• Please review our updated Terms and Rules here

pdp 11/04 assembly from parts

L

Lou - N2MIY

Veteran Member
Joined
Apr 1, 2008
Messages
1,306
Location
Albuquerque NM / Potomac MD
For anyone interested, I've uploaded a few pics of a recent project that has been going well so far. I've been putting a pdp 11/04 back together from parts. I do not have the front panel, so built my own. The machine is working with 32kW of core running RT11 off of RX01 floppy. The next steps involve making it pretty, putting it in a rack, then populating the unibus with various fun options I have accumulated. This is a good point to start looking through the album:

http://www.vintage-computer.com/vcforum/album.php?albumid=77&attachmentid=4322

Special thanks are in order to Tim for providing many of the parts.

This is about 80 hours of work after cleaning, assembling, building, troubleshooting, etc. This is my very first unibus machine and I have been learning a lot.

Lou
 
Wow! Good work. Having worked with an 11/70 PSU I know the regulators can be a pain. If it was an OEM machine, It might not have had an operators console.

Is the operators console any different from an 11/23 console?

Also, is that a VAX I see in the background?
 
Pontus,

Since you bring up working on the regulators, I will elaborate on two especially painful parts. First, the transformer is in a separate part of the chassis and the harnesses to the regulators are pretty short. This makes moving a regulator out of the chassis to a convenient to work on location while having it energized pretty difficult - in a fully assembled machine. Fortunately, this machine had been torn down so far that the power distribution PCB had been removed from the main chassis. Only the bits necessary were on the bench as I worked on the regulators.

Second, there is a power diode in a TO5 case that I cannot find a cross for. It's used in a few of the regulators, including the H745. It's an NAE model NSR8117. I have not been able to find a datasheet on this diode anywhere. NAE was a small diode manufacturer in Lynn, Mass. that was bought up in the 80's by Diodes Inc. After poor financial performance, the Lynn plant was shut down in 1990. Of course the current Diodes Inc. website knows nothing of this particular diode. If anyone out there has an NAE databook on their shelf and can scan the datasheet for this diode, I would be greatly appreciative!!!

I usually like to fix everything, but I had one left over H745 that I scavenged the good NSR8117 diode out of. I would really like to make this H745 whole again as a good spare.

As for the console, the model list does say that this 11/04-JH was supposed to ship with the KY11-LA. So, it did have one once....

In the background is an empty option rack from an 11/780. I was given that by Tim when he was cleaning out before he moved. It was empty when he got it. It is really strong. It can only fit five 10.5" tall devices (one less than an H960) but it looks strong enough to hold five RA80s! I plan to move the 8/e, RX02, two RK05s and the TU60 into that rack (eventually....) Rearranging the equipment in the racks is not as fun as troubleshooting...

Lou
 
Lou - N2MIY said:
Pontus,
In the background is an empty option rack from an 11/780. I was given that by Tim when he was cleaning out before he moved. It was empty when he got it. It is really strong. It can only fit five 10.5" tall devices (one less than an H960) but it looks strong enough to hold five RA80s! I plan to move the 8/e, RX02, two RK05s and the TU60 into that rack (eventually....) Rearranging the equipment in the racks is not as fun as troubleshooting...
Lou
Click to expand...

Sorry, can't help you with the diodes :(

But I think your equipment will look nice in the VAX-rack!
 
Don't think so--that's a 7-segment display. Lou's looking for a TO5 power schottky diode. Lou, can you probe a bit and determine the contents of the case? (i.e. diode direction). Since power schottky is commonplace nowadays, it's quite probable that you can find a substitute.
 
Chuck,

The case is the cathode, both pins are the anode. I have uploaded the schematic here: http://www.vintage-computer.com/vcforum/album.php?albumid=77&attachmentid=4333

It looks to me like a snubber for the output filter coil. In this application, it would make sense to be a Schottky diode for speed.

I suppose a cathode case TO220 power Schottky diode could work here mechanically also if it had sufficient ratings. Any suggestions?

Lou
 
Last edited:
Chuck,

MBR1035 looks good. It is case cathode TO220, plenty of current and good voltage ratings. Looks like a winner. It will go on my next DigiKey order. Thanks!

Otherwise, we now have an RL11 installed and and RL01 connected. Installed another DL11 for an emulated TU58 (DD: ). Both are also bootable. Still no XXDP yet though. I need to make a small monitor bootable TU58 image.

Lou
 
Last edited:
Lou - N2MIY said:
Chuck,

MBR1035 looks good. It is case cathode TO220, plenty of current and good voltage ratings. Looks like a winner. It will go on my next DigiKey order. Thanks!

Otherwise, we now have an RL11 installed and and RL01 connected. Installed another DL11 for an emulated TU58 (DD: ). Both are also bootable. Still no XXDP yet though. I need to make a small monitor bootable TU58 image.

Lou
Click to expand...

Take a look at: http://www.ak6dn.dyndns.org/PDP-11/TU58/ for bootable TU58 xxdp images with diagnostics. I have created image files with basic CPU diags for 11/34 and 11/44, but not 11/04. I could make an 11/04 basic CPU diags image if there is interest.

