• Please review our updated Terms and Rules here

Minc-23

glitch

glitch

Veteran Member
Joined
Jan 31, 2010
Messages
5,383
Location
Central VA
So we found a MINC-23 at the MIT Flea (full album: http://imgur.com/a/Hov9T#0 ):



Came with what you see (MINC-23 chassis + RX02), a VT-105, and a pile of cables! It was used at the Mount Washington Observatory, most likely taking inputs from various bits of weather-monitoring equipment. The folks from the observatory have a pile of manuals and documentation and, as luck would have it, will be in Troy on Friday anyway.

Unfortunately, it appears to have had a family of mice in it at some point, as you can see from the "before the cleaning" and "in the card cage" pictures. Lots of mouse filth. The cards cleaned up OK, but I haven't extracted/cleaned the backplane yet, which is where the real filth is. It appears to have oozed down through the backplane and out of the bottom of the MINC case, so I'm sure that will be *real* pleasant to get clean. Hopefully it's not ruined.

I pulled the cards tonight and cleaned the mouse filth off, couldn't have been there too long because it came off with hot water, dish soap and a tooth brush with little scrubbing. There was very minimal corrosion, mostly at the point where the tin traces became gold edge fingers. I dug out my 11/03 chassis and tested the CPU, RAM and DLV-11J which all appear to work. The RAM (64 KW/128 KB parity MOS) was one of the filthier boards, so that's promising!
 
Last edited:
Where’s the link?, I have seen mouse piss eat the traces right off the card so your lucky if it all cleaned up and the other thing is all the family of Qbus cards for the 11/23 are common and available so if you had to replace them it’s no big deal I think the most valuable part is the case and big things like the RX drive and power supplies, and the opportunity to know the past history of the system.
 
Qbus said:
Where’s the link?
Click to expand...

The image itself is the link to the album.

Hopefully the mouse filth hasn't destroyed the backplane. I haven't dug the manuals up yet, since I should be getting paper copies on Friday -- is the QBUS portion a standard DEC item, or something special for the MINC? I'll post the board numbers tonight, but it appears to be a pretty standard MINC-23 with the addition of a 4-port async mux.
 
Would imagine if it’s a KDF-11A (8186) it’s a standard Qbus system and will work with an RT-11 boot disk but never have seen a Minc myself so that’s all speculation. Have seen my share of 11/23 systems both DEC and others and the KDF-11 is a great card and plays well with almost everything.
 
Hi Glitch!

glitch said:
The image itself is the link to the album.
Click to expand...

I must be missing something obvious -- when I click on the image I go to its imgur page, but there's no album available. I can't seem to find it. I'd love to see the rest of the pictures.

Regardless, it's a great system, and I'm excited you got it. I had a surplus one in college, and it provided a great deal of learning experiences about DEC gear!

-Twylo
 
Twylo said:
I must be missing something obvious -- when I click on the image I go to its imgur page, but there's no album available. I can't seem to find it. I'd love to see the rest of the pictures.
Click to expand...

Huh, I guess it takes me there being logged in and the uploader, or something. Anyway, here's the link:

http://imgur.com/a/Hov9T#0

I'll edit the first post to provide it as well.
 
I was able to fully disassemble the chassis and clean it last night. The backplane was pretty awful, but again it seems the mouse filth hadn't been there too long. It still smells a little funny if you get your nose close to the slots (but who does that?). Took about an hour of scrubbing and picking "stuff" out of the slots with the end of a heavy zip tie.

While I had the backplane out, I tested the power supply. It looks like the +5 and +12 rails are spot-on, but the -12 rail is actually somewhat positive (+3V). The -12 rail's level changes depending on what the meter is set to (presumably the effective load it presents the power rail). -12 won't affect the QBus board set I have since they're all +5/+12 cards, but I assume it's used by the lab modules. Does anyone know if this is "normal" for a completely unloaded -12V rail?
 
