HP-25C Eumulator

[PANAMATIK]

(01-07-2015 09:54 PM)brouhaha Wrote:  

(01-07-2015 12:54 PM)Chris Chung Wrote:  The flickering should resemble the real thing. Eric's code follows the ACT unit to every single flag in the registers which included the display on/off and segment controls. And my code follows his code very closely.

Not 100% sure, don't have a really calculator to compare w/. It appears so when I compare w/ a youtube video.

In Nonpareil ("Eric's code"), I made a serious effort to simulate the instruction-cycle behavior of the HP ACT, including how the display is refreshed one digit per instruction cycle. I did not specifically compare for exact matching of the display flickering, so there might be some differences, though in casual use it seems about the same.

As far as I've seen, Chris has done an excellent job of faithfully adapting the code to the microcontroller.

With the parallel out 74HC164 it is not possible to simulate the exact display behaviour, because the original HP-25 segment driver activated not one digit at a time, but only one segment in one digit at a time. But the human eye cannot distinguish that. The new ACT, also based on "Eric's code", multiplexes the display very exactly and I cannot see any slightest difference.
Bernhard

[Chris Chung]

(01-07-2015 10:37 PM)PANAMATIK Wrote:  

(01-07-2015 09:54 PM)brouhaha Wrote:  In Nonpareil ("Eric's code"), I made a serious effort to simulate the instruction-cycle behavior of the HP ACT, including how the display is refreshed one digit per instruction cycle. I did not specifically compare for exact matching of the display flickering, so there might be some differences, though in casual use it seems about the same.

As far as I've seen, Chris has done an excellent job of faithfully adapting the code to the microcontroller.

With the parallel out 74HC164 it is not possible to simulate the exact display behaviour, because the original HP-25 segment driver activated not one digit at a time, but only one segment in one digit at a time. But the human eye cannot distinguish that. The new ACT, also based on "Eric's code", multiplexes the display very exactly and I cannot see any slightest difference.
Bernhard

I agree, w/ the shift register's buffering effect, there is hardware constraint that dictate the differences. My level of "faithful" emulation can only goes to where it had to be ended. I.e. not using same led driving circuitry.

I did tried and explored various option when I design this. If I had used an 28 pin AVR, we don't really need to use a shift register. It was actually a better choice for the project. Except that my design goal is low power and I am more familiar w/ the TI chip. This project to me is more of electronic hobbyist than of HP enthusiast.

The reason I did this project is I had some success doing a TI DataMatch / Sinclair Scientific h/w emulator and in the process searched all the available microcode emulators on the internet. And I found nonpareil, thought I might just make another toy for my daughter to use at school (the Sinclair Scientific's trig function will fail her math tests, I think).

It would be a good exercise for someone to move this to an 328. This would reduce the hardware to just one MCU and allow inclusion of the 34C. Anyone?

(01-07-2015 08:20 PM)walter b Wrote:  

(01-06-2015 03:59 PM)walter b Wrote:  Please count me in for one as well. I might use it in a nonfunctional Woodstock. Ooh, and my soldering skills leave plenty of space for improvement

... thus I'll be grateful for an assembled unit.

d:-)

Mistook that you want to "practice" soldering. My bad

V3 3rd batch list
rposperi x3, Marcus (k), Geoff, jebem (k), walter, Den (8 units)

[RMollov]

(01-08-2015 12:56 PM)Chris Chung Wrote:  V3 3rd batch list
rposperi x3, Marcus (k), Geoff, jebem (k), walter, Den (8 units)

I'd like to get one assembled unit too
Cheers

[BobVA]

(01-07-2015 05:58 PM)Chris Chung Wrote:  Updating list. 2nd batch all allocated and shipped.

V3 2nd batch list
John, SlideRule, Brouhaha X2, PANAMATIK, Etienne, Lourens, viaPM1, viaPM2, BobVA, viaPM3 (11)

Arrived safe and sound - thanks! (The new silk screen does look very nice.)

[rwiker]

(01-10-2015 09:04 PM)BobVA Wrote:  

(01-07-2015 05:58 PM)Chris Chung Wrote:  Updating list. 2nd batch all allocated and shipped.

V3 2nd batch list
John, SlideRule, Brouhaha X2, PANAMATIK, Etienne, Lourens, viaPM1, viaPM2, BobVA, viaPM3 (11)

Arrived safe and sound - thanks! (The new silk screen does look very nice.)

I got mine today - just the day after I bought batteries in hope of getting the kit sometime next week :-)

[PANAMATIK]

I got mine today here in germany. But I don't have batteries and will get them on monday, I forgot buying, despite your reminding.

It's smaller than I thought and it looks very, very nice. The smallest three in one HP calculator ever.

Bernhard

[Herman]

Add me to 3rd batch for 1 kit.

Thanks

[Chris Chung]

Thanks for comments on the 2nd batch V3 units.

