NP-41 Emulator (may be)

[Chris Chung]

(09-28-2015 01:01 AM)Helix Wrote:  

(09-22-2015 01:05 PM)Chris Chung Wrote:  I will keep you posted on the progress and I welcome your comments / questions / suggestions.

I don't know if my comment is useful, but on your PCB diagram the "shift 7" writing seems to be FF instead of SF.

Thanks for the catch, will correct for the next PCB design. I am thinking of separating the keypad and the display + MCU into two PCB units. This will save PCB fabrication cost and potentially keypad upgrades.

(09-27-2015 11:16 PM)Zweb Wrote:  

(09-27-2015 10:58 PM)matthiaspaul Wrote:  So, you and Waon are working on a Nonpareil port and calculator hardware based on a microcontroller of ST's STM32 family, whereas Chris is working on a hardware based on Texas Instruments' MSP430FR6989, both to emulate the HP-41C in the end?

Greetings,

Matthias

Yes, exactly.

We have a couple of LCDs in Waon's hand so that Zweb and Waon can work on an alternative version. I did some digging and I believe they have good success on another calculator project, ArithMax.

I think them China student team has the edge of being accessible to fast and inexpensive parts, manufacturing services.

I am sure my design will lack a lot on speed. But hopefully my choice of MCU will beat the STM32 design on lowest power RAM retention (at 0.0 nA :-). I understand that Zweb will be using STM32L series w/ LCD segment driver so that would also be a one chip design focus on power savings.

(09-27-2015 10:17 PM)Benoit Maag Wrote:  

(09-22-2015 11:00 AM)Waon Shinyoe Wrote:  Here is a comparison of the original 41C screen and the newly built one:

Beautiful ! - it even looks better than the original (more neutral colors and better contrast)
I would be interested in buying a few if available (say 5)

Currently we have 10 sample units (of which at least one has been tested faulty), the plan is to make sure it will work on our design (via prototyping) and we will order production units. I haven't test them except Waon did drove them w/ DC. I still cannot believe we are able to get them fabricated in such cost and time, so I have to multiplex them and make sure everything is good when I receive them.

The production units can only be ordered at batches of 300 pcs. I envisage that the prototyping will take 4-6 weeks. The cost on the production units is very affordable, say cost of a cup of coffee (not StarBuck, more like Timmy's). Still there are the upfront engineering cost, shipping from factory, absorption cost on extra units, etc, etc. If there are enough interest, may be Waon can setup something later as it could be less expensive to ship from China.

During our discussion w/ the manufacturer, Waon wants it to look more closely to the original, the manufacturer I guess has already some templates for 14 segments so we opt for what they suggested. I just need them to be at the best contrast and use their proven layout would be safer.

[jjohnson873]

Hi Chris...very nice display and HP41C emulator in hardware & firmware. I'd be interested in at least one or 10! (is that binary or decimal?). Not sure if I missed it in the thread, but....is there any chance that you're making the electrical connections on the new HP41C display backward compatible with the old HP41C display? There are many, many, many HP41C's (41V's & 41X's) that have damaged displays and would simply need a new one to be fully functional.

Btw....the NP25 went over very well at HHC2015 and there are many new users.
Cheers! Jim J.

[Helix]

(09-28-2015 01:29 PM)Chris Chung Wrote:  Thanks for the catch, will correct for the next PCB design. I am thinking of separating the keypad and the display + MCU into two PCB units. This will save PCB fabrication cost and potentially keypad upgrades.

So, if corrections can be made, here are additional suggestions:

- The "E-" "E+" and "CLE" writing should be replaced by "Σ-" "Σ-" and "CLΣ", like in the last versions of the NP-25: http://www.hpmuseum.org/forum/thread-217...l#pid31814
- Also, the "R v" writing to simulate the "R↓" key is not very pleasing. In the NP-25 of the link above, the writing is much better.
- "x<y?" should be replaced by "x<=y?". I understand that space was limited on the NP-25 PCB, but there is plenty of room in the NP-41.
- your PCB image is small so I can be mistaken, but I see "e²" instead of "ex" (x superscript), and a dot instead of a coma for the ALPHA dot writing.

