The Museum of HP Calculators

HP Forum Archive 13

[ Return to Index | Top of Index ]

HP34C "2" disease, flashing, voltages, schematic - long
Message #1 Posted by Ellis Easley on 24 June 2003, 5:01 a.m.

(*)Note: I edited this message to correct an error. When I measured the output current of the constant current supply, I connected the milliameter between pin 10 and pin 6, not pin 10 and pin 4. 6/24/2003 3:00 pm (**)Corrected another error - output voltages measured relative to pin 6 and not pin 4. 6/24/2003 6:40 pm

Now that I have an earlier 34C working, I see what the "2" disease is! I would describe the condition this way: whenever you press the "2" key when there is someplace for a 2 to go in the display, or there is a 2 in the display and you press any key, the minus sign and possibly various commas flash momentarily. If you type ten digits other than 2, filling the display, then pressing "2" doesn't cause flashing. If you type a number in scientific notation without any 2's in the mantissa, you can make the flashing start and stop by shifting a 2 through the exponent; in this case the flashing commas are limited to the exponent digits. My newer unit doesn't do this "2" flashing at all.

The other flashing my older unit does is to flash random commas and the minus sign during execution of built-in functions. Calculating the gamma function (! of non-integers) of negative numbers with large magnitude seems to take the longest. My newer unit is completely dark while executing a built-in function. You can make the calculator work longer with the Integrate function, but then both units flash random digits when the user function subroutine is being evaluated.

I measured the power supply outputs and current drain with no load (power supply module on the bench) and in circuit (calculator assembled). I made the no load measurements with two different input voltages to see the effect of the voltage and current regulation circuits in the power supply module.

First, I connected two NiCads in series to the power supply module, (-) to pin 6 and (+) to pin 4 (see the netlist below to see how I numbered the pins). Both pin 7 (the standby supply that powers the 8 pin IC closest to the power supply module) and pin 8 (which supplies power to the other two 8 pin ICs and the 40 pin IC) measured 7.12V relative to pin 6 (**). Pin 10 of the power supply module, which goes to pin 8 of the 40 pin IC and which is driven by a constant current circuit in the power supply module, measured 4.62V and when connected to pin 6 (*) through a milliameter, it sourced 39 mA. The current drain from the battery with no load was 8 mA and when the constant current supply was loaded, the current drain was 106 mA and the voltage on pins 7 and 8 increased to 7.46V.

Next, I added another NiCad, increasing the input voltage from 2.58V to 3.88V. The no load current drain increased to 9 mA. The voltage on pins 7 and 8 was 7.14V. The voltage on pin 10 was 5.22V. Connecting a milliameter between pin 10 and pin 6 (*), I measured 38 mA, the voltage on pins 7 and 8 increased to 7.28V and the battery current drain increased to 76 mA. So the no load output voltages on pins 7 and 8 and the current sourced by pin 10 change very little with a 50% increase in input voltage.

The constant current supply is similar to the one in the Classic recharger. It works by maintaining a voltage drop equal to the emitter-base junction voltage of a silicon transistor (~.6V) across a 16 ohm resistor: (0.6V)/(16 ohms)=(0.0375A) or ~38 mA. The circuit in the Classic recharger (which provides the current that recharges the battery) uses a 13 ohm resistor and provides ~50 mA.

Next, I assembled the calculator and measured the voltages that are accessible that way. I couldn't get to the pins connected to the constant current supply because of the molded plastic frame but I could measure the two voltage supplies. With power supplied by two NiCads, the voltage on pin 8 of the 8 pin IC closest to the power supply module measured 6.82V and the voltage on pin 8 of the other two 8 pin ICs measured 6.52V, both relative to pin 7 of the 8 pin ICs (which are connected to the negative terminal of the battery).

I also measured the voltages on the other pins of the 8 pin chips with the calculator idle and running a program. Pin 6, which goes to the unknown jumper on the power supply module and to pin 14 of the 40 pin chip, measure 5.33V either way. If the unknown jumper were installed, it would connect the pins to ground. Pin 5 measures 0.65V idle and 0.88V running. Pins 4 and 3 measure 0.86V either way. Pin 2 measures 0.86V idle and 1.35-1.41V running. Pin 1 measures 0.04V idle and 0.3-0.5V running. All were measured with a digital voltmeter with respect to the battery negative terminal (pin 7 of the 8 pin IC's). I'll have to fire up one of my many scopes and see if some of those "quiet" lines have low duty-cycle signal activity on them.

Other than the power supply, the schematic is just connections between the pins of the IC's, LED and keyboard that can be communicated better as a netlist. I've typed that up and include it here:

;HP34C Main Circuit Assembly (solderless version) netlist-

;component reference designators and pin numbering IC1=8 pin chip with standby power (closest to the power supply module) IC2=second 8 pin chip IC3=third 8 pin chip IC4=40 pin chip PS=power supply module with the flex cable solder pads on its PCB numbered from left to right as viewed from the rear. Note: pin 1 only goes to the recharger connector, it doesn't go to the main circuit assembly. The pads where the flex cable contacts the main circuit assembly flex circuit are PS.2 - PS.10. LED=LED display with pins numbered from left to right as viewed in normal use PRSW=Program-Run switch, pin 1=common, pin 2=closed on Run OOSW=On-Off switch, pin 1=common, pin 2=closed first when switched to "On", pin 3=closed second when switched to "On" KBR=keyboard rows numbered in standard keycode format KBC=keyboard columns numbered in standard keycode format for rows 3-7, offset by -1 for rows 1-2 AC=AC connection from battery contact assembly to power supply module (return is bat-) (doesn't go to main circuit assembly)

;memory chip connections v+standby = PS.7,IC1.8 v+ = PS.8,IC2.8,IC3.8,IC4.12,PRSW.1 bat- = PS.6,IC1.7,IC2.7,IC3.7,IC4.13,LED.22; LED.22 is minus sign cathode, anode of ??? diode with cathode on LED.1 (goes to psin) mem6 = PS.5,IC1.6,IC2.6,IC3.6,IC4.14; memory bus, ??? jumper connects to bat- if present mem5 = IC1.5,IC2.5,IC3.5,IC4.15; memory bus mem4 = IC1.4,IC2.4,IC3.4,IC4.16; memory bus mem3 = IC1.3,IC2.3,IC3.3,IC4.17; memory bus mem2 = IC1.2,IC2.2,IC3.2,IC4.18; memory bus mem1 = IC1.1,IC2.1,IC3.1,IC4.19; memory bus

;control signals run = PRSW.2,IC4.20; Program:open,Run:switch connects to v+ radix = PS.9,IC4.9; comma:open,period:jumper on power supply module connects to v+

;another power supply module output iled = PS.10,IC4.8; 38 mA constant current supply - for LEDs?

