Threaded Mode | Linear Mode

Claudio L. · 09-12-2016, 08:14 PM

(09-12-2016 04:19 PM)Nigel (UK) Wrote: I've managed to obtain a second-hand HP-50g in decent condition to run newRPL on. I've flashed it and my HP-49g+ with the latest firmware.

The 49g+ power consumption is unchanged from my previous measurements - 4mA when turned off; about 42mA when the calculator is on.

The 50g draws \(29\,\mu\rm A\) when turned off. However, to my great surprise it draws 48 mA when turned on and doing nothing. Is this the expected power consumption? Has the latest firmware caused a problem, or do I have an "odd" machine?

Nigel (UK)

I didn't measure myself yet, but it's supposed to be around 12 mA, based on old posts.
In a couple of days I'll update the firmware, I've changed much of the power management of the SD card, hopefully that consumption will come down. There were a couple of bugs I fixed:
* The SD card controller was kept ON all the time, while it was supposed to be powered off when not in use.
* Some pins were incorrectly programmed thinking the controller was off, but the controller was working, this could've caused some current leaks.
* Now I'm sending ACMD 42 to the SD cards to disconnect an internal pull-up resistor on DAT3 that some (all?) cards have. This could've also caused some leaks.

Also, now SD card insertion/removal is detected and an interrupt is issued, so pins are reconfigured accordingly at the proper moment.

Stay tuned for the update.

JDW · 09-12-2016, 11:50 PM

When I flashed my 50g with newRPL (for the sake of fixing my SD card problem), I noticed that an icon flashes in the upper right corner all the time. But why? That takes CPU time. Why not kill the unneeded on-screen fluff and save even more current?

Also, why are so many icons gray and flickering in newRPL? I looks more like a problem than a feature.

Claudio L. · 09-13-2016, 02:29 AM

(09-12-2016 11:50 PM)JDW Wrote: When I flashed my 50g with newRPL (for the sake of fixing my SD card problem), I noticed that an icon flashes in the upper right corner all the time. But why? That takes CPU time. Why not kill the unneeded on-screen fluff and save even more current?

It's a heartbeat indicator, used for debugging, to know if the CPU crashed. As long as there's a heartbeat, the code is executing. It hardly takes any CPU to toggle the annunciator, but in any case it will be removed eventually, it's just a debugging aid for the time being.

(09-12-2016 11:50 PM)JDW Wrote: Also, why are so many icons gray and flickering in newRPL? I looks more like a problem than a feature.

They are gray because it uses grayscale mode. Grayscale is done in hardware by cycling the pixels on/off according to a certain duty cycle. This is all hardwired in the LCD controller and I can't change it.
The flicker I guess depends on your eyes, the type of environment light you have and perhaps your hardware. I have 2 50g's and don't notice much flicker at all. Fluorescent light can make it more visible (as the light itself is flickering at 50 or 60 Hz depending on where you live, it causes interference between the light cycle). Incandescent, LED or sun light help reduce/eliminate the flicker.
But I use mine at the office all day with fluorescent lighting and don't notice any flicker at all. Is anybody else seeing excessive flicker?
Also, what do you mean by "icons"? There's only a couple of gray lines on the screen, but no icons at all. Indicators in the status area are all black (except for the new SD card indicator, it's the only gray one).

JDW · 09-13-2016, 02:37 AM

Claudio,

Thank you for the explanation about the heartbeat indicator.

I use "icon" and "indicator" interchangeably. But for me to be more specific would mean to reflash my 50g with newRPL and check again. I don't remember everything specifically from memory. When I have the time, I will do that and post back here.

But here in Japan, it's pretty much all fluorescent or, more recently, LED. That's true at home and at the office. In fact, even when I first came here back in 1994, I was shocked at how fluorescents were everywhere, with nary an incandescent to be found. Electric power costs more here than in the US, so perhaps that is why. People here don't seem to appreciate warm lighting so much, especially if it's expensive. But I see warm colored LED bulbs on Amazon Japan now, so maybe that will change over time.

Claudio L. · 09-17-2016, 06:55 PM

September update!

Changes in this version:
* Much improved SD card support:

Minimal mount time (approx. 140 ms at slow speed), including on large SDHC cards.
Various compatibility improvements with non-standard implementations
Interrupt-driven card insertion and removal detection
Safe notification and exception when card is removed with open files
Automatic flush of any pending write operations after 3 seconds if inactivity
Indicator in status area to show SD card status
Added support for FAT32 hints, proper mount/unmount.
Various internal bug fixes

* Complex mode re-written from scratch to improve mathematical correctness when dealing with infinities

Please test all operations in complex mode (-103 SF to enable) and report any bugs found.

