In this wiki you can store howto's what to know etc..
About using and preserving vintage HP computers.
|
Re: HP 9821 - display and tape transport issues
#VintageHPComputers
The 'die without notice' applies to all TTL chips made by NI in the early 1970s. I had to replace dozens of ICs in my various 98xx calculators and peripherals and in 99% they where made by National Instruments. If I get a defective 98xx board I first check all NI chips by means of a logic comparator. NI seemed to have problems with long term stability of their TTL ICs in the 1970s. The dying already began in the 1990s and still continues. Some times I thought about exchanging all NI chips prophylactic, but that would cause a lot of work in my ~20 devices. Moreover, some chips are hard to find these days.
|
|
|
|
Re: HP 9821 - display and tape transport issues
#VintageHPComputers
A lot of transport problems are caused by a bad working EOT sensor, the LDR’s used in the system to detect the end of tape deteriorate and the resistance changes permanent to a low value. The system then won’t detect the end of tape or thinks end of tape has arrived when it hasn’t, with all kind of nice tape behavior as result. The LDR and Lamp are place left from the tape head and accessible when you remove the tape door, you can check the working of the LDR with a multimeter the resistance should vary between 100-1K and >1M roughly (light – no light condition). Other problems with the tape boards are the 7403 (1820-0269) made by NI they die without notice 😉 They’re used in several boards but the most important is the control board. Another problem early multi-layer boards suffer from via contact problems causing intermittent or permanent errors, boards I know off are mostly memory and memory control boards (M-register, T-register, display) dual-layer boards typically don’t suffer from this only multi-layer boards.
Hope it helps a little, -Rik
Van: VintHPcom@groups.io <VintHPcom@groups.io> Namens mariuszmleczko@...
9821A finally opened
|
|
|
|
Re: Error 128 - Line too long during get
Anders,
I've attached a zip file containing a HP9122 LIF Disk image file that contains the following: hp9122lif_YATZ9816.bin VOLUME LABEL: B9836 FILE NAME PRO TYPE REC/FILE BYTE/REC ADDRESS DATE TIME YATZ9816 PROG 82 256 16 2-Feb-20 18:15 YATZ9816a ASCII 82 256 98 2-Feb-20 18:16 584704 of 626688 bytes free. The "YATZ9816" file is a LOAD'able version of the program. The "YATZ9816a" file is a GET'able ASCII version of the program. Let me know if you need something else. Enjoy,
|
|
|
|
Re: HP 9821 - display and tape transport issues
#VintageHPComputers
mariuszmleczko@...
9821A finally opened
Here you can see photos of all the boards. Some of them are not the best quality (sorry). https://groups.io/g/VintHPcom/album?id=240360 So far I've repaired the lever for paper feed. Next thing to do is a tape transport. O-rings looks OK without any dents, cassette pulling strength and speed seems to be OK as well. But it not behaves properly.... 1) Rewinding.: - when I execute "<>REW" command, it is rewindig but only for about one second and then stop immediately. Here you can see video (few times executed <>REW command on execute keypress). https://groups.io/g/VintHPcom/files/HP%209821A/rewinding.mp4 2) Program load "<> LDF 0".: - after few tape turns when clear leader is finished it starts to behave strangely. electromagnets on the both sides flip back and forward, then tape finally is rewinded back to begginnig. This test was made on origin HP cassete which is OK in 99%. https://groups.io/g/VintHPcom/files/HP%209821A/program%20load.mp4 3) Mark command: - when I run "<>MARK 1, 5"; command it's running and running....don't stop. Do you have any idea what shall I check first ?
|
|
|
|
Re: Error 128 - Line too long during get
The Ascii format seems to be more complicated than that. A line like "10 PRINT PI*2" gets an FF tagged to the end, but not a line that ends with a quote, ie "20 PRINT "RMB" "
|
|
|
|
Re: Error 128 - Line too long during get
Can you pretty please post the LIF file you have?
I also found out that you need a FF at the end of every line when converting. It works fine for other programs now, but not for that YATZY thing :(
|
|
|
|
Re: Error 128 - Line too long during get
I haven’t a problem reading it, I used the .td0 file to recreate the lif disc. And used my viper card (82324B) to load the file in memory and create the .txt file. I’ve run the program on the viper and it runs fine.
-Rik
Van: VintHPcom@groups.io <VintHPcom@groups.io> Namens Anders
Yes, but all lines are shorter than that. Im a stumped.
|
|
|
|
Re: Error 128 - Line too long during get
Yes, but all lines are shorter than that. Im a stumped.
|
|
|
|
Re: Error 128 - Line too long during get
Paul Berger
The 86/87 Operating and Basic Programming Manual, available at hpmuseum.net, states the maximium line length you can enter from the keyboard is 159 characters. I would expect that this would apply to lines read from a file as well. Paul.
