HP Forums

Full Version: HP 9830 font
You're currently viewing a stripped down version of our content. View the full version with proper formatting.

Got a TMS4100 or TMS4103 character ROM?
Me neither. If you have a working 9830A,
please validate my guesswork at

	http://jovi.net/9830A 
http://jovi.net/9830A/guess-1.html
http://jovi.net/9830A/guess-1.jpeg
thanks!
		Peace
--Devon
/~\
\ / Health Care
X not warfare
/ \

Dubya won the digital vote
Kerry won the popular vote

Devon,
take a look at my HP9800 emulator http://sourceforge.net/projects/hp9800e, which also emulates the 9830 display. The char layout is exactly identical to the real thing.
In the source file HP9830Mainframe.java you will find the bit masks for all characters.

Achim

Thank you!

The real 9830A font looks much better.

My display is totally dark now, what command could I type blind to get a beep or other sign of life?

		Peace
--Devon
/~\
\ / Health Care
X not warfare
/ \

Dubya won the digital vote
Kerry won the popular vote

PS: First it lost an input bit (numeral 2 looked like 0, etc.) then it lost an output pixel (a black row across the entire display so numeral 4 looked like 1 with a longer barb and a broken stem, etc.) and now it is entirely black. Perhaps not display related this time as typing B E E P <EXECUTE> is silent but again, thirty years later I don't remember clearly if that should beep.

I believe the 'BEEP' command is in the Advanced Programming ROM. If you don't have that, it won't beep.

Do you have a 9866 printer? If so, try a command to make that print something.

Unfortunately, a totally blank display can be due to a fault almost anywhere in the machine. The display is scanned by the firmware, so for it to work the processor, memory control, ROM, RAM, PSU, display sections must all be working. Even a problem with the keyboard, cassette interface or I/O backplane can corrupt the I/O bus and blank the display.

What electronic test gear (and experience) do you have? I can talk you through the logic and suggest tests (and even show you how to trace the CPU microcode using a logic analyser), but I will warn you, these machines are not simple.

I removed the three tape control cards after my 9830 went comatose, I *think* I tried B E E P <EXECUTE> before opening it, no sound, only signs of life are the fan, power light and tape sensor light. I finally uncovered test points hidden under a lever atop the power PCB, all good.

I borrowed a dual trace scope, as expected no signal on any TMS4100 output, pin 15 Vgg -12V good, pin 17 Vss +12V unreachable without a 28-pin DIP clip but good at the big decoupling cap as is +5V, disinclined to solder test leads to PMOS.

Could be CPU boot failure, I/O wedged, hmmm...
alarming burning smell turns out to be the scope,
sure hope I can return it in working condition.

		Peace
--Devon
/~\
\ / Health Care
X not warfare
/ \

Dubya won the digital vote
Kerry won the popular vote

PS: TMS4103JC DIP-28 presumably identical to TMS4100

           +-----__-----+
o1 | 1 28 | i7 or EN or GND ???
- | 2 27 | i1 or b6
o2 | 3 26 | i2 or b5 <-BAD0
- | 4 25 | i3 or b4
BAD1<- o3 | 5 24 | i4 or b3
- | 6 23 | i5 or b2 <-BAD0
o4 | 7 22 | ce
- | 8 21 | cd
BAD1<- o5 | 9 20 | cc
- | 10 19 | cb
o6 | 11 18 | ca
- | 12 17 | Vss +12V
o7 | 13 16 | i6 or b1
GND | 14 15 | Vgg -12V
+------------+

BAD0 one input bit stuck on zero as delivered
(dealer claimed "excellent working condition"
but knew it needed repair, proof evident
in eBay photo once you know what to look for)
BAD1 one output row stuck on black the day after
BAD2 9830A comatose but for fan the week after

i6-1 ASCII in according to one source
b1-6 ASCII in according to another source
ca-ce one-of-five column select
o1-7 column pixels out

PPS: My HP 9830A has a working 9866A thermal printer and

  • (far back)
    fan, AC power, PCB with printer header & I/O slots

  • (back: back-to-front)
    GyGn 09830-66551 REV A power supply test points hidden under levers
    YGy 11275-66584* 8809L REV B ?RAM?
    YGy 09830-66584* 5020-8304 REVA Jul 10 1973 ?RAM?
    OGy 09830-66583 REV-A 44-pin 5/32" top edge connector
    RGy ? mysterious marks like AEN 44-pin 5/32" top edge connector
    GyR REVA ?ROM? 7x4?xDIP-24 marked 1818

