Found - 9845 Light Pen


Precaud
 

While looking through a drawer looking for an old adapter, I came across the light pen for my 9845B. The engineering programs I used back then didn't support it, but I think I used it once with some Basic Users Club program just to see how/if it worked.

If anyone is interested in it, let me know. I have no idea what these things are worth, so I'm open to suggestions.

John


Ansgar
 

That's pretty good news, since Francois Lanciault (francoislanciault@...) has currently started a project for building a replica for that light pen device. But without having a working one on his own, this is a tough task. I recommend that you get in touch with him, since we 9845 friends are all looking forward to the outcome of his project.

By the way, if you have any history stories to share around the 9845 (e.g. on the Basic Users Club program) you are welcome to do so!

-Ansgar 


Lee A. White
 

John

I am interested. I just started work on restoring a pair of 9845Bs and a 9845A I recently acquired - all dead as received. I was a design engineer at HP Loveland CPD in the 70s and worked on the 9830 and the 9845. I am in the USA.

Lee


Ansgar
 

Guess everyone in this group is interested (including myself), however I'd suggest to give the reconstruction of the light pen priority, independent who later calls the item his own. So we all can take profit of it. So please coordinate.

Just my 2 cents.

-Ansgar


François Lanciault
 

Well of course I am interested ! I wouldn't be trying to reproduce it I wasn't.

As Ansgar explained, if you are kind enough to lend it to me for a while, it would help a lot the reconstruction attempt. I would take measurement and I would return it to you, undamaged, so you can sell it to whoever you want.

But if I can buy it from you, it would almost guaranty a successful reconstruction as I would open it to confirm a few obscure electronic points and be able to know the actual PIN photo diode used. It would also allow me to reproduce the actual printed circuit board so the replacement would fit inside a same size 3D printed case.

Let us know what you think.

François

On Jul 1, 2018, at 22:52, Precaud <jbau@centurylink.net> wrote:

While looking through a drawer looking for an old adapter, I came across the light pen for my 9845B. The engineering programs I used back then didn't support it, but I think I used it once with some Basic Users Club program just to see how/if it worked.

If anyone is interested in it, let me know. I have no idea what these things are worth, so I'm open to suggestions.

John




Precaud
 

Hi Lee,

Thanks for your contribution to this amazing machine, what part did you work on? Looks like you're first to speak for the light pen. I'm guessing you'll coordinate with other members on the reconstruction?

As to stories, I bought a 9845T in 2000, and then traded it in on the bit-slice with fast mono top within a year (HP had very liberal tradein policies back then). My girlfriend and friends thought I was crazy borrowing over 20 grand, to buy a *what*? I still have that unit and it still works fine. It is the most productive and reliable engineering tool I ever owned. 16 years of daily use, often 24/7, and not a single thing ever went wrong with it! It only gets booted up a couple times a year now.

The bit-slice was a godsend. It did an FFT in Basic faster than the original unit did running Frank Key's FFT binary. (I never figured out how to make binaries, but did a lot of assembler programming, mostly I/O related.)

Along the way I picked up spares of anything 45-related, so I have lots of stuff to go through. I'm pretty sure most of the BUC printouts were thrown out already, but I'll have another look. I'm in slim-down mode, so a lot of this stuff has to find a new home.

Cheers.


Precaud
 

Hi François,

Lee was the first speak so I have to honor that. Perhaps you can coordinate with him? If he approves, you could be an intermediary in the process.

  John


Steve Leibson
 

OK Lee,

I've been racking by brain trying to figure out if I know you. I was at CPD in Loveland and Ft. Collins from 1975 to 1980. I worked on the 9878A I/O expander, the 98036A serial I/O card, the 98225A systems programming ROM for the 9825, the bulletproof die-cast case for the 98034B HPIB card, and the unreleased 98037A analog I/O card. I also did emergency meatball surgery on the original 9845A I/O backplane so the product could ship on time.

For the life of me, I can't dredge up a "Lee" in the CPD lab from my rapidly aging brain. Do we know each other? (Apologies in advance for not remembering you.)

