VAX 4000-300 DSSI dies?

[dtih]

I've had this VAX since 2001 and over time I installed more memory and 2 RF73 DSSI disks, it came with a TF85 tape drive. I installed VMS in 2001, last time I powered it up wa 2016. It was running just fine, except for a smell it would begin generating.

Powered it on the other day and passes POST tests.

KA675-A V4.8, VMB 2.12
Performing normal system tests.
66..65..64..63..62..61..60..59..58..57..56..55..54..53..52..51..
50..49..48..47..46..45..44..43..42..41..40..39..38..37..36..35..
34..33..32..31..30..29..28..27..26..25..24..23..22..21..20..19..
18..17..16..15..14..13..12..11..10..09..08..07..06..05..04..03..
Tests completed.

However, none of the disks spin up (the led's on the drives light up temporarily, then go dark)

>>>show dssi
DSSI Bus 0 Node 6 (*)

DSSI Bus 1 Node 7 (*)

It doesn't even see the tape drive!

Here is what I've done:
1. Remove and reseat all drives, cpu, memory and power supply.
2. Measure voltages at TF85 drive, 12V OK, 5V OK.
3. Replace 2 RIFA capacitors in the H7864 power supply. (thanks to Aaron Jackson for his site - https://aaronsplace.co.uk/blog/2018-06-04-vax-4000-power-supply.html ) Smell is gone!
4. Put Qbus minimum load board back in.

Have I crossed the Rubicon with DSSI devices? Is there any hope?

 

[rjarratt]

Could you be missing a DSSI terminator? Have you tried with just one drive installed? There is a Ready switch on the drives have you checked those? It's odd that the drive lights go dark though, can't remember seeing that happen before.

 

[dtih]

I did try that, putting in one drive at a time. The terminator question is not resolved. The DSSI drives in the cabinet are on channel 0 and I put a terminator on the DSSI plug on the far left side of the cabinet. The light in it was quite faint, I will swap it with one on the CPU/Console panel which lights up quite bright.

 

[DECtechGuy]

While it's not a pedestal unit, on my VAX 4000-100A there's this blue wire that snakes off the power connector down to the motherboard. I believe it's the wire referred to in some documentation as "POK to ISE" or "ACLO to ISE"

I don't know if the RF73's have this though.

 

[Terri Kennedy]

Could you be missing a DSSI terminator? Have you tried with just one drive installed? There is a Ready switch on the drives have you checked those? It's odd that the drive lights go dark though, can't remember seeing that happen before.

This is in a BA440, so the drives have full operator control panels?

If so, the normal behavior is all lights come on at power-on and then go off after a few seconds, assuming the write protect buttons aren't on and the ready buttons aren't on either (DSSI drives are always spinning, but releasing the ready button puts the heads over the landing zone which you should always do on these systems because they're old. I have a drive that works fine except that the RCT is unreadable because of a power-down glitch while the drive was still in the ready state.

You can easily make a DSSI to PC-type drive power adapter cable - you need a Molex Mini-Fit 0039014051 shell, 5 0039000077 pins and a regular PC-type 4-pin Molex "Y" adapter. With the operator control panel connected to the drive and powering the drives this way (I keep a PC ATX type supply around that is permanently set to be "ON" for this purpose) you can bring the drives ready with the OCP and they will do their full "popcorn popping" self-test. That will lwt you know if the problem is in the drive or the power supply.

LEDs on DSSI terminators vary greatly in brightness. If it's lit at all, it's good. Running without a terminator seems to blow the termpwr fuse. On a 4-slot front panel like most BA4x0 VAXen use, these are on the inside of the front panel. Look for 2 green objects that look like 1/4W resistors. They go into pin sockets so they're easy to replace. I believe they're picofuses. On the BA430 (VAX 4000-200 with the 2-slot front panel) I forget if they're on the CPU board or the front panel. I can go look in a day or two if anyone cares. Bonus: There's another one of these fuses on the KZQSA (SCSI) card, if you have one.

Despite what DEC claims, DSSI drives in the BA4x0 are not hot-swap. That's a sure-fire way to damage the power supply on systems this old. The symptom is that the power supply shuts down when you try it, until eventually the PSU won't turn on again without repair / replacement.

I don't know if the TFxx tape shot shares a power slot with the rest of the DSSI slots - it probably does as I run my systems with 4 RF73s and no tapes.

