WP34s Repurposing Journey PhotoJournal

[MarkHaysHarris777]

Greetings,
The purpose for posting tonight is to start a photo-journal of sorts documenting my repurposing experience converting the HP30b business calc to the now famous WP34s community scientific programmable calc. My hope is to encourage other first-time repurposing DIY folks like me, let folks know the kinds of things I ran into, and to share ideas (materials, parts, construction techniques, &c. For most folks these days the main issues are going to be 1) flash cable, 2) main board pcb mods (crystal, caps), and 3) pc systems issues.

I have opted to mod the calc (crystal and caps), and I will also be constructing my own flash cable box, flash clip, and IR mods. I am not going to be repeating the on-line doc other than to make clarifications where I think the doc could be clearer, or where I think other noobs like me would find it helpful. I have most of the materials I'm going to need including the HP30b(s) (man was that a lot of research and cross checking) but some of it has not arrived yet. I have started construction on the cable box and flash clip.



I made my flash clip (Katie clips included) using pogo stick (compressed spring header pins) mounted in modeling wood (bass wood) with the comm cable soldered directly to the pogo pins. The Katie Clips are soft aluminum sheet, and the construction is wood glue and tape (keep it simple, keep the cost down).



I got my pogo pins from Sparkfun for about $0.95 per pin. There are different kinds of heads... I chose the bullet shaped pointed pin for better alignment and better electrical contact. I am new to working with these pins. If you use flux they solder well. The pin is a compressed spring (pogo stick) tube which is used in this project to connect to the gold pads on the 30b pcb through the back cover (for connecting the unit to the PC to download the WP34s flash image. There are six tiny holes through the back cover (I'm leaving mine, discussed later) and so I'll be using six pins. I purchased a couple of extra in case I screw up one or two.



I am leaving the tiny holes in my unit. This is easier, and will help to align the pins during flash without the possibility of shorting. i ran into a problem, however, with this approach; easily solved with a jewels rasp (tiny engraving tip for Dremel tool). HP tries to make sure that their flash cable does not get plugged in backwards (the female backplate is keyed, the clip slots are different widths, and three of the tiny holes in the backplate are smaller. I used a very fine, tiny, jewelers rasp (engraving tip) in my fingures (by hand, no motors!) to make the holes all the same size (just a flea eyebrow hair larger than the pointed heads on the pogo pins. I used the same rasp for making the very snug holes in the bass wood to hold the pogo pins in place. The pins I'm using are 33mm in length (1.3 ") so it is important to have aligning holes holding the pins at the top, and aligning holes holding the pins at the bottom (we want NO shorts).



I also made a 'keeper' (not shown) that is simply a small bass wood block with six holes milled into it to 'keep' the pins aligned, straight, and clean when not in use. My Katie Clips are designed for 3mm of compression when the clip is plugged in. The tension needs to be firm to prevent noise in the comm line, but not sooo firm that the clips pop out! I had a little trouble with this, so getting the clip at the right angle, and getting the compression tension correct were somewhat of a challenge for me. (I have been told that some people just hold the pins down with one finger during the flash; with the pointed pins probably would work). I plan to keep my hands free messing with the PC, and for better control over noise in the comm connection.

Soldering to the pins was easy (use flux, its our friend) and I used an old disk drive ribbon cable for the comm line (just trimmed it up with a pen knife). The wiring is arbitrary but I did take care to diagram in my notebook the wiring from the colored wire out the J connector labels (and meaning) to make things easier when I get the cable box and power supply finished.

PS The trick to mounting the pins in the balsa (or basswood) block is to make the holes manually with a jewelers rasp checking frequently for 'straight' and 'firm'. The pins need to be absolutely straight AND the tips need to be 2mm apart (almost no variation will work) at rest in order for the pins to line up with the tiny holes through the backplate. I'll be honest, this took me most of the evening to accomplish (I do not have a dremel press).

PSS While I'm working on the cable box I'm also spending some time in the WP34s manual (with the mac 34s emulator) learning the ropes)... its an exciting little unit.

Cheers

[StephenR0]

I must say, that's a very creative way to do it. Very interesting!

[John Galt]

Nice work Mark! Looking forward to your progress.

The pins I used are 16mm long. They're also 1/10 as expensive. I'm sure yours will work fine though. You probably won't lose as many of them as I did :-p Thumbnail sized pictures here.

Bending is the biggest problem. Unless they're straight, the pins will bind and become useless. The "keeper" as you call it will minimize that possibility. I used a 2mm IDC connector and the pins simply plug into that (it takes a firm push). No soldering required. A "keeper" was still necessary though. I built mine from three thicknesses of leftover header strip, with its machined pins removed.