On 2020-02-02 10:00 a.m., Anders wrote:
Yes, problem seems to be that it barfs on line 40 with line too long. Does the 98x6 have a maximum line length when reading ascii program files? Also, there needs to be an FF at the end.
|
|
|
|
Re: Error 128 - Line too long during get
Yes, problem seems to be that it barfs on line 40 with line too long. Does the 98x6 have a maximum line length when reading ascii program files? Also, there needs to be an FF at the end.
|
|
|
|
Re: Error 128 - Line too long during get
The ascii file yatz.txt
-Rik
Van: VintHPcom@groups.io <VintHPcom@groups.io> Namens Anders
Here is the file anyway!
|
|
|
|
Re: Error 128 - Line too long during get
Here is the file anyway!
|
|
|
|
Re: Error 128 - Line too long during get
// a2rmb.cpp : Defines the entry point for the console application. // #include "stdafx.h" #include "stdio.h" #include "string.h" FILE *infile; FILE *outfile; char buffer[1000]; char temp[10]; int rc; int length; unsigned char hi, lo; int _tmain(int argc, char* argv[]) { rc= fopen_s(&infile,argv[1],"r"); if(rc != 0) { printf("Invalid infile name %s", argv[1]); return 0; } rc = fopen_s(&outfile,argv[2],"wc"); if(rc != 0) { printf("Invalid outfile name %s", argv[2]); return 0; } // memset(buffer,0,500); while(fgets(buffer,500,infile)!=NULL) { //fprintf(outfile,"%s",buffer); length = strlen(buffer); if(buffer[length-1] == 0x0d) { buffer[length-1] = 0; --length; } if(buffer[length-1] == 0x0a) { buffer[length-1] = 0; --length; } hi = length/256; lo = length % 256; fwrite(&hi,1,1,outfile); fwrite(&lo,1,1,outfile); fwrite(buffer,1,length, outfile); } fclose(infile); fclose(outfile); return 0; }
No idea why this will nt post correctly though..
|
|
|
|
Re: Error 128 - Line too long during get
Yes, the other solution is using a Viper card with a 9121 connected. Read the file and use SAVE ”xxxx” to write the ascii version to the dos disk.. Or install HPBASIC for Windows -Rik
Van: VintHPcom@groups.io <VintHPcom@groups.io> Namens Anders
That is the probem. That utility does not work on more recent PCs, so I use, libdisk
|
|
|
|
Re: Error 128 - Line too long during get
That is the probem. That utility does not work on more recent PCs, so I use, libdisk
libdsk1512\dskconv.exe" -otype raw BAS2SYS.TD0 BAS2SYS.BIN This has woerked well for most disks, but fails on this one. No idea why? OTOH, writing a small C program that takes an Ascii file and produces one with Pascal style strings is trivial. I will see if I can do that instead.
|
|
|
|
Re: Error 128 - Line too long during get
Anders,
Convert the td0 to hpi and your problem should be solved. https://hp9845.net/9845/downloads/software/Teledisk-to-hpi-utility.zip Or recreate the disk from the td0 image with teledisk on a DOS pc.
-Rik
Van: VintHPcom@groups.io <VintHPcom@groups.io> Namens Anders
This is what I get when I try loading YATZY from an image located at: https://retrocalculator.com/software/#YATZ9816. There are two files in the image, YATZ9816 which is a PROG file and YATZA which is an ASCII file, but when I look at the actual disk image they both look like tokenised BASIC.
|
|
|
|
Error 128 - Line too long during get
This is what I get when I try loading YATZY from an image located at: https://retrocalculator.com/software/#YATZ9816. There are two files in the image, YATZ9816 which is a PROG file and YATZA which is an ASCII file, but when I look at the actual disk image they both look like tokenised BASIC.
If I extract the YATZ_9816 file from the tarball and put it in the image, using HPDir, then it refuses to load. It is a straught ASCII-file with 0A as delimiter. When I look at other files I have saved as ASCII, they appear to be stored as BASIC (or Pascal) strings, ie an integer for the length and then the actual characters. How do I go about getting a program file saved as ASCII with UNIX line endings onto a LIF image so that I can GET it on the 9816? I tried converting the TD0 file to a straight image, but I cannot load from that either.
|
|
|
|
Re: Troubleshooting a 9845C
Dave,
If the mainframe is the problem, it’s most likely caused by bad ROM’s and then you have to replace them by a ROM board. https://hp9845.net/9845/projects/romboard/
-Rik
Van: VintHPcom@groups.io <VintHPcom@groups.io> Namens Rik Bos
Dave,
The 9845C uses the same graphic subsystem as the 9845B option 200 so it is possible to use a 98780A monitor instead of the color 98770A monitor. The color monitor is a very complex peace of work with 3 independent power supplies build in. Because of its complexity and power hunger it has 2 fans one on the outside one on the inside. It is also very sensitive for overheating, if the power transistors won’t get enough cooling they will be destroyed. Most of the time this will result in a short and destroy the driver transistors also. First thing to do is building a turnon fixture to start the 9845C without a monitor to narrow down the problem. https://hp9845.net/9845/tutorials/testrom/images/turn-on-fixture-schematics.png https://hp9845.net/9845/tutorials/testrom/ If the system doesn’t beep it’s the monitor.
