PDA

View Full Version : Compaq Portable power on Issues



cbolvin
August 31st, 2014, 02:42 PM
About time I joined . I've been "lurking" off and on for a few years.

Anyway, One of my two Compaq portables decided it would refuse to boot properly. It's an early dual diskette version (No hard drives). It had been working up until this week when I was using it to compare with a recent acquisition of a Portable Plus and to start on restoring the keyboard (disintegrated keypads on both units). I added the Portable Plus to my existing collection that includes the Portable I (Dual diskette) a Portable II and two Portable III's (all working) and a fully restored dual diskette NorthStar Horizon with Televideo 910 terminal and my Hercules system with working vintage 3174, 3278 and 3287 gear.

When I turned the power on, I got a fan and nothing more, no screen, nuttin. Also the system light on the main board didn't light. So I'm thinking a bad board has affected the PS voltages. So I started by pulling the generic memory expansion/Clock/IO board and tried again.

This time I get the system light but the unit sounds a constant low tone for a minute or two followed by a low motor-boating sound followed by a constant high tone. The Drive 0 LED also comes on. All this with no screen or cursor. None of the diagnostic manuals seem to address this so I thought I'd throw it out for the group. Same thing when I pull the remaining boards (diskette controller and Video board). Not sure if I'm looking at a power supply issue here or not since the tones and diskette seem to indicate otherwise.

Any clues here?

Thanks!

Chuck

tkc8800
August 31st, 2014, 04:06 PM
I recently got hold of a Compaq Portable I and had very similar issues with the screen not coming on. You would power it up and the fan could be heard but no cursor, it would also activate the A drive in the normal way. I noticed that it happened intermittently, especially after a warm re-boot. This lead me to check the connectors between the CRT and analog video board. I suspect on my machine, it is a bad solder joint on one of the connectors because after pressing down on all the connectors the problem has gone for now. I may pull the video board if it happens again to check the solder joints. This is a common problem with CRT video boards, I have two other systems that required re-soldering of the connectors on the video board.

cbolvin
August 31st, 2014, 04:46 PM
Thanks for the tip! Sounds like a good idea. I'll give it a go tomorrow. Hopefully someone will have some insight on the remaining problem.

Thanks!

cbolvin
September 2nd, 2014, 03:09 PM
Well, some progress!

The screen is working. It turns out the brightness needed to be turned up all the way. Looks like the master brightness needs adjusting. And I did find a video pin with a cracked solder joint so I re-flowed all of them (thanks again for the tip!).

The POST codes are coming up 101 and 301. I already know the unit has keyboard issues but the 101 indicates an I/O ROM error. I'm not sure if the ROM is actually bad or what. I do get a prompt to insert a diskette so I don't think the entire system board is bad.

Could this be just a failed EPROM I wonder?

Chuck

modem7
September 2nd, 2014, 11:07 PM
but the 101 indicates an I/O ROM error.
According to tables 5-1 and 5-2 in chapter 5 of the manual at [here (http://www.minuszerodegrees.net/manuals/Compaq%20Portable_Plus_286%20-%20Maintenance%20and%20Service%20Guide.pdf)], a 101 has a slightly different meaning amongst the variations of Compaq Portable.

tkc8800
September 3rd, 2014, 01:44 AM
Well, some progress!

The screen is working. It turns out the brightness needed to be turned up all the way. Looks like the master brightness needs adjusting. And I did find a video pin with a cracked solder joint so I re-flowed all of them (thanks again for the tip!).

The POST codes are coming up 101 and 301. I already know the unit has keyboard issues but the 101 indicates an I/O ROM error. I'm not sure if the ROM is actually bad or what. I do get a prompt to insert a diskette so I don't think the entire system board is bad.

Could this be just a failed EPROM I wonder?

Chuck

Good to hear the screen is working. Table 5-2 of the manual states "stuck key" as a 3xx error code. This is highly likely, I did a good clean out of the degraded foam in the keypads, then was able to test the keyboard using the bare pcd and a test key.

cbolvin
September 3rd, 2014, 05:04 AM
According to tables 5-1 and 5-2 in chapter 5 of the manual at [here (http://www.minuszerodegrees.net/manuals/Compaq%20Portable_Plus_286%20-%20Maintenance%20and%20Service%20Guide.pdf)], a 101 has a slightly different meaning amongst the variations of Compaq Portable.

