HP 71b repair advice - won't power on with batteries or on AC adaptor

[KimH]

(07-13-2021 01:55 AM)dmh Wrote:  

(06-11-2021 06:40 AM)KimH Wrote:  Pretty much what I expected it would look like.

Some gentle application of vinegar to get rid of all the corrosion (watch out for possible bridging between traces and components, use a loupe).

The “flange” for the zebra is probably held by corrosion as well, again vinegar is you friend.

Then on to the soldering…

I have repaired at least a handful that way, none failed to work.

Tip: use two different colors of wire alternating, it helps avoiding mistakes in the sequence.

Hi KimH,

So I opened my other dead 71b and it is very clean with only slight damage to the zebra and barely any corrosion and have soldered the 2 halves together with wire wrap wire but still no go.

Do you have any other suggestions / advice?

Thanks

Hi there,

Without having anything in front of me it is really hard to give suggestions here.

I have looked at my notes and I have repaired 6 machines which were „like yours“ and all of them came to life, I have yet to see one which was completely dead.

Note that even the tiniest of damage on the zebra can be fatal, the contact is critical. The display-part which is welded also has zebras, which might be your issue, but as said, I have yet to see one of those situations.

One test which could be done is to see if there is power on the front-module ports when having batteries. The first and last pads/contacts are those where the power can be found. The battery voltage or a bit less (around 5 volts) should be available even when OFF.

[teenix]

Hi,

I've been looking at the 71 for Darren. All boards were cleaned up and look ok. I've added jumper wiring to bypass the zebra connector, and checked it. The power supply rails have battery voltage appearing around the circuit boards. Nothing inside appears to be getting warm with prolonged power applied.

As far as I can see, there is not much else that is user repairable in these machines. The only thing I thought of is a faulty keyboard or a problem with one or more IC's.

I can see a very small current when the battery is connected which disappears when not. This would suggest that it is not fully powering up after the ON button is pressed.

I wonder why they didn't just put a simple pin and socket connector instead of the zebra style. I don't think it would have affected manufacture assembly.

cheers

Tony

[KimH]

(07-13-2021 05:03 AM)teenix Wrote:  Hi,

I've been looking at the 71 for Darren. All boards were cleaned up and look ok. I've added jumper wiring to bypass the zebra connector, and checked it. The power supply rails have battery voltage appearing around the circuit boards. Nothing inside appears to be getting warm with prolonged power applied.

As far as I can see, there is not much else that is user repairable in these machines. The only thing I thought of is a faulty keyboard or a problem with one or more IC's.

I can see a very small current when the battery is connected which disappears when not. This would suggest that it is not fully powering up after the ON button is pressed.

I wonder why they didn't just put a simple pin and socket connector instead of the zebra style. I don't think it would have affected manufacture assembly.

cheers

Tony

Hi Tony,

The fact that it draw a small current is actually (probably) good news. The Static Ram should draw somewhere about 5 uA

my next step would be two fold.

Measure the actual current when the ON is pressed, does it launch the CPU & BUS

If it does, I would take a loupe and check for whether the corrosion has shorted something

Regards
Kim

[teenix]

I fiddled with it again this morning and checked the waveforms with my oscilloscope. They matched all the service manual pictures including the 32KHz clock signal so I thought the display or drivers were damaged.

To my surprise when I looked at the display it was working. Why this is so - not sure and nothing obvious, but it seems to be ok now and still going after reassembly.

One thing I did notice when looking at the 2.5Mhz clock signal into the CPU it was working when the ON switch was down, then stopped when the ON button was released.

If I remove the power completely and discharge the supply rails, then turn it back on the clock signal stays on all the time with "Memory Lost" on display.

All the keys appear to work including [ON] and [f] [OFF].

I'm not familiar with the 71 operation, so hopefully it will be ok. They must have had big shirt pockets back then for this beastie :-)

cheers

Tony

[dmh]

(07-14-2021 12:41 AM)teenix Wrote:  I fiddled with it again this morning and checked the waveforms with my oscilloscope. They matched all the service manual pictures including the 32KHz clock signal so I thought the display or drivers were damaged.

To my surprise when I looked at the display it was working. Why this is so - not sure and nothing obvious, but it seems to be ok now and still going after reassembly.

One thing I did notice when looking at the 2.5Mhz clock signal into the CPU it was working when the ON switch was down, then stopped when the ON button was released.

