WP34s flashing procedure

[Dieter]

Since late 2012 I am using a HP30b with WP34s firmware. Updates to current versions work without any problems, I am using the genuine HP programming cable with a computer running XP SP3, equipped with a serial port. The update procedure is performed as described in the 3.2 manual.

However, I had a problem when I tried to flash the 30b for the first time, following the respective procedure. It simply did not work, and the problem has been discussed in a thread in the old forum. At that time, Marcus von Cube was so kind to flash the device for me. Since then, all further updates worked fine.

Now I was trying to flash a new 20b – and I encountered the same problems as more than two years ago. I tend to think it's not the cable since it works fine for firmware updates, and also communication between the emulator and the physical device (transferring programs and data between both) works flawlessly using this cable. So I wonder if there may be an error in the procedure described in the manual:

Quote:A1. Remove the black battery door of your calculator. Connect the programming cable to the programming port of the calculator and to your computer.
A2. Start MySamBa. Pick the port you use for your cable and select the file you want to transmit. The window may look like: (picture)
Do not start sending yet! Just leave the window open.
A3. Press ON/CE to turn the calculator on.
A4. Hold down the ERASE button on the cable (do not release it until step A7).
A5. Press the RESET button on the cable. It will turn the calculator off.
A6. Press ON/CE to turn the calculator on again.
A7. Release ERASE now. Press RESET to turn the calculator off again.
A8. Press ON/CE to turn the calculator on again. It will look dead. Do not worry – see the note above.
...

Everything works as described until (and including) step A7. Pressing RESET at that point indeed turns the calculator off. However, the following ON/CE in step A8 does not (!) make the calculator look dead – it simply turns on as usual.

So I still wonder if there is an error in the description, or if I am doing something wrong, or if there may be a problem with the cable although it works fine for all other purposes.

Any ideas?

Dieter

[matthiaspaul]

From your description it looks as if either the ERASE switch or the ERASE wire is broken. You could easily check this using a multimeter.

Greetings,

Matthias

[Dieter]

(02-26-2015 10:29 PM)matthiaspaul Wrote:  From your description it looks as if either the ERASE switch or the ERASE wire is broken. You could easily check this using a multimeter.

Fine. But how?

Dieter

[Marcus von Cube]

(02-26-2015 08:10 PM)Dieter Wrote:  

Quote:A1. Remove the black battery door of your calculator. Connect the programming cable to the programming port of the calculator and to your computer.
A2. Start MySamBa. Pick the port you use for your cable and select the file you want to transmit. The window may look like: (picture)
Do not start sending yet! Just leave the window open.
A3. Press ON/CE to turn the calculator on.
A4. Hold down the ERASE button on the cable (do not release it until step A7).
A5. Press the RESET button on the cable. It will turn the calculator off.
A6. Press ON/CE to turn the calculator on again.
A7. Release ERASE now. Press RESET to turn the calculator off again.
A8. Press ON/CE to turn the calculator on again. It will look dead. Do not worry – see the note above.
...

This looks overly complicated to me. I've programmed a bunch of calculators for other members of this board the following way:

1. Press and hold ERASE.
2. Press ON on the calculator.
3. Press RESET
4. Press ON again.
   

Steps 1 and 2 are enough to erase the flash and reset the boot bit. Step 3 turns the calculator off, step 4 turns it on again with the boot loader enabled.

[matthiaspaul]

(02-26-2015 10:35 PM)Dieter Wrote:  how?

Select "diode check" or "resistance measurement" and connect the multimeter to the programming cable's ERASE and VCC pins. The multimeter should display "open connection" or "infinity". If you then press and hold the ERASE button, the multimeter should indicate a short circuit (close to 0 ohm).

Greetings,

Matthias

EDIT: Changed GND to VCC. Thanks, Barry.

[Dieter]

(02-26-2015 10:53 PM)matthiaspaul Wrote:  Select "diode check" or "resistance measurement" and connect the multimeter to the programming cable's ERASE and GND pins.

I know how to handle a multimeter, but how should I know which of the six pins on the cable is the ERASE resp. GND pin? So which of the contacts should get connected if the ERASE button is pressed?