;LED anodes antr = LED.21,IC4.3; top right digit segment ant = LED.20,IC4.2; top digit segment antl = LED.19,IC4.1; top left digit segment anm = LED.18,IC4.4; middle digit segment anb = LED.17.IC4.5; bottom digit segment anbl = LED.6,IC4.36; bottom left digit segment ancom = LED.5,IC4.40; comma anbr = LED.4,IC4.39; bottom right digit segment anper = LED.3,IC4.38; period anms = LED.2,IC4.37; minus sign anode (cathode is on LED.22, goes to bat-)

;LED cathodes, keyboard rows(numbered in standard keycode format) ca1 = LED.16,IC4.21,KBR.7; digit 1 (leftmost), keyboard row 7 ca2 = LED.15,IC4.23,KBR.6; digit 2, keyboard row 6 ca3 = LED.14,IC4.26,KBR.5; digit 3, keyboard row 5 ca4 = LED.13,IC4.28,KBR.4; digit 4, keyboard row 4 ca5 = LED.12,IC4.29,KBR.3; digit 5, keyboard row 3 ca6 = LED.11,IC4.30,KBR.2; digit 6, keyboard row 2 ca7 = LED.10,IC4.32,KBR.1; digit 7, keyboard row 1 ca8 = LED.9,IC4.33; digit 8 ca9 = LED.8,IC4.34; digit 9 ca10 = LED.7,IC4.35; digit 10 (rightmost)

;keyboard columns(numbered in standard keycode format rows 3-7, offset by -1 rows 1-2) kc0 = IC4.31,KBC.0; "A","X<>Y"(column 1) kc1 = IC4.27,KBC.1; "B","GTO"(column 2),"ENTER","-","+","*","/"(column 1) kc2 = IC4.25,KBC.2; "GSB","STO"(column 3),"CHS","7","4","1","0"(column 2) kc3 = IC4.24,KBC.3; "f","RCL"(column 4),"EEX","8","5","2","."(column 3) kc4 = IC4.22,KBC.4; "g","h"(column 5),"CLX","9","6","3","R/S"(column 4)

;power switch psin = OOSW.1,PS.4,LED.1; + input to PS, LED.1 is cathode of ??? diode with anode on LED.22 (connected to bat-) bat+ = OOSW.2,PS.3; makes first when switch turned "On" chrg = OOSW.3,PS.2; makes second when switch turned "On", to short out 8.2 ohm current limit resistor on power supply module

;AC adapter (doesn't go to main circuit assembly - goes only to power supply module) acsource = AC.1,PS.1

Now all that's left is a schematic of the power supply module. I will attempt to write a description of the circuits on the power supply module with enough details to draw the schematic.

HP34C Power Supply Module circuit description-

The battery negative terminal and the AC adapter share a common return connection on pin 6 of the power supply module; this is also the return for the main circuit assembly. The other side of the AC adapter is connected to pin 1, which isn't connected to the main circuit assembly. Pin 1 goes to the anode of a rectifier diode whose cathode goes to pin 2. Pin 2 goes to an 8.2 ohm, 1 watt resistor which goes to pin 3. Pin 3 goes to the battery positive terminal. This completes the circuit which charges the battery when the unit is turned off.

Pin 4 is the input to the switching power supply and is connected to the common terminal of the On-Off switch. The switch has two other terminals: one makes contact with the common terminal first when the switch is turned to "On" and it is connected to pin 3, the battery positive terminal. The other switch terminal makes contact with the common terminal second when the switch is turned to "On" and it is connected to pin 2; its purpose is to short out the 8.2 ohm battery charging current limiting resistor when the unit is on. This completes the description of how power from the battery and the recharger is connected to the power supply input.

Pin 5 only goes to a pad which can be connected by a jumper (not installed) to pin 6, the battery negative terminal. On the main circuit assembly, pin 5 is connected to pin 6 of the 8 pin IC's and pin 14 of the 40 pin IC. The purpose of the jumper option is unknown. Pin 7 provides standby power to pin 8 of the 8 pin IC closest to the power supply module. The voltage on pin 7 is supplied by two sources: first, it is connected to the battery positive terminal to provide memory backup when the unit is off - through a diode (anode to the battery positive terminal, cathode to pin 7) on my older unit, or through a 10K resistor on my newer unit. To support this, there is a 22 mf capacitor from the battery positive terminal (cap + lead) to the battery negative terminal to provide backup power while changing the battery. The second source of power for pin 7 is a voltage regulated output of the switching power supply which also provides power to pin 8 for the other chips on the main circuit assembly. Both pins are connected to the supply through separate diodes (anodes to the voltage regulated supply, cathodes to pins 7 and 8). When the unit is on, the voltages on pins 7 and 8 are approximately +7V. Pin 9 only goes to a pad which can be connected by a jumper to pin 8. On the main circuit assembly, pin 9 goes to pin 9 of the 40 pin chip. When the jumper is installed, the calculator uses period as the radix mark; when it is not installed, comma is used. Pin 10 is connected to a constant current supply which sources approximately 38 mA and is powered by the switching power supply. On the main circuit assembly, this supply goes to pin 8 of the 40 pin chip. I believe this current source is used to drive the LED segments. This completes the description of the terminals of the power supply module. What remains to be discussed is the switching power converter, the regulated voltage supply and the regulated current supply.

