Easy fix on an IBM 5106 unit (for IBM 5100)

[voidstar78]

I have an IBM 5106 (tape drive unit) that wasn't responding, it kept giving ERR 005 (tape not inserted). I put this off for a while, dreading that it was an electrical problem. Turns out, no, just a minor mechanical problem.

With the cover off, and the tape "fully" inserted, you can see it wasn't pushing in the contact switches that tell the system that the tape is inserted. I recall the MIM talking about adjusting for this with the two screws shown in the photo below:





But adjustments on this one exposed side wasn't good enough. You have to remove the board (one screw, then just pull it out gently sideways). The description in the MIM manual makes this seem more complicated than it is... Just loosen all 4 screws (loosen, not remove -- half a turn to one turn is enough). Then insert a tape (the MIM describes some fancy proxy tape insert for this -- probably just the metal structure of a DC300 tape). Bias towards making sure the side with the contact switches is all the way in (the tape should not "rock"/tilt while inserted). Re-secure the 4 screws (hold the tape in place to be sure). While the 4 screws are loose, there are little tabs below them as a spacer (2 layers) -- holding the tape in position while re-securing the screws ensures those tabs are where they need to be. Then, of course, put the control board back in :D




And with that, this 5106 is working just fine for me!

REWIND E40
MARK 3,2,1,E40
SAVE 1,E40
LOAD 1,E40

Those commands all worked just fine now, no ERR 005.

At least, it is working on the IBM 5100 just fine. I'm not able to get this 5106 to work with either my Type 1 or Type 2 IBM 5110 system. The 5110 will rewind the tape, but won't mark or load (gives ERR 004, which is a non-specific "tape error"). Maybe something changed in the 5106 hardware between when the 5100 and 5110 were released. OR, my other theory is the software in the 5110 may have tighter timing or something -- resulting in it being more picky about the specific type of tape being used?? I know that's a stretch (but I have an internal 5110 E80 tape deck that is giving the same issue, ERR 04 -- but on my 5100, both the internal E80 and this 5106 E40 are both working fine).


Speaking of tape, I have a question! I'm using 3M tapes that have never been opened. But a few of these "non-opened" cases (that were sealed), the tape wasn't on the reels. Here is an example of a "new out of the wrap and box" example of what I mean:



Fortunately 3x other tapes in the set were all on the reel, and those are the ones I'm using to verify the tape units are working. But the 4x of the ones I've opened were like the example above - was that a normal thing, and there is some obvious way to wind the tape back onto the reel? I tried for awhile -- including removing the two screws on the backside of the metal support of the tape (but more than those 2 screws secure it closed, so I didn't figure out how to actually open these DC300 cases yet).

BTW, while these tapes have BOT markers - the 5100 can still "eat" a tape (so I've learned). If the tape is rewind "too far" (maybe manually), then issue a REWIND command -- it can fly off the reel by itself. I'm not sure of the correct wording, but there are two BOT markers (single dot and then 3 sets of double dots). If you rewind before the 3rd double dot, that might be a problem. In any case, all I can say is on one of the tapes, I had it come off the reel by itself during a rewind.

Just curious if there is an actual correct way to get these back on the reel. I only tried for about 10 minutes trying to thread it back on the wheel (bottom left of image above), but without success.

 

[voidstar78]

One other nit thing while I'm talking about these tape units.... I have 3x of these. On the first one that I got, I thought the black metal structure behind the tape unit was damaged (by having a small chunk missing from the top). I thought "that's weird, maybe it was damaged during shipping"

Then the 2nd unit had the same thing. That's weird.

And I notice yet again on this 3rd unit!! (see red arrow below in the image below)

This is looking down, but the black "bar" is actually a very thick and heavy metal plate.

So, now I'm thinking this nicked edge is just part of the process in how these plates were cut/prepared? Either intentional in how the pieces are cut (from a grid of plates or a larger plate?), or a post effect in how the pieces were tested??

Anyhow, if anyone comes across this in the future with their 5100/5110 or 5106, don't consider it damaged -- it's normal :D

 

[voidstar78]

Just to comment a little bit more about the IBM 5110 (which is not yet working with this 5106)...

I've confirmed the IBM 5110 Type 2 (that does NOT have an internal tape unit) is missing hardware components to have the external 5106 attached. Those components are on the IO BASE that is the "2nd half" of the Processor card (the one connected by the rubber piece with a couple crosslinks). You normally can see these components in any photos, since they are obscured by the main Processor card itself. Even with the 5106 connected, if I issue a REWIND E40 or other tape command, the (stock) 5110 Type 2 just says ERR 13 (device not found).