--Steve Leibson



On 7/2/2018 12:12 AM, thermoman@... wrote:
John

I am interested. I just started work on restoring a pair of 9845Bs and a 9845A I recently acquired - all dead as received. I was a design engineer at HP Loveland CPD in the 70s and worked on the 9830 and the 9845. I am in the USA.

Lee

-- 
Steve Leibson

Phone (Cell): 408-910-5992
Phone (Home): 408-292-4930


Please feel free to link to me on LinkedIn


History site: www.hp9825.com


 

Yep, totally agreed.

-Rik




On Mon, Jul 2, 2018 at 3:15 PM +0200, "Precaud" <jbau@...> wrote:

Hi Lee,

Thanks for your contribution to this amazing machine, what part did you work on? Looks like you're first to speak for the light pen. I'm guessing you'll coordinate with other members on the reconstruction?

As to stories, I bought a 9845T in 2000, and then traded it in on the bit-slice with fast mono top within a year (HP had very liberal tradein policies back then). My girlfriend and friends thought I was crazy borrowing over 20 grand, to buy a *what*? I still have that unit and it still works fine. It is the most productive and reliable engineering tool I ever owned. 16 years of daily use, often 24/7, and not a single thing ever went wrong with it! It only gets booted up a couple times a year now.

The bit-slice was a godsend. It did an FFT in Basic faster than the original unit did running Frank Key's FFT binary. (I never figured out how to make binaries, but did a lot of assembler programming, mostly I/O related.)

Along the way I picked up spares of anything 45-related, so I have lots of stuff to go through. I'm pretty sure most of the BUC printouts were thrown out already, but I'll have another look. I'm in slim-down mode, so a lot of this stuff has to find a new home.

Cheers.


Ansgar
 

I started to describe how binary programs can be built on my website (see http://hp9845.net/9845/software/binaries/). Still want to write a real tutorial on how HP, SSS etc. created their binaries by use of the Assembly Dev & Exec ROM, another one of my pending tasks.

The top of the art was to write a binary which controls both PPU and LPU, by the way. The Assembly Dev & Exec ROM runs programs on the PPU only.

-Ansgar


Lee A. White
 

John
 
Over-riding my own lust for tech toys and HP ones in particular and admitting to being occupied at the moment with trying to determine exactly what I have just received and then trying to make what I have operational, I think it is in the best interests of the group as a whole that Francois have access to the light pen now. I am willing to cooperate fully in making that happen some how. I would hope that it can lead to everyone understanding the little bugger and to every one who wants a light pen having at least a replica that does work. I would like to have the light pen at some point, but when is not important. If you agree, you should plan on sending it directly to him now. Given that I don't know anyone here personally or even know where they are located, some details will have to be put together. It is probably best that you and Francois work out those details together. Let me know how you want to proceed.
 
My first impression is that this is a great group. I hope that I will be well received and welcome here.
 
And, if you really want to hear an old man ramble on about what he did back when and what it was like to be at HP Loveland CPD at that magic time, I can do some of that too.
 
 
Lee

From: Precaud
Sent: Monday, July 02, 2018 8:21 AM
Subject: Re: [VintHPcom] Found - 9845 Light Pen

Hi François,

Lee was the first speak so I have to honor that. Perhaps you can coordinate with him? If he approves, you could be an intermediary in the process.

  John


Lee A. White
 

I was a physicist and a mechanical engineer with significant interest and education and experience in electronics and instrumentation and computers and software when I interviewed with HP. I was initially hired by HP Loveland to investigate and resolve reliability and production related problems with the 9810, 20, 21, 30, and 66 product lines. The employment offer included the promise of a design engineering position on the next product design cycle for the replacement for the 9830/66 - the 9845.
 
The 9830 was suffering unacceptably high failures rates for multilayer boards - particularly the memory boards. All products were suffering from poor workmanship in soldering with blow holes and cold solder joints prevalent. The 9830 had display chip intensity variations that looked terrible when those chips were placed side by side in the 32 character long display. The 9866 printer had similar dot intensity variations from character to character across a page. There was only one source and one color for the thermal print paper. Keyboard switches were problematic, key caps were falling off at one point, and there were a number of other assembly problems that were impeding production rates and causing misses in production schedules. I became an advisor to other HP division on multi layer PCB production and solderability, conducted seminars, and rewrote part of the IPCA600B printed circuit workmanship standards. I am proud to had helped to make the 9830 a reliable machine and multilayer boards a reasonable design option.  I had resolved those production issues by the time the 9845 program was initiated and as promised I was moved into the design group to work on QWERT - as the 9845 was code named.
 