[Chris Chung]

(09-28-2015 11:59 PM)jjohnson873 Wrote:  Hi Chris...very nice display and HP41C emulator in hardware & firmware. I'd be interested in at least one or 10! (is that binary or decimal?). Not sure if I missed it in the thread, but....is there any chance that you're making the electrical connections on the new HP41C display backward compatible with the old HP41C display? There are many, many, many HP41C's (41V's & 41X's) that have damaged displays and would simply need a new one to be fully functional.

Btw....the NP25 went over very well at HHC2015 and there are many new users.
Cheers! Jim J.

I just received the sample LCDs from Waon two hours ago. Will test drive them tomorrow, digit by digit as my PCBs are not ready yet. The current design have the MCU MSP430FR6989 drive the segments directly.

I don't know how the original HP-41C drives the LCDs. From the photos it appears that the LCD is made into a module. We can do the same by using a dedicated driver IC, like the NXP PCF8545 and the likes. They are driver by SPI or I2C so there may be another MCU needs to do the interfacing. I am not sure if there will be a lot of demand on these though.

Glad to hear people playing w/ the NP25C. I was building one this weekend and thought it would be better to move to smd, both the same PCB estate and to eliminate the need for the shift register. For smd packaging, there is an extra 8 pins on the MSP430G2553. Not sure if this will happen though, as my interest is now on the 41.

For the LCD availability, see my previous note on the plans. But 1st I need to make sure the samples are good to drive.

[Helix]

(09-27-2015 12:51 PM)Zweb Wrote:  Here is layout of the STM32 version of NP-41 calculator!
First trial, may buggy, but should work.

This design seems to use much bigger keys than Chris Chung's version. I'm afraid that the small shift writings will be easily masked by the keys.
I think also that the figures and the - + x / operations should have bigger writings, as Chris Chung did, or use bold characters, to make the keyboard more readable.

[rprosperi]

(09-28-2015 11:59 PM)jjohnson873 Wrote:  Hi Chris...very nice display and HP41C emulator in hardware & firmware. I'd be interested in at least one or 10! (is that binary or decimal?). Not sure if I missed it in the thread, but....is there any chance that you're making the electrical connections on the new HP41C display backward compatible with the old HP41C display? There are many, many, many HP41C's (41V's & 41X's) that have damaged displays and would simply need a new one to be fully functional.

Btw....the NP25 went over very well at HHC2015 and there are many new users.
Cheers! Jim J.

Chris - If you are keeping a list yet, please add me as well; while I'd certainly like more than 1 (eventually), 1 from an early batch is fine, I don't want to deny others also wanting one.

Jim did a great job explaining and showing off the NP25! In addition to many new interested buyers, to those that were seeing it for the first time, it was a shock to see such a capable and functional device in such a simple-looking pkg. One old-timer there simply did not believe it was what was described until I literally showed it to him and put it in his hands. After the shock wore off, he did not want to put it down. It continues to make impressions, just like the new NP41 will I'm sure.

Thanks to both Chris and Jim for this great device.

[Chris Chung]

(09-29-2015 09:55 PM)rprosperi Wrote:  .....

Chris - If you are keeping a list yet, please add me as well; while I'd certainly like more than 1 (eventually), 1 from an early batch is fine, I don't want to deny others also wanting one.

Jim did a great job explaining and showing off the NP25! In addition to many new interested buyers, to those that were seeing it for the first time, it was a shock to see such a capable and functional device in such a simple-looking pkg. One old-timer there simply did not believe it was what was described until I literally showed it to him and put it in his hands. After the shock wore off, he did not want to put it down. It continues to make impressions, just like the new NP41 will I'm sure.

Thanks to both Chris and Jim for this great device.

NP41 is far from complete. I am glad that NP25 got all these attention

I did a test rig set-up to test the LCD glass from Waon. Here is a video.

