• Please review our updated Terms and Rules here

IBM 5100 System

Bob, APL (or "that Iverson language", as one of my co-workers used to call it somewhat scornfully) is a very powerful language, but also quite unconventional. It uses its own character set (although some compilers will accept composite substitutes) and is very heavily array-oriented and is probably one of the tersest languages out in the wild. It gathered a fevered following in the late 60s and seems never to have shed it.

APL exists for most modern platforms. I worked for a time with a vector supercomputer and APL was one of the supported langauges (along with FORTRAN) back in the day.

Children learning it as a first computer language evidently do better than adults who already know programming in more traditional languages. Or so I've heard.
 
Bob, APL (or "that Iverson language", as one of my co-workers used to call it somewhat scornfully) is a very powerful language, but also quite unconventional. It uses its own character set (although some compilers will accept composite substitutes) and is very heavily array-oriented and is probably one of the tersest languages out in the wild. It gathered a fevered following in the late 60s and seems never to have shed it.

APL exists for most modern platforms. I worked for a time with a vector supercomputer and APL was one of the supported langauges (along with FORTRAN) back in the day.

Children learning it as a first computer language evidently do better than adults who already know programming in more traditional languages. Or so I've heard.

Thanks Chuck. I had never even heard of APL until I started to do some research on the 5100. I looked at some examples of APL code and it doesn't look very intuitive (but not everything needs to be intuitive). I certaintly wouldn't set out to learn APL on a whim- the learning curve seems like it would be very steep. I have enough trouble remembering which day to take out the trash!
 
It was really there to have mainframe compatibility/development. That system when it came out had a huge selling point of being almost a mainframe on your desk which was pretty damn amazing at the time.
 
I recall seeing the 5100 at NCC, probably in 1980 at Anaheim. Given that 8-bit microprocessor machines had been in circulation since about 1975 (when the 5100 was trotted out), it was a bit underwhelming for the price and the tiny display. I just tended to lump it in with the HP85 and 9800 units, as well as the Wang systems (e.g. 2200) which preceded it by several years.

In 1975, BASIC wasn't all that popular and APL was downright obscure. Now, if the 5100 ran FORTRAN, that would have been interesting. But FORTRAN was already available on the PDP8 for a comparable price.
 
A very long time ago I read an interesting article on the development of the prototype for the 5100, the SCAMP (Special Computer APL Machine Portable) from 1973. Two things that stuck in my mind about it was that it used audio cassettes for storage, and that it was built into a briefcase with curved edges so it would rock when knocked over. References to the SCAMP on Google are mostly the same regurgitated piece of text about it using the PALM processor and so on, but no further details. I've been searching the net high and low for a picture of this groundbreaking machine but surprisingly, have not been able to find any. However I just happened to find a poor quality scan of that article from PC World October 1983 (gosh, was it that long ago!) buried in this PDF, starting at page 26: http://archive.computerhistory.org/resources/access/text/2011/03/102679788-05-01.pdf There is another article about it a bit further on, too.
Given its historical value you think it would be much better known than it is.

Steve
 
Last edited:
A very long time ago I read an interesting article on the development of the prototype for the 5100, the SCAMP (Special Computer APL Machine Portable) from 1973. Two things that stuck in my mind about it was that it used audio cassettes for storage, and that it was built into a briefcase with curved edges so it would rock when knocked over. References to the SCAMP on Google are mostly the same regurgitated piece of text about it using the PALM processor and so on, but no further details. I've been searching the net high and low for a picture of this groundbreaking machine but surprisingly, have not been able to find any. However I just happened to find a poor quality scan of that article from PC World October 1983 (gosh, was it that long ago!) buried in this PDF, starting at page 26: http://archive.computerhistory.org/resources/access/text/2011/03/102679788-05-01.pdf
Given its historical value you think it would be much better known than it is.

Steve

I had seen this article somewhere as well and found a few pics of the machine, which is in the Smithsonian.img_0029.jpgimg_0030.jpgimg_0031.jpg

The SCAMP machine was clearly the inspiration for the 5100 and even the front panel layout looks very similar. I recently found a picture of an IBM 3767 Communications Terminal from about 1974: http://en.wikipedia.org/wiki/IBM_3767. It's interesting, but not surprising, to see the similarities between the two machines. The tractor feed looks identical to the one on the 5100's printer.

slimbob
 
Last edited:
I recall seeing the 5100 at NCC, probably in 1980 at Anaheim. Given that 8-bit microprocessor machines had been in circulation since about 1975 (when the 5100 was trotted out), it was a bit underwhelming for the price and the tiny display. I just tended to lump it in with the HP85 and 9800 units, as well as the Wang systems (e.g. 2200) which preceded it by several years.

In 1975, BASIC wasn't all that popular and APL was downright obscure. Now, if the 5100 ran FORTRAN, that would have been interesting. But FORTRAN was already available on the PDP8 for a comparable price.

