Woodstock HP-27 ACT woes...

[John Garza (3665)]

I'm working on an HP-27.

I've read the old articles on Woodstock ACT chip problems. There are obvious differences in pinouts/etc. so you have to take those old articles with a grain of salt. But the theory seems sound - for some reason, internal electrical noise renders the machine useless. Has anyone found the root cause of this? Is this some sort of internal chip degradation, or maybe external capacitors, micro-cracks, etc ? And is there an accepted 'standard' fix? The old articles mention adding resistors and capacitors to thwart some of the noise - but I think that may have limited effect. If the chip continues to degrade, it may reach a point where filtering won't work.

This particular machine appears to have 2 different clock signals, both with noise. On a scope, the noise looks like 3 or 4 different signals with different amplitudes and phases superimposed on the main signal. The machine was completely dead - no display. But by putting a 25K resistor across one clock line to ground allowed me to use the keyboard, stack, arithmetic (I presume the ACT chip). However trig functions, stats, etc. are still a no go (I presume the ROM/RAM chip for the CORDIC tables, etc. ?)

-John

[Nick]

I'm not familiar with the cause other than reading mention that operating a unit with a missing or bad battery could contribute.

Quote:Many ACT chips of the “Woodstock” series HP calculators were damaged while using its power supply without batteries, due to a design flaw of the original power switching unit. - http://www.panamatik.de/html/hp_calculat...r_kit.html

Here are some links:

• Thread: ACT available [hpmuseum.org]
• HP Calculator Repair Kit [panamatik.de]
• Thread: Woodstock battery modification
  

Yip,
Nick

Edit: Added Quote; Battery Mod

[PANAMATIK]

(09-12-2016 12:20 AM)John Garza (3665) Wrote:  ... However trig functions, stats, etc. are still a no go (I presume the ROM/RAM chip for the CORDIC tables, etc. ?)

-John

I remember I had a HP-25C calculator, which worked fine except for calculating trig functions. When analyzing the problem I found a single defective ROM bit in the 2048x10 bits ROM code, which caused the problem. By replacing the ACT with newACT using internal ROM mode everything worked fine again.

Bernhard

[John Garza (3665)]

WOW!
The new ACT chip looks great. And a GPS option too! In addition to the HP-27, I have three 25C's all in good shape - but I just might sacrifice one for the new chip. Is Panamatik still taking orders, it it via Paypal, or send a check? All I see is an email address on their website.

-John

[PANAMATIK]

(09-13-2016 12:20 AM)John Garza (3665) Wrote:  WOW!
The new ACT chip looks great. And a GPS option too! In addition to the HP-27, I have three 25C's all in good shape - but I just might sacrifice one for the new chip. Is Panamatik still taking orders, it it via Paypal, or send a check? All I see is an email address on their website.

-John

You cannot take an HP-25 ACT for repairing an HP-27. They are not compatible, one is PMOS, the other NMOS.

For ordering a newACT just mail and you will get info about payment via paypal.

Bernhard

[jupiters_spot]

I'm curious how one goes about determining that the ACT is the root of the problem. I have exactly one HP-27 that's intermittent, so it's hard for me to measure known-good values and compare to my problematic calculator.

I do have a couple of HP-21 calculators, but from discussions on Panamatik.de it seems the CMOS technology differs on the 27, so again, not a good "apples-to-apples" comparison.

[teenix]

(09-26-2016 06:41 PM)jupiters_spot Wrote:  I'm curious how one goes about determining that the ACT is the root of the problem. I have exactly one HP-27 that's intermittent, so it's hard for me to measure known-good values and compare to my problematic calculator.

I do have a couple of HP-21 calculators, but from discussions on Panamatik.de it seems the CMOS technology differs on the 27, so again, not a good "apples-to-apples" comparison.

A stab in the dark but...

Semiconductors will surely break down over time and it is possible that a single bit (ie transistor) in the thousands that make up the ROM, ACT or other chip circuitry works "mostly" and due to temperature and possibly other effects, that rogue bit may fail to operate properly causing intermittent behaviour.

Could even be a bad discrete component like a capacitor or old carbon resistor that is part of the circuitry, they too can begin to fail over time.

Some "gunk" that's managed to get on the circuit board from a cup of coffee that was spilt ten years ago. Over the years of handling and vibration, an old solder connection on a component has become a partial dry joint.

Very hard to fix a problem like this without having specialised test gear or some plug in replacements to find the culprit. A CRO probably won't pick up internal chip faults.

An hour in the fridge or a brief time under a hair dryer might expose a temperature problem, but maybe unwise to stress out aging (perhaps irreplaceable) components further.

My old HP-45 played up about a year after purchase and I would never have thought to try and fix back in those days. HP did though, and it looked brand new when they returned it.

cheers

Tony

[John Garza (3665)]

... and don't forget the horrible battery contacts and springs that corrode...

Most 'intermittent' behaviour is related to that. If it acts weird with a good power supply securely connected, then it's most likely the ACT.

-J