Don
AK6DN
 
Last edited:
Thanks Don,

I will use the 11/34 image as a starting point. I will swap out the F??????.BI? routines with the G??????.BI? for the 11/04 processor. I have spent plenty of time with Will's TU58 emulator and have built various images over the years. It will be a good time savings for me to start with your 11/34 image though. If I think I have made something respectable, I'll send you a copy to include with the rest of your files.

Lou
 
I went ahead and generated an 11/04 base CPU plus memory and M9312 diagnostic TU58 image as 1104_1.dsk and posted it (simple for me to do since I have a bunch of perl/shell scripts that extract diagnostic files from the XXDPv25 master RL02 disk image and build a custom TU58 image).

You might also look at the TU58 emulator links on that page. I rewrote (basically from scratch) the original Univ of Rochester TU58 emulator by Dan Ts'o and cleaned up the code. With the timing switches on to make it behave like a real (ie, slow) TU58 it even passes the DEC TU58 diagnostic ZTUUF0.BIN.
 
Well, we've got RXDP floppies made now from the DDDP. I used Will's TU58.exe emulator. I will have to try the emulator on your site someday, Don. I was hoping that it would work on a DOS PC, but alas it needs Windows. I use Will's on a dos laptop (dec HiNote, of course.)

With the ability to run diagnostics, I connected one of my old RL01s and finally repaired it. The upper head was electrically dead. I had a set of heads and positioner from Tim, so I replaced the upper head only and did a head alignment. My one and only good RL01 pack (also from Tim) works just fine. All errors were reported as soft by the exerciser routine. Here's a fuzzy picture of the setup now: http://www.vintage-computer.com/vcforum/album.php?albumid=77&attachmentid=4370 . Here is a fun closeup of the RL01 with the guts hanging out : http://www.vintage-computer.com/vcforum/album.php?albumid=63&attachmentid=4371 .

This system is coming together. 11/04 with 32kW of core, RX01 floppy (256kB - RX: ), RL01 disk pack (5MB - DL0: ), emulated TU58 (DD: ) running RT11 V5.0. I have the DL: boot rom in my M9312, so no more manual bootstrap entry! Now I need to finish a cosmetic front panel and put it all in a rack. Later I will connect my TU60 from my 8/e for the hell of it (I have TA11 controllers) and a DEUNA (the Kent State small rt-11 tcp/ip package will run on this hardware with this little memory.)

Lou
 
Last edited:
I have a bunch of RL01 disk packs if you want (free for cost of shipping, I have original boxes). Unfortunately both my RL01 drives seems busted :( So I have no way of testing right now. Also I'm uncertain what is on the packs, so if you could make images of them that would be good.
 
Pontus,

Thanks for the kind offer, but I'm afraid that the postage would be killer. There should be plenty of packs on this side of the pond still waiting to be rescued.

Anyway, you should fix your RL01 drives. What is wrong with them?

Lou
 
Lou - N2MIY said:
Pontus,

Thanks for the kind offer, but I'm afraid that the postage would be killer. There should be plenty of packs on this side of the pond still waiting to be rescued.

Anyway, you should fix your RL01 drives. What is wrong with them?

Lou
Click to expand...

I should! I haven't investigated further. I have two, both show the same behaviour. When I press the load button the drive spins up but the load light doesn't come back on. One never worked and the other worked up until a week or two ago :(

I did remove the SLU card to get a look at the cable. (I tried removing just the cable but It wouldn't budge). Perhaps I touched something I shouldn't have, but that sounds unlikely.
 
Pontus,

Well, a lot is working to just get so far as having the load button light come on. Now, after the drive spins up, do the heads attempt to load? They should come out, go just past the outer guard track, and seek track zero. Then, the ready light will come on. This is all easy to watch with the back cover up and the head amplifier board in the service position. This is shown in the picture in the earlier post.

Of course you know that the load light will not come back on until the pack has spun down. Is your ready bulb good? Is your fault bulb good (do you get a fault?)

On my drives with the bad upper head (head zero), they would shoot out, past the outer guard track, past all the data cylinders, past the inner guard track, and slam into the travel stop. Then the fault light would come on. Why, because they never found any track on initial load and just kept on going.

For anyone else out there, I should mention that this drive is extemely educational to work on. Doing a head alignment teaches how stuff really works. The track density is so low (2.5MB per surface on a 14" disk) with your own hand and a screwdriver you can advance the heads and "see" the centers of the tracks on the scope. No human could ever have this level of interaction with a 2TB 3.5" drive.

Lou
 
Last edited:
I'm certain that the bulbs are good, since they come on when you first power the drive up. I'm not sure the heads come out, I think I would hear them bump into the travel stop if that was the case.

I'll pop the hood and take a look when time allows.
 
Finishing touches for 11/04

Finishing touches for 11/04

Today I finally finished my nice home-made KY11-LA panel and fitted it to the front of the chassis. Here is the finished unit: http://www.vintage-computer.com/vcforum/album.php?albumid=77&attachmentid=4413

And here is a closeup of the switch / logo panel : http://www.vintage-computer.com/vcforum/album.php?albumid=77&attachmentid=4412.

I didn't label it as an 11/04 as such. I might upgrade it to an 11/34 eventually, so I wanted to leave the front panel generic. Now all I need to do is make sheet metal top and bottom covers, since the originals are long gone.

Lou
 
You must log in or register to reply here.
Back
Top