Dieter

[rprosperi]

(02-26-2015 11:05 PM)Dieter Wrote:  

(02-26-2015 10:53 PM)matthiaspaul Wrote:  Select "diode check" or "resistance measurement" and connect the multimeter to the programming cable's ERASE and GND pins.

I know how to handle a multimeter, but how should I know which of the six pins on the cable is the ERASE resp. GND pin? So which of the contacts should get connected if the ERASE button is pressed?

Dieter

Dieter - look here - this shows the port, but you can easily understand the corresponding cable pins

[Dieter]

(02-26-2015 11:09 PM)rprosperi Wrote:  Dieter - look here - this shows the port, but you can easily understand the corresponding cable pins

Thank you. I now see the pinout is also mentioned in the manual (p. 203 in edition 3.2). Now, if the ERASE button actually doesn't work (haven't tested this yet), is there another way to get the firmware into the 20b? Otherwise the cable seems to be fine: updates are no problem and the RESET button seems to work.

Dieter

[rprosperi]

(02-26-2015 11:17 PM)Dieter Wrote:  

(02-26-2015 11:09 PM)rprosperi Wrote:  Dieter - look here - this shows the port, but you can easily understand the corresponding cable pins

Thank you. I now see the pinout is also mentioned in the manual (p. 203 in edition 3.2). Now, if the ERASE button actually doesn't work (haven't tested this yet), is there another way to get the firmware into the 20b? Otherwise the cable seems to be fine: updates are no problem and the RESET button seems to work.

Dieter

I am no expert on this, but there was recently a discussion on this and think the gist was that erasing the flash is not needed every time, you can simply write the new version over the old, skipping the erase step, which I suppose is why can have success reflashing when you update your 34S.

The above may not be fully correct, but rest assured, if wrong, it will be corrected in a very few minutes. I love that about this place.

[John Galt]

(02-26-2015 11:17 PM)Dieter Wrote:  . Now, if the ERASE button actually doesn't work (haven't tested this yet), is there another way to get the firmware into the 20b?

Yes. Once the original firmware has been erased, it is no longer required to use the Erase function again. Updating firmware is described on p. 171 of Walter's printed manual (Firmware v. 3.2). No Erase required... unless something goes wrong. I don't know what could go wrong, or if the Erase function would help if it did.

Edit to add: Marcus von Cube explains here: http://www.hpmuseum.org/forum/thread-295...l#pid26371

[walter b]

(02-26-2015 10:52 PM)Marcus von Cube Wrote:  This looks overly complicated to me.

Maybe. But it covers everything.

d:-)

[BarryMead]

(02-26-2015 10:53 PM)matthiaspaul Wrote:  

(02-26-2015 10:35 PM)Dieter Wrote:  how?

Select "diode check" or "resistance measurement" and connect the multimeter to the programming cable's ERASE and GND pins. The multimeter should display "open connection" or "infinity". If you then press and hold the ERASE button, the multimeter should indicate a short circuit (close to 0 ohm).

Greetings,

Matthias

I believe that the "Erase" button shorts the "Erase" pin to "VCC" not "GND". At least that is what the instructions say in the manual.

[Dieter]

(02-27-2015 02:00 AM)John Galt Wrote:  Yes. Once the original firmware has been erased, it is no longer required to use the Erase function again. Updating firmware is described on p. 171 of Walter's printed manual (Firmware v. 3.2).

Yes, this works without any problems for updates. But you already said it: "once the original firmware has been erased".... It isn't in a brand new calculator that gets re-flashed for the first time. That's exactly my problem.

I'll check the cable later this weekend.

Dieter

[Dieter]

(02-26-2015 10:52 PM)Marcus von Cube Wrote:  This looks overly complicated to me. I've programmed a bunch of calculators for other members of this board the following way:

1. Press and hold ERASE.
2. Press ON on the calculator.
3. Press RESET
4. Press ON again.
   

Steps 1 and 2 are enough to erase the flash and reset the boot bit. Step 3 turns the calculator off, step 4 turns it on again with the boot loader enabled.

Just to be sure: RESET turns the calculator off, and pressing ON in step 4 turns it on again.
Should the display now stay blank or does it show "0.00" ?

