WP 34S will not turn on

[brettpim]

I made a WP 34S a few years ago. I went to use it recently after it being in my drawer for over a year. It would not turn on. I have put in new batteries and still no luck. I have searched the forums here for similar problems and looked in the troubleshooting part of the manual but found nothing. Does anyone have suggestions?

[BarryMead]

When the batteries get low the microprocessor that controls the calculator can get
confused. Did you press the reset button (behind the little hole) when you put in
the new batteries?

[brettpim]

I have tried that but it has not helped. On page 148 of the 3.1 manual it says that the ON button "shall turn it on, as long as the batteries are good – even if you do not see anything on the LCD." If it is indeed turning on, what could cause the LCD to not show anything if it has simply been sitting in a drawer in its case?

[BarryMead]

I don't know how your situation could arise. If you don't want to discard the calculator, then the next course of action is to find a friend with a flashing cable
and attempt to re-flash the firmware from a cold start using the button sequences on
the flashing instructions in the manual. Since you can't be sure of the current draw of the calculator when the display isn't working it would be a good idea to attempt the flashing procedure with a completely FRESH set of batteries.

[brettpim]

I have PL2303 USB-Serial Port adapter which I used to flash the WP 34S originally. I am powering the calculator with a fixed voltage power external supply at 3V. I have used the reset procedure on page 150 of the manual to erase the firmware and have then tried flashing many times but with no success. I am using

wp34sflashcmd calc_xtal_full.bin /dev/ttyUSB0

with read/write permissions for ttyUSB0

The calculator draws 130 mA as soon as I attach the power supply and pressing the on button does not change this, nor does resetting the calculator. When I short the Vcc and the Erase terminals from the 6 pads the calculator then draws around 730 mA and again this does not change when I press rest nor when I press the on button. Is it correct that the amperage does not change whether the calculator is on or off?

Is there some way to verify if my PL2303 is functioning?

Is there some way to determine the internal state of the calculator?

[brettpim]

I got my hands on a second USB to serial adapter and tried again with that and still no success. I was able to test the adapters in two ways:

1) using RS232 mini-tester I saw the transmit light go on when I ran

wp34sflashcmd calc_xtal_full.bin /dev/ttyUSB0

2) using both cables, I was able to send and recieve signals miniterm.py

So I think the USB to serial adapters are good which would seem to indicate that it is the calculator that is having problems. Does anyone have ideas on how to start troubleshooting the calculator?

thanks

[BarryMead]

(12-05-2017 05:51 PM)brettpim Wrote:  I have PL2303 USB-Serial Port adapter which I used to flash the WP 34S originally. I am powering the calculator with a fixed voltage power external supply at 3V. I have used the reset procedure on page 150 of the manual to erase the firmware and have then tried flashing many times but with no success. I am using

wp34sflashcmd calc_xtal_full.bin /dev/ttyUSB0

with read/write permissions for ttyUSB0

The calculator draws 130 mA as soon as I attach the power supply and pressing the on button does not change this, nor does resetting the calculator. When I short the Vcc and the Erase terminals from the 6 pads the calculator then draws around 730 mA and again this does not change when I press rest nor when I press the on button. Is it correct that the amperage does not change whether the calculator is on or off?

Is there some way to verify if my PL2303 is functioning?

Is there some way to determine the internal state of the calculator?

The current draw of the calculator should go way down to a few micro amps when you press the reset button, and not go into the mA range until after you press the "ON" button. It seems clear that the "Boot Loader" portion of the firmware is corrupted, and that the only way you will be able to resurrect your calculator is to use a hardware flashing device known as a jtag programmer. One can be purchased for about $43 US
and I found a source a while back on this forum. The device plugs into and gets its power from a standard type B USB Printer Cable, and attaches to the JTAG pins of the calculator. I don't own one of these devices, but others
have reported being able to fix dead calculators (those with corrupted bootloader firmware) using this device. The device has other uses as well. It allows you to hardware debug programs and set breakpoints etc.
I found the web page of where to order one I was thinking of. Here it is:

A company named OLIMEX makes a cheaper open-source software version of the ARM JTAG programmer that sells for only $43.17 (39.95 euros). It can be ordered from this web site:
https://www.olimex.com/Products/ARM/JTAG/ARM-USB-TINY/

You may also need the ARM-JTAG-20-10 Accessory to adapt the JTAG pin count to that of the calculator. This accessory costs $4.95 more.

[brettpim]

(12-06-2017 12:56 AM)BarryMead Wrote:  It seems clear that the "Boot Loader" portion of the firmware is corrupted, and that the only way you will be able to resurrect your calculator is to use a hardware flashing device known as a jtag programmer.

OK, thanks for the diagnosis. I will begin reading about recovering from a corrupted bootloader.

[brettpim]

(12-06-2017 12:56 AM)BarryMead Wrote:  others have reported being able to fix dead calculators (those with corrupted bootloader firmware) using this device.

I have searched the forum for combinations of "corrupted", "bootloader", "JTAG", "WP 34S" but I have not found any discussion of this procedure. Are there resources you know that discuss this?

thanks

[rprosperi]

(12-06-2017 02:51 AM)brettpim Wrote:  

(12-06-2017 12:56 AM)BarryMead Wrote:  others have reported being able to fix dead calculators (those with corrupted bootloader firmware) using this device.

I have searched the forum for combinations of "corrupted", "bootloader", "JTAG", "WP 34S" but I have not found any discussion of this procedure. Are there resources you know that discuss this?

thanks

Did you search with the forum search tool or google?

The forum search tool is all but useless, so just compose your search and add "site:hpmuseum.org" (without the quotes) to limit google's search to just this domain, you'll get much better results.

