EEVblog #903 - HP85 Vintage Computer Teardown

[EEVblog]

1979, a screaming 613KHz clock, killer graphics, and it's a Hewlett Packard, it doesn't get much better than this!
Inside the classic HP85 Scientific / Engineering Professional Personal Computer.

[gus789]

Great tear down, Dave.

I think the tape roller is not supposed to be coated with that brown liquid gunk. The rubber probably decomposed/melted over time. I've used a vintage HP network analyzer (8510?) that had one of these tape drives and the tape failed in exactly the same way.

Cheers

[EEVblog]

I think the tape roller is not supposed to be coated with that brown liquid gunk. The rubber probably decomposed/melted over time. I've used a vintage HP network analyzer (8510?) that had one of these tape drives and the tape failed in exactly the same way.

Ah, thanks. I just assumed it was some sort of lube.

[kwass]

You can "rebuild" the capstan as I described here: http://www.series80.org/Articles/capstan-repair.html
The tape cartridge you have is filled with the goo the decomposed rubber and will probably never work again.  However these cartridges can still be found new on ebay and elsewhere.  Look for 3M DC1000 tapes.

[rch]

I remember using one of these machines (possibly a previous model) for data entry and statistics in physiology research in the mid 1970s.  I seem to recollect it had some interpreted language beyond just calculator functions, was it a form of BASIC?  But it also had provided routines for statistical analysis.  Printer paper worked quite well as a display, even for editing programs.  Seemed pretty impressive for a desktop machine at the time!

[nfmax]

That sounds a bit early for an HP85. I remember seeing a demo HP85 in about 1980 (looking to replace the Tektronix 4051's and 4052's we were using). Maybe it was an HP9830A http://www.hpmuseum.org/hp9830.htm? That was a BASIC programmable calculator-cum-computer, with an LED display and a separate printer that stacked on top of it. The 9825 http://www.hpmuseum.org/hp9825.htmwas a scaled-down version with a built-n printer.

[rch]

That sounds a bit early for an HP85. I remember seeing a demo HP85 in about 1980 (looking to replace the Tektronix 4051's and 4052's we were using). Maybe it was an HP9830A http://www.hpmuseum.org/hp9830.htm? That was a BASIC programmable calculator-cum-computer, with an LED display and a separate printer that stacked on top of it. The 9825 http://www.hpmuseum.org/hp9825.htmwas a scaled-down version with a built-n printer.

On reflection, I am sure you're right.  I don't think it had a CRT.   As to which of the above it was, I honestly can't remember.   I think it had a rather narrow format  printer next to it rather then integral, but I can't be certain after all this time.  So not strictly on-topic;  but they were impressive machines at the time.

[max666]

So what's the magnetic component, and why is it magnetic?

[bktemp]

So what's the magnetic component, and why is it magnetic?

It's the linearity coil:
http://www.repairfaq.org/sam/deflfaq.htm#dshlc

[netdudeuk]

That's a very nice computer.  I wish I had it.

I programmed one of these for a living in 1981.  It had the serial interface and the ROM drawer.  I wrote the software used to drive the electronic cricket scoreboard at Headingley, Leeds, where the third Cornhill test took place, England vs Australia (photo here - https://www.the-newshub.com/cricket/the-headingley-ashes-test-1981-the-ian-botham-test-match).I joined the company at the last minute, had to start coding from scratch and I was debugging the code during the five days.  I spent my time in the score box with the two scorers and it was a historic match.

I did the same for the rugby scoreboard.

[max666]

So what's the magnetic component, and why is it magnetic?

It's the linearity coil:
http://www.repairfaq.org/sam/deflfaq.htm#dshlc

Using a near saturated coil to compensate for dampening in the deflection coil. Neato!
Thank you very much, that page you linked will make for an interesting read, I'm sure.

[craigh]

Cool Dave!  Thanks for showing us the HP85.  I'm amazed how clean it was inside for such an old instrument.

I programmed one of these back in 1980 to run my bench setup for testing microwave radio transponders.  The HP85 had a GPIB interface plug-in that controlled the signal generator and other test equipment I was using.  I remember enjoying setting up the graphing functions to display the frequency response of the microwave radio's filters.  Pretty geeky stuff for a young electronics guy like myself at the time.

BTW - the tape cartridge is a DC100.  A previous poster said it was a DC1000, but I assume that was a typo.  At the time I also worked on some data logging equipment.  They used DC300 tape cartridges (a bit larger than the DC100) to store the logged data.

[open loop]

Now I really must investigate the loft, I think I have two of these :-). I may even have some of the original manuals. The last time I saw one being was about 2004ish and I think it was running some very old test software. And yes I remember seeing these fail when the tape capstan turned to goo it was impossible to read the tapes.