(01-10-2015 09:34 PM)rwiker Wrote:  

(01-10-2015 09:04 PM)BobVA Wrote:  Arrived safe and sound - thanks! (The new silk screen does look very nice.)

I got mine today - just the day after I bought batteries in hope of getting the kit sometime next week :-)

Bob got units from both batches and will notice the silkscreen difference.

The silkscreen on this batch did looks better than the 1st batch. I compared and this batch appears to have "thicker" font prints. I am at the mercy of my PCB fabricator and does not know why it is not consistent. May be they have introduce new machines. Actually I spent good time to draw all the math symbols for the V2 PCBs and they got messed up and cannot print them, you can see the different from the pictures on page 1 or 2.

(01-10-2015 11:37 PM)PANAMATIK Wrote:  I got mine today here in germany. But I don't have batteries and will get them on monday, I forgot buying, despite your reminding.

It's smaller than I thought and it looks very, very nice. The smallest three in one HP calculator ever.

Bernhard

Yes, It meant more as a toy than something to use for everyday work. Another reason for the weird (and smaller) size is that it cost less to produce the PCBs.

This shipping is faster than expected. Could be that it's after Christmas? I would imagine the kits are also getting into their builders hands. There are 3 kits in this batch and I had wrote up advise for the single kit ordered on the 1st batch, So kit builders for 2nd batch can also use. I will re-type as follows


"There are some build tips for my "other" calculator which also applies to the NP-25

http://forum.43oh.com/topic/5524-tms0803...d/?p=49241

The tactile buttons must be bend by using a pair or long nose pliers. The holes are still a bit large so it fits loosely and you might need to align the buttons while soldering one side of it.

The buttons are SMD parts and I found that using thru-hole PCB allow me to "fit" them on the PCB so there is no need to use stencils and I can have a "flat" bottom.

I would recommend you to solder the ICs before the displays. ICs need to be cut short on one side. The LED displays are very tricky to align properly and they are designed to be tilted at a angle (see datasheet found in sparkfun product page). I would solder everything and then "friction fit" the leds w/o soldering them. Turn it on and based on the display use a pair of twitters to adjust the final position / angle of the leds before soldering them.

If I got a chance to re-do this, I am going to try to get everything on one side for easier assembly."

The link to 43oh you will not see photos unless you are a member. But please do read the text I post there (as simpleavr) especially about aligning the LED and IC parts (pin 1 position).

Another tip that I just developed doing the 2nd batch assembly is to find yourself a perfboard, and test fit the LED modules on it before fitting them onto the NP-25 PCB. Perfboard will have larger holes. You can adjust the pins, set the LED module on the perfboard, rock it back and forth, lift it up and down. When you reach a point that no pin is resisting when you lift it out of the perfboard, the alignment is at its best. This will allow way easier final fitting and better consistency on all the 3 led modules.

Please feel free to post any question regarding the build.

Also update for next batch.

V3 3rd batch list
rposperi x3, Marcus (k), Geoff, jebem (k), walter, Den, RMollov, Herman (10 units)

[Chris Chung]

For kit builders, the original video has assemble information at the beginning of the video. Which may be helpful to you.

Please note that in the video I solder the LEDs 1st, then the ICs. I suggest now to solder the ICs 1st, LEDs last as the LEDs are the most tricky parts.

http://youtu.be/WwPM0slZro4

[jebem]

Hi, Chris,
Thanks for putting me on the waiting list for the V3 NP-25.

Could you publish photos of the V3 PCB on both sides (with and without components), please?
That gives 4 photos in total.
This would help the kit builders like myself.

Thanks.
Cheers,
Jose

[Algorhythm]

Two kits for me please.

Cheers!

[pcfu44]

Hi, Chris
I would like to get one assembled unit if that's still possible.
Thanks!
Paul

[Bernd Grubert]

Hi Chris,
Please put me on the list for an assembled unit.

Best regards
Bernd

[Chris Chung]

(01-12-2015 09:14 PM)jebem Wrote:  Hi, Chris,
Thanks for putting me on the waiting list for the V3 NP-25.

Could you publish photos of the V3 PCB on both sides (with and without components), please?
That gives 4 photos in total.
This would help the kit builders like myself.

Thanks.
Cheers,
Jose

I don't have "blanks" now. They are all gone. I did have photos of an assembled unit closed up on the major components (i.e. not the tactile buttons).

* I made image bigger so we can see more detail

Front view below, apart from the LED modules, there are only 4 passive components. From left to right, a 104 (100nF), 47k pull-up, 102 (1nF), and 1k. Components are marked so for the caps just match 104 and 102 markings. For the resistors you have to read the colored rings or use a multimeter.