In my case it's actually probably referring to the System ROM since I don't have an Option ROM installed.

cbolvin
September 3rd, 2014, 05:10 AM
Good to hear the screen is working. Table 5-2 of the manual states "stuck key" as a 3xx error code. This is highly likely, I did a good clean out of the degraded foam in the keypads, then was able to test the keyboard using the bare pcd and a test key.

I have the keyboard already disassembled in preparation for rebuilding both my units and ran the last test with just the bare board connected and still got the 301. I'm going to try and swap the other keyboard as a quick test since most of those keys still work. However, I'm going to ignore the 301 for the time being and concentrating on the 101. This keyboard may be legitimately bad.

cbolvin
September 3rd, 2014, 09:27 AM
I pulled the EPROM and popped into my programmer and was able to read it but the verify failed. That would explain the 101 code. I'll need to replace the EPROM. I don't have a fresh copy of the BIOS or a way to erase the EPROM. I'm going to pull the EPROM from the Portable Plus and burn a new EEPROM and try it.

cbolvin
September 5th, 2014, 10:50 AM
This is interesting. I pulled the BIOS rom from the other portable and it appears to be a mask rom, not an EPROM and not one that I can find a data sheet on.

AMD AM9265DPC. Before I can read it I need to what settings to use on my programmer.

3pcedev
September 5th, 2014, 02:06 PM
http://bitsavers.informatik.uni-stuttgart.de/pdf/umc/UMC_Data_Book_1986.pdf

Have a look at that UMC IC handbook from 1986 in the link above. If you look at the cross reference table it shows that the AM9265 is pin compatible with the UM2366A. If you then search in the PDF for the UM2366A you will find the datasheet on page 23 of the PDF. Seems like a very standard ROM; your programmer should be able to read it.

Good luck!

cbolvin
September 5th, 2014, 02:29 PM
http://bitsavers.informatik.uni-stuttgart.de/pdf/umc/UMC_Data_Book_1986.pdf

Have a look at that UMC IC handbook from 1986 in the link above. If you look at the cross reference table it shows that the AM9265 is pin compatible with the UM2366A. If you then search in the PDF for the UM2366A you will find the datasheet on page 23 of the PDF. Seems like a very standard ROM; your programmer should be able to read it.

Good luck!

Out STANDING! I didn't think to look at bitsavers! The sheet says the UM2366A is pin compatible with the 2764 which should work. Now to give it a go.

Thanks for the assist!

Chuck

cbolvin
September 10th, 2014, 02:23 PM
Well, it's not the ROM. I trieda fresh ROM with two different BIOS versions and get the same behavior. I tested the old ROM and was able to get a good read and checksum so I'm thinking it's been fine all along. Turns out I used an incorrect procedure on the programmer. What I did notice is the beep sounds are the same even when the ROM socket is empty.

So next step was to verify voltages. The voltages are all there but not the most robust I've seen. +5 is actually +4.4, -5 is -4.4, +12 is more like +11 and -12 is -11. The same regardless if the system board is connected or not.

My other Portable I voltages are better. +5 is about +4.8 and +12 is closer to +11.5. Better but not dramatically better. This is with System Board disconnected. That unit powers up and runs fine. It's a Plus so it also has the 20MB hard drive.

The voltages on the problem unit are out of tolerance (5%) according to the maintenance and service guide, but not dramatically.

Anyone know how picky these system boards are?

Chuck

cbolvin
September 18th, 2014, 02:04 PM
Doesn't appear to be power supply either. I connected the bad system board to the known good power supply on the other Portable and same symptoms. I tried a POST card and evidently the Compaq doesn't display post codes on the bus. It however shows that one particular card, the no name clock/ram multifunction card causes the +5 line to shut down.

So I started looking at possible system board shorts. I thought I may have had a shorted tantalum and unsoldered one leg of all the tantalums and tested again. Same problem. I tested each of the tantalums and they seem to be ok since they do react like an electrolytic should when you use an ohmmeter on them.

