HP Forums

Full Version: HP 9825a
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3
It's been sitting on my desk for the last three years. Works except for the tape drive of course but I have done nothing with it.

So today while the hot water tank in the house was being replaced with an on-demand-tankless system I thought I would take it apart.

What a dream to work on. All modular and plenty of room. To date I have:

1) Removed all boards and cleaned all connections and applied contact enhancer to all plug connectors

2) reseated non-soldered ICs.

3) removed printer, tape drive and led display modules to allow access to the tape/display/printer bezel.

4) removed the tape/display/printer bezel to be shipped with the tape drive module to Larry Atherton for refurbishment to modern QIC tapes.

5) removed keyboard and then removed keyboard PCA from keyboard.

6) soaked keyboard in luke warm water and a small amount of liquid detergent to solve sticky keyboard and bouncy keys agitating also. After five thorough rinses keyboard is now being air dried via forced air. It will be stored for awhile in a drying box full of drying silica pouches. Then it will be treated with deoxit contact enhancer via the small pores on the reverse side of the keyboard. This was a must as one of the F keys was generating a looped error 30 message due to being stuck down/on.

7) the outer shell has been cleaned, third party labels removed.

8) it should look pretty new when reassembled. Of course it should work perfectly.

Have taken a series of pictures of the diassembly and will document them later.

This is the chicklet keyboard machine with 15036 bytes of RAM. I have the string, advance programming and IO modules.

Have a question though. The keyboard PCA is attached to the keyboard and keys by a series of tiny screws on the reverse of th keyboard PCA. After removing these I thought I would have access to the keys and contacts. But no, the PCA seems to be glued or fixed permanently to the keyboard assembly. Has anyone taken the chicklet keyboard apart for servicing?

Will be in London at HPPC this Saturday so I will be able to talk to Tony Duell in person about it.

Cheers, Geoff
Well here are the diassembly pictures:

The top is off, only two screws!

[Image: IMG_4443_zps3hnuacyb.jpg]

The components removed:

[Image: IMG_4451_zpsoobty5cd.jpg]

The tape drive which is going south for updating (modifying) to new QIC tapes:

[Image: IMG_4453_zpsgipaoo8b.jpg]

The printer which is fully functional but will be disassembled, cleaned and lubricated:

[Image: IMG_4452_zpsr5uzg3b8.jpg]

The display (alpha numeric LED dot matrix):

[Image: IMG_4459_zpsxrdbmpoj.jpg]

The memory boards:

[Image: IMG_4442_zpsyhwh4uc6.jpg]

To date I have cleaned:

- the outer shell.

- all connectors.

- removed third party labels.

- dissembled most of the printer and reassembled.

Then I looked at the keyboard:

In situ:

[Image: IMG_4450_zpsji9kfidt.jpg]


[Image: IMG_4461_zpsejhno2g1.jpg]


[Image: IMG_4462_zpschrsolgt.jpg]

which had a few sticky keys (the old coffee, two cream and sugar sticky: no, not by me!). This is the low profile 'chicklet' keyboard. In fact on occasion at start up I got "error 30" in a repeated beeping loop; which indicated an undefined function key. This was pinpointed to the f2 key being stuck down and when wriggled was okay. Now the first generation were all screws; in figure three the reverse side of the keyboard the solder dots (heatstakes) were ALL screws. Of course that makes diassembly a dream if not tedious. Unfortunately, I have the 11 screw and all remaining are solder heatstakes. So a very tedious desolder for diassembly.

After referring to two experts, Tony Duell in London (thanks Tony) and Larry Atherton (tape drive and all around 98xx expert in California, thanks Larry) the advise was to desolder to diassembly the keyboard sandwich. So I disregarded these two advises with the following caveat: if what I did did not work then a full diassembly would follow: so be advised, my option is not the advised method.

Now I have done this with separate PC keyboards and many heat staked HP calculator keyboards. I washed the keyboard in warm soapy water letting it soak and agitating the keys with a large paint brush. Both sides. This was followed by a luke warm rinse (many, many times). A full drain of the keyboard after each rinse. I let the water run over the keyboard for a length of time to remove any soapy residue. The reverse of the keyboard contains pores under the keys and extra pores allowing for water to drain.

I placed the keyboard on edge and allowed a fan to blow over the obverse for a day then the reverse. Occasionally using a hair dryer on medium to augment the flow. I placed the keyboard in a large box packed with 4 by 4 inch cloth silica drying pouches and covered it with the pouches. I then left for a four day trip.