However... You can take the Processor and BASE IO from a 5110 Type 1 (the version with the built in internal tape designated by E80) and stick it in a Type 2. I don't think the IO board (in SLOT A) is any different between these models, and I also don't think code in the Executive or Common ROS (at least as far as tape support goes-- they could be different for other reasons). I'm just saying that's what I did: I took the Processor+BaseIO board from my Type 1, put it into the Type 2, and now the Type 2 no longer says ERR 013 and it can rewind the 5106 tape (REWIND E40).


That said, I still haven't successfully done a MARK command on the 5110 Type 1, or with the 5106 attached (which is a necessary step to basically format the tape, to then save/load files on it). If anyone has a still-working tape drive for the 5110 (specifically the 5110, not the 5100) - internal or external, I'd be curious to hear about it -- and whether it is working with an IBM DC300 tape, or a 3M one.

 

[krebizfan]

Congratulations on getting some of the tape working.

Vaxbarn has webpage on refurbishing QIC tapes but that page refers to research done by posters here. You may recognize the names. https://www.vaxbarn.com/2-uncategorised/755-qic-tape-data-recovery

 

[voidstar78]

Just odd it is working with the older system (5100), but then that same tape drive isn't working with the newer (5110) system. It's listed as a valid configuration for the 5110. Appreciate that it works at all, but still a bit of a mystery :D


Thanks for the link, it has notes that confirm the two screws are the only things necessary to open up the tape (for repairing the reel issue). Just needed a little more pressure on the case.

 

[voidstar78]

Recap...

5106 is working fine with the 5100 system (both E80 for internal, and E40 for external).


When attaching the 5106 to the 5110 Type 1 (that has an E80 internal tape also), on power up: I get to boot-up sequence letter "N" and then the PROCESS CHECK stays on and system stops the startup sequence.

Per the MIM, sequence "N" is related to checking the Feature ROS (which is the disk sorting hardware). I do have that, but the system has always booted up fine without it (and the MIM also suggest that if there is a failure at this step, to just remove the Feature ROS -- as it is not essential to the system and applies only when you have a disk drive attached). In any case, whether the Feature ROS is inserted or not, I get this same startup behavior when the 5106 tape drive is attached (even if the 5106 is powered off).

If I disconnect the 5106, the 5110 Type 1 starts up fine as it always has.


Two thoughts on things to try:

(1) Maybe the external IO Card in slot A is bad? I can try swapping it out with one from the other 5110's. That card is a little harder than normal to swap, since it is bolted in and have to be careful about the two translucent cables attached at the top.

(2) Maybe the terminator is different for the 5110 ? I don't think so, since last I recall I compared all the external IO pins between the 5100 and 5110 in the MIMs, and they are exact same pin outs (IIRC). But can double check that. In any case, I never came across any actual description of how the terminator is wired - so I think I'll open it and find out. [ this is the terminator that connects to the back of the 5106 tape drive itself, to indicate the "end" of the attached accessories to the 5100/5110 system ]

 

[voidstar78]

I think it's #1, the IO Card is bad. Swapping the IO card is difficult because not only in it bolted in with a bracket on the chassis, but the two connectors are glued to the card (glued on both sides, I don't want to risk chiseling that apart). So swapping the IO Card means opening up the entire power supply and swapping out the entire External IO panel (along with the card). Not difficult, but takes space having two 5110's fully opened, etc. -- just time consuming. And I also suspect the IO Card because I placed all the other cards into another 5110, and it is booting up fine (with the 5106 attached).

And... I'm using DC 6150 tapes!! I'm going to look around for some actual DC300 or DC600 tapes, to try those instead (isn't using a higher number DC tape like using a 120 minute audio tape? the reliability becomes iffy?). The tape rollers on the 5110 are deteriorated, so that may be also why using the tape isn't so reliable (have to hold it in place at just the right pressure/angle). But the rollers in the 5100 and 5106 are perfect.

Will try the new tapes whenever they arrive, see how that goes. Meanwhile, I'll still open that terminator, since I don't think what wires that they are crossing has yet been documented.

