Error,

Yes, on-on-on-on = address 0. Switch 1 is the least significant bit, switch 4 is the most significant. On the newer encoder cards, switch 1 has no effect, as the boards need to be addressed to an even address.

As for the tests in the previous message, these show that the new encoder card (addr 0) is working correctly, and all other cards appear at the expected address.

You could swap the addresses of the encoder cards to check the other card. Since you were having the relay clattering problem in LinuxCNC, I would at least run the DIOtest mode of the diagnostics to test the master DIO card. First, remove the SSR connectors (ST3 and ST4) to avoid clattering whatever these DIO boards control. Set the watchdog jumper on all DAC boards to OFF to allow this test to run. Make sure the E-stop loop is connected to the master DIO board, so the green LED is lit. Then run :

sudo ppmcdiags 378 diotest

it should give the same "no error" reports as the commtest. The red LEDs on the DIO board should flicker, this test loads random values to all SSRs and then reads back the setting to confirm it received the setting correctly. If this test runs OK, you could try it with each DIO board set to address 6 (making sure no other board is set to addr 6). The diagnostic only tests the DIO board at addr 6. If this all works OK, then it is seems likely that the sparking relay is causing so much electrical noise that it is resetting all the logic. I would try testing it with that relay disconnected to see if everything else works, and if so, then you will have to apply some transient suppressors or something to that relay.

Jon

edit

So if you hold the card upside down so that pin one ins on the right. The switch settings look like this.

No, you have an error on address 2 - it should be 4-3-2-1 as on on off on

I think all the others are right.

Does it matter where the cards are on the back plane or just the address?

Here are the results from testing the I/O cards I have in this machine.

drive.google.com/file/d/0B86EwtSOzbrjSDM...V1E/view?usp=sharing

Does it matter where the cards are on the back plane or just the address?

No, all backplane slots are the same. Only the address switch is needed to select the board address.

Jon

Here are the results from testing the I/O cards I have in this machine.

The bus test looks OK.

OK, to test the DIO cards with the diotest function, it will only test the DIO board at address 6. It will allow another DIO board to be at address 7, and set that one to slave to the one at addr 6. But, if you have more than 2 DIO boards, it doesn't know to select those ALSO to slave mode. So, I think the diag will only work with one or 2 DIO boards in the system, not 4. So, you will have to pull the extra boards out at least far enough so they are not plugged into the backplane. It may be simpler to just test one at a time.
Whatever DIO board is at address 6 will have to have the E-stop loop made to be a closed circuit, so the green LED is lit on that board. And, the DAC board in the system has to have the watchdog jumper set to off for the test to work.

Jon

I did test them one at a time and the results are listed with serial number of the card in that zip file.

The green and red lights are all on they do not flicker.

Hmmm, this is really weird! Now, some of the older DIO boards did not have the readback feature in the firmware, so this result would be expected. But, there is one much later board (SN0041) that should have the latest firmware, so the readback on that one should be working.

When you power up the PPMC motherboard, do all the red lights come on immediately? That should not happen, the red LEDs should all be off, as the board should be in the power-up reset state, and in E-stop.

If you halt the diotest with ctrl/c, are all the red LEDs lit, or do they have some random pattern with some lit, some dark? (The diotest diagnostic sends random patterns to the SSRs and then reads back to see if the values match what it sent.) It goes so fast that the LEDs may appear all lit when the diag is running, but they are actually flickering on and off about 50%, but probably too fast to see.)

Jon