I'm running out of things to check for since there doesn't seem to be a troubleshooting procedure that addresses this problem. I may end up just using the unit for parts at this point. Crap...

H-A-L-9000
September 19th, 2014, 10:20 AM
- Voltages too low: Are you sure it's not your multimeter?
- beep sounds? Could it be that there are multiple ROMs in the computer and you only found one? Are these from the speaker?

Unfortunately it seems not so easy to find a photo of the mainboard.

cbolvin
September 19th, 2014, 01:35 PM
- Voltages too low: Are you sure it's not your multimeter?
- beep sounds? Could it be that there are multiple ROMs in the computer and you only found one? Are these from the speaker?

Unfortunately it seems not so easy to find a photo of the mainboard.

This is a Version 1 board with only one ROM which I've already replaced as part of the testing. If you look at my initial posts, I referenced the sounds and the POST codes that show up if the display board is plugged in. 101 and 301 (System ROM and keyboard). I've fixed the keyboard but still encountering the ROM issue. I'm beginning to think there's a blown IC but at this point I'm not sure where to start without schematics or a procedure.

The board has been tested with another Portable with a known good power supply that is fully functional and the board still fails so it's not power supply at this point.

H-A-L-9000
September 19th, 2014, 02:04 PM
Guess it's time for the SuperSoft ROM.

Can you confirm?
- The sound is from the PC speaker, not a fan or drive motor
- In case this BIOS does a memory test/count, is the sound audible during this timespan or does it start later? (if it starts later it may be the code execution running out of control, otherwise the timer or PIO responding to bus accesses they shouldn't)

3pcedev
September 19th, 2014, 04:21 PM
Guess it's time for the SuperSoft ROM.



+1 for the Supersoft Rom. As I found out it is not always 100% definitive with its results; however it will at least point you in the right direction. http://www.minuszerodegrees.net/supersoft_landmark/Supersoft%20Landmark%20ROM.htm

cbolvin
September 20th, 2014, 07:22 AM
Guess it's time for the SuperSoft ROM.

Can you confirm?
- The sound is from the PC speaker, not a fan or drive motor
- In case this BIOS does a memory test/count, is the sound audible during this timespan or does it start later? (if it starts later it may be the code execution running out of control, otherwise the timer or PIO responding to bus accesses they shouldn't)

Sounds are definitely from the speaker. It's not hard to tell the difference between a fan or disk noise and speaker tones ;)

There is no BIOS memory count in a Compaq Portable I. That didn't come into play until the 286's AFAIK. The sounds start the instant the power is turned on. Low tone for several seconds followed by a kind of a growl for a short time then high tone from then on.

The POST Analyzer card I have does at least indicate all the voltages are there along with clock signal, IRDY and FRAME signals on the bus.

The SuperSoft ROM sounds interesting. Will it work in a non-IBM machine like the Compaq Portable I ?? The Portable I uses 2764 EPROMS so would I need the special adapter? I wouldn't think so but???

cbolvin
September 20th, 2014, 09:51 AM
Well, SuperSoft ROM to the Rescue! A big thank you for the suggestion H-A-L-9000 and 3pcedev!

I programmed up a 28C64B EEPROM, plugged it in and it took off running!

Looks like a dead keyboard controller. I have a known good keyboard on the machine and it's still throwing the error. And that seems to be the only error! I'm letting it run for a few hours to make sure there aren't any other problems once the machine warms up.

Now to figure out where the issue is as I don't think Compaq uses an actual keyboard controller but sends serial keyboard data to an 8255 on the system board.

Anyone have the system board keyboard connector pin-outs on the Portable I? I'm seeing +12V on one of the pins and that doesn't sound right. Unless Compaq uses 12V to power the keyboard. All the PC and XT pin-outs I see show +5.

20549

modem7
September 20th, 2014, 05:54 PM
Now to figure out where the issue is as I don't think Compaq uses an actual keyboard controller ...
Neither does the IBM PC nor IBM XT.

The test is described on page 26 of the diagnostics manual, however, to be considered when reading that manual is that it covers both the PC and the AT versions of the diagnostics.
In the PC version of the diagnostics, the use of 'KEYBOARD CONTROLLER' is misleading. I think that 'KEYBOARD SUBSYSTEM' should have been used instead.