I was assigned to Gary Egan's group under the supervision of Ray Cozzens. Key assignments were the integration of the computer, the display monitor, and the printer into one desk top machine; the mechanical design of the printer (but not the print head itself); the cooling of and the thermal management of the entire machine including the power supply; design for shock and vibration integrity and environmental testing; over all reliability assessment and failure rate determination and warranty cost estimation; and manufacturability in general. I came up with the overall configuration of a monitor up on towers, easily removable, a printer under, dual fans in the rear, and a dual cooling fan system that sucks - instead of blows. The final cosmetic design of the case was done primarily by an industrial designer from corporate on temporary assignment to our group, and the actual case parts design was done by another ME in our group, but the over all configuration and look came from me. That dual fan system was memorialized in a cartoon due to the fact that the first version was - well - pretty noisy to put it nicely.  So, I spent a lot of time in the recording studio at Loveland - yes HP had one - doing noise tests to get it quieted down. Air flow testing presented challenges too, pitot static probes, hot wire anemometers, thermocouple grids, propeller air speed meters - nothing gave reliable and repeatable results. Ultimately, it was done with a sophisticated home brew instrumentation package - a very long tube made from dry cleaner bags taped together and a stop watch to determine time required to fill it. Component temperature monitoring got interesting when the department secretary (a very attractive lady to say the least) stepped in front of the IR camera we were using - interesting hot spots! Thermal print paper tracking (it was a pretty high resolution graphics printer - straight lines were a necessity) showed significant problems when lots of dots were printed on one side and few on the other. Thermal paper works by melting a wax coating and allowing two pigment chemicals to mix and react and create a dot. When you shut off the print head heat that creates the dots and the wax solidifies, it sort of glues itself to the print head. The paper drive system has to break that paper loose from the print head. That is where the staccato snapping noise of a thermal printer comes from. When you do that in an unbalanced way across a page, the paper wants to swerve toward the side with the most printed dots destroying tracking and making curvy lines in the process. The power supply was reportedly the highest power density supply ever made by HP, and cooling that power supply was indeed a challenge - and it had to be done quietly too. 
 
Prior calculator cases were made from aluminum extrusions and castings and sheet steel - good shielding and strong but heavy. The 9845 case was molded from a new plastic - a foamed Noryl from GE as I recall. That brought in a new problem - shielding. A new metal spray process had to be developed and proven. I did not do it, but that print head design was pretty amazing too. At a time when a 2-1/2 inch wafer was pretty big - that monolithic print head was 8 inches wide! None of this mechanical stuff was very routine handbook engineering. Think about how big and how heavy a 9866 is. Now imagine being in a meeting and being told your assignment is to shrink that printer and stick it inside a desk top calculator about the size of a 9830/66 - and oh by the way, you need to put a monitor on top of it too - exactly where the printer usually goes - and the paper has to be easy loading and self feeding too, and we want twice as many dots and graphics too - and be able to blacken the page, and it has to be real quiet! Yeah - right - and you want to be in production in how many eons?
 
Now, when ever I read about computers all I ever read about is how brilliant and wonderful all the EEs were. No doubt, they were all that and more. However - NO ONE EVER EVEN MENTIONS THE MEs - the guys who kept it all together under shock and vibration, got it all crammed into the little itty bitty box, forced air to circulate in what little volume was left (EEs tend to think every little nook and cranny is just a dandy place for another component), made it pretty to look at, light enough to move, quiet enough to live with, shielded enough not to radiate EMI-RFI, and kept it from melting! Not to mention making it something the factory could actually build. So, raise one for the MEs some time.
 
Lee
 
 
 
 
 
 

From: Precaud
Sent: Monday, July 02, 2018 8:15 AM
Subject: Re: [VintHPcom] Found - 9845 Light Pen

Hi Lee,

Thanks for your contribution to this amazing machine, what part did you work on? Looks like you're first to speak for the light pen. I'm guessing you'll coordinate with other members on the reconstruction?