  • (middle right: back-to-front)
    YBn 09810-66514 REV A ?CPU?
    OBn 09810-66513 REVA
    -- -- 12-pin edge socket
    RBn 09810-66512 REV C 09810-66512*
    BnBn 09810-66511 REV B

  • (middle left: back-to-front)
    GnR 09830-66525 Rev B ?internal ROM?
    BuR 09830-66526* Rev A
    BkR 5020-6884 Rev A ?option 271 perhaps?
    BkR --empty--
    BkR --empty--

  • (front left door stickers)
    OPTION 271 PLOTTER CONTROL
    5856 WORDS READ-WRITE-MEMORY

  • (front left: top-to-bottom)
    --empty--
    11279B ADVANCED PROG I
    11274B STRING VARIABLES
    11283B PRINTER CONTROL
    --empty--

  • (front right)
    tape cassette deck

  • (front middle: back-to-front)
    BnBu ? speaker
    RBu 09865-66562 REV B Jan ?2?6 1973
    OBu 09865-66563 REV-C read/write bd. jan 29 1973?
    YBu 09865-66564 REV B motor control bd. jan 31 1973?
    RY 09830-66542 11-3-?72? display logic board with ?losing? TMS4100 character generator ROM

  • (far front)
    ? ? display board with 8x4x5x7 LED
    ? ? keyboard and AC power switch

Bk=Black Bn=Brown R=Red O=Orange Y=Yellow Gn=Green Bu=Blue V=Violet Gy=Gray W=White
(no asterisk) metal mask
* stickered
** rubber stamped
*** scratched

You can make the 9830 beeping by forcing an error condition. Try to enter blind: 1 / 0 EXECUTE

Achim

Yes, the TMS4100 and TM4103 have the same pinouts (I worked from the 'wrong' data sheet when I produced schematics for this machine, they're available from the Australian Site (one page missing?) or on the HPCC CD-ROM).

The boards in the 9830 are :

At the very back, the I/O backplane with the 9866 interface circuit (a simple-ish parallel port) on it.

09830-66551 PSU

11275-66584 expansion RAM (loads of 1103 DRAMs)

09830-66584 standard RAM (ditto, they're actually much the same PCB)

09830-66583 Memory Data (T register, etc)

09830-66582 Memory Address (M register, address decoder)

(Can't remember, but full width in front of those) : Main ROM

09810-66514 CPU Data Path (registers, ALU, etc)

09810-66513 CPU Control (Microcode ROMs, etc)

09810-66512 CPU Clock (and I/O state machine)

09810-66511 CPU I/O Interface (I/O register)

09830-66525 Extension ROM selector (address decoder)

09830-66526 Extra ROM (that wouldn't physically fit on the main ROM board!)

In front of that are 3 slots for internal ROM options, like your plotter ROM.

09830-66561 Tape Interface (and beeper, etc, that's the one with the speaker on it)

09865-66562 Tape Controller (shift register, etc)

09865-66563 Read/Write (for the tape drive)

09865-66564 Motor Contol

09830-66542 Display Driver (that's where this character generator ROM is)

09830-66541 Display

The keyboard is a total of 4 PCBs, all with 09830-6653x numbers. The keyboard itself, the encoder board that screws to the bottom of it, the interconnect board that plugs into those 2, and the paddleboard wired to it that links to the main backplane.

A bit of troubleshooting info. First of all, remove the following from the machine -- it'll give a display without it :

The I/O backplane (leave the 12 pin edge socket on top 'floating', it just wires to the 9866 connector)

All 4 tape boards (including the one with the speaker on it).

The internal extension ROMs (your Opt 271 board)

The ROM cartridge cage and all cartridges (4 screws and pull it out)

The keyboard paddleboard (but leave the power switch harness connected, obviously)

My guess is you'll still not get a display, but...

Pull the expansion RAM PCB. Try the machine with just one RAM board in the front of the 2 RAM slots -- in fact try each of your 2 RAM boards there. Address decoding is on the Memory Address board, so there are no links to set or anything.

At this point, you've got 'fun'. It's a bit serial machine, whcih means there are few repetitive signals to look at. I think you will need schematics to go much further.