Now would also be a good time to give the backplane a good washing. I used a commercial detergent called "Alcojet" that was recommended to me by a former DEC engineer, good for removing organic compounds. It proved safe and effective when I was cleaning out my PDP-11/35. Here's a before-and-after of my backplane modules (after picking out mouse/fly/spider/god-knows-what droppings with a dental pick and washing in Alcojet):

before_and_after.jpg
 
Yep, cleaning the backplane involved leaving it under a running hot-only tap for maybe 5 minutes (the water ran brown out of it for the entire time!), then adding Dawn dish detergent, scrubbing with a toothbrush, rinse, more dish detergent, scrubbing with a long-bristle chip brush, an extended rinse, then shaking it to remove the big drops of water. I picked out the bits of straw and cardboard down in the slots with the end of a heavy zip tie. I then used a bus grant card to push a piece of re-usable paper towel (the tough stuff that doesn't rip when wet) into the slots, and finished with the hot air gun. The before and after basically look like your image! It was completely nonserviceable before cleaning.
 
The MINC-23 is now booting XXDP from emulated TU58! RetroHacker_ and I have been working on figuring out the RX02 situation. XXDP reports that the only RXV21 controller we have (the one from the MINC) is probably damaged. I've got another one on the way, on loan from a fellow MARCHin.
 
I also just got a MINC-23 as well as a MINC-11 with the A/D, Clock, D/A, Digital in and Digital Out. Neither came with Qbus CPUs, memory or I/O. I have been able to get a CPU, Memory and DLV11-J running and have a BDV11. Using Jorge Hoppe's PDPGUI I have been tinkering with a M8186 Dual 11/23 CPU using M8059-KF (256 KB Dual) and it communicates fine. I also have a M8189-YB (11/73 Dual) that works fine. My problem is I/O.

The RX02 I have has drives that are not turning (Index hole doesn't move on inserted floppy), but I've decided to leave that alone till later.

I have used TU58em on a Raspberry PI with my 11/83 system, but can't seemed to get that to work on the DLV11-J (I did learn that a MINC's 9 pin RS232 connector pin out is very different than a normal 9 pin RS232 connector).

What I really want to get running is an Emulex UC07 with a SCSI2SD. I have a pair of SCSI2SD cards running great on the 11/83 with a UC07 running RSX11M+ V4.6. This UC07/SCSI2SD would let me load a range of disk images available on the net and is key to what I want to accomplish with the MINC.

I tested the new UC07 with a SCSI2SD with RT11 V5.7 on the 11/83 with a newly made SCSI cable and it works perfectly. When I put it in the MINC, I provide the 5 volts for the SCSI2SD from the voltage test probe sockets on the BDV11. I can run FRD and config the NVRAM but when I run option 7 the list available drives it finds none. I can key in the UC07 boot but it does not find drives on the SCSI bus. If I take the UC07, SCSI cable, and SCSI2SD device up to the BA23 based 11/83, it boots RT11 perfectly.

I don't have +12 connected to the SCSI2SD card on the MINC but I don't think it uses it. I have the UC07 just under memory so I don't think it is DMA grant, and there are grant cards in all slots to the DLV11-J. The one big question I have is would the UC07 need Q22 wired in order to see the drives on the SCSI bus? Also, I'd like to find a MACRO-11 version of a DU boot. One other observation. The halt and boot switches on the BDV11 work ok, but all of the RED Leds are on all the time. Voltages on the MINC are all normal.

I'm about ready to wire wrap the Q18 bus to Q22 to see what difference that makes. Also, my BDV11 does not have the ECO E 22 bit termination but I could also do that if it mattered.

Any thoughts or suggestions would be much appreciated. Also, I am looking for any MINC software such as the RT11 FEP package or better yet the more rare RSX11M FEP software.

Thanks,
Mark
 
Hi Mark,


if you haven't already, grab a copy of ftp://bitsavers.informatik.uni-stuttgart.de/pdf/dec/pdp11//LSI-11_Systems_Service_Manual_Aug81.pdf. It has all sorts of configuration and refresh information.

If you have the UC07 plugged into the MINC QBus, running FRD and it does not see the SCSI2SD drives after configuring NVRAM I doubt the QBUS DMA has anything to do with it. The problem is that the test probe sockets will not supply enough current to drive this device. Those sockets will have a current limiting resistor. The voltages may look fine when idle, but when active the current draw increases and the voltage for the SCSI2SD will drop. Wire it up or find a old desktop PC PS and give it a good supply.

Just a note of caution. It is possible to configure these systems to map the lower 2K Bank 7 as memory instead of IO. For those of us running real time acquisition using RT11SJ or FB, having 30K total was common, especially when running an LSI 11/02. It will cause problems for XM or other monitors.

If you have still having problems (sorry captain obvious here), but can you supply the following information.