Upon return the keys all clicked perfectly, no sticking. I introduced deoxit gold into each pore thereby coating the key contacts with contact enhancer and contact corrosion inhibitor. I have subsequently reassembled the computer minus the tape drive and now have a clean, perfectly functioning keyboard saving the dreaded desolder step. With the tape drive removed the start up error 43 occurs: "tape transport failure", which makes sense and can be interpreted as "where is the tape drive?". Proving the keyboard is functional it was removed and it is back in the drying box until the tape drive is updated.

This machine came with the systems programming, string advanced programming and general IO and extended IO ROMs plus the 15kb memory option. It was fully functional except for the taped drive.

The LED display is wonderful and the HPL language is intuitive to programmers. Can't put it in the flight bag though!

Cheers, Geoff

You might remember this picture taken a 2008, same computer, just sitting waiting for attention:

[Image: office3.jpg]
I also like the modularity and cleanness of design of HP machines of the time. I used to have a HP-9820, it went to some other forum member when I moved "out" of university. It was still working after 40 years Smile. Gotta love these machines !
(03-15-2017 05:49 PM)Geoff Quickfall Wrote: [ -> ]The LED display is wonderful and the HPL language is intuitive to programmers. Can't put it in the flight bag though!

If you could, would you be able to use it? Would you have a suitable power outlet?

I remember seeing a picture in a sailing magazine, late '70s or early '80s, showing the navigator's station in some high-end racing yacht. Amidst the expected equipment on such a boat -- shortwave, VHF, depth sounder, compass, radar, Loran, etc., there was an HP-9825. I was floored; who would have an 110 or 220V outlet to power such a machine on a sailboat? I guess I wasn't thinking big enough. :-)
(03-20-2017 01:57 AM)Thomas Okken Wrote: [ -> ]If you could, would you be able to use it? Would you have a suitable power outlet?

Just plug an automotive power inverter into the cigarette lighter on the B777's flight deck. It's right next to the 8-track player. Smile
Another chapter in the book?

- Pauli

Yes, what a dream to work on! Things to do, upgrade memory (15k now). Maybe 24k would be nice as the 32k option interferes with rom accessories.


-the B737-200 had the eight track quadrophonic system.

-the B767 had a cassette deck option.

-the b777 is CD audio stereophonic.

-the b787 is thumbdrive.

I have a Bose A20 Bluetooth so music for the long Pacific hauls where we have a lot of radio silence, data received via satellite. That way best of the Rolling Stones on the way down and the best of CCR on the way home :-)

Thomas and Dave:

We have three 110-120 volt 60hz outlets on the flight deck, usually used to charge the IPad that contain our manuals, flight planning, weather and with wifi these have dynamic real time update capability. The inverter is built into the aircraft so yes, useable on the flight deck. The largest system I have had on the flight deck is the compucorp 326 with tape drive although I did transport an HP logic analyzer to London for Tony Duell.


Thought about that Paul, but Larry Atherton has done sooooo much work, documentation and etc. He deserves first rights to a book/manual. Maybe an appendix.
(03-20-2017 06:15 PM)Geoff Quickfall Wrote: [ -> ]-the B737-200 had the eight track quadrophonic system.

-the B767 had a cassette deck option.

-the b777 is CD audio stereophonic.

-the b787 is thumbdrive.

The Phoenix, the warp drive ship in First Contact, had some sort of mini-disc player, so you might have to upgrade your music collection down the road.

Geoff </George>,

As one of the original 9825 team, it warms my heart to see you spending so much time on the successful restoration of a 9825a. Thanks for the photos and the explanation. That a 40-year-old machine continues to work is a real testament to the people who brought that computer into being (not me, I'm the I/O card and Systems Programming ROM guy).

Hello Steve!

George = Geoff. :-)

Thanks to people like you I get to restore the 9825a! I have had this sitting on my office table for a few years. Then, at HHC2016 Don Morris gave a prize winning talk on the HP 35 to HP 9825.

That inspired me to take it apart, clean and repair. It programs beautifully and can't wait to reinstall the tape drive after Larry modifies it.

Cheers, Geoff
Just uploaded a keyboard test video showing the response the keyboard has since the "wash". This was not just a wash as one would apply to the standard detached keyboards. The caveat here is that the keyboard is the sandwich type similar to many HP keyboards composing of:

  1. Keyboard bezel
  2. Keys
  3. Clear plastic film
  4. Insulator
  5. Contact components
  6. Keyboard PCA

The keyboard is a sandwich which has about 100 soldered heat stakes. If the wash did not work I was gong to remove the stakes and deal with the assembled keyboard. If the wash worked, well, excellent. It did work as seen in the video. Prior to the wash I got "error 30" and repeating beeps indicating a stuck function key at start up. The beeping error message looped even after a reset and, as well some keys bounced.