The power converter transformer is wired as an autotransformer with a primary winding and two secondary windings connected in series with the primary winding. An additional independent winding provides feedback for the oscillator that drives the primary winding. The power supply input (from pin 4) is connected to the top of the primary winding (= the bottom of the secondary windings) and the bottom of the primary winding is connected to the collector of an NPN transistor with its emitter connected to the battery negative terminal (pin 6). The base of the transistor is biased on by a 680 ohm resistor from the power supply input. One end of the feedback winding is connected to the battery negative terminal and the other end is connected to the base of the NPN transistor through a series network of a 2000 pf capacitor and a 47 ohm resistor. The end of the higher secondary winding in series with the primary winding, with a voltage equal to the power supply input voltage (the battery positive terminal voltage) plus the sum of the voltages across the two windings, is connected to a rectifier diode anode. A 22 mf capacitor (33 mf on my newer unit) is connected from the diode cathode (cap + lead) to the battery negative terminal. This capacitor positive lead is connected to the anodes of the two diodes that supply the regulated +7V to pins 7 and 8 and also to the cathode of a zener diode. The anode of the zener is connected to a 1 mf capacitor (+ lead) with its other lead connected to the battery negative terminal. The anode of the zener is also connected to a 1K resistor which goes to the base of a smaller NPN transistor with its emitter connected to the battery negative terminal. The collector of this transistor is connected to the base of the oscillator transistor. When the output voltage of the supply rises high enough to cause the zener diode to conduct, the smaller transistor turns on and shunts some bias current (provided by the 680 ohm resistor) away from the base of the oscillator transistor, reducing the amplitude of the oscillation and the output voltage. This completes the description of the switching power converter and the voltage regulated supply.

The remaining tap of the transformer - the junction of the two windings in series with the primary winding - is connected to the anode of a rectifier diode. A 22 mf capacitor is connected between the cathode of the diode (cap + lead) and the battery negative terminal. This provides a supply of something less than 7V for a constant current source. The diode cathode and the cap + lead are connected to a 16 ohm resistor which goes to the emitter of a PNP transistor with its collector connected to pin 10, the current regulated output. The base of the transistor is connected through a 2.2K resistor to the battery negative terminal; this biases the transistor on. A smaller PNP transistor has its base connected to the other transistor's emitter, its emitter connected to the other end of the 16 ohm resistor, and its collector connected to the other transistor's base. When the voltage across the 16 ohm resistor exceeds approximately 0.6V, current flows through the smaller transistor's emitter-base junction and it turns on, shunting current away from the larger transistor's base and reducing its collector and emitter current. Since 0.6V divided by 16 ohms = .0375 amps, the circuit reaches equilibrium with about 38 mA flowing out of the larger transistor's collector towards the battery negative terminal.

Edited: 24 June 2003, 7:44 p.m. after one or more responses were posted

      
I dont like "2 disease" either
Message #2 Posted by Norm on 24 June 2003, 1:44 p.m.,
in response to message #1 by Ellis Easley

I dont like "2 disease" either, but, there doesn't seem to be any cure.

Our latest best guess is that there are 2 kinds of software code, so the problem is in the 8-pin memory chip that holds the firmware then.

That might allow one or two "miracle cures" however the supply of donor patients is inherently low.

            
Re: I dont like "2 disease" either
Message #3 Posted by David Smith on 24 June 2003, 4:59 p.m.,
in response to message #2 by Norm

The problem could also be in the 40 pin chip. There was a Keynotes description on the HP65 display system that talked about how the display was formatted by various byte codes in a particular internal register. Certain illegal operations left weird values in the register and could cause funny displays. If the 34C display circuit is similar, then a design bug might be a likely cause of the symptoms.

                  
Re: I dont like "2 disease" either
Message #4 Posted by Ellis Easley on 24 June 2003, 5:54 p.m.,
in response to message #3 by David Smith

They definitely changed something because my (slightly) newer, soldered 34C doesn't have the "disease". There are differences in the chips. Do you think there would be a difference between "1820-2162 A" (in my 1934A unit) and "1820-2162" (in my 2051S unit)? These are the 40 pin chips, both made by AMI, packaged in Korea, with datecodes 7906 and 8047 respectively. I've never noticed revision letters on HP part numbers (and these seem to be progressing in the wrong direction!) I would think that for custom chips especially HP would assign a new part number, even if there is no change in fit or function, if only to keep things straight with the vendor - look at all the apparently equivalent ACT chip P/N's for the Woodstocks. Another chip difference is in the 8 pin chip with the backup power supply. In the older unit it is marked "1LB5 002", in the newer unit "1LB9 02". I would have assumed this chip was primarily RAM but I realize it could also contain ROM. The other two chips have almost the same part numbers in both units but vary slightly in the leading zeros in the suffix - the older unit has "1MA1 14" and "1MA1 0015", the newer one has "1MA1 14" and "1MA1 15". Looking at all these 8 pin DIP part numbers, it looks like they dropped leading zeros sometimes to fit in the significant digits - I know chip vendors will only put so many digits on custom parts for your purposes, since they need room for their own tracking numbers.

                        
HP-33E "2 disease"
Message #5 Posted by Jon on 25 June 2003, 2:46 p.m.,
in response to message #4 by Ellis Easley

I had an HP-33E without the disease. Years ago I had to repair battery contacts and replace flex connector for wires. It was my first time so I made several soldering tests (I didnīt know which was + or - wire). Once I made it work it had the "2 disease".

I donīt know the reason but it is something I damaged

                              
aahhhhhhhh HAH! "2 disease" is damage ...
Message #6 Posted by Norm on 26 June 2003, 1:36 a.m.,
in response to message #5 by Jon

Hi Jon, This is valuable data. It says perhaps ESD into the large 40 pin chip could cause "2 disease".

When Luiz finally swaps firmware chips (the 8-pin chip that holds the program) it will be an important piece of data.

If swapping the firmware does NOT change the problem, then the real problem is damage in the 40-pin chip that runs the display.

We can't forget how ancient these chips are, I mean, fer cryin' out loud, late 1970's. They didn't have a clue how to protect something from static damage. If we can prove that "2 disease" is a static-induced problem, then we will have to do all our work grounded out with alligator clips, everything on tinfoil, even the solder iron and the roll of solder grounded out, and stuff aluminum foil into your socks and attach alligator clips

(then U get up in a hurry to answer the phone and wow, what a circus!)

ALSO PLZ NOTE that I am suspicious that just running a solderless spice calculator with a dirty connection might be enough to induce "2 disease". IF YOUR SPICE CALCULATOR gets the least bit unreliable, and its solderless, STOP using it because it might damage it if used with loose connections.

There is a LOT of spice calculators afflicted with these problems, and that cuts down the pool of available units.

So we need to resolve what is the real story.