If I remove the power completely and discharge the supply rails, then turn it back on the clock signal stays on all the time with "Memory Lost" on display.

All the keys appear to work including [ON] and [f] [OFF].

I'm not familiar with the 71 operation, so hopefully it will be ok. They must have had big shirt pockets back then for this beastie :-)

cheers

Tony

Great work!!!

Wait until you see the size of the 75C/D :-)

[KimH]

(07-14-2021 12:41 AM)teenix Wrote:  I fiddled with it again this morning and checked the waveforms with my oscilloscope. They matched all the service manual pictures including the 32KHz clock signal so I thought the display or drivers were damaged.

To my surprise when I looked at the display it was working. Why this is so - not sure and nothing obvious, but it seems to be ok now and still going after reassembly.

One thing I did notice when looking at the 2.5Mhz clock signal into the CPU it was working when the ON switch was down, then stopped when the ON button was released.

If I remove the power completely and discharge the supply rails, then turn it back on the clock signal stays on all the time with "Memory Lost" on display.

All the keys appear to work including [ON] and [f] [OFF].

I'm not familiar with the 71 operation, so hopefully it will be ok. They must have had big shirt pockets back then for this beastie :-)

cheers

Tony

That's great news - congratulations on your first repair of the TITAN

The behavior of the clock running when you press a button (ON in this case) is normal. You may know this, the CPU has 3 states of power - DeepSleep (OFF), Idle (waiting for input, display on) and Running (well, when it actually does something).

Might be too much information, but heck - a LEX-Program called KEYWAIT manages to have the CPU go into Idle when a/your program is pausing and waiting for input "Enter Temp:" for example. Without the KEYWAIT function the CPU would gobble power while waiting, nifty little detail which ought to have been part of the original design.

The fact that Memory Lost has the CPU running is a bit odd, it maybe be doing something in the background for a few seconds then go idle. I will try this next time I have some time.

BTW - didn't Darren have 2 machines

[dmh]

(07-14-2021 08:21 AM)KimH Wrote:  

(07-14-2021 12:41 AM)teenix Wrote:  I fiddled with it again this morning and checked the waveforms with my oscilloscope. They matched all the service manual pictures including the 32KHz clock signal so I thought the display or drivers were damaged.

To my surprise when I looked at the display it was working. Why this is so - not sure and nothing obvious, but it seems to be ok now and still going after reassembly.

One thing I did notice when looking at the 2.5Mhz clock signal into the CPU it was working when the ON switch was down, then stopped when the ON button was released.

If I remove the power completely and discharge the supply rails, then turn it back on the clock signal stays on all the time with "Memory Lost" on display.

All the keys appear to work including [ON] and [f] [OFF].

I'm not familiar with the 71 operation, so hopefully it will be ok. They must have had big shirt pockets back then for this beastie :-)

cheers

Tony

That's great news - congratulations on your first repair of the TITAN

The behavior of the clock running when you press a button (ON in this case) is normal. You may know this, the CPU has 3 states of power - DeepSleep (OFF), Idle (waiting for input, display on) and Running (well, when it actually does something).

Might be too much information, but heck - a LEX-Program called KEYWAIT manages to have the CPU go into Idle when a/your program is pausing and waiting for input "Enter Temp:" for example. Without the KEYWAIT function the CPU would gobble power while waiting, nifty little detail which ought to have been part of the original design.

The fact that Memory Lost has the CPU running is a bit odd, it maybe be doing something in the background for a few seconds then go idle. I will try this next time I have some time.

BTW - didn't Darren have 2 machines

I do have another 71b :-)

The first one I opened was in worse condition - battery terminals corroded/broken and upon opening it there was more corrosion (not just under the zebra connector) so it needs more work.

The second / now working device was almost spotless inside and out. The zebra had corroded only a few millimetres.

Might not be as lucky with the second one.

[vassilisprevelakis]

(06-07-2021 12:29 PM)Etienne Victoria Wrote:  - Later models with the cost saving, not soldered, el cheapo, plasticky, Hp41-like, so-called "Zebra strip" : these have less clearance to insert and fold a flex cable so I would advise to solder top to bottom wires directly.

I repaired my HP-71B with failed "zebra strip" the same way (by soldering wire-wrap wires between top and bottom) in the early 2000s (more or less 20 years ago -- time flies) and its still working fine.

It is a scary procedure though!

**vp
http://www.series80.org