• Please review our updated Terms and Rules here

M8189 bootstrap error LED's D1 and D4 on

tradde said:
So was the M8029 in slot 5 before you added the RL02? Or did you move it? Could it be that the vector is still off from what it should be? It appears to be correct, but what is that "resistor like thing" pictured for V2 of the vector?
Click to expand...
The M8029 was moved from slot 3 to slot 6.
 
So, that is interesting.

In the 'none working' case the M8029 was in slot 3.

In the 'working' case no cards were in slot 3.

Is slot 3 faulty I ask myself?

You could prove this by removing the Rl controller from slot 4 and inserting the M8029 into slot 4.

I have had backplanes where the contacts internal to the connectors (on the backplanes) had become damaged.

Dave
 
That is exactly why I asked. Seems he moved the card in question and it went from non-working to working. I've seen stuff get down into these slots and cause issues. Good reason to blow them out just in case. Put some De-Oxit in there too maybe to clean the contacts?
 
Yep, always do one thing at any one time.

It always takes you longer - but it saves you getting thoroughly confused!

Dave
 
I noticed too in the MAP output after you moved the RLV11 and RXV21 boards that the address block 775130..775136 disappeared, and 777170..777176 and 774400..774416 appeared ...
 
Yes, if you change more than one thing and something fails or starts to work, you don't know which made it fail or fixed it. Been burned by that many a time. Now you have to step back and do one or the other things you did, but nothing else. Confused? That seems to be my natural state at time. Thoroughly confused takes more. :)

Could something wrong in socket 3 cause that address to show different? Thus giving this board a different address? Seems quite likely. Maybe a shorted or broken finger?
 
daver2 said:
So, that is interesting.

In the 'none working' case the M8029 was in slot 3.

In the 'working' case no cards were in slot 3.

Is slot 3 faulty I ask myself?

You could prove this by removing the Rl controller from slot 4 and inserting the M8029 into slot 4.

I have had backplanes where the contacts internal to the connectors (on the backplanes) had become damaged.

Dave
Click to expand...

New results:

Slot 1: M8189 (ABCD)
Slot 2: M8067 (ABCD)
Slot 3: empty
Slot 4: M8029 (AB)

1684175717417.png
 
AK6DN said:
I noticed too in the MAP output after you moved the RLV11 and RXV21 boards that the address block 775130..775136 disappeared, and 777170..777176 and 774400..774416 appeared ...
Click to expand...
Don, I can repeat the "non-working" case: No more 777160

Slot 1: M8189 (ABCD)
Slot 2: M8067 (ABCD)
Slot 3: M8029 (AB)

1684176046726.png
 
In slot 3 you get device 777130 and in slot 4 you get device 777170.

There is a faulty Data/Address line 5 in slot 3. BDAL05 pin BJ2. Have a look at the internal pins of the slot 3 'B' connector (solder side of the card that would plug into the connector).

Dave
 
daver2 said:
In slot 3 you get device 777130 and in slot 4 you get device 777170.

There is a faulty Data/Address line 5 in slot 3. BDAL05 pin BJ2. Have a look at the internal pins of the slot 3 'B' connector (solder side of the card that would plug into the connector).

Dave
Click to expand...
I have tested the rest of slots. All working except for the slot 3. I will look into the backplane to see if I can find damage on slot 3.
 
>>> I have tested the rest of slots. All working except for the slot 3.

You were really unlucky then! Been there, done that, got the t-shirt!

Right, next thing is to get it booting from Don's emulator :)...

Have you got a QBUS continuity card to install in slot 3?

Dave
 
daver2 said:
>>> I have tested the rest of slots. All working except for the slot 3.

You were really unlucky then! Been there, done that, got the t-shirt!

Right, next thing is to get it booting from Don's emulator :)...

Have you got a QBUS continuity card to install in slot 3?

Dave
Click to expand...
What is the QBUS continuity card ? what happens if I don't have it installed ?
 
A QBUS continuity card is (usually) a dual width card that "passes through" the interrupt and DMA daisy-chain signals.

It permits you to have 'holes' in the QBUS that don't (technically) have 'real' cards installed into them.

It is (basically) an empty PCB with a handle with a couple of PCB tracks linking critical signals.

This is why you can't have any empty QBUS slots between the CPU and any other QBUS card that either (a) generates interrupts or (b) generates DMA requests. Your M8029 falls into this category (i.e. it definitely generates interrupts as we have to configure an interrupt vector onto it with the wire links).

Having nothing in slot 3 (between the CPU and RAM - and the M8029) will result in the disk controller not working. You either have to get slot 3 working correctly, or insert a QBUS continuity card (on the basis that these pins of the slot 3 connector are not also damaged).

See here for more information: https://gunkies.org/wiki/M9047_grant_continuity_card

Dave
 
I work with mostly Unibus stuff so know of the G7273 and the smaller G727 cards. I never knew Qbus had these. And I am guessing they are not quite the same between the G7273 and M9047? I have a ton of the little cards, and a few of the bigger G7273 cards. I have none of the M9047 though or I'd offer him one.
 
They improved the QBUS equivalent over the UNIBUS ones (but that is only my opinion of course). Bigger cards = more expensive cards though!

I remember trying to squeeze my hands between UNIBUS slots to remove those little cards...

Dave
 
Oh, I scraped my knuckles way too many times trying to insert or remove the small cards. I hate them. Thus why I grabbed a few of the big ones when I had the chance.
 
daver2 said:
They improved the QBUS equivalent over the UNIBUS ones (but that is only my opinion of course). Bigger cards = more expensive cards though!

I remember trying to squeeze my hands between UNIBUS slots to remove those little cards...

Dave
Click to expand...

What was worse tho is that the DEC UNIBUS backplanes normally came from the factory with wire wrap jumpers on the pin side across the BGx and NPGx lines.
So if you wanted to put a card in a slot that was non-memory, you needed to go to the backside, find the jumper wires on the slot you wanted, and unwrap them.
Over time I think most backplanes have now lost all their jumper wires, and nobody goes to the trouble of putting them back if you remove a card.
You just track down a grant card for the slot.

In fact, in my PDP-11/34 DD11-PK and DD1-DK 9 slot backplanes I just removed all the jumpers when I got them and bought a pile go G7273 dual grant cards.
On a few cards (DL11-W comes to mind) I had to add a small jumper wire across the NPGx in/out pins since the card itself was missing those (an oversight IMHO.

And the little cards only did the interrupt BGx propagation, not the NPGx, which was in another block. The larger G7273 handles both (as well as being easier to insert/remove).

(Ok, back to the original thread now ... enough of a diversion).
 
Last edited:
Hopefully he can fix the issue with slot 3 easily. Maybe just a pin that's stuck out. Or something wedged into the slot keep the pin from making contact. I'd look in there with a very bright light to see if there is anything that can be seen. Or possibly a bad solder joint on the connector? Seems less likely.
 
You must log in or register to reply here.
Back
Top