                                    
One isolated, confirmed case, so far...
Message #7 Posted by Vieira, Luiz C. (Brazil) on 26 June 2003, 3:09 a.m.,
in response to message #6 by Norm

Hi Norm, guys;

maybe I'm wrong, but that would be such a tremendous coincidence that amongst the so many MOS transistors in an average complex digital chip, we have all and all the same set being burnt in certain circumstances and most of all, it does not alter anything but a selectable blinking dots/commas/minus_sign when a highly selected digit, in this case "2" and no other one else, is present in the display and another digit is entered or an operation is performed.

I have seen cases when digital I.C.'s loose part of its operating capability or become completely "off" when submitted to unacceptable operating environment. MOS are likely to suffer more damage because of its own nature. But it's still hard for me to accept that such a highly selected lost of operating characteristic with no other perfomance degradation, not a single keyboard freeze or the like, is simply a hardware problem instead of a firmware problem. I'd actualy like to know how many Spices with this behavior work like this since the day they were 'born'. I have no doubts about Jon's experience, I am sure it happened, I'm not questioning this fact, I just cannot understand such a highly "specialized" hardware problem.

If we consider that part of the existing HW is actualy too sensitive and when fails causes this exact behavior, then it's a matter of time to expect that all Spices that have been serviced to be like this, and there's nothing we can do about it except to feel lucky enough that our calculators never need service.

Analog IC's may show this sort of "partial" performance degradation, and they may run for a long time with internal componentes partialy burnt. Digital IC's are likely to show random "problems".

I have a stuck HP38C that locks the display to 0.00 everytime it comes to life. Sometimes it shows Pr Error when I switch it off and on again. I am waitng for a chance to test another power supply board on it, and I have one HP33C that does nothing. This HP38C is a soldered unit and I am thinking seriously about removing all of its chips and swap one by one with another HP38C so I can see what happens. The other HP38C has no problem with "2" digit.

I know and understand Norm's concerns. I accept the fact that he does not like this behavior and I do not disagree with him. But to be honest, for as long as [STO][ENTER] returns [-8.8.8.8.8.8.8.8.8.] instead of [ Error 9 ] to the display, I'm confident using any Spice. And I don't care much having a "2" disease calculator: I have three Spices like this.

My thoughts.

Luiz C. Vieira - Brazil

Edited: 26 June 2003, 3:17 a.m.

                                          
Luiz, I used to design CMOS chips
Message #8 Posted by Norm on 26 June 2003, 10:34 a.m.,
in response to message #7 by Vieira, Luiz C. (Brazil)

Luiz, I used to design CMOS chips, and it is possible for the weakest link in the chain to consistently snap.

They may have used a certain transistor to cause the blanking, and maybe that transistor gets fried from ESD. Maybe the last logic gate is hooked to a very large power transistor that runs the LED's. I'm not saying it DOES, I'm saying that it COULD be that way.

While I would be happy if it is just firmware (maye I can swap a chip and make it work properly?) you cannot presume it is firmware until you have run a comparison. A swap between 2 units would offer some proof. Otherwise nobody knows and you can't fairly say that it is the firmware.

                                                
Re: Luiz, I used to design CMOS chips (edited)
Message #9 Posted by Vieira, Luiz C. (Brazil) on 26 June 2003, 12:51 p.m.,
in response to message #8 by Norm

Hello, Norm;

Quote:
They may have used a certain transistor to cause the blanking, and maybe that transistor gets fried from ESD. Maybe the last logic gate is hooked to a very large power transistor that runs the LED's. I'm not saying it DOES, I'm saying that it COULD be that way.

In this case, would the blank be selective or fail in any case the gate is activated?

Taht's what I mean. I did not design CMOS chips, but I also designed actual MOS-based circuits, and taking care of ESD 20 years ago was a matter of protection level and costs. Today, MOS protection is achieved inside chips.

Spices come from a time where MOS protection was achieved outside chips, and we cannot see much of this protection in the circuit board. In fact, there is nothing but chjips and power supply board.

I just wonder of the fact this "disease" is so selective. If blanking is the issue, and should it occur in many circumstances and not in only one, I'd fairly agree with you. I also strongly agree with the fact that Jon's experience is extremely revealing and adds lots of new information. When I read his post I thought exactly what you wrote, but I was not completely sure if this defective condition woudl be selective as well.

(additional part)

UNLESS there are additional gate and MOS transistor just to blank this particular blinking in new units and they are easily burnt, what would cause exactly this behavior and nothing else. Then we have a disease that allows a firmware problem if "blanking" hardware fails (Wow! Where did I gat this?)

(finish additional part)

So, my friend, I was just enhancing what I see against what you see. I agree with you, maybe I'm having my own doubts and I should consider them in another perspective. Do you understand this? In Brazil we have the expression "Devil's Lawyer" meaning the guy that's always trying to find "reasonable doubts"... The pri..y scumbag that never allows others to agree with anything. I'm feeling as if I'm playing this rule right now... :-(

Best regards, my friend.

Luiz C. Vieira - Brazil

Edited: 26 June 2003, 4:04 p.m.

                                                      
Re: Luiz, I used to design CMOS chips (edited)
Message #10 Posted by Jon on 26 June 2003, 5:13 p.m.,
in response to message #9 by Vieira, Luiz C. (Brazil)

Hello everybody:

I donīt understand anything about electronics but I have repaired the battery contacts of several spices and I can say that when I repair a calculator that nobody has opened they normally donīt have the "2 disease". When I see that the calculator has been serviced by unexperienced people (itīs easy to see it) they always have that problem (four calculators, 100% of my experience, in that condition of different models 37E, 34C, 33E and 32E). Has anybody similar experience?

Cheers

Jon

                                                            
A few questions:
Message #11 Posted by Vieira, Luiz C. (Brazil) on 26 June 2003, 5:25 p.m.,
in response to message #10 by Jon

Hi, Jon;

once again, thak you for your valuable contribution. May I ask you a few questions?

1 - except for your own Spice, that became sick after being serviced, did you check if any other one became sick as well?

2 - are these serviced units solderless or soldered type?

3 - Did you observe if they have a different number (quantity) of 8-pin DIP IC's?

We are trying to track down the possible origin of this "2" disease, as you have read, and I see one thing and Norm sees other thing. We're lucky because this is never going to interfere in our long lasting friendship... will it, Norm?