 

[stepleton]

I don't think the number maps well onto the length of the tape, and I wouldn't draw inferences about reliability either. I expect 6150 tapes are likely to work just fine if they are working fine now --- better, in fact, since the media will be newer and therefore further away from perishing (as it will probably be doing in most DC300 tapes you can get your hands on).

A more important difference between DC300 and DC6150 media is the magnetic coercivity of the tape.

One way to think about coercivity is to use paper material as an analogy. Imagine you have a fountain pen: if you touch it to a paper towel, the ink is drawn out of the fountain pen quite rapidly and you get a big mark fast. By contrast, a dense piece of cardstock is nowhere near as absorbent: you would have to dwell the pen a lot longer to make the same kind of mark you can make on a paper towel.

We could say that the cardstock has a higher "coercivity" to the ink than the paper towel does. (But note: coercivity is really only about magnetics --- nobody uses it for ink and paper!) Depending on your needs, you might choose one kind of paper over another. If you only wanted to brush the paper gently with your pen, then the low-coercivity paper towel is the choice for you. But if you want to make lots of fine marks, you'll avoid the paper towel's blotting action and choose the high-coercivity cardstock.


OK, now it's time to think about magnetic media. DC300 media is a bit more like the paper towel. This was a good choice for the early 1970s: since the important markings on magnetic tape are mainly the transitions between the polarity of the magnetic "marks" on the tape, the recording tape head is constantly swinging the polarity of its magnetic field back and forth as it writes. Because the DC300 media is so "absorbent", the tape head doesn't have to use such strong magnetic fields, which means that the electronics don't have to switch back and forth between particularly strong currents going in opposite directions.

Now, a few years later, people want to store more information on the tape, and also we become better at making tape heads and the electronics that drive them. We can't keep using the old media, though: we need to make finer marks on the tape, so it's time to switch from paper towel to something less absorbent. We therefore choose media with higher coercivity when we design the DC6150.


It would be reasonable to worry that the record heads IBM 51xx tape drives are not strong enough to make a sufficient impression on higher-coercivity DC6150 media. Empirically, however, it seems not to matter. This is about all anyone can tell you at this point: 6150 tapes seem to work. If there's a way that 6150 recordings aren't as reliable as DC300 recordings, it must take years for it to show up. I've never seen it.


For more details, refer to the chart on PDF page 5 of https://www.qic.org/html/standards/9x.x/qic95-101e.pdf showing the coercivity of various types of tape media. You can see that coercivity is measured in Oe, short for oersteds --- you know, one dyne per maxwell. Note also that DC6150 tape is rated to move faster than DC300 tape. That's one reason to think it won't be like old 120-minute tape cassettes after all: the 6150 tapes could actually be tougher than their shorter, more absorbent ancestors!

 

[voidstar78]

Alright, I'll try a variety then (non-3M brands too). Nothing against 3M, but maybe in the variety, something will work. (I only have 1x 3M 6150 left that I've marked working, all the rest have issues)

And once MARK'd for the 5100, I don't think the same tape can be read by a 5110 or vice versa (haven't 100% confirmed that, but I thought I also read that somewhere -- perhaps something like the 5100 is more EBCDIC oriented and the 5110 more ASCII oriented, if that makes sense? {such as how it encodes data to tape}). [ actually maybe it was the other way around, the 5100 had something weird, and 5110 used EBCDIC? ]

BTW, anyone familiar with "Z Code"? Is that an IBM thing? I was thumbing through some of the physical manuals, and found these z-codes listed in a table in the back of the SY31-0581-0 IBM manual (Language Support Maintenance Information Manual). Not sure if that's been digitized yet. These z-codes are just a different set of values like ASCII.

 

[krebizfan]

The 5110 BASIC Reference Manual page 47 suggests that some files can be shared between 5100 and 5110, some files can only be created on the 5100 but readable on the 5110 (like BASIC source tapes*), and some files created on the 5100 can not be read on the 5110. Since all 5100 files will be on tape, that requires tapes to be readable on the 5110.

I find it amusing that the 5110 manual adds an extra file type that the 5100 manual lacks but the 5110 can't read that new file type.

* But the 5110 can store BASIC files on tape so I am a bit confused by this.

 

[voidstar78]

Here's the "glued in connectors" for the Slot A IO card. It has two "translucent" cables that wrap around to the external IO ports at the back of the system.

The little haze below the connectors on the PCB is glue, on both sides. I tried pulling them a bit, but didn't budge at all and I didn't want to risk damage. Probably a hair dryer and some time would soften them up, but I won't do that just yet. Want to try a few things with fresh new tapes first, probably be a week.