One thing that really stands out to me about the 5100 series in general is the high cost vs. performance ratio of these machines. The company that originally purchased the 5100 system I have paid around $12,000 for it in 1976 or 77. That works out to something like $45,000 in 2012 dollars, although the president of the company put the figure at $35,000. I'm not sure what the actual figure is, but a quote I recall was "it was obsolete before we paid it off".

slimbob
 
I tried but cannot remember myself and just did a little bit of looking but did not find anything. Is that transport using cassettes that are only driven by the capstan? And internal gears in the cassette that drive the two reels together? Or is there some way that the reel tables are driven by an external motor? The more I think about it somehow think the only thing that moved the tape was the capstan and somehow the reels were locked together.

I need to examine the cartridges more closely, but I know that the capstan drives a roller that protrudes slightly from the cartridge but doesn't touch the tape directly. The roller drives an elastic rubber belt that surrounds the two tape reels and rotates them. There is no other mechanical drive for the two reel, just the belt. I haven't checked the belt's on the old cartidges I have, but my guess is that the belts have lost their elasticity and probably won't drive the reels properly. I understand that you can replace an old belt with one from a newer cartridge (there is at least one company still making these cartridges). I may give this a try because apparently, without proper tension on the rollers the tape can just unspool inside the case, which isn't a good situation. If I can duplicate the original tapes I have (and its a big if at the moment) I won't need to use the originals, so I may be able to use one new belt to temporarily fix all of the cartridges I have long enough to copy them. I'll give it a shot anyway.
 
Cool, watching the tape drive work is a thing of beauty! Epically when you're running the Basic training course. Like I said before don’t know how but the course just requires you to do a LOAD 00 and then maybe RUN and the whole thing takes off and runs from there. Don’t have any idea on what would be required to copy tapes because from what I recall there were no type of tape utilities and have to wonder if there was a requirement for formatting the tapes or if they built the track index as they went, unlike any type of modern storage device don’t think they had anything like a file access table or index of any kind and remember the one system I use to use for power and efficiency calculations that I used track 00 to print a menu of what was on the other tracks. I never saw a system with a second tape drive so have to wonder if you second tape drive comes with anything that may allow it to do a copy or dub function?
 
Cool, watching the tape drive work is a thing of beauty! Epically when you're running the Basic training course. Like I said before don’t know how but the course just requires you to do a LOAD 00 and then maybe RUN and the whole thing takes off and runs from there. Don’t have any idea on what would be required to copy tapes because from what I recall there were no type of tape utilities and have to wonder if there was a requirement for formatting the tapes or if they built the track index as they went, unlike any type of modern storage device don’t think they had anything like a file access table or index of any kind and remember the one system I use to use for power and efficiency calculations that I used track 00 to print a menu of what was on the other tracks. I never saw a system with a second tape drive so have to wonder if you second tape drive comes with anything that may allow it to do a copy or dub function?

There is a tape to tape function that would allow copying the contents of one cartridge to another, from one drive to a second. This is something I need to study because my impression is that you have to load some of these functions from a tape cartridge first, then execute the function (or command- not sure on the correct terminology). This would mean that I will have to locate a tape among the ones I own that has the proper command on it and load it into the machine before I can even try to copy my tapes. I could be totally wrong about this as I have not had time to study all of this in detail. Hopefully it will turn out to be much more straight forward than it seems now.

As I recall, if you did not have a separate auxiliary tape drive you could copy an entire tape by using another function which, when executed reads the tape into memory until the memory is filled, then stops while you exchange tapes. It then writes the data to the second tape, after which the tapes are switched again and the process is repeated until the entire contents of the original cartidge have been copied to the second cartridge. This doesn't sound like a fun process to me because the standard amount of memory was 16k and part of this is used to store the copy function when it is loaded from cartridge (if I understand correctly). So, each copying step would transfer less than 16k to the second tape, and an entire tape is around 200k. Using a second drive, the contents of the first tape is apparently read and written to the second tape directly. I don't know if there is anything in the way of error detection or correction-I need to study this more.

Also, regarding the tape tracks, my impression is that one track is used for data, and one is used for what I think are called labels, or basically a file name for whatever data is to be stored. So, when you typed load 00 (in your example), the 00 was a label for whatever was to be loaded and the drive searches for the 00 label on the label track (or whatever it is called) and stops at that point. If you have typed run the computer executes the program associated with that label. Again, this is just my impression from a first pass through a lot of information, so I may be off on some of this.

slimbob
 
It's been a while since I updated this thread and for what it's worth I thought I would give a brief overview of what I've been up to for the last few months.

One of my goals from the beginning has been to archive the original IBM tape cartridges I have, and any others I might come across. The original cartridges I had all suffered from brittle or broken drive belts, and in many cases the belts had become fused to the tape itself where the two elements had been in contact. In these cases the belt could literally rip the magnetic coating from the tape. I found that a small amount of distilled water applied between the two surfaces with a small brush helped to separate them without damage. After any excess moisture dried out I was able to replace the broken belts with belts removed from NOS cartridges. Most of the cartridges found on eBay are at least 20 years younger than the originals, but completely useable. There are some tricks to installing the new belts and tensioning the tape- but it's not rocket science.