That's the biggest advantage of having different views of the same problem when brainy, reasonable and thoughtful guys are involved: everyone learns a bit more. If it not for Norm, this "2" disease was never an issue in here. Thanks to him we're involved in this "crusade".

Best regards and thanks again, Jon.

Luiz C. Vieira - Brazil

                                                                  
Re: A few questions:
Message #12 Posted by Jon on 27 June 2003, 11:57 a.m.,
in response to message #11 by Vieira, Luiz C. (Brazil)

Luiz:

1) I have that experience just with my own calculator 2) All of them where solderless 3) The 34C had three chips, 33E and 37E had two and I am not sure about the 32E.

I think I took notes of the identification number of some of them.

Make me know if you need them

Cheers

Jon

                                                                        
If you can, please...
Message #13 Posted by Vieira, Luiz C. (Brazil) on 27 June 2003, 3:44 p.m.,
in response to message #12 by Jon

Hey, Jim; thank you.

I would like to compare the ones you noted with the ones I have. If others who own Spices can also post their ID # and add the fact they have the "2"disease or not, maybe we can find a "common place".

Our crusade starts now... Come you all!

Luiz C. Vieira - Brazil

                                                                              
Here you have the id numbers
Message #14 Posted by Jon on 28 June 2003, 10:59 a.m.,
in response to message #13 by Vieira, Luiz C. (Brazil)

Hello Luiz:

These are the identification numbers of the chips. As you know there are three holes and I have represented as 1) the place closest to the display, then 2) and 3).

With “2disease” calculators

HP-33E (my own one, in the beginning it was ok) 1) empty 2) 80212 3) 79384

HP-37E 1) empty 2) 80012 3) 790228 (may be 79228)

HP-34C 1) 79364 2) 79334 3) 79324

I can’t find any note of the 32E

I have notes also of two calcs that hadn’t this problem (they had never been serviced).

HP-38C 1) 81221 2) 81212 3) empty

HP-33C 1) 81261 2) empty 3) empty

And I can’t find more notes

Cheers

Jon

                                                                                    
Thank you, Jon! ... and Hey, Norm!
Message #15 Posted by Vieira, Luiz C. (Brazil) on 28 June 2003, 11:33 a.m.,
in response to message #14 by Jon

Hi, Jon;

first of all, forgive me using another name instead of yours in my last post; I thought I should ask for excuses and not to edit it.

Second: I can take the numbers of at least eight Spices, some with the "2" disease, others without. I'm posting later because I'll first collect all data then I'll put it all together.

Our crusade is going fine...

Hey, Norm; can you add some ID # too? Let's put all we can at once!

Best regards.

Luiz C. Vieira - Brazil

                                                                                          
Hello Luiz,
Message #16 Posted by Norm on 2 July 2003, 5:27 p.m.,
in response to message #15 by Vieira, Luiz C. (Brazil)

Hello Luiz,

I am glad you are still trying to understand "2 disease".

I believe it is still the #1 predator and destroyer of the limited population of spice calculators.

I have a single SPICE calculator that has "2 disease". Unfortunately it is fully assembled. Please do not think me lazy if I dont take it apart right now to get numbers, but its a big job to take apart and not right now. I will take it apart after it has a rubber keypad ready for it.

It's serial number is 2051S37556 .

Thank you Luiz for sending the black rubber key pad. I have already taken it to an industrial rubber supplier, though they could not source it exactly. I notice that the smell does not match to rubber. Rubber is stinky, the HP item is not. I think it would be good to get an exact match.

Now I am going to mail it to a company in Minnessota and see if they can match it.

GOOD NEWS.... I do know how to PUNCH HOLES into these materials. U go to the dentist and you beg him for a "dental dam punch". They use rubber sheets that they stick over the tooth while they work on it. The dental dam punch will punch a little hole in the rubber, and its about the right size. Oh, the dental dam material? Not bad, but its too sticky (no white powder) and it stinks too much.

I hope to have more information in another couple weeks. Im not taking the 34C apart yet, because although I am in good pursuit of the answer, I still dont have replacement black rubber key cushions. - Norm

                                                                                                
You're welcome.
Message #17 Posted by Vieira, Luiz C. (Brazil) on 2 July 2003, 11:56 p.m.,
in response to message #16 by Norm

Hey, Norm;

thank you confirming the rubber pad is already with you. As you mentioned about the smell (I never noticed, but you're right: it smells almost nothing) and you're suspicious it's not a rubber, what about a blend of different, raw material? There are some sort of mixed material fairly used today, maybe there is something that resembles it.

Another fact: there is another type of actual rubber pad used in soldered-type spices. It's thin and has thicker caps molded under each key. I thought about sending this one too, but it is different of the one you needed, that is used in the solderless Spices.

About the "2" disease, you're right: I should reason about all aspects, and the fact that an extra "blanking" circuit was added later and this extra MOS is easily burnt out is an issue, of course. I was observing something that I'm not sure if is a valid fact; if you have the means to check if it is true, I'd be glad. The LED's blanking time after a digit-entry key is pressed in a "2"-disease immune unit is larger than the time observed in a "2"-disease "infected" unit in the same circumstances, i.e., infected units' display blinks faster when digits are entered (did I write it correctly?). If this is considered electronically, I can see two possible causes: there is an extra circuit to ensure the extended "delay", or it is augmented by SW or frimware.

I was beginning to think the "2"-disease was no longer an issue because only Ellis Easley continued the thread I asked for IC's ID #. I know that opening Spices is not a "piece o'cake" operation; and I am sure many interested users are waiting till the chance the calculator needs to be opened so they can check for serial numbers. Then I thought about asking for another request: would Spice owners check for construction type, serial # (first four digits and origin identifier), "2" disease occurrence and, if applicable, any other "misoperation"? 'That would be a starting point, at least to "count" as many "2"-disease occurrences as possible. This way, there's no further need to open the calculator and we can track if after an specific "datecode" for each Spice type (and manufacturing location) the "2"-disease appeared. And Norm, not finding your name in related threads seems to let them meaningless, if you understand me. You announced the "2"-disease, I just suggested the name. And everybody here knows you are a LED/34C addicted :)

I think that's all for now. And Norm, the keypad is a gift, remember? If you send it back to me I'll be p... off! B^(

Best regards, my friend.

Luiz (Brazil)

Edited: 3 July 2003, 12:43 a.m.