As to stories, I bought a 9845T in 2000, and then traded it in on the bit-slice with fast mono top within a year (HP had very liberal tradein policies back then). My girlfriend and friends thought I was crazy borrowing over 20 grand, to buy a *what*? I still have that unit and it still works fine. It is the most productive and reliable engineering tool I ever owned. 16 years of daily use, often 24/7, and not a single thing ever went wrong with it! It only gets booted up a couple times a year now.

The bit-slice was a godsend. It did an FFT in Basic faster than the original unit did running Frank Key's FFT binary. (I never figured out how to make binaries, but did a lot of assembler programming, mostly I/O related.)

Along the way I picked up spares of anything 45-related, so I have lots of stuff to go through. I'm pretty sure most of the BUC printouts were thrown out already, but I'll have another look. I'm in slim-down mode, so a lot of this stuff has to find a new home.

Cheers.


Jack Rubin
 

Thanks for the story - and your work on the 9845. Maybe you and Dyke could work with Ansgar to add a few more pages to 9845.net?
There can never be too many stories from the "glory days"!

I'll definitely see you to another round! Want to come join us in Chicago for VCF-Midwest in September? We'll save you a spot as a speaker!

Best,
Jack


Ansgar
 

Lee,

thanks for the great insights. You will be a real win for this group. I guess it was one of the special cultural things at HP that EEs ranked top of the MEs, at the end of the day. But you did a phantastic job in all areas, chapeau from my side. The design of the 9845 was much ahead of its time.

I am planning an exhibition for a very special edition of the 9845 on this year VCFB by the way. It is what I call TEMPEST edition. See http://hp9845.net/9845/hardware/9845b/#tempest. Must have been a horror vision for every ME to see what security requirements can make of the original design, but it might be of interest anyway for every ME, too.

-Ansgar

P.S. You are welome to provide me with a (time period correct) photograph of you, and I'll add it to the team overview on my site :-)


Jack Rubin
 

And of course, I should have suggested that Dyke hook up with _Steve_ to do a page at hp9825.com. Sorry, too much excitement about the amount of new old info that has shown up over the last couple of weeks.

Jack


François Lanciault
 

Thank Lee. The community will hopefully benefit from this (if I can do my part of the job)

I will return the light pen when asked for or as soon as the replacement is functional whichever comes first. I will obviously post every progress on this board.

Best regards,
François


On Jul 2, 2018, at 19:05, Lee A. White <web2464p@...> wrote:

John
 
Over-riding my own lust for tech toys and HP ones in particular and admitting to being occupied at the moment with trying to determine exactly what I have just received and then trying to make what I have operational, I think it is in the best interests of the group as a whole that Francois have access to the light pen now. I am willing to cooperate fully in making that happen some how. I would hope that it can lead to everyone understanding the little bugger and to every one who wants a light pen having at least a replica that does work. I would like to have the light pen at some point, but when is not important. If you agree, you should plan on sending it directly to him now. Given that I don't know anyone here personally or even know where they are located, some details will have to be put together. It is probably best that you and Francois work out those details together. Let me know how you want to proceed.
 
My first impression is that this is a great group. I hope that I will be well received and welcome here.
 
And, if you really want to hear an old man ramble on about what he did back when and what it was like to be at HP Loveland CPD at that magic time, I can do some of that too.
 
 
Lee

From: Precaud
Sent: Monday, July 02, 2018 8:21 AM
Subject: Re: [VintHPcom] Found - 9845 Light Pen

Hi François,

Lee was the first speak so I have to honor that. Perhaps you can coordinate with him? If he approves, you could be an intermediary in the process.

  John


Precaud
 

Wow, great story, Lee. Thanks for sharing it, and for your great work. Besides being functionally appropriate, the design layout of the 9845 resonates with something in the unconscious, like a  first glimpse of an emerging archetype. Of all the computers of that era, the 9845 *looks like* what it should be. Accomplished through the marriage of imagination and solid engineering.

I live in a town (Santa Fe, NM) that, for as long as I've been here. celebrates the artist and the creative process. While that is a good thing, their definitions of "artist" and "creative process" are very narrow, confined almost exclusively to the classical visual arts; painting, drawing, sculpture, photography, etc. What consistently passes by them unappreciated is the creation of entirely new classes of objects as they emerge and find their place in our world. To me, this is where the *real* creativity exists. It's too bad our culture does not choose its heroes from this type of person.

