97S low battery Indicator with no display

[John Garza (3665)]

Interesting problem. I did a printer refurb on a 97S that was otherwise working well - except the paper advance problem. When I assembled everything only the low battery indicator was lit, no friendly 0.00 in the display. I tried multiple battery packs (all good), and even tried the charger for a couple of secs. Still no good. I did double check my connections/disconnections with the post pins during the work, and then triple checked after the problem arose. No issues there. I worked on an anti-static mat and bench humidity was 61%. With the cover off, I noticed a transistor was overheating on the main logic board. Markings are a Motorola 931 4-668. I take it as an internal HP part number. That looks to be Q2 according to the 97 service manual. However the component layout diagram in the manual does not match my machine. Just wondering what could have taken out the transistor, and if there is anything else that might be wrong with the unit causing the transistor to fail. Anyone have experience with this sort of thing in Topcats? -J

The internal arrangement.



The transistor in question:

[Hiwi]

Hi John,

i can't see the attached pictures. Is the transistor on the mainboard or on the piggypack PCB for the printer?

I had a similar problem caused by a overload of the printer mechanism, the head was mechanically blocked.
Is the printhead moving (Paper advance)?

The display shortly turns off when the printer is working. TRACE switch is in position NORMAL?
The 3 contacts of the Paper advance switch are in correct postition ? I had one HP97 where the colors did not match the service manual.
Ralf

[EdS2]

(It should be possible to open those images in a new tab - they are on an http site and some browsers don't like to show mixed content.)

[Hiwi]

Ah Ok, that worked. The transistor is on the mainboard.

[John Garza (3665)]

Sorry for the confusion over the images. This forum won't let me add more than one attachment to the post, so I had to use links.

The print-head isn't blocked. I can rotate the gears by hand and make the print-head move across the full range of travel.

I checked the wires going to the paper advance switch - and the post/socket connections leading to the mainboard. They match the colors/positions on a known good 97. However, I did notice the paper advance switch was touched at some point in the past - crude solder job. And one wire is hanging by one strand of wire. Looks to be making contact though. I'll clean that up tonight. Maybe it is a bad connection there.

TRACE is on MAN

And there is no paper in the machine. I need to get the thing ON first before I can load paper. I'm trying things for only a few seconds with power on in the hopes the overheating transistor isn't dead or something else won't get fried.


-J

[Hiwi]

Regarding the adv knob: 2 wires are connected when idle, open when pressed, then the other pair is connected.

Do you have a multimeter on hand? I would check the transistor before turning on again. There could be a faulty tantalum nearby. Or one of the zeners.

I willl check the service manual and look what this transistor does. And most of the HP-97 transistors can be replaced with standard common NPN or PNP transistors e.g. 2N3904


Ralf

[AndiGer]

First of all: Get the extended version of the service manual here
https://literature.hpcalc.org/items/1127
There is the also the newer layout of the main PCB

The printer PCB is mounted correctly - you surely double checked (and show in the picture).

Transistor 4-xxx is NPN and equals to HP part no. 1854-0xxx
(3-xxx is PNP, 1853-0xxx resp.)

[John Garza (3665)]

Well... it's not the wires for the paper advance switch. I cleaned up the solder job with clean secure connections, and checked again (4th time) the wires were where they should be. When I earlier rotated the printer gear manually to check for travel, I left it near the middle position. Applying power still shows a low batt LED but no digits. But the print head did seek the home position on it's own.

Also, wondering if the fault is not the transistor that's overheating. Maybe something else is causing it to draw too much current. I'm considering CR9 as it's involved in setting the oscillation rate for Q2. Maybe if Q2 is over-driven it will generate heat.

Has anyone published a cross-reference between HP part numbers and their standard equivalents? Of course there are custom ICs in here, but I mean the non-custom components.