                                                                                                      
under the microscope
Message #18 Posted by Norm on 3 July 2003, 4:18 a.m.,
in response to message #17 by Vieira, Luiz C. (Brazil)

by the way , Luiz, the black rubber keypad piece is interesting under the microscope. I have just a 7 to 30 power stereozoom microscope. Very nice for modest but detailed inspection needs.

The white dust is very significant. There are a LOT of little white flecks, and they are embedded quite well into the black material.

The rubber material has "talcum powder" on it. This so its not too sticky. But the talcum powder comes off very easy, and there isn't as much of it. Under the microscope, the flakes look about the same.

It's just that the talcum powder is deeply embedded in the genuine material. I think this is so that it remains slippery.

I am sure we will find the match still. And get some bulk material, so we can punch holes in them with the dental-dam-punch.

                                                                                                            
brief comment (and brief post)
Message #19 Posted by Vieira, Luiz C. (Brazil) on 3 July 2003, 6:51 a.m.,
in response to message #18 by Norm

Hi, Norm;

just passed through my mind: would this talcum powder be a sort of solid lubricating so the rubber will resist friction activity and rests longer? Voyagers' ordinary rubber pad gets rough easily, maybe this is the key.

Just a brief thought. Maybe other sort of material with embeeded solid lubricating would fit fine.

Luiz (Brazil)

                                                                                                
2051S 34C
Message #20 Posted by Ellis Easley on 3 July 2003, 10:18 a.m.,
in response to message #16 by Norm

My 2051S unit is soldered, has no "2" behavior, under the keyboard it has the rubber sheet with the disks molded in for each key. The rubber under the decimal point key has been cut through by the key.

                                                                                                      
cut thru at the decimal point
Message #21 Posted by Norm on 3 July 2003, 1:01 p.m.,
in response to message #20 by Ellis Easley

This key-pad layer sounds like the orange-yellow one.

It's made of cellular foam, with molded-in high-spots at each key.

There are at least 2 types shipped with spice calculators, the orange-yellow one, and the one of black rubbery material.

I was immediately concerned that the cellular foam material hasn't got enough strength at the point-contact, where the pressure is too high. Your statement makes this concern more genuine.

The orange-yellow will be great on day #1 when U open the box, but after that, it will 'go away' where its needed, at the point of contact where the pressure is high.

Therefore, obviously, the black rubbery material is stronger and a better selection.

Disappointing if they switched to the orange-yellow material last. This shows a continuum of bad engineering decisions with our almighty 1970's HP Gods (solderless circuit boards, "2 disease", orange-yellow key-cushion material that is destined to fail, etc etc).

                                                                                                            
Re: cut thru at the decimal point
Message #22 Posted by Ellis Easley on 3 July 2003, 2:16 p.m.,
in response to message #21 by Norm

No, mine has the dense black rubbery material. I think its original owner just used a lot of fractional numbers! But I think the differences between my 2051S unit and yours are interesting: mine has no "2" behavior and yours does; mine has the rubber pad and yours doesn't. Is yours a soldered unit? Mine is. If yours is not a soldered unit, has no rubber pad and has the "2" behavior due to different firmware and not chip degradation, then why do these two units have the same S/N prefix? We know the YYMM datecode is not exact but rather should reflect a revision level. How can these two units be so different? Is it possible a previous owner swapped insides with another unit? (I'll allow that possibility on my unit, I can't vouch for it before I got it!) Of course, it could be that the insides were swapped by HP, during repair. All I can say about mine is that the datecodes on the chips are in line with the datecode in the S/N prefix (8047, 80282,8033?,80182), at least it would appear none of those chips were made after the 51st week of 1980.

                                                                                                                  
you lost me totally, ESD precautions for SPICE
Message #23 Posted by Norm on 3 July 2003, 6:26 p.m.,
in response to message #22 by Ellis Easley

you lost me totally, but I think I will take a moment and reply anyway, in order to maintain the flow of information.

I know there is a quest to determine which date codes have "2 disease" and which date codes don't etc etc and I am not objecting because you never know what you will learn. You might learn something interesting.

as to me, I dont have date codes available (no open units right now, said units having 2 disease).

Personally I think it is an irreversible symptom caused by some sort of ESD damage when we service them. I believe there is an extra-frail portion of the integrated circuit and it pops and the display doesn't mute properly after that.

It's not just "I think this" but also that other people report that they had a healthy unit which showed "2 disease" after they had serviced it.

THE WAY TO FIND OUT IS TO HAVE A HEALTHY UNIT, AND A DISEASED UNIT, and just swap chips and see which chip is responsible. I think this would be the best test right now.

WHERE THIS WILL GO IS THAT WE MUST USE EXTREME ESD PRECAUTIONS WHEN SERVICING A SPICE CALCULATOR. You would use aluminum foil underneath the service area, grounded out to an earth ground. YOU would be grounded out by stuffing aluminum foil in your socks, and connecting with alligator clips. YOUR SOLDER IRON must be grounded out, to the tip, and checked with an ohm-meter. ONLY THEN would you dare to open up a SPICE unit, else it will get "2 disease".

What I just described is an example of "hard grounding" where the metal is just fractions of an ohm. Some of that soft blue-plastic or pink-plastic anti-ESD bag material is also an option....... it is common to use meg-ohm resistors in the drain paths for ESD precautions. BUT IF IT WAS my calculator and my money (and it is, next time I open one) then I would favor the hard grounding thru aluminum foil.

And don't be doing all this if some dude with a ham radio antenna is going to key up and pump 20 watts into your workspace. You want to create a truly electrically neutral environment. The aluminum foil would be plentiful, the ground wires would be trustworthy and short. And it would all go to an earth ground, such as your building ground which presumably drives into a stake outside.

                                                                                                                        
the rubber key cushion for SPICE
Message #24 Posted by Norm on 3 July 2003, 6:33 p.m.,
in response to message #23 by Norm

There are really 2 big problems when dealing with SPICE calculators, IMHO. These problems are worth dealing with, because these are the coolest of the classic HP calculators IMHO (i know, i know, everybody else likes 41C, yeah yeah, fine, keeps demand down on the ones I like).

Problems with spice:

1. The risk of "2 disease" especially if triggered by a loose connection in the solderless units. Also, the need to have successful anti-ESD precautions when opening it, so that the calculator can be worked on without inducing "2 disease"

2. The calculators need a rubber key cushion inside, but many of them either have none, OR, they have got an orange/yellow foam cushion that isn't going to be as good as a more durable material.