Having rebuilt all of my original cartridges it was time to try them out. I was surprised and very happy to find that every tape I had rebuilt was readable and seemed to work well in the tape drives. I was able to read the contents of each tape, but I decided not to run the tapes any more than necessary until I can copy them. I was hoping to find a tape cartridge among the ones I have called the "Customer Support Cartridge". This cartridge contains the commands used to perform tape to tape copying and is something I will have to find before I can archive my tapes.

While checking the tapes I noticed a few things about the operation of the tape drives that were a little troubling. Neither drive seemed to lock the tape in position against the capstan. This meant that I had to actually push the tape against the capstan with my hand. The internal drive was fairly quiet, but the 5106 auxilary drive would emit a loud squealing every other minute or so. Rather than take any chances I decided to pull the drives out and look them over. The tape positioning problem turned out to be caused by broken guide rollers underneath the tape itself. There are 4 hard rubber, spring mounted rollers that the tape rides on as it is inserted into the drive. On the bottom of each tape are two recesses cut into the heavy aluminum plate that forms the core of the cartridge. Normally, when the cartridge reaches the proper position two of the rollers push up into the recesses and hold the tape in position against the capstan. On both of my drives the rubber rollers had become brittle and broken away from the plastic hubs they were mounted to. Luckily, I found that a size 00 rubber faucet washer is identical in size to the rubber originally used for the rollers. The washers fit the plastic hubs perfectly as well, so the tape positioning problem seems to be solved.

The squealing noise from the 5106 drive was almost surely caused by dried up grease in the spindle bearings. There are 6 small R series bearings in each drive. 4 R4-ZZ bearings for the forward and reverse drive spindles, and 2 R2-ZZ bearings for the capstan spindle. On the squealing drive forward and reverse spindles rotated roughly, and a closer look showed that the grease in two of them was completely dried up. I decided to replace all of the bearings with new ones since I want to be able to rely on the drives, and after sitting idle for 30 years the condition of the bearings was questionable. I will save the old bearings, but for now, the new bearings give me a little peace of mind.

Getting the internal drive out of the 5100 required me to open up the "gate" that the backplane and cards are attached to. I had held off doing this before because typically the foam cushioning under the cards is in very deteriorated condition- ie "a mess". This turned out to be the case. The foam deteriorates into a black, tarry, sticky mess that gets on everything. It took me two hours of careful work with q-tips to carefully wipe the goop from the top of the cards and everywhere else it had attached itself. I found that the foam was attache to the computer housing with some type of peel and stick adhesive which was easily peeled away. The foam consists of two 1/8" layers of what I believe was polyurethane foam (similar to conductive foam), with a thin plastic sheet between them. The sheet looks like it may have some antistatic coating on it. If anyone knows anything about this foam "sandwich" material or who made it I would like to find out more about it. As a replacement for the original foam I bought a 1 X 1 ft. X1/4" piece of HD conductive foam which should substitute for the original very well.

I am in the process of finishing up this work and should have the computer and drives together and working again soon. I completed some work on the 5103 printer, which is probably worth mentioning. I had two problems with the printer. One of the hard plastic brackets that mount the tractor feed had broken, and no replacements are available. I was able to repair the broken plastic with a product called Plastex. All I can say is the product worked as advertised and repaired the broken plastic so the part is 100%. Another problem I had was that a plastic gear and it's metal shaft were missing from the ribbon drive. I ended up using a plastic resin casting kit from the hobby shop. I made a mold out of silicone rubber and used a two part urethane casting resin to duplicate the original gear. I actually made several just in case, and I must say that the results far exceeded my expectations! For the shaft, I used a stock piece of steel rod and used a Dremel tool and a file to put a flat and a small groove for a snap-ring on one end. The original rollers had also disappeared, and I found that faucet washers of the exact size made perfect replacements. That's two uses for faucet washers on my 5100!

I'm attaching a few pics that illustrate some of what I have gotten into here. I may be able to claim the prize for most boring post, but hopefully something here will prove helpful to someone. I'd love to hear from anyone else who has experience working on these machines.

slimbobIMAG0246.jpgIMAG0257.jpgIMAG2158.jpgIMAG2146.jpgIMAG2420.jpg
 
Last edited:
Not boring at all, it was seriously interesting. Helps keep the hope for a 5100 up!.
patscc

Thanks patscc. When I get the drives finished and the system running again I will try to do a video showing the operation of the tape drives in action. They are fascinating to watch! I have been really impressed by the straightforward, simple, and robust design of these things- really a pleasure to work on. I've gotten to know one of the members of the 5100 design team (he wrote the tape drive microcode) and he told be the tape units were known as BLT's, for the guys that designed them- Bob, Larry, and Tom. Hat's off to those guys- they did a great job!
 
Slimbob what did you use to clean the exterior surfaces of your IBM 5100? Is there a particularly effective solution (like toothpaste for example) to erase some of the more heavy markings? Thanks.
 
Wow.. this is a gem! I suppose that there are few of 5100 that still work well.
 
Back
Top