Ran out of time today to open up the terminator gizmo. I'm pretty sure it's just a couple patch wires across a few pins. Will get to it later this week.

 

[voidstar78]

Not sure why the bottom side of the terminator has a separate panel. I guess if you remove this panel (two bolts inside), the terminator can still be used, while also the internals can be probed and accessed - so maybe that's why it's there, expansion or some diagnostics?



And as suspected, just 4-wires of concern. But it's a tad bit more than just some patch wires -- the components on the backside of the little boards, unfortunately the label on those parts aren't exposed (on either side; the black stack side is 5-stack, while the other side is 3-stack deep).

 

[voidstar78]

Ok, confirmed, the IBM 5110 Type 2 as-is can't have a 5106 or internal tape drive added to it. It's more than just a ROS(ROM) difference - physical hardware components are missing from the BASE IO board that sits next to the main Processor/Controller card (BASE IO board is different than the EXTERNAL IO card that sits in Slot A). When I tried adding either an internal or the 5106 external tape to the 5110 Type 2, any command to use the tape just returns ERR 013 (device not found). No surprises about any of that.



So how about turning a 5110 Type 2 (no-tape) into a Type 1 (has tape)? I attempted that by just taking ALL the cards from the Type 1 and putting them into the Type 2. And that actually works - sort of. I can REWIND the tapes, but I can't actually MARK them - doing so gives ERR 003 (which is a general "tape error"). The tapes I'm using do work in the IBM 5100, so I don't think it is actually a tape-media error (although that said, the 5110 could have timing differences in its code that make it more sensitive about the tape media? kind of a stretch, but its in the realm of possibility -- even if using the same code, there could be subtle timing differences in hardware {or just the aging of the hardware itself?}).

HOWEVER, I think the issue is actually this: In the 5110 MAP, near the beginning I remember coming across jumpers settings... ("X1X" means Type 1, and "X2X" means Type 2).



Indeed, the Executive ROS card for the IBM 5110-X1X (Type 1) does have that extra 4th jumper set. I got that when moving the Executive ROS from a type 1 into the type 2... But how about the auxiliary tape unit stuff, on the left???

I assume "Auxiliary Tape Unit Adapter Card" is referring to the card inside the 5106..... So, ok, just rotate that jumper!! Well... My 5106 external tape unit has no such jumpers. Per the diagram above, the jumper should be between the 5th and 6th chip in from the top. I don't want to actually pull the card out (since it is the only known-working-5106 that I have), but looking down the board in the gap available, I don't see any jumpers anywhere on the board.




So I looked at the part numbers of tape-related devices across the 5100 and 5110, and there are a variety of Auxiliary Tape Adapter part numbers..

For 5100:
Auxiliary Tape Adapter
1606988
1607134

For 5110:
Tape Adapter 5106
1607134
1607144
1607146

There is one overlapping part number, but the presence of different numbers also makes me think it is possible that some early 5106's aren't compatible with the later 5110's..

I'm not sure which part number I have for the 5106. The 5106 (that I have) doesn't have a single obvious part number for the unit as a whole (e.g. on the outside of the case or bottom -- there is a part number on the black brace of the tape mechanism itself, but that number doesn't match any of the above -- and unclear if it refers to the 5106 as a whole, or just that internal tape mechanism). The cables inside the 5106 also have part numbers (and with dates from '75 to '77). On the inside of the 5106 enclosure itself, it has this date shown below: early 1977 (Valentine's Day in fact) -- which is a date before 5110's were even released -- which stands to reason that *this* 5106 didn't have the jumper pins installed (and associated tracing) to make it "future compatible" with the 5110's that came out in 1978.


(above: 5106 tape unit likely assembled in 1977, using some cable parts from 1975, and sold for use on an IBM 5100 that was built c. 1976 for use at a hospital)


Above: part number posted near the tape mechanism doesn't match the numbers listed in the MAP - interesting that these are handwritten and "TESTED"; just reminds me of the stamps and QC markings on an old 1886 piano. A much earlier manufactured product... Curious why the IBM label above isn't using stamps (like the ones you tap into an ink base).





So, nice lesson to know that don't for granted that just any ol'e 5106 tape unit will work with any IBM 5 1/2/3 00. Which also means figuring out any modern emulator replacement to the 5106 will work a little differently between the 5110 and 5100. Bummer.