Even the black rubber ones, I measured at .012" thick and this isn't very much material. Tentatively .022" thick would make for a gentler and quieter keypad.

WELL, typical Neoprene rubber isn't a good substitute (like a tire innertube). It's stinky, and the talcum powder isn't bonded permanently like the real HP material. The real HP material has the dry lubricant embedded in real deep, as seen under the microscope.

Somebody said go to Home Repot and buy pond-liner. Well, bad news, the only stuff I saw was black PVC material, which really isn't very similar at all. And worse than that, I had to buy 10' x 10' in a box that included a ceramic frog with a spout in his mouth, and an electric water pump.

HERE'S THE LATEST UPDATE: I am approaching a company that routinely makes cut rubber parts, I am asking them to find a genuine match to the material, AND, I am even considering purchasing of tooling so that they could stamp me a few hundred and be done with it.

Don't rule this out, guys, a few hundred bucks and you can have whatever rubber HP parts your heart desires. You don't have to sit there hand-crafting rubber feet that look crummy when you are done. Pay the money and get the real thing. And after a year or 2, make the tooling money back by selling these parts to those who need them on the HP Museum.

- Norm

                                                                                    
Re: Here you have the id numbers
Message #25 Posted by David Smith on 28 June 2003, 2:44 p.m.,
in response to message #14 by Jon

The numbers that you gave are manufacturing date codes. The chip numbers should be something like ILA010, ILN123, etc.

                                                                                          
Re: Here you have the id numbers
Message #26 Posted by Jon on 29 June 2003, 2:39 p.m.,
in response to message #25 by David Smith

Oh!. I took those numbers just to avoid mistakes when placing the ICs again after repairing the contacts. So it is a worthless information.

I will take the correct identification in the future. Thanks

Jon

                                          
Re: One isolated, confirmed case, so far...
Message #27 Posted by David Smith on 26 June 2003, 5:29 p.m.,
in response to message #7 by Vieira, Luiz C. (Brazil)

I would not recommend unsoldering the chips to try them in a different Spice machine. I have attempted several cross-species Spice transplants and some will work and others will not... even among swaps between the same species. I think that some firmware revisions may not be compatible with some CPU revisions. Also the Spice CPU chips are definitely not as cross-compatible as the Woodstock ACT chips.

                                                
support the war against "2 disease"
Message #28 Posted by Norm on 26 June 2003, 10:45 p.m.,
in response to message #27 by David Smith

Only you can help fight the war against "2 disease".

Across the country, precious spice calculators are being damaged and maimed by this terrible ailment, and so far, there is no cure.

But if you will pick up that telephone, and give generously, the doctors may yet be able to find a cure to "2 disease".

Give generously. The spice calculator you save, might be your own.

                                                      
Re: support the war against "2 disease" Plus Calculator Commercials
Message #29 Posted by Michael F. Coyle on 27 June 2003, 10:25 p.m.,
in response to message #28 by Norm

Quote:
Only you can help fight the war against "2 disease".

Across the country, precious spice calculators are being damaged and maimed by this terrible ailment, and so far, there is no cure.


Great. All we need now is Sally Struthers to do the commercial...

"This neglected calculator is sitting in an attic...won't you help? For only pennies a day you can give it the batteries and health care it needs to become a useful, productive member of society. Twice a year you'll get your very own printout from your calculator. Adopt either an HP or a TI. Please help save a poor hungry calculator today!"

And on a slightly more serious note...

Back in the 1970's Rockwell ran a radio ad for their calculators. It was done by the "Dick and Bert" team who did many hilarious commercials at the time. The commercial was based on a (sung) jingle, which to the best of my recollection went something like:

"Rockwell Calculators / They really can't be beat / They've got big green numbers / And little rubber feet."

Does anyone else remember this commercial? And has any other calculator been the subject of a radio or TV ad?

- Michael

                                                            
big green numbers and little rubber feet
Message #30 Posted by Norm on 28 June 2003, 1:30 a.m.,
in response to message #29 by Michael F. Coyle

yeah there was a song, like a happy-go-lucky little radio jingle.

kind of like some college guys singing their campus fight song ...

"its got big green numbers ... and little rubber feet"

See, I think they knew more about making calculators THEN than they know about it now.

What's happening presently, is the people at the company don't use their own product. So what do they care if the little rubber feet stick to the desk, or not. What do they care if the display has big clear numbers or not. They dont use it so they don't care.

I was out changing the oil on my 8 horsepower Bearcat chipper. Now, the thing is, the Briggs & Stratton engine doesn't quite mate to the mount nicely, and w/o some extra plumbing (a horizontal 4" long pipe stub) you wouldn't be able to drain the oil out of the engine without it clearing the engine mount.

So from the factory, these bozo's put the 4" pipe stub, with a REDUCER pipe so it has almost no flow, and then they put a cap on the end.

So..... go ahead and try removing the cap.... the whole pipe turns instead !!!! useless. They don't even use their own product, thats what I am getting at. The only cure was to go on down to the hydraulic pipe shop and buy fancy hydraulic fittings to redo the job PROPERLY with aftermarket fittings. This is the total antithesis of mass production, and the dumb part is, these lunacies are multiplying daily.

People should use their own products, thats why they realized that it should have little rubber feet, and a big display easy to read. Presently, our CEO MBA types would think that insane. I know, I have met these people personally. They are out of their skulls. They dont think it matters whether the product is useful or not. They only think in terms of satisfying a certain amount of product to the market at a certain price.

They count on Joe Six Pack to buy it w/o realizing its junk, and then he gets to put up with it until he chucks it out (because it DIDNT have little rubber feet, or because you can't change the oil w/o repiping the pipe stub, etc etc)

the answer here is to REFUSE to buy product that is not correctly designed, and the trouble is there is so much affluence and way too much production in our culture, that people don't refuse, they buy it anyway even if its junk. So stop doing that, its a bad habit.