[jebem]

I searched like this:
JTAG 30B 20B 34S 31S site:www.hpmuseum.org

and it came out with a few threads on 20B/30B machines:
HP30b flashed with WP31 no longer responding
Fully-modified 20b
HP30B/20B Repurposing Project
[WP34S] Flashing Issues
[WP 34S] I finally bricked it

In my humble opinion, unless one is willing to deep dive on these micro-controller chips and therefore spend money on a JATG developer system, probably it would cost less money (and time) to just buy another calculator.


On a different note concerning HP 30B current consumption:
The values you measured are way above what is expected here.
Maximum current values when operating or when flashing the machine are just a little above 20mA (0.02 Amp), no matter what state the calculator is in.

If you measured currents in the hundreds of milliampere, then definitely one or more components are defective in the calculator. Those current values are like a short circuit for any CR2032 coin cell battery.

[BarryMead]

(12-06-2017 12:08 PM)jebem Wrote:  If you measured currents in the hundreds of milliampere, then definitely one or more components are defective in the calculator. Those current values are like a short circuit for any CR2032 coin cell battery.

I believe jebem is right. If I remember correctly the HP-30B will have a few micro amps of current draw after pressing reset, but only about 8 mA after you press
the "ON" button if the boot loader is intact. I have never seen current draws as high as you report even on calculators with corrupted boot loaders. It is entirely possible that your calculator is damaged beyond any means of repair even with a JTAG programmer. I can't imagine any "NORMAL" state the calculator could get into that would draw 130 mA let alone 730 mA. Perhaps you should ignore my advice and listen to jebem (buy a new calculator). It may be more trouble than it is worth to try to repair or recover your old one.

[BarryMead]

You had asked a question about the WP-34S (HP-30B) boot loader. The boot loader on this device is not located in a "PROTECTED" area of flash memory as it is with many other flash programmable devices. Unfortunately the boot loader is merely a section of the normal flash memory. Each time you "PROGRAM" the flash memory of the HP-30B (WP-34S), you are overwriting the boot loader as well as the rest of the flash memory that define the behavior of the calculator. This is why "EACH ATTEMPT" to update the firmware has a risk of bricking the device, and why it is so critical that the battery be new, and the cables be wiggle free and solid. If you attempt to program the flash memory and the process is interrupted, you will end up with a bricked calculator. So the only thing special about the JTAG method of programming flash memory is that it can be achieved even without the cooperation of boot loader firmware which reads the serial port interface. On the Swiss Micros DM-15L the boot loader is in a special "PROTECTED" area of memory that cannot be erased using the USB serial interface, so one cannot brick the calculator by a failed attempt to program flash memory as you can with the WP-34S. This is all a moot point if your calculator has a high-current lockup hardware problem, as even a JTAG programmer will probably not be able to fix the issue. It is also worth noting that using such a programmer requires a considerable learning curve and skill level that may not be easy for most users of the Calculator. I happen to be a digital electronics engineer and software developer, who uses these kind of programming devices every day (for Atmel processors not arm processors), so I may have made it seem easier than it really is to use such equipment. For that I apologize. I attempted to be helpful, but I do not want to create financial or technical hardships for others.

[BarryMead]

Latch UP. Most modern CMOS, and NMOS devices have what are called "Input Protection Diodes" to prevent electrostatic shock from damaging the circuitry. These diode pairs provide a low impedance path (diode on) when the voltage at an input goes above VCC or below GND. The configuration of these diode pairs is very much like that of an SCR (Silicon Controlled Rectifier) and can "Latch Up" into a high current state if the battery voltage ever falls below 2.3 volts (when a 3 Volt input exceeds VCC by .6 to .7 volts). Anyway if the calculator ever gets into this "Latch Up" state the only way to get it out of this state is to remove the batteries and short the VCC and GND pins of the battery together for a few seconds to bleed off any residual charge held in capacitors.
It might be worth a try to remove all power from the calculator and short out the battery terminals for a few seconds before attempting to re-apply power. If the current draw returns to a few micro amps then you may yet be able to salvage your calculator.
Although even then it might still be less trouble to ship it to someone who owns a JTAG programmer or buy a new one. Best of luck, Barry

[jebem]

Thank you for sharing your good knowledge here, Barry.
Very detailed and useful and helpful information.

I didn't meant to contradict you with my post, but just complement it as i don't know what is the OP background knowledge in electronics. Indeed he mentioned to have used an external fixed 3V power supply but it is easy to have more than that applied to the machine depending on the quality or precision of the used equipment.

Again, thank you from my side.
I allways enjoy and learn from your posts.

[brettpim]

I would say that I am of medium comfort/skill with electronics. I own the external power supply because I have an old microscope that was missing its power supply, but some years ago I did solder a JTAG interface onto a wifi access-point and rescued it from bad firmware so I might be up for this project.

I did just buy some custom USB boards for the WP34S and used one to make a WP34S from another HP-30b I bought. Would I be able to a USB board to fix the board or does it really require a JTAG?

thanks
brett

[marsman2020]

(02-28-2020 12:59 AM)brettpim Wrote:  I did just buy some custom USB boards for the WP34S and used one to make a WP34S from another HP-30b I bought. Would I be able to a USB board to fix the board or does it really require a JTAG?

thanks
brett

The USB board is an FTDI USB-Serial adapter, similar to the Prolific unit you already have. It attaches to the serial port on the microcontroller, not the JTAG port.

There are lots of inexpensive JTAG hosts available these days, all the way down to using the $5 Raspberry Pi Zero as a host - https://github.com/synthetos/PiOCD/wiki/...TAG-Dongle

I don't have any direct experience using JTAG with the HP30b though.