I don't know if the OP is even in the US, but if they're near NYC I'm glad to help and can bring an RF73 and PSU (or they can bring theirs to me). I still have a 4000-200. The 4000-705A (which is a closer config to the OP's -300) went to the LSSM last month.

 

[dtih]

I'm in the US, near Wash, DC. If I pull one of the RF73 drives and power it up on the bench, what is likely to happen? Will it spin up and reach operating condition? What should I look for? Does the DSSI drive wait for a signal from the bus to tell it to go on line? [I think I have the required power connector/adapter somewhere]

I'll keep the NYC offer in mind, I may be in that area Sept 22 + or -. This week I'll be heading for VCFMW.

If I can't get the DSSI drives to work I did put the S-Box Emulex UC07 in the VAX and could use that as a boot path. My question is what modern SCSI drives could be used with this old controller?

 

[Alegend]

If I can't get the DSSI drives to work I did put the S-Box Emulex UC07 in the VAX and could use that as a boot path. My question is what modern SCSI drives could be used with this old controller?

Every single-ended 50-pin SCSI old drive I have connected to UC07's was recognized and worked fine, various models from DEC, Seagate and Quantum. Also a few 68-pin drives using a $2 adapter.
I recommend you consider a ZuluSCSI or SCSI2SD board instead. Old spinning rust can die on you in an instant and the noise of whining bearings is no great pleasure either.
-Alon.

 

[dtih]

The sane thing to do is to use the ZuloSCSI or SCSI2SD, but I miss the sound of spinning disks, especially those DSSI disk - they let you know they were working! I'm curious to try that $2 adapter.

 

[Bitly]

The sane thing to do is to use the ZuloSCSI or SCSI2SD, but I miss the sound of spinning disks, especially those DSSI disk - they let you know they were working! I'm curious to try that $2 adapter.

We need a Zulu with a speaker to output a high pitched whine! And clunking sounds when it seeks

CW

 

[DECtechGuy]

I'm in the US, near Wash, DC. If I pull one of the RF73 drives and power it up on the bench, what is likely to happen? Will it spin up and reach operating condition? What should I look for? Does the DSSI drive wait for a signal from the bus to tell it to go on line? [I think I have the required power connector/adapter somewhere]

I'll keep the NYC offer in mind, I may be in that area Sept 22 + or -. This week I'll be heading for VCFMW.

If I can't get the DSSI drives to work I did put the S-Box Emulex UC07 in the VAX and could use that as a boot path. My question is what modern SCSI drives could be used with this old controller?

I'd expect the drive would start up, load its on-board OS, and then fail its POST as it can't reach the host. This would be indicated if both the power and fault LEDs turn on and stay on.

My reading of your initial post doesn't make it clear to me (this is mostly likely my fault ), but have you tried out a single drive startup with either one? I wonder if things will behave if a drive is totally out of the VAX.

 

[Terri Kennedy]

I'd expect the drive would start up, load its on-board OS, and then fail its POST as it can't reach the host. This would be indicated if both the power and fault LEDs turn on and stay on.

My reading of your initial post doesn't make it clear to me (this is mostly likely my fault ), but have you tried out a single drive startup with either one? I wonder if things will behave if a drive is totally out of the VAX.

No, it will post and you can get it to go ready if you have the front-panel ribbon cable for controls and the right power cable. Remember, these are basically baby HSCs with lots of on-board smarts.

 

[DECtechGuy]

No, it will post and you can get it to go ready if you have the front-panel ribbon cable for controls and the right power cable. Remember, these are basically baby HSCs with lots of on-board smarts.

You're right I miss read the document (RF Series ISE User Guide) I was reading, specifically what I have in bold from page 2-28:

POST
The other diagnostic program that runs at system initialization
is called power-on self-test (POST). The POST firmware tests the
hardware on the drive module and in the HDA, and achieves
almost complete coverage.

POST Coverage
The only hardware not verified by the POST is a small part of
the DSSI area (shaded part of Figure 2–16). This part of the
module is verified by the ISE’s ability to communicate with the
host.

 

[dtih]

The DSSI disk and tape are on the same DSSI channel, and the fact I can't see either one of them makes me suspicious of the CPU. The SHAQ communicates with these device through the backplane. I'm going to remove and reseat the CPU again.

 

[Terri Kennedy]

The DSSI disk and tape are on the same DSSI channel, and the fact I can't see either one of them makes me suspicious of the CPU. The SHAQ communicates with these device through the backplane. I'm going to remove and reseat the CPU again.