I will arrange with François for the delivery of the pen to him.

I have been informed off-list that the light pen is a rare bird and I should price it appropriately. I would appreciate some input.

I am on a "glide-path" into retirement, and will most likely be relocating to a smaller residence at some point. My son, who one day be left with all of this stuff I've collected, has made it very clear that he doesn't want to deal with it, and I need to pare it down. And so I am using this process to compact not only my inventory, but also my electronics test bench, which is populated almost entirely by instruments made prior to 1990. While they are all functionally superb, best-in-class at the time, some of them need to be replaced with modern, more compact equivalents. That's what I'll be using the funds for.

  John


François Lanciault
 

Hi Lee,

Interesting read indeed. I am a ME that works in the antenna group of a company that makes satellite and satellite components. So I know what you are talking about! Did you designed the 9845 light pen body ?

François


Lee A. White
 

Francois
 
No, I did not work on the design of the light pen so my ability to help you is very limited. That project was done after I left HP, but it was under consideration while I was still there. As I vaguely recall, the main issue was how to find the location of the pen when the screen area under it was dark. One would think it would be an easy task to find where a lighted pixel is by measuring light output vs time at the pen tip and then relating that to the beam location, but finding the location of the pen tip when the pixels at that location are not lit up is a whole different problem. Then too, the face glass of the CRT is pretty thick and the anti glare film and the extra glass laminated panel on top of that make it even thicker. It is a fair distance from the probe tip to the phosphors and that caused concerns about parallax errors. Would the angle the probe was held at be a problem too? I seem to recall that there were also some concerns about phosphor persistence for lit up pixels as well, but it is all very very vague after all those years and I would not put much credence in my recollections.
 
I also seem to recall a discussion about having a uniform background light gray screen so that all pixels would be lit while probe location was being determined, but that might meant giving up the ability to make a truly dark area on the screen. Then there was a thought about interleaving a background screen with the image screen lighting up the entire screen after every so many scans and letting the eye integrate it out. There were also discussions about using some kind of an electromagnet field detection scheme but again, what do you do when the beam is off and there is no beam current? Some one suggested maybe making a clear digitizer to put over the screen. Early stages of projects were like that, lots of ideas were kicked around until one was chosen as worth pursuing. Then, some one would go out and try to make a proof of concept demonstrator - usually a vey crude one at that - but one that would show how it might be doable and that would point out the problems in attempting to do it. It will be interesting to see how they resolved all these issues and to see how they actually did it in the production version. 
 
I have never used one either, so any one care to comment on how well it actually worked?
 
Lee 

Sent: Wednesday, July 04, 2018 5:55 PM
Subject: Re: [VintHPcom] Found - 9845 Light Pen

Hi Lee,

Interesting read indeed. I am a ME that works in the antenna group of a company that makes satellite and satellite components. So I know what you are talking about! Did you designed the 9845 light pen body ?

François


Ansgar
 

It works pixel accurate due to the concept of moving a tracked cursor. In fact, the principle behind is not that far away to a computer mouse. With a mouse, you move a cursor with the relative movement measured against the surface of the table. With the 9845 light pen, a cursor is moved with the movement across the screen. Once you get close enough to the screen (which is about two or three inches), the beam detection locks in and the cursor has its starting position. You control the position of the cursor by the movement of the light pen, even from a distance of one or two inches from the surface of the screen, so that neither exact absolute position error nor parallax error really become an issue. The fine resolution of the detection light pen circuit is high enough to reliably control pixelwise move of the cursor. The cursor marks the position on the screen which is used as the source for gathering positioning data, not the absolute position of the tip of the light pen. The cursor at the same time works as a light source, which is available also on dark parts of the screen.

The light pen can also be used as a direct positioning device without a cursor, however the results are not as accurate and reliable as in combination with a cursor. That mode is used e.g. for picking a menu field. In fact the light pen is very sensitive and returns position data even on dark areas as long as the monitor is being switched on (there is always a signal), but it is less reliable.

-Ansgar