  1. Firstly I soaked the keyboard in like warm soapy (liquid dish detergent) water to remove the coffee and possible sugar residue.
  2. I used a long fine bristle brush and agitated the keys from the top (obverse) of the keyboard removing any contamination from the top of the clear plastic film.
  3. I repeated the agitation to the underside (reverse) working the bristles into the 'pores' on the keyboard. I have used this technique successfully on many heat staked HP families.
  4. the soapy wash was followed by many (approx. ten) thorough rinses and immersions in warm water.
  5. the keyboard was dried overnight by leaning it gainst a support and allowing the water to 'drain'.
  6. comoressed air was blown through the keys and the reverse of the board. At the same time a hair dryer, set to warm, not hot augmented the canned air.
  7. the obverse side of the keyboard sat inclined in front of a fan for the next 2 days, the reverse side for another two days. Again augmented with warm air from the hairdryer on occasion.
  8. it was then placed in an enclosed box resting on a bed of silica dying pouches with similar pouches covering the top and left for a week. The pouches are 3 inches by 2 inches in cloth sacks and contain silica drying pellets.
  9. the board was removed and deoxit gold was introduced to each pore on the reverse, backing each key and injected into the pore via a syringe. This enhances and preserves the contacts.

The result as you can see is a fully functional, non bouncy and non sticking keyboard with a perfect feel:

The 'cleaned keyboard'

Ensure video quality is set 720.


Edited due to poor typing skills.
Well I am obviously having fun here.

Here is a simple program running on the 9825a demonstrating prompt, error test, looping. Right out of the manual, it is prompting for a value V, then testing for negativity, re-prompting for positive value or taking square root of the positive value:

9825a running simple program

I will be increasing the RAM by 8KBs and then, with the tape drive modified I will try some heavier programming and storing. More fun!
(03-22-2017 02:30 AM)Geoff Quickfall Wrote: [ -> ]Well I am obviously having fun here.

I assume, the fun will delay a certain ... book?

But earnestly: I guess that more than one feel quite jealous following this thread!


(03-22-2017 01:19 AM)Geoff Quickfall Wrote: [ -> ]The result as you can see is a fully functional, non bouncy and non sticking keyboard with a perfect feel:

I think your printer must be a fake. The video contained no sound of the annoyingly loud paper advance mechanism. :-)
(03-22-2017 11:54 PM)cruff Wrote: [ -> ]
(03-22-2017 01:19 AM)Geoff Quickfall Wrote: [ -> ]The result as you can see is a fully functional, non bouncy and non sticking keyboard with a perfect feel:

I think your printer must be a fake. The video contained no sound of the annoyingly loud paper advance mechanism. :-)

That is the fun of incremental drive, the printer in the 9825/35 is about the loudest I have ever heard, but before stepper motors became really common, incremental drive (nicknamed "incredible drive" by some of the service reps I knew), where the common way to achieve stepped motion, lots of card punches and reader used that kind of a mechanism to move the cards, but they where not made of plastic and where generally faster and quieter than the one in this printer.
Ha, the new silent version. Was testing YouTube direct from iPad video to YouTube and forgot to check 'the extremely loud printer button!

I was fortunate with this computer as all electronics and modules checked out correctly except, of course the tape drive.

Someone told me a 9835 rom drawer can plug into the side of the 9825a providing a similar operating system and 'Basic' language instead of HPL.

Anyone tested this?

That should be 9831 not 9835. The 9831 is the same hardware as a 9825A but with a different OS it is the same BASIC as the 9830 probably designed as a stop-gap BASIC machine between 9830 and 9845. The 9835 came after the 9845A and runs the same BASIC as the 9845 but does not support graphics. The processor in the 9835/45 is an enhanced version of the processor used in the 9825/31. I have a OS module for a 9831 here but I have not tried it as I have a 9825T so I would need a memory card for a 9825A and heal a couple traces on the processor board that are cut on the 9825 B&T models. I do plan to do this, it is in the queue with a bunch of other things.
Thanks for the info. So next I should look for a 9831 ROM drawer just to play around. At the moment the 9825a is quite capable and HPL is not difficult.

The HP9825 printer story is here:

Thanks Steve,

I have that pinned to my homepage on the iPad and home computer.

Larratherton has advised me that the mods to the tape drive are completed and it tests okay. Also he is making up a stat tape, another with extra programs and two blanks for me.

Will get programming when it is returned.

Pages: 1 2 3
Reference URL's