Claudio L. · 09-17-2016, 07:08 PM

I thought this needs some clarification:

New SD card indicator meaning.

* SD Grayed out: Card inserted but file system not mounted yet (never used).
* SD or HC in black: Card inserted and mounted. File system is clean and card can be removed safely.
* SD or HC in inverse (white on black): File system has pending write operations or a program has at least one file open for writing. It is NOT safe to remove the card in this state. After idling for 3 seconds, the system will flush all pending operations and revert to "normal" (black on white).
* SD or HC in gray inverse (white on gray): File system still has open files but all write operations were completed. It is safe to remove the card in this state. However, it's best to make sure your programs close all files properly.

Nigel (UK) · 09-19-2016, 04:03 PM

I've flashed my 49g+ and 50g with the latest firmware and measured their current consumption. I see no change from the previous version - the 49g+ still draws several mA when turned off, exactly how much depending on whether the SD card is in, out, or simply resting in the slot without being engaged. The 50g is fine when turned off but is still drawing 44-48mA when turned on.

Nigel (UK)

matthiaspaul · 09-19-2016, 06:48 PM

Isn't there a 49g+ owner who (also owns a 50g and therefore) could lend his (old) calculator to Claudio for a couple of weeks? This would allow him to systematically change the configuration of port pins and measure the power consumption. It would also allow him to compare the 49g+ and 50g directly while performing the tests. I guess this would considerably speed up the process of identifying that unknown hardware difference causing the problem, and thereby grow the newRPL install base and help the (new)RPL community in general. :-)

Greetings,

Matthias

(PS. I don't own a 49g+, otherwise I would volunteer to lend my calculator.)

Nigel (UK) · 09-19-2016, 07:25 PM

(09-19-2016 06:48 PM)matthiaspaul Wrote: Isn't there a 49g+ owner who (also owns a 50g and therefore) could lend his (old) calculator to Claudio for a couple of weeks? This would allow him to systematically change the configuration of port pins and measure the power consumption. It would also allow him to compare the 49g+ and 50g directly while performing the tests. I guess this would considerably speed up the process of identifying that unknown hardware difference causing the problem, and thereby grow the newRPL install base and help the (new)RPL community in general. :-)

Greetings,

Matthias

(PS. I don't own a 49g+, otherwise I would volunteer to lend my calculator.)

A very sensible suggestion. Claudio, do you want to PM me with your contact details? I can certainly live without my 49g+ for more than a few weeks, if needed!

Nigel (UK)

Claudio L. · (This post was last modified: 09-19-2016 08:57 PM by Claudio L..)

(09-19-2016 07:25 PM)Nigel (UK) Wrote:
(09-19-2016 06:48 PM)matthiaspaul Wrote: Isn't there a 49g+ owner who (also owns a 50g and therefore) could lend his (old) calculator to Claudio for a couple of weeks? This would allow him to systematically change the configuration of port pins and measure the power consumption. It would also allow him to compare the 49g+ and 50g directly while performing the tests. I guess this would considerably speed up the process of identifying that unknown hardware difference causing the problem, and thereby grow the newRPL install base and help the (new)RPL community in general. :-)

Greetings,

Matthias

(PS. I don't own a 49g+, otherwise I would volunteer to lend my calculator.)

A very sensible suggestion. Claudio, do you want to PM me with your contact details? I can certainly live without my 49g+ for more than a few weeks, if needed!

Nigel (UK)

Sounds like a good idea, however I don't see myself being able to work on that in the next few weeks. Measuring current on real hardware requires me to be at my place with all the equipment setup, versus my usual approach of "let's code a few lines wherever and whenever I find a few minutes to kill".
Here's what we'll do: I'll setup everything, and investigate how to bring the 50g from those 48 mA down to the 12 mA where's supposed to be. Then I'll let you know I'm ready and I'll borrow your 49G+ to tweak that one and ship it back as quick as I can. It may be January before I'm ready, though, but we'll get it done.

Edit: I can't believe there's still some 49G+ in good shape, mine fell apart so quickly!

Francois Lanciault · 09-20-2016, 01:18 AM

(09-19-2016 04:03 PM)Nigel (UK) Wrote: I've flashed my 49g+ and 50g with the latest firmware and measured their current consumption. I see no change from the previous version - the 49g+ still draws several mA when turned off, exactly how much depending on whether the SD card is in, out, or simply resting in the slot without being engaged. The 50g is fine when turned off but is still drawing 44-48mA when turned on.

Nigel (UK)

I will try to find time to check newRPL power consomption on my HP-49G+ in the next few days. Maybe it is a problem specific to Nigel unit ???

