left is those stinking pins and they were really lousy to begin
with.

[ramegoom]

I replaced the pin headers that attach the RAM board, and reflowed the solder connections on the two connectors on the board. Also checked for tension on each pin. Trouble is, I don't have an E1000 machine to test the board. I tried putting E2000 eproms into it and plugged it into my E2000, but it didn't work. So, apparently the CPU's are not forward-compatible. I'm going to use the shotgun approach next. I can't find anything glaring wrong; regulators are all up to correct voltages, capacitors test OK, no weird looking solder joints, and current draw on the board is normal. Dhellis, I might just go ahead and send it to you to have a shot at it.

[dhellis]

I replaced the pin headers that attach the RAM board, and reflowed the solder connections on the two connectors on the board. Also checked for tension on each pin. Trouble is, I don't have an E1000 machine to test the board. I tried putting E2000 eproms into it and plugged it into my E2000, but it didn't work. So, apparently the CPU's are not forward-compatible. I'm going to use the shotgun approach next. I can't find anything glaring wrong; regulators are all up to correct voltages, capacitors test OK, no weird looking solder joints, and current draw on the board is normal. Dhellis, I might just go ahead and send it to you to have a shot at it.

I don't charge for testing, except for return postage and if there is something wrong I would contact you before making
any repairs. Shipping address and contact information is located at   http://www.coinslots.com/our-thanks/

[rjpohl]

I would suggest you send both boards to Dan so he can test them both.  It's good piece of mind to know both are good.
Dan's repaired a few for me, he's fast too.
Bob

[dhellis]

I would suggest you send both boards to Dan so he can test them both.  It's good piece of mind to know both are good.
Dan's repaired a few for me, he's fast too.
Bob

Thanks Bob and yes, send the MPU board and IO board, shipping costs are going to be about the same, especially if
you are using flat rate.

[ramegoom]

I'll get the board sent to you via UPS. Also, I'm going to send a couple other things you could use as well...:-)

[dhellis]

I'll get the board sent to you via UPS. Also, I'm going to send a couple other things you could use as well...:-)

When sending the boards, include the ram boards and eproms as well along with the IO board
that way I can test everything as a set.