Scored a KDJ11-A

[smp]

I just scored a KDJ11-A processor on eBay:

https://www.ebay.com/itm/Vintage-Di...2F%2BA62K8%3D&orig_cvip=true&nordt=true&rt=nc

I intend to try this new processor in my existing tiny PDP-11/23 system:

H9821-BA backplane and card cage
KDF11-AA CPU
DLV11-J 4 port SIO
MSV11-LK 256KB/128KW memory
Emulex UC07 SCSI interface

Power-on Reset board from here:
http://www.heeltoe.com/index.php?n=Pcbs.Qbus-por

MeanWell RT-125B power supply

and, of course, a hefty fan!

I have the KDJ11-A Users Guide, and I see that I can strap the processor to start up in ODT (just like my current KDF11-AA). Other than that, is there anything else that I ought to be aware of? I don't see any strap to indicate this is only an 18 bit system. Is there anything that I ought to do prior to giving it a try?

Thanks very much, in advance, for any thoughts or advice!

smp

 

[smp]

Received my KDJ11-A yesterday, and I'm giving it a try today...

I set the straps with all removed and then inserted W3, so I have W3 in and W7 out. This gets me into ODT when I power up the system.

When I type in the bootstrap for the Emulex UC07 and I type 0G, the ODT light goes out on the processor, but the UC07 does not respond, the system seems to just hang. I tried to boot into the UC07 setup code, but again the UC07 does not respond and the system seems to just hang.

I tried installing both W8 and W9 to disable both the Wakeup and the BEVNT, but the system hangs up the same way for attempting to boot from the SCSI disk as well as boot into the UC07 setup code.

I suspect that simply getting the processor into ODT does not indicate that the processor is fully functional. Does anyone have any experience getting the KDJ11 to boot properly with an Emulex UC07? Any advice on something else to try in ODT to see If I can get my system going with this processor?

Thanks very much, in advance, for any thoughts or advice!

smp

 

[wa2flq]

Received my KDJ11-A yesterday, and I'm giving it a try today...

I set the straps with all removed and then inserted W3, so I have W3 in and W7 out. This gets me into ODT when I power up the system.

When I type in the bootstrap for the Emulex UC07 and I type 0G, the ODT light goes out on the processor, but the UC07 does not respond, the system seems to just hang. I tried to boot into the UC07 setup code, but again the UC07 does not respond and the system seems to just hang.

I tried installing both W8 and W9 to disable both the Wakeup and the BEVNT, but the system hangs up the same way for attempting to boot from the SCSI disk as well as boot into the UC07 setup code.

I suspect that simply getting the processor into ODT does not indicate that the processor is fully functional. Does anyone have any experience getting the KDJ11 to boot properly with an Emulex UC07? Any advice on something else to try in ODT to see If I can get my system going with this processor?

Thanks very much, in advance, for any thoughts or advice!

smp

My standard gear is a KDJ11-A and UC07. My KDJ11 Jumpers are only W3 and W5. The diagnostic is pretty good at detecting problems local to the CPU (except for cache I think). Most likely the problem is elsewhere.

UC07 (G143R Firmware) only SW2-5 on. SW1 all open. Jumpers D-E, K-L, M-N, T-U and Terminators installed. Autoboot is off. Not getting to the FRD is strange, as I believe the UC07 talks directly to the console, w/o using the CPU.

Before you try the 0G, suggest you read the first 10 or so locations starting at 0 and report them to see if they contain a valid bootstrap.

Jerry

 

[cruff]

If I recall from the microcode listings, the startup diagnostics verify the internal CPU data paths and ALU are functioning, and that the console SLU hardware responds. More detailed diagnostics require a working external ROM. Presumably you do not have any breaks in the bus request/grant chain?

 

[smp]

Hi Jerry,

My UC07 matches with yours exactly. I set my KDJ11 straps to match yours, too.

When I power up, I get into ODT, as I should. I type in the bootstrap:

000204 <- typed when ODT started
@17772150/000000
17772152/0054000 30003
@/000400 40000
@RS/000000 304
@R0/006010 0
R1/001020 17772150
@?
0/000240
0000002/000415
0000004/000000
0000006/000000
0000010/000000
0000012/000000
0000014/000000
0000016/000000
0000020/000000

And it booted up!

This time, I was sure to type in the full addresses, rather than the shorter ones I used for my KDF11. Maybe that was it, maybe it was plugging and unplugging a couple of times. I don't know.

Thanks very much for your advice, and attention!

smp

 

[wa2flq]

Congratulations.

ODT on the KDJ11A requires 22bit addresses…even in 18bit backplanes. I just use 77777xxxx, as it drops any excess leading bits. The value 240 (NOP) is found at the beginning of most DEC PDP11 disk bootstraps. Its an indication that block 0 has loaded from the disk.

See Section 3.6 and 3.6.3 page 3-8 in EK-KDJ1A-UG-002_KDJ11-A_CPU_Module_Users_Guide_2ed_Jun86.pdf.

Jerry

 

[smp]

<...snip...>
See Section 3.6 and 3.6.3 page 3-8 in EK-KDJ1A-UG-002_KDJ11-A_CPU_Module_Users_Guide_2ed_Jun86.pdf.

Thanks very much (again) Jerry. I have EK-KDJ11A-UG-001... What you pointed out shows up in the same sections on the same page. Is there any real differences in the 2nd edition?

Thanks again for all your help!

smp

 

[wa2flq]

Happy to share what I know.

Appendix B is expanded to include some information on the ASH ASHC instructions. There were some differences in the DCJ11-AC chips from Rev 04a to 08 (bug fixes).
Appendix C was added and discusses configurations items including FPJ11-AA compatibility and read/modify/write instructions compatibility warnings for non-DEC Peripherals compared to the 11/23.

Not doubt there are other errata in the document, but the above is what jumps out at me.

Jerry