                                                                  
Re: big green numbers and little rubber feet
Message #31 Posted by Ted on 28 June 2003, 10:18 a.m.,
in response to message #30 by Norm

Dear Norm, I always read your tirades with tremendous pleasure, because I generally agree with you all the time. The problem is, that it is much worse than even you portray it. I thought it was mostly me, but the business world seems to be populated by idiots on all levels- and it is getting worse each year. There is a saying here in Holland for it: "you have to be light enough enough to float" i.e.: if you know too much, you'll never get to the top. My neighbor is a management expert, and earns tons coaching companies about restructuring, etc. One day he proudly showed me a powerpoint presentation of one of his classes. "We make products for the clients" Happy faces...."Our goal is to make good products" people hard at work. He gets paid $ 1,500 a day to teach this stuff, can you believe it? It was unbeleivebly depressing. Companies don't make products for customers anymore, their only goal is to make their stocks go up before the year's bookkeeping closes, and then worry about the next year. The HP's we all love are icons not only of their kind, but represent the very best of human effort to be productive, be the best and improve/ add to the world around us. It sounds a bit too lofty, perhaps. But that is my personal opinion, for what it's worth.

                                                                        
yeah i see what U are saying, Ted,
Message #32 Posted by Norm on 28 June 2003, 2:55 p.m.,
in response to message #31 by Ted

yeah, i see what U are saying, Ted.

They bring in some consultant, who is kind of like a teacher from kindergarten. Pulls up a powerpoint spread sheet on some digital projector, and a bunch of dumb color slides, and says 'we have to meet the demographics of our marketplace'
we need bright smiley faces and happy people

and its just so much horse manure. Yeah I do know what U are saying. I was saying they don't use their own products (they don't) so they don't care if it works. U are saying they do it all with microsoft color slides and a projector.

Yeah it is totally depressing, because they are idiots and also because they have all the money. I just say 'these are not the people who could build and fly a rocket to the moon'.

Enjoyed your comment 'you have to be light enough to float'. Don't think the original Mr. Hewlett and Mr. Packard would agree with that, not in the least. They wanted intelligent technical people at the top. JAPAN has a societal culture that prizes engineering, and their companies have technically competent people near the top......that's why they do all the innovating now (imagine Carly with a business model that revolved around building 6' flat screen TV's for the world market... requiring massive vision, planning, and world-class technologies and clean rooms, etc etc hah! She just wants to sell junk PC's like any other bunch of MBA idiots in town are doing.

We should export our MBA's with their powerpoint slides and their happy smiley faces and their 'market demographics' to Japan so that they would become infected with the disease, reduce their economy to 3rd world, so that we would no longer have a trade deficit with Japan.

                                                                              
Re: yeah i see what U are saying, Ted,
Message #33 Posted by Ellis Easley on 28 June 2003, 3:08 p.m.,
in response to message #32 by Norm

Norm, are you aware that Japan's economy has been in recession for ten years? I that the fault of American MBA's?

                                                                              
Re: yeah i see what U are saying, Ted,
Message #34 Posted by Ted on 28 June 2003, 3:17 p.m.,
in response to message #32 by Norm

Better English translation of that Dutch phrase: 'the 'light' people get to the top' i.e. if you are really good, they get scared of you, and don't allow you a position of any importance. The airheads get to high positions. HP was just the opposite thank god!

                                                                                    
light people to the top
Message #35 Posted by Ellis Easley on 28 June 2003, 3:35 p.m.,
in response to message #34 by Ted

In English there is a saying, "Cream rises to the top." Now, cream does that because the fat content makes it less dense, but the meaning of the old saying is that just as cream is better (more valuable, more nutritious in the old sense), "better" (more capable) people rise to positions of authority. Nowadays, skim milk is better for you!

I understand that in Japan they say, "The nail that sticks up is hammered down." This saying is supposed to encourage people to conform - the last thing one is supposed to do is stand out! But of course, even in Japan, somebody ends up in charge. BTW, I thought in Japan they built houses, at least, without nails?!!!

                                                                  
It's my deal, Carly!
Message #36 Posted by Ellis Easley on 28 June 2003, 3:11 p.m.,
in response to message #30 by Norm

Am I the only person not surprised to hear that Norm has a Fargo-syle wood chipper? Just kidding, Norm, I envy you - I'm trying to get along with a little electric chipper-shreddder - emphasis on the "shredder".

                                                                        
Bearcat chipper
Message #37 Posted by Norm on 28 June 2003, 8:01 p.m.,
in response to message #36 by Ellis Easley

Are U telling me that U are running sticks and branches thru your office paper shredder ? Better quit or it wont last very long.

Definitely recommend something with an 8 horsepower motor, its enough for backyard work. Its painful how much it cost. Like $1600 and then add on the 'double V-belt option' so that U got 2 belts not one so the belt wont burn, there goes another $150 and then add on sales tax.

I mean what, I can't drive this thing on the freeway but it costs more than a decent used car.

It helps a little if U buy it in a no-sales-tax state.

I'd say just bite the bullet and get one. As long as you have a significant source of branches to grind up, then you wont ever have to buy beauty bark again. I dare say the output looks prettier than beauty bark. That's because its a golden brown color with white flecks, whereas beauty bark is like coffee grounds. You get some of your money back by not buying beauty bark anymore.

                                                                              
Re: Bearcat chipper - an additional use
Message #38 Posted by Dave Shaffer on 29 June 2003, 10:54 p.m.,
in response to message #37 by Norm

Is the chipper an acceptable alternative for calculator "fixing" to the railroad tracks? (Calculator in, plastic mulch out)

                                                                                    
Re: Bearcat chipper - repair method
Message #39 Posted by Norm on 30 June 2003, 2:04 a.m.,
in response to message #38 by Dave Shaffer

Hello Dave Shaffer, I have never tried repairing a calculator with my 8 horsepower Bearcat Chipper.

However, I think this is an outstanding idea, and I would be very agreeable to trying out repairing calculators in this way.

I would like for HP to send me 30 to 40 of the HP-49G. I will personally repair each one using the Bearcat Chipper. At the outlet, I should be able to get beautiful calculator mulch. This mulch could be used to decorate around mailbox posts, along walkways, and next to the streetlamp posts.

I would think the mixture of tinfoil, bits of fiberglass, calculator buttons, and chunks of plastic would add exceptional beauty to any landscaping program.

ONE THING THOUGH.... be sure to install the finer sized chipper screen .... don't want any big chunks getting thru (like the entire corner of a 49G could be unsightly).

I have also heard that the mafia uses chippers to 'repair' people, such as those who owed money and could not pay it back.

- Norm


[ Return to Index | Top of Index ]

Go back to the main exhibit hall