François

Francois Lanciault · 09-20-2016, 01:52 AM

Hi Claudio,

I have some odd "soft menu" behaviour I would like to report. Maybe it is normal I don't know. This is with the latest firmware but it was probably there before, I don't know.

Lets say I have the "units" soft menu in the top row and my variables in the lower soft menu (the six pack one). I am at root level (HOME).

I create a directory with CRDIR. The directory shows in my variable. I press the soft menu key to go in that directory. The status window updates with the path, the UNITS soft menu is still there and functional. I can go deeper in the tree and the UNITS soft menu stays there.

Now if I power down the calculator and power up again. I am still in the directory I selected but the UNITS soft menu has disappeared. Why ? On the standard firmware the chosen soft menu stays when doing a power OFF/ON.
Also with the top soft menu blank if I UPDIR up to home, it stays blank and not functional . But if I do a power OFF/ON at the HOME level the UNITS soft menu will show back.

Now if the UNITS menu is called while in the subdirectory, this is another matter. It will shows at power up in the sub directory, but not always...

It looks like the system check in the current directory for some information as to what to display in the soft menus upon power up.

As I said, maybe it was always like that, but it would be nice if the soft menus would survive a power cycle, even when the current location has changed.

Another odd behaviour that I am sure everyone knows about: the status area is always blank upon a power ON. It spring to life when a key is pressed but not every keys! Operator keys like + / SIN SQRT will not show the informations in the status window, but number keys and ALPHA letter will.

As always, thank you for the last update. I have check the complex mode and so far have nothing to report. The SD indicator is also a nice addition.

François

Claudio L. · 09-20-2016, 02:27 PM

(09-20-2016 01:52 AM)Francois Lanciault Wrote: Hi Claudio,

I have some odd "soft menu" behaviour I would like to report. Maybe it is normal I don't know. This is with the latest firmware but it was probably there before, I don't know.

...

It looks like the system check in the current directory for some information as to what to display in the soft menus upon power up.

Thanks for reporting this!
Settings like flags, number formatting and menus are stored in the hidden directory '.Settings'. A bug was creating this directory inside the current dir instead of root, therefore creating individual "per-directory" settings. What you experienced was that on power-up, it was recalling the menus you had open last time you used that directory.
On second thought, it could be a good thing to have the ability to override flags, number formatting, etc when you enter a directory, but it can be extremely confusing so for now let's just have global settings and call this bug "fixed".
We may revisit this topic in the future as an improvement.

(09-20-2016 01:52 AM)Francois Lanciault Wrote: Another odd behaviour that I am sure everyone knows about: the status area is always blank upon a power ON. It spring to life when a key is pressed but not every keys! Operator keys like + / SIN SQRT will not show the informations in the status window, but number keys and ALPHA letter will.

This was fixed in the latest update, and it works well on my calc, but since some code commits might have been lost, I'll have to double check.

(09-20-2016 01:52 AM)Francois Lanciault Wrote: As always, thank you for the last update. I have check the complex mode and so far have nothing to report. The SD indicator is also a nice addition.

Good to hear at least results are coming up well. Especially tricky to deal with infinities.

RMollov · 09-21-2016, 10:36 PM

(09-12-2016 04:19 PM)Nigel (UK) Wrote: The 49g+ power consumption is unchanged from my previous measurements - 4mA when turned off; about 42mA when the calculator is on.

The 50g draws \(29\,\mu\rm A\) when turned off. However, to my great surprise it draws 48 mA when turned on and doing nothing. Is this the expected power consumption? Has the latest firmware caused a problem, or do I have an "odd" machine?

Nigel (UK)

I'm not sure what these figures mean. Are you suggesting that 49g+ draws batteries flat even when turned off? Mine (it is 49g+) has been off for a lot longer than a year with the SD card in and batteries were OK.

And yes, Claudio, I've got one in near mint condition partly because it was not used much at all - I didn't like the whole concept of that calculator.

Cheeers

Nigel (UK) · (This post was last modified: 09-22-2016 10:52 AM by Nigel (UK).)

(09-21-2016 10:36 PM)RMollov Wrote: I'm not sure what these figures mean. Are you suggesting that 49g+ draws batteries flat even when turned off? Mine (it is 49g+) has been off for a lot longer than a year with the SD card in and batteries were OK.

Cheeers

Yes, that's exactly right. I started doing the power consumption measurements when I noticed that the batteries in my 49g+ flashed with newRPL seemed to keep going flat, even though the same calculator with HP ROM 2.15 had not drained its batteries in several years of occasional use. I believe that one other 49g+ newRPL user has reported the same issue earlier in this thread.

I got my HP 49g+ many years ago, shortly after it was released. I seem to remember that early versions of the ROM had a battery draining problem that went away when HP issued an update - was it ROM 1.20? Maybe the two problems have a similar cause.

