Need help with M307A (Integrating One Shot) for TC08

[gnupublic]

The M307A drives me crazy. I'm working on two TC08 tape controllers at a time. One is working (so lala). So I change the Flip Chips for getting the second to work also.
The second one has some defect FlipChips. One of them is the M307A, a one shot delay. This has to produce 2 delay times, one is 140ms, the other 70µs.
There are some test instructions for producing and measuring the time constants in the TC08 handbook.
With the M307B FlipChip from the first controller I get both times. But when I change to the M307A Flipchip, I don't get 140ms (70µs seems OK). I testet all parts on the board, but found nothing to be defect.



E2 9601

0+4 = pin 8 Output and pin 6 inverse Output
1= Input pin 1+2
2= Input pin 3
3= Input pin 4

There is the short XSTA L signal that triggers the beginning of the delay. It's going high to low during the longer positive pulse, so that the 9601 is triggered and starts the delay signal. The picture is from the working M307B.

From the drawing (thanks to Vince) I tried to pull up the SN7401 (open collector E3 pin 10) with a resistor of about 1K to VCC. Then I get the 140ms delay time signal!
But this is not very stable. I think the IC's (Fairchild 9601) are OK, it seems to tell me there is a problem with the working point.
Can anybody give me a hint how to debug this?

Thanks
Volker

 

[vrs42]

E2 9601

0+4 = pin 8 Output and pin 6 inverse Output
1= Input pin 1+2
2= Input pin 3
3= Input pin 4

Hmm...that's not the pin-out for E2 that I have on record for the M307A.
The connections to E2 differ between the A and B revisions.

I show pins 1 and 2 go to L2 and K2 respectively.
Pin 3 is tied to 3.3V, as derived by the voltage divider R11/R12.
Pin 4 is goes to J2.

E2 pins 3 and 4 are swapped in the B revision.

Are you saying that L2 and K2 are tied together in the TC08?

Your photo looks right to me for B, which you also say is working. Is there a similar picture for the non-working case?

Vince

 

[gnupublic]

Yes L2 and K2 are tied together, also E2=N1, U1=S1, V1=P1, A1=C1 through the backplane.

The picture above was from the working M307B.

Today I made some changes, like in the datasheet of 9601/DM9601:



This is very very simple, but now it works. I get the delay time. When I run os8diskserver it is possible to "DIR DTA=0" or "RU DTA= PFOCAL" from the tape. I even can write and boot a tape.

To do this I preared a socket with missing pin 11 and pin 13.



Here a small picture of my extender cable.



So I think the 9601 is OK, but I still don't understand why the original circuit does not run.

Have fun
Volker

 

[vrs42]

So I think the 9601 is OK, but I still don't understand why the original circuit does not run.

I agree your 9601 seems fine. I suspect a discrete, or perhaps even more likely a bad solder joint, short or similar issue with the PCB itself. I've seen "dirt" establish paths that should not be there, and also scratches that opened traces.

Vince