-J

[Jonas Sandstedt]

Hello,

I think you can find some clues in

http://www.panamatik.de/Taschenrechnersammlung.pdf

from page 96 and forwards, in particular page 101. Some components are common for HP-19c and HP-97, in particular for the printer. The HP-19c has a less advanced printer which does not detect when out of paper and the head drive worm gear is unidirectional, so there are less components.

Best regards
Jonas

[Hiwi]

Hi John,
can you measure some voltages?
VSS,VB and VGG?
6V, 5V, -11V approx.

Ralf

[John Garza (3665)]

VSS: slowly rises from zero to 0.03V
VB: slowly rises from zero to 0.06V
VGG: -5.8V
Vbattery: 5.12V (Voltage of battery out of circuit)

Note: VSS and VB may have risen a little farther, but I only ran the power for 5 sec for each reading.

Looks like something is draining the voltages across the board.
Maybe C6 ? But then that wouldn't explain the transistor overheating.
Maybe if the base-emitter junction was faulty in Q2, the battery would drain to ground through it, thereby causing heat ??

-J

[Hiwi]

Hi John,
bad news. You can check first all discrete parts (diodes, capacitors).

I your case the service manual advices us to change ICs on the mainboard until VGG is good.

I would disconnect VGG line part for part until the reaseon for heavy load is found.

Ralf

[AndiGer]

John,

Did you also service the card reader? Is the ribbon cable misaligned with the connectors inside the connector? If so, it could have produced a short and a diode was blown in this case. I had that once but could see the smoke and this way I knew which part had failed when only the low bat light was on.

Maybe it's only the perspective of the picture but to me it seems the ribbon is not centered ...

Andi

[John Garza (3665)]

I had serviced the card reader years ago. But I did remove the J8 ribbon cable to check the gear/roller. It was not centered, but didn't look to be making the wrong connections - I reseated it with better alignment. But still the problem persists. There was no smoke or smell when I first turned it on and found the low battery display. What diode was affected? I did visually inspect all the diodes for signs of trauma and found none.

[AndiGer]

Check CR5, RHS of the torroid - even mounted should be possible. Edit: use diode tester of DMM
If necessary replace by 1N4150

Andi

[John Garza (3665)]

CR5 and Q2 both test good. So maybe Q2 is over-driven. I looked at the signal on the base of Q2 and it looks to be a sloppy waveform at 550KHz. A reference known good HP-97 showed around 50KHz with a more squarish wave. Maybe Q3 or CR9 aren't keeping the oscillation in range? The service manual says it should be 20KHz, but my 97 is working fine at 50. So the 97S is oscillating much faster than it should be!

-J

   

[John Garza (3665)]

Those sharp valleys made me think of diode trouble. So I rechecked my work. Luckily I got in the habit long ago of documenting everything I do step by step with pictures and notes. I found I goofed! I confused CR5 and CR9. Turns out CR9 is good and CR5 is bad. It conducts both ways. So hats off to AndiGer ! It was the J8 cable to the card reader that shorted and killed the diode. I have a replacement 1N4150 on order. Whew! Thanks everyone for your advice and comments. -J

[AndiGer]

I‘m happy I could help you. Thx for feedback
Andi

[Duane Hess]

(02-28-2023 01:51 PM)John Garza (3665) Wrote:  .........................
... omitted ..........

the battery would drain to ground through it, thereby causing heat ??

-J

* EDITED *

From your posts, it appears you've found a solution. However, no display, low battery LED only, heating and battery drain is similar to the following thread, so thought mentioining it, due to similarity, might be useful to someone:

https://www.hpmuseum.org/cgi-sys/cgiwrap...read=17322
(note comments by Tony Duell)

and a related post where, depending on the service manual you use, some items are (erroneously) labelled differently:

https://www.hpmuseum.org/forum/thread-16655.html

The posts don't mention the transistor getting hot (Q2?) but the big resistor on the lower-right of your 1st picture -- the resistor with the yellow, purple, gold bands.

Thought I would mention it for whatever its worth.