A list of the qbus cards you have in the system and how they are arranged in the system.
The QBUS backplane model.
The monitor type of RT115.7

Regards,
Jerry
 
Jerry,
Thanks so much for your help!! I had not thought about the SCSI2SD pulling those test point voltages down during activity. I had a meter on them at first and it looked ok, but your point about when SCSI activity starts the voltage dropping is very likely what I'm experiencing. In fact, your comment about testing with a small PC power supply was a great one and I just tried that and FRD now sees the disk!!

I tried running the UC07 boot procedure but it didn't work, however I am elated that the UC07 can see the disk and I'll just need to figure out a bootstrap that can be loaded into memory or get some 2716 EPROMs programmed for the BDV11 with a DU boot.

Also, thanks for the link to the LSI-11 manual below. I downloaded the things in the MINC folder there but have not seen that manual. It does look like it has some very useful info.

The RT11 monitor I am trying to boot is RT-11FB V5.07 It is on a RD54 .dsk file on the microSD card. It boots ok on the 11/83.

The Qbus cards and order at this time are:

M8186 11/23 (Dual with MMU and FPP chips)
M8059-KF (256 KB Dual)
M9047 Grant
Emulex UC07
M9047 Grant
M9047 Grant
M8043 DLV11-J
then all MINC slots are empty
M8012 BDV11 in its designated last slot

I do not know the Qbus backplane model but with ODT I can see and change memory above 177777 at for example 300000 so I think it is an 18 bit backplane. The name
plate says MINC-23.

Thanks again,
Mark
 
glitch said:
The MINC-23 is now booting XXDP from emulated TU58! RetroHacker_ and I have been working on figuring out the RX02 situation. XXDP reports that the only RXV21 controller we have (the one from the MINC) is probably damaged. I've got another one on the way, on loan from a fellow MARCHin.
Click to expand...
I fixed it -- today! So this is what XXDP test "ZRXFB0" was spitting out:

Code: 
CZRXFB0 DVC FTL ERR  00027 ON UNIT 00 TST 011 SUB 000 PC: 003476
 CSR BITS - LGC TST
  CSR- ERROR
      REG ACTUAL=005520
      REG EXPECT=005560

  POSSIBLE FAILING "FRU'S":
            INTERFACE - M8029
            CONTROLLER - M7744
            INTERFACE->CONTROLLER CABLE

  UNIT#0 RXCSR=000000 RXESR=000000 CMD=000000 ->
CZRXFB0 DVC FTL ERR  00027 ON UNIT 00 TST 011 SUB 000 PC: 003476
 CSR BITS - LGC TST
  CSR- ERROR
      REG ACTUAL=004000
      REG EXPECT=004040

  POSSIBLE FAILING "FRU'S":
            INTERFACE - M8029

Bit 5 on the CSR was not high when it was expected to be. This is the "DONE" bit of the CSR. I chased the problem probably the wrong direction -- I confirmed the "RX DONE" signal was getting into the board and through a couple gates, then due to other unrelated errors in the data register tests later on decided it had to be a problem on the internal tristate bus on the RXV21. So, I replaced the DC005 QBus transceiver that handled it, the DC006 that touched bit 5, the 74LS299 shift register that touched bit 5, and the 8097/74LS367 that touched bit 5. No changes. So I went back to chasing the "B DONE" signal, and found a dead 74LS174 at E18.

At least it wasn't the fuse PROM at E24! That was the next option. Probably ought to pull one and back it up...
 
glitch said:
12, but who's counting? :P
Click to expand...
Glitch,
Congrats on your MINC!

To give an update on my MINC, I did program a set of 8 2716 EPROMs that provided the boot code of an 11/23+ on the dual width 11/23 so it boot DU disks. I ran for many years with two SCSI2SD cards and the UC07. That way I could make disk to disk backups.

Some years later I put a QBone in a MINC module space that was perfect since it was widely spaced and one doesn’t need the grant card spacers to prevent the QBone from touching other cards. This gave not only MSCP disks but also RL02, RK05, and RX02/01 emulators that made it easy to ftp a disk image off bitsavers and read it. Later updates of the QBone software had the transparent RL02 magic that allows one to ftp a file to a Linux directory on the BeagleBone and it appears on the RT-11 RL02 while RT-11 is running!!

Best,
Mark
 
You must log in or register to reply here.
Back
Top