In the PC version of the diagnostics, the test is probably doing what the POST does, as shown [here (http://www.minuszerodegrees.net/5160/keyboard/5160_keyboard_startup.jpg)].

cbolvin
September 21st, 2014, 04:29 AM
Agreed. In this case I've had two known good keyboards attached and the Supersoft ROM behaves the same way so I would think we can eliminate the actual keyboard itself.

I went and grabbed the Sams Photofact on the Portable and see where the keyboard interface comes into the main board. There's a 74LS125AN (U90) thats in the the path between the keyboard and the 8255. There's also a 74LS322AN (U49). In my case when I checked the pins on U90 I see that it's a quad bus buffer and three out of the four buffers are used. Two buffers show +5 on the A,Y and G leads but one shows +5 on the A and Y leads but 0 on the G lead (pin 13). This buffer I believe is the data in from the keyboard (J1-1). The G pin is coming from the 8255. If this chip is blown it would prevent the keyboard from responding.

The other possibility I see is U49 (74LS322AN). It's in the path as well to the 8255.

I marked the sections on the schematic related to the keyboard and U90 and used a green line to show the path from U90 to the 8255.

20566

H-A-L-9000
September 21st, 2014, 12:10 PM
I'd vote for the 8255 as it also plays a role with the PC speaker and may cause the noise. It's also on the data bus and thus may have a chance to cause the 101.

It could also be whatever chip is connected to Pin 6 of the 8255. Or there's a broken trace for Pins 5 or 36...

cbolvin
September 21st, 2014, 02:03 PM
Interesting. The 8255 was on my radar but what is it about Pin 6 that intrigues you? It's connected to U32 (74LS138 Decoder). That particular line (PPICS) is the output of one of the decoders that uses RA5,6 and 7.

20571

H-A-L-9000
September 22nd, 2014, 09:18 AM
Interesting. The 8255 was on my radar but what is it about Pin 6 that intrigues you? It's connected to U32 (74LS138 Decoder). That particular line (PPICS) is the output of one of the decoders that uses RA5,6 and 7.

20571

It's theoretically possible that a broken decoder maps the 8255 to incorrect addresses.

cbolvin
September 22nd, 2014, 10:23 AM
You may be on to something on the 8255. I dug up my old logic probe and found that 8255 Port B might be dead as pins 18 through 25 (PB0 through PB7) appear to be floating as the logic probe can't detect anything, not high, not low and not pulsing.

Same for U90 (74LS125) pin 13 which feeds PB6 (pin 24) of the 8255.

H-A-L-9000
September 22nd, 2014, 12:15 PM
Yes, the port B should be configured as output. The 74LS125 is a driver with "switch off" input so it can well be that the probe doesn't pick up anything depending on this input.

cbolvin
September 22nd, 2014, 12:53 PM
Yes, the port B should be configured as output. The 74LS125 is a driver with "switch off" input so it can well be that the probe doesn't pick up anything depending on this input.

But should it be floating? Pin 13 of the 74LS125 is also floating which doesn't seem right.

H-A-L-9000
September 23rd, 2014, 10:09 AM
I agree with you, outputs shouldn't be floating, they should be at a defined level.
The schematics are not well readable so I can't say anything about U90. But the outputs of the LS125 are special as they can be high, low, and floating (tristate).

cbolvin
September 24th, 2014, 08:41 AM
Something to think about so I took another look at the LS125 to 8255 relationship. I found logic charts in the Sams Schematic docs.

If you look at the 8255 (U42), the port B pins should not be floating. The logic chart shows Pins 18 to 24 Low and pin 25 High. But the circuit diagram shows Pin 24 High which would match what U90 Pin 13 is supposed to be.

Pin 13 of the LS125 (U90) should be High.

Looks like your initial call on the 8255 appears to be right on the money.

cbolvin
October 26th, 2014, 08:08 AM
Thought I'd update on the repair status. I recently started a new job and it's had me really busy. This weekend was the first free time I've had.

But the good news is the main board lives! I pulled the 8255, installed a socket and popped in a new chip and the board sprang to life and passes all the Supersoft tests.

Only thing left to do is replace all the tantalums and keyboard pads but the portable is back!

Thanks for all the help!