In addition to that, check the cabling between the backplane and the DSSI slots, and between the DSSI slots and the DSSI 0 out connector on the bulkhead to the left of the slots. I forget if it's on the front or back side of the backplane, but the IPB and service manual on Bitsavers should have the info.

I have a spare 4000-500 CPU card and a spare 4000-705A CPU card, but they're both at the LSSM. The 4705A wants different memory, and I'm not sure if your backplane needs to be upgraded to deal with the 500-series CPU. The 4000-200 I have here is Q-bus (aside from doing DSSI on the C/D pins). So I don't have anything here or within easy reach that would let you test your CPU's SHAC.

Also, as someone else suggested, remove all DSSI devices except one (I suggest removing the tape drive and one of the disks) and see if the remaining disk shows up.

Lastly, you should have three DSSI terminators - two on the CPU cover plate and one on the bulkhead to the left of the slots. It's the one to the left of the slots that's most important - that's DSSI 0 Out (the other end is terminated on the CPU module). The two on the CPU module are DSSI 1 In and DSSI 1 Out, and that bust doesn't go up to your internal DSSI slots.

 

[dtih]

It's sort of a math problem, the BA440 box has 4 DSSI slots and I have 3 DSSI devices, if I do one device at a time that becomes 1 slot for tape drive, 2 disks into 4 slots....

I did power the VAX up with no devices, and got past POST. The LED in the DSSI terminator on the box left hand side seemed a little brighter, but still dim compared to those on the CPU door.

The only success I had was putting the disk DIA0 (a bootable VMS volume) into the right most slot, where the tape drive was.
>>> show dssi
DSSI Bus 0 Mode 0 (DIA0)
- DIA0 (RF73)

Moved the disk to the next slot and the next and the drive isn't detected. However, the more I powered it on it seemed to making DSSI disk calibration noises and the ready light came on.
Put the tape drive in by itself, hoping that being a simple tape drive, it had to show up in the DSSI device list. Result, it didn't.

Put DIA0 back into the tape slot and now the computer doesn't see it at all. These drives haven't been powered up for a couple of years. If this was football I would punt.

 

[dtih]

Here is picture of the problem. Wonderful machine, I wish it worked, once it did.

 

[Terri Kennedy]

It's sort of a math problem, the BA440 box has 4 DSSI slots and I have 3 DSSI devices, if I do one device at a time that becomes 1 slot for tape drive, 2 disks into 4 slots....

I did power the VAX up with no devices, and got past POST. The LED in the DSSI terminator on the box left hand side seemed a little brighter, but still dim compared to those on the CPU door.

The only success I had was putting the disk DIA0 (a bootable VMS volume) into the right most slot, where the tape drive was.
>>> show dssi
DSSI Bus 0 Mode 0 (DIA0)
- DIA0 (RF73)

Moved the disk to the next slot and the next and the drive isn't detected. However, the more I powered it on it seemed to making DSSI disk calibration noises and the ready light came on.
Put the tape drive in by itself, hoping that being a simple tape drive, it had to show up in the DSSI device list. Result, it didn't.

Put DIA0 back into the tape slot and now the computer doesn't see it at all. These drives haven't been powered up for a couple of years. If this was football I would punt.

These drives are pretty robust. I bought 2 NOS ones a year or so ago, and when powered up they had 3 power cycles and under 20 hours on them, format dates in 1993, and were in sealed DEC foil bags. So if anything was going to gunk up in them, it would have.

All you need is the console to talk to the controller in the drive, and you can run VERIFY and DRVEXR to make sure they're working.

I'd start with a minimum config with one disk, get it going solidly, then repeat. I think the TF drives had DSSI bridge boards and are probably more flaky due to that.

 

[Terri Kennedy]

Here is picture of the problem. Wonderful machine, I wish it worked, once it did.

One other thought - pull out all of the expansion cards - you don't need any of them for this, and there's at least one mystery one in there. Tape cardboard over the open slots to maintain airflow.

 

[dtih]

I'm trying. It seems like the DIA0 drive wants to come to life, the more I run it, the more it makes those calibration noises and the ready light goes green.

 

[DECtechGuy]

There's hope for them yet!

And I can second the general robustness of these DSSI drives. The pair of RF35s in my 4000-100 are still spinning.