Dieter

[walter b]

Dieter, nimm einfach Ablauf B (den fürs USB-Kabel). Dann bist du unabhängig von den Knöpfen an deinem Kabel.

(Just a little hint)

d:-)

[matthiaspaul]

(02-27-2015 03:49 AM)BarryMead Wrote:  I believe that the "Erase" button shorts the "Erase" pin to "VCC" not "GND". At least that is what the instructions say in the manual.

That's right! Thanks for the correction - this is also what is described in the processor's manual.

Greetings,

Matthias

[matthiaspaul]

(02-27-2015 07:12 AM)Dieter Wrote:  

(02-26-2015 10:52 PM)Marcus von Cube Wrote:  This looks overly complicated to me. I've programmed a bunch of calculators for other members of this board the following way:

1. Press and hold ERASE.
2. Press ON on the calculator.
3. Press RESET
4. Press ON again.
   

Steps 1 and 2 are enough to erase the flash and reset the boot bit. Step 3 turns the calculator off, step 4 turns it on again with the boot loader enabled.

Just to be sure: RESET turns the calculator off, and pressing ON in step 4 turns it on again.
Should the display now stay blank or does it show "0.00"?

If the erasure worked and no new firmware was flashed yet, it should stay blank, as there is no longer any firmware code around which could cause the display to show "0.00" - you just erased it.

The default boot loader is stored in ROM, not flash, and thus it cannot be erased. However, the default boot loader is generic to the processor and doesn't know anything about peripherals, therefore, it doesn't know that the processor is running in a calculator and that an LCD is attached, and even if it would, it would not know how to handle it.

Greetings,

Matthias

[MarkHaysHarris777]

Dieter, Walter is giving you good advice, use the 'update' sequence not the 'erase' sequence. In that way you need not worry about your cable buttons, whether the strobe is working, etc.

What you are going through is the primary reason for why I put the instrumentation into my flash-box. My unit did not update early this morning. But, I noticed that the Tx light was on solid... I said to myself, "that's not supposed to happen". The FTDI converter got honked and needed to be reset. Without the lights, I would just be guessing, "why didn't it work". I unplugged the cable and reseated it; voila... worked fine the next time after reset.

I have also had the problem where the ON switch did not make. In that situation the 'update' will fail, of course. Because the display is normally blank during this process without instrumentation you really don't know whether the beasty came on or not. When the unit comes on the 'reset' pin J32 goes high. If you monitor that pin somehow you will know whether your unit came 'ON' when you pressed the 'ON' button.

Cheers,
marcus

[MarkHaysHarris777]

(02-27-2015 10:57 AM)matthiaspaul Wrote:  

(02-27-2015 03:49 AM)BarryMead Wrote:  I believe that the "Erase" button shorts the "Erase" pin to "VCC" not "GND". At least that is what the instructions say in the manual.

That's right! Thanks for the correction - this is also what is described in the processor's manual.

The original cable 'shorts' the 'erase' pin to 'high' Vcc; but that's not the right way to think about it. What is supposed to happen is that the 'erase' pin gets a nice clean 'strobe' from 'low' to 'high' and if that doesn't happen (spikes, contact bounce, ESD... ) the worst thing that could happen is that the unit's I|O will get fried, otherwise the 'erase' will fail. Since using de-bounce circuitry my flash operations have become more reliable; not perfect mind you ... still 3 out of 4 isn't bad /

Cheers,
marcus

[Dieter]

Thank you very much to all for your suggestions. After reading your replies it seems that the following procedure should work even if the programming cable's ERASE button does not work:

1. Use a paper clip or something similar to connect the ERASE and VCC contact on the calculator's JTAG pad.
2. Press ON/CE to turn the device on.
3. Remove the paper clip and attach the programming cable.
4. Press the RESET button on the cable. The calculator should turn off.
5. Press ON/CE once again. The calculator should turn on, but the display will stay blank.
6. Transfer the new firmware with MySamba.
7. Press RESET on the cable, then disconnect it.
8. Press ON/CE on the calculator. It should wake up as a WP34s.
   

Do you think this will work?

Dieter