Nigel (UK)

Edit: It was ROM 1.22 that fixed the battery life problem - see https://www.educalc.net/281487.page . I don't remember any details, unfortunately.

RMollov · (This post was last modified: 09-23-2016 03:43 AM by RMollov.)

(09-22-2016 08:01 AM)Nigel (UK) Wrote:
(09-21-2016 10:36 PM)RMollov Wrote: I'm not sure what these figures mean. Are you suggesting that 49g+ draws batteries flat even when turned off? Mine (it is 49g+) has been off for a lot longer than a year with the SD card in and batteries were OK.

Cheeers

Yes, that's exactly right. I started doing the power consumption measurements when I noticed that the batteries in my 49g+ flashed with newRPL seemed to keep going flat, even though the same calculator with HP ROM 2.15 had not drained its batteries in several years of occasional use. I believe that one other 49g+ newRPL user has reported the same issue earlier in this thread.

I got my HP 49g+ many years ago, shortly after it was released. I seem to remember that early versions of the ROM had a battery draining problem that went away when HP issued an update - was it ROM 1.20? Maybe the two problems have a similar cause.

Nigel (UK)

Edit: It was ROM 1.22 that fixed the battery life problem - see https://www.educalc.net/281487.page . I don't remember any details, unfortunately.

Sorry, I actually meant 49g+ with ver. #2.15, not newRPL - for the latter I know from experience. The figures you mention (29μA vs. 4mA) suggest it still draws alot more than the 50g when off. And still lasting years?

Cheers,

Nigel (UK) · 09-23-2016, 07:40 AM

(09-23-2016 03:40 AM)RMollov Wrote: Sorry, I actually meant 49g+ with ver. #2.15, not newRPL - for the latter I know from experience. The figures you mention (29μA vs. 4mA) suggest it still draws alot more than the 50g when off. And still lasting years?

Cheers,

My measured figures are all for newRPL; I never had any battery problems with either machine when running ver. #2.15 so I never made any measurements. Apologies if this wasn't clear - we seem to have been communicating at cross purposes!

Nigel (UK)

RMollov · 09-23-2016, 11:23 AM

(09-23-2016 07:40 AM)Nigel (UK) Wrote:
(09-23-2016 03:40 AM)RMollov Wrote: Sorry, I actually meant 49g+ with ver. #2.15, not newRPL - for the latter I know from experience. The figures you mention (29μA vs. 4mA) suggest it still draws alot more than the 50g when off. And still lasting years?

Cheers,

My measured figures are all for newRPL; I never had any battery problems with either machine when running ver. #2.15 so I never made any measurements. Apologies if this wasn't clear - we seem to have been communicating at cross purposes!

Nigel (UK)

OK then, fair enough now Wink

. I already shared my findings about abnormally high power consumption of the newRPL firmware earlier in this thread.

Regards,

Bruno · (This post was last modified: 10-03-2016 06:38 PM by Bruno.)

Hi,

is there a few note anywhere that describe the pre-requisite and how to compile the newRPL ROM please ?

I tried a 'make rom.bin' under the /newrpl-source/firmware directory,
but make can't find a directive to build fw_base.c

Thanks

EDIT : The project seems to be created with Qt, I suppose the makefiles require some modifications to work without Qt ?

Claudio L. · (This post was last modified: 10-03-2016 09:46 PM by Claudio L..)

(10-03-2016 05:20 PM)Bruno Wrote: Hi,

is there a few note anywhere that describe the pre-requisite and how to compile the newRPL ROM please ?

I tried a 'make rom.bin' under the /newrpl-source/firmware directory,
but make can't find a directive to build fw_base.c

Thanks

EDIT : The project seems to be created with Qt, I suppose the makefiles require some modifications to work without Qt ?

The steps are quite simple, the following should work on Linux/FreeBSD:
* Install gcc or Clang compiler.
* Install Qt with Qt Creator (for easier coding).
* Install arm-none-eabi-gcc and arm-none-eabi-binutils
* Install libelf-devel package (if it's not already installed).
* Open Qt Creator, first you need to build the tools: go to the newrpl/tools directory and open the project elf2rom, choose the "Release" version, edit the make steps to add a "make install" step at the end.
* Do a build-all, look at the output to make sure the executable was copied properly.
* Now do the same with the other tools.
* Now go back to the main directory, open the project newrpl-comp.pro. Same thing (select Release, add a make install step, then build-all).
* Now close all, open newrpl-fw.pro, edit the .pro file as needed to change the folders where you have arm-none-eabi-gcc (they change according to version, look at the file).
* Hit build and everything should work fine, if there are no errors, you should see the last line "Binary newrplfw.bin written".