The contrast is not as good as shown in Waon's photo. As I am properly multiplexing them at 1/6 duty cycles instead of static DC. I am using resistor dividers to get the voltage bias needed (the glass is of 1/3 bias, so I need to generate 1, 2 and 3V). There is shadowing as you can see on the video. I hope with a proper LCD driver I can fine tune the bias voltage and get better results. This set-up is used to make sure every segment is good and individually addressable (via multiplexing) on the sample pieces.

[Steve Simpkin]

Very cool Chris!

[Massimo Gnerucci]

It is always nice to look at those 14 segments: brings back memories... :)

[Chris Chung]

MEMORY LOST.

I had over-estimate my ability to hand solder SMD components (this one at 0.5mm pitch). After a failed PCB and a destroyed MCU ($15 dollar value), I managed to assemble my backup MCU (only ordered 2) and get it running.

There are a few missed pins when I turn on the power.

   
Another soldering session have all the segments light up.
   

I can't get the MCU charge pump working, so for now I am suck w/ no contrast control. I am not sure where the shadow segments came from, could be software or hardware. Hope I will be able to figure that out. The poor soldering may also contribute to ghosting.

Will be waiting for SMD capacitors to come in and try the on chip charge pump again. I have not yet soldering and code the keys yet. Still using UART (usb-serial) and a terminal for input. So the UART works and will allow us to download programs in the further. I am going to work w/ this rig and sort out the firmware 1st. Then have another PCB design. I feel that I need at least another engineering / prototype design before finalizing.

The most worry now is the quality of how I am driving the display. The ghosting is just too much. Have to rule out these possibilities.

. bad soldering.
. lack of precise bias control (due to no charge pump).
. software / firmware not optimized (timing and other control).
. LCD itself?

[Zweb]

Happy to see that you have had it working. About the annoying LCD shadowing, I may help to determine if LCD itself is good by using my PCB. I'm currently in HK so I can start working on that before tomorrow. A little curiosity, if the charge pmup is not working, how did you get the voltage needed?

[rprosperi]

(10-03-2015 09:58 PM)Chris Chung Wrote:  MEMORY LOST.

Cool. Possibly the happiest I've ever been seeing a 41 display show MEMORY LOST.

[Chris Chung]

(10-03-2015 11:15 PM)Zweb Wrote:  Happy to see that you have had it working. About the annoying LCD shadowing, I may help to determine if LCD itself is good by using my PCB. I'm currently in HK so I can start working on that before tomorrow. A little curiosity, if the charge pmup is not working, how did you get the voltage needed?

I was thinking about asking you to compare notes on the LCD, so just take your time on it and add a note here.
The MCU I am using can use 1) DVCC, 2) External Supply, and 3) Internal charge pump to drive the LCDs.

I cannot make 3) work yet as I don't have the right SMD capacitors so I am using 1) DVCC.
There are also many controls that I need to explore in driving the segment. I may need to build some kind of rig just to work on optimizing the LCD contrast. As the LCD is 6-MUX, the contrast will be lower than a static or 2/4 MUX. I just hope it will end up clearly readable w/ no ghosts.

[Chris Chung]

(10-04-2015 12:37 AM)rprosperi Wrote:  

(10-03-2015 09:58 PM)Chris Chung Wrote:  MEMORY LOST.

Cool. Possibly the happiest I've ever been seeing a 41 display show MEMORY LOST.

For me, the 1st MSP430FR6989 I fired is the happiest (and most costly) chip that I toasted. As it made path to a workable unit. I see it is an expensive SMD soldering practice kit.

Still I need to find better ways to do this, but not too costly thought, May be an entry level hot air gun will do.

[jebem]

(10-03-2015 09:58 PM)Chris Chung Wrote:  MEMORY LOST.



The most worry now is the quality of how I am driving the display. The ghosting is just too much. Have to rule out these possibilities.

. bad soldering.
. lack of precise bias control (due to no charge pump).
. software / firmware not optimized (timing and other control).
. LCD itself?

Hi, Chris,
The LCD ghost can be caused by excessive high scan frequency (timings in your shortlist above).
Did you try to reduce it to a value as low as possible (without flickering)?