[Kryten 2X4B]

Great video Dave, as you noted the board interconnects are fragile and you do have to be careful with them. As others have said the capstan in the tape drive turning to goo is quite an issue. Also as said previously there are a number of web sites that detail ways to fix the problem, I've tried most of these and they work for the HP85 as only moderate tape speeds are used. The same drive is used in the HP 5420A/B and 5423A spectrum analysers and early versions of the 8510 Network Analyser. In the 5420 and 23 the tape speeds are much higher and the fixes I've tried have never been reliable for the long term, however there is or was an ebay seller that has a tyre that fits the capstan once the goo is removed, and so far it hasn't failed.

[kwass]

BTW - the tape cartridge is a DC100.  A previous poster said it was a DC1000, but I assume that was a typo.  At the time I also worked on some data logging equipment.  They used DC300 tape cartridges (a bit larger than the DC100) to store the logged data.

Not a typo.  The DC1000 is a longer version of the DC100 and is also easier to find on ebay, etc..  I'm pretty sure that the HP-85 used DC100/DC1000 tapes not DC300 tapes, read this: http://www.hp9825.com/html/qic_and_the_dead.html

[roli]

I absolutely love vintage computer teardowns!

[Richard Head]

Dave, that yellowed plastic case is crying out to be bleached white again.

[Halcyon]

Dave, thank you! Yet another awesome vintage computer video. These are by far my favourite. 

[joeqsmith]

Thanks for putting this video together.   

[valvedoctor]

Dave, that yellowed plastic case is crying out to be bleached white again.

I actually like the bromide patina. I'd leave it as it.


Sent from my iPhone using Tapatalk

[KerryW]

We had one of those on the ship I was on (in the Navy) for a while in 1977.  It must have been a pre-production unit, but it had the internal (white) CRT and the thermal printer.

I don't know why WE had it, I don't think any of us knew squat about computers.  Mostly we played Star Trek on it.

[joeqsmith]

Great video Dave, as you noted the board interconnects are fragile and you do have to be careful with them. As others have said the capstan in the tape drive turning to goo is quite an issue. Also as said previously there are a number of web sites that detail ways to fix the problem, I've tried most of these and they work for the HP85 as only moderate tape speeds are used. The same drive is used in the HP 5420A/B and 5423A spectrum analysers and early versions of the 8510 Network Analyser. In the 5420 and 23 the tape speeds are much higher and the fixes I've tried have never been reliable for the long term, however there is or was an ebay seller that has a tyre that fits the capstan once the goo is removed, and so far it hasn't failed.

Can you find any more information on the part you found?  Thanks.

[Kryten 2X4B]

I had thought that the ebay seller had run out of these but I just checked and he appears to have stock again. I have no connection with this person, but I have purchased an number of these replacement tyres and as yet I haven't had a failure. Follow this link to the ebay listing: http://www.ebay.com/itm/HP-85-8510A-B-Tape-Drive-Replacement-Capstan-/371734921300?hash=item568d20c054:m:mVP6BHFFm0KZ2eUGOHe1U_A

[KeepItSimpleStupid]

Dave:

You have to get the IEEE-488 interface for it.   I programmed the HP85 and HP86 computer.

It was the ERA in which the floppy disk drives, printers, plotters all connected via  the HPIB (IEEE-488) interface.

The ink jet printer was like $150.00 USD and the special coated paper was like  $50.00 USD per box.  I forget the price of the print cartridges.

We did write a Kermit protocol for the HP86.

[Hawkuletz]

Hello all!

I have an HP-85 (well, 2 in fact, but one had the simple fuse problem) with another power supply problem. I picked them up as non-working from a highschool that received them as a donation from Boston University (it says so on the case, in a message burned with the soldering iron). It appears that someone has forgot to set the voltage switch and plugged them in (line voltage here is 230v). So one of them had both fuses blown (mains and the one inside the power supply), the other had only blown the mains. Inside the PS it was the switching transistor (I think) that protected the fuse by destroing itself along with its driver transistor.

The transistor used for switching is a 2N6317 and the other one (I think it's its driver) is marked with 113M and 3-449. I know it's a PNP one (from measuring the one in the working PS), but not much more. I have seen a schematic in Dave's film, but I couldn't find it. All I found is the HP-85 repair course that doesn't include power supply schematics. Can anybody point me to that schematic diagram?
Since I don't have those particular transistors (well, I'm not even sure what the other one is), I was thinking if maybe they could be replaced either with 2 equivalents or maybe both of them with a single MOSFET.

Thank you in advance

[hoofie]

I used one of this in 1989/1990 to do my final year EE Project.

It involved comparing a Spice model of a transistor against a real junction on a silicon wafer.

The HP85 was connected via GPIB to a Parametric Analyser which was then connected to a wafer probe station with a single wafer on it.

I recall it was all done in BASIC but was quite powerful.

I got a Distinction for it so all was good !