BE VERY CAREFULL LETHAL VOLTAGES EVEN WHEN THE UNIT HAS BEEN SWITCHED OFF! The big capacitors keep the power for some time, it takes them a few minutes to discharge fully.
Monitor fault finding, first download the service manual for the 98770A monitor and the diagrams. (I’ve printed the diagrams on A3 that works for me) Next remove the outer casing of the monitor, place the monitor on the 9845C and connect both to the main power. Switch on the 9845C and look at the power leds on the left side of the monitor, just only the rightest LED should light, all other should be off. If more led’s are lighting, remove main power, take a small break go to the toilet etc.. Remove the cover of the power and remove the power supply board, check the power transistors (separate board), rectifier board and capacitor board for the obvious. You should probably replace all the ELCO’s for new ones, after 30-40 years they’re worn out and developing high ESR. (Maybe do this first before initial power on.) When you replace a power transistor also check the drivers on the main board, If the power transistors keep dying probably one of the small driver transformers has gone bad.
Hope it helps -Rik
Van: VintHPcom@groups.io <VintHPcom@groups.io> Namens Dave via Groups.Io
I recently, after a long search and much patience, managed to acquire an option 200 power supply for my option 200 9845C. I finally had a chance to power it up, after verifying voltages on the supply.
|
|
|
|
Re: Troubleshooting a 9845C
Dave,
The 9845C uses the same graphic subsystem as the 9845B option 200 so it is possible to use a 98780A monitor instead of the color 98770A monitor. The color monitor is a very complex peace of work with 3 independent power supplies build in. Because of its complexity and power hunger it has 2 fans one on the outside one on the inside. It is also very sensitive for overheating, if the power transistors won’t get enough cooling they will be destroyed. Most of the time this will result in a short and destroy the driver transistors also. First thing to do is building a turnon fixture to start the 9845C without a monitor to narrow down the problem. https://hp9845.net/9845/tutorials/testrom/images/turn-on-fixture-schematics.png https://hp9845.net/9845/tutorials/testrom/ If the system doesn’t beep it’s the monitor.
BE VERY CAREFULL LETHAL VOLTAGES EVEN WHEN THE UNIT HAS BEEN SWITCHED OFF! The big capacitors keep the power for some time, it takes them a few minutes to discharge fully.
Monitor fault finding, first download the service manual for the 98770A monitor and the diagrams. (I’ve printed the diagrams on A3 that works for me) Next remove the outer casing of the monitor, place the monitor on the 9845C and connect both to the main power. Switch on the 9845C and look at the power leds on the left side of the monitor, just only the rightest LED should light, all other should be off. If more led’s are lighting, remove main power, take a small break go to the toilet etc.. Remove the cover of the power and remove the power supply board, check the power transistors (separate board), rectifier board and capacitor board for the obvious. You should probably replace all the ELCO’s for new ones, after 30-40 years they’re worn out and developing high ESR. (Maybe do this first before initial power on.) When you replace a power transistor also check the drivers on the main board, If the power transistors keep dying probably one of the small driver transformers has gone bad.
Hope it helps -Rik
Van: VintHPcom@groups.io <VintHPcom@groups.io> Namens Dave via Groups.Io
I recently, after a long search and much patience, managed to acquire an option 200 power supply for my option 200 9845C. I finally had a chance to power it up, after verifying voltages on the supply.
|
|
|
|
Re: Troubleshooting a 9845C
François Lanciault
Hi Dave,
toggle quoted messageShow quoted text
What is the health of the color monitor power supply ? On the 9845C, the PSU inside the monitor is usually the "hard to solve" problem. You should check the 98770 service manual and the 98770 schematics for more information and for how to check if the PSU is working correctly. At least, there are no RIFA capacitors on the 98770 monitor PSU (at least none on mine). If you know for a fact that your monitor PSU is ok then you could try to swap boards if you have another working 9845. If not then maybe the internal ROMs have gone bad and should be replaced. Good luck with this project ! At least you have a chance to make it work. I am missing to many parts in my 98770 monitor to even start thinking about fixing it. :-( François
On Jan 31, 2020, at 17:12, Dave via Groups.Io <dfnr2@...> wrote:
|
|
|