• A square pad on the PCB means the pin 1 position. Under the LED module, there is one pin marked w/ a white dot, and it's the pin 1 position, so that pin should go into the square pad.
• The 100nF is just a bypass cap for the MCU.
• Next should be a 47k (yellow + purple + black + red + brown) pull-up resistor. For the last few units / kits I had 56k (green + blue + black + red + brown) instead as I had no 47k left. The pull-up value is not that important.
• The 1nF and the 1k are more important as they form a RC delay circuit for the shift register. use a multimeter or match the ring colors w/ the photo here. (the 1k should have brown rings on both sides, i.e. brown + black + black + brown + brown = '1' '0' '0' * 10^'1', '1'% tolerance)
  

Moving to the back side there are only 2 ICs to worry about. Just follow this orientation. Some ICs have a dot on pin 1. But most DIP device uses a notch to indicate the top of the chip, and pin 1 is always the top-left pin. Again, the square pad on the PCB is pin 1.


The following illustrates what you need to do with the tactile buttons. You will need a pair of long nose plier and bent the smd pins. Then fit them into the holes on the PCB. This batch of PCB has slightly larger holes and they wobbles a bit. You can use a ruler for easier alignment (see a photo I post earlier). You can also see the white dot (pin 1) for the LED module here.


This is a side view of the test unit I used to verify all the parts for the kits. The LEDs are friction fitted (not soldered) and I had sockets for the ICs so everything (well, the expensive parts) is removable.
It works well on its own. So if you are considering savings the parts for future projects, you can do the same.


OTOH If you are soldering them all down, you will have to cut one side of the LEDs (and the ICs) short so they are inserted flush.
If you are soldering the ICs 1st, you can cut one side of the pins after your solder.
You can then friction fit the LED modules, turn on / test the unit before trimming down pins on the inner side and do final soldering.


I emphasized the final step as the double side component design, once soldered, the LEDs and ICs are permanent and there is no fixing. So It is better to fit, align, adjust, before final soldering.

The LED modules are tricky, the pins are difficult to bent / adjust. It took me 10+ units to get a hold of it. And I apologies to the early buyers as they were fitted worse than the recent units (including my own units).

[Etienne Victoria]

I have received my Eumulator today.

It looks much prettier than on the posted pictures and is actually very usable!

Chris, thank you very much for the great work you put in this project and also for the kind greetings message :-)

Very happy. I'll start my quest for an adequate packaging...

Happy 2015 to all!

Etienne

[jebem]

(01-13-2015 02:22 PM)Chris Chung Wrote:  

(01-12-2015 09:14 PM)jebem Wrote:  Could you publish photos of the V3 PCB on both sides (with and without components), please?
This would help the kit builders like myself.

(...) I did have photos of an assembled unit closed up on the major components (i.e. not the tactile buttons).
* I made image bigger so we can see more detail (...)

It is crystal clear now. Excellent instructions.
Thank you!

Do you have any estimate delivery schedule for the next batch?

[Sanjeev Visvanatha]

Chris:

Thank you so much for the assembled unit. It came in the mail today, and I was very excited to see the personal greeting! Is there any recommended packaging?

Sanjeev

[Chris Chung]

(01-13-2015 10:25 PM)jebem Wrote:  Do you have any estimate delivery schedule for the next batch?

I got notified that the PCBs are on the way. That means another 10-14 days before I received them. I can prepare and ship kits in 1-2 days. For assembling, I usually find the adjacent weekend to work on them.

I am receiving 11 boards for the 3rd batch. This batch has a change of smaller tactile button hole for better fit / alignment and I do not ordering the 4th batch before confirming the change works correctly.. So the 4th batch will be ordered once I successfully assembled a unit from the 3rd batch.

V3 3rd batch list expect shipping late January
rposperi x3, Marcus (k), Geoff, jebem (k), walter, Den, RMollov, Herman (k), jebem (k) (11 units)

V3 4th batch list expect shpping mid Feburary
algorhythm x2 (k), pcfu44, Bernd (4 units)

(01-14-2015 02:23 AM)Sanjeev Visvanatha Wrote:  Chris:

Thank you so much for the assembled unit. It came in the mail today, and I was very excited to see the personal greeting! Is there any recommended packaging?

Sanjeev

When I design this, the units were meant to be "exposed" and attached to a lanyard as conference badges. Since it's very close to credit card size, you may want to try leather / soft credit card holders. Guenter had demonstrated this here http://www.hpmuseum.org/forum/thread-217...l#pid23976

[John R]

Hi Chris,

I got mine in the mail today. It is even nicer than I imagined -- worth every penny (and more)!

Thanks so much for the time and effort you have put into this, and for making it available to all of us.

Cheers,
John

[iceman]

Hi Chris,

Got it yesterday. Built it today, very nice kit. Mounted bubbles at 20 degrees as suggested in HP data sheet. Used machine tool sockets.

Thank-you very much Chris.

(Pictures in email)