[Chris Chung]

(10-04-2015 01:49 PM)jebem Wrote:  

(10-03-2015 09:58 PM)Chris Chung Wrote:  MEMORY LOST.



The most worry now is the quality of how I am driving the display. The ghosting is just too much. Have to rule out these possibilities.

. bad soldering.
. lack of precise bias control (due to no charge pump).
. software / firmware not optimized (timing and other control).
. LCD itself?

Hi, Chris,
The LCD ghost can be caused by excessive high scan frequency (timings in your shortlist above).
Did you try to reduce it to a value as low as possible (without flickering)?

Thanks, I will try that. I will be having a busy work week so may not be able to work too much on it. Anyway, I will give updates when there are advances.

[Chris Chung]

Tried to lower the frequency but did not help much. I did get rid of a lot of the ghosting by lowering the supply voltage from 3.3V to 2.9V.



The LCD were ordered to have VLcd at 3.0V as I intend to run the calculator off a 3V coin cell, which I was doing on the photo, except I add a potentiometer to drop it further to 2.9V. When I get the charge pump to work, I can make it software configurable.

It is readable but does not have high contrast, as it's 6-MUX so I will have live w/ that. There are still a few odd ghosted segments with probably was introduced by my poor soldering.

I will start to implement the keyboard scanner.

/EDIT embedded image doesn't seem to work. upload this one
   

[jebem]

(10-06-2015 12:32 AM)Chris Chung Wrote:  The LCD were ordered to have VLcd at 3.0V as I intend to run the calculator off a 3V coin cell, which I was doing on the photo, except I add a potentiometer to drop it further to 2.9V. When I get the charge pump to work, I can make it software configurable.

It looks good enough to me.
How about the annunciators? Are it readable?

As a last resort, a simple voltage 2-resistor network could be used instead of a potentiometer, as long as the adjustment is not critical (variances between different machine components).

[Chris Chung]

(10-07-2015 01:29 PM)jebem Wrote:  

(10-06-2015 12:32 AM)Chris Chung Wrote:  The LCD were ordered to have VLcd at 3.0V as I intend to run the calculator off a 3V coin cell, which I was doing on the photo, except I add a potentiometer to drop it further to 2.9V. When I get the charge pump to work, I can make it software configurable.

It looks good enough to me.
How about the annunciators? Are it readable?

As a last resort, a simple voltage 2-resistor network could be used instead of a potentiometer, as long as the adjustment is not critical (variances between different machine components).

I have to get the charge pump to work as it will guarantee the when the battery ages, it won't affect the contrast. So far no know, it could be my PCB, will have to spend time on it.

The annunciators / icons are good, I just have not map them w/ the simulator code. I am working on the keys now, almost done and there is a lone pin that I need to re-flow.

I definitely need to invest on something to solder the LQFP. Leaning towards a simple upside down iron re-flow trick. Will see.

[Chris Chung]

Got the keys working, will need more timing adjustment (scanning) as they are sometimes not very responsive.

I am running this on a partially depleted coin cell at around 2.85V. A 3V battery will produce too much ghosting on the LCD.





Will be designing a new PCB to move on. The charge pump still fails after 1/2 second. It drops to 1.66V instead of staying at 2.8V. I had tried different values and types of capacitors and still cannot get it to work. I had also ripped the capacitor pad off after excessive re-flows. Will include more options in a new PCB design for charge pump trials (also allows custom biasing to see if contrast can be improved.

Also found out that the current design w/ the use of a port 9, I cannot assign the "ON" key the role of h/w interrupt after getting into deep sleep. So the current unit does not have a good power saving sleep scheme.

I am just showing the video here as there is a lot of interest for it. My next PCB will just contain the MCU and LCD, plus a 9-10 pin connector to a keyboard. This will save cost at the current stage.

The final unit (for this PCB) is a bit large for my liking. I was indecisive when doing the LCD and had chose to match the HP-41C LCD size. It appears that based on that the product is more likely a desk unit (a slim one though) instead of a pocket unit.