[Dave Frederickson]

(01-27-2015 05:49 PM)John Galt Wrote:  Nice work Mark! Looking forward to your progress.

The pins I used are 16mm long. They're also 1/10 as expensive. I'm sure yours will work fine though. You probably won't lose as many of them as I did :-p Thumbnail sized pictures here.

Bending is the biggest problem. Unless they're straight, the pins will bind and become useless. The "keeper" as you call it will minimize that possibility. I used a 2mm IDC connector and the pins simply plug into that (it takes a firm push). No soldering required. A "keeper" was still necessary though. I built mine from three thicknesses of leftover header strip, with its machined pins removed.

I think your "keeper" is similar to mine in that it's part of the plug and aids in alignment. Marcus' keeper protects the pins when not in use.

[John Galt]

Ah yes, I see that.

Bending might not be as much a concern if his pins have a sufficiently large diameter. I'm sure it will work.

The foam tape on yours and mine also helps, in that you and I have to compress the foam slightly to insert the clip ends. When released, the foam expands and the spring pins maintain a little more force than they would if there were no foam to compress. It's a slight improvement over Katie's idea.

[Dave Frederickson]

(01-27-2015 06:30 PM)John Galt Wrote:  Ah yes, I see that.

Bending might not be as much a concern if his pins have a sufficiently large diameter. I'm sure it will work.

The foam tape on yours and mine also helps, in that you and I have to compress the foam slightly to insert the clip ends. When released, the foam expands and the spring pins maintain a little more force than they would if there were no foam to compress. It's a slight improvement over Katie's idea.

Ideally only the pogo pins should have to compress. It took some tweaking of the clip to get it right.

[John Galt]

(01-27-2015 06:37 PM)Dave Frederickson Wrote:  It took some tweaking of the clip to get it right.

It certainly did! If I'm not mistaken, you can press your connector all the way home, until the plastic contacts the circuit board. When releasing it, the spring pins will push everything away from the board, but that movement should be as little as possible. My initial prototype traveled too much and became intermittent.

Another advantage of your idea is that it is impossible to compress the pins to their limits. They will all be exerting approximately the same force.

My popsicle stick and rubber band idea would never have worked...

[Dave Frederickson]

I should mention that the cable assemblies are available for FREE from Samtec. One end of a double-ended cable can be disassembled to get the second housing.

[MarkHaysHarris777]

Well, today I have a functional flash-box, albeit rudimentary at this point. It doesn't have its power-supply regulator yet, nor flops, nor lights (so no bells and whistles yet) but its on its way; it talks to the 30b, and theoretically will facilitate a successful flash; I'm stuck for the moment because my Dynex USB -> Serial TTL converter does not appear to work with MySamba. I get the message 'unable to connect' even though the comm link is working, loop-back works, and erase, power-on reset is working fine. The unit will turn on and off (as testing indicates) but its a paper weight until my FTDI unit gets here. The view of my flash-box is from the rear; the box triples as working platform making it easy to get to the power button, facilitates flashing, and becomes the carrying storage case for the cables, clip, spare batteries, &c.



I'm mounting everything in a cigar box (a lot of my projects end up in those) because they are easy to work with, and because they are cheap.

This unit will either have an LM117, or a D cell battery pack... haven't decided yet. So, still to make the lights work, provide power from the PC, get the buttons working with flops, and mount the cradle that will hold the 30b atop the flash-box.



So, for the moment I have a paper weight that will power-on and power-off ! :-}

On the brighter side, the overlays are here (and look great) and I've had a few hours to hone my soldering skills before I take a crack at the crystals and caps.

Cheers,

[MarkHaysHarris777]

Greetings, I received my FTDI adapters this afternoon in the mails. They worked fine with MySamba with a very pleasant surprise that may interest some of you EE DIY kinda guys -- so read on for some tips and caveats.



First, the really fabulous surprise with this FTDI based USB -> Serial adapter--- its dual voltage; yup... 3.3v and 5.1/ ... so if you want to power the 30b from the PC without an LM117 you can open this adapter and move the Vcc line from the 5.1 volt pinout to the 3.3 volt pinout. The Vcc line has been moved in this photo. When the adapter arrived it only had four wires through the cable and the 3.3v volt line was open. With the 30b connected (batteries out) the Vcc drops to a very nice 3.0 volts; and there is enough current to flash!

Now for a caveat and workaround. i successfully flashed two units with this adapter, after checking loopback. The first time it worked without a hitch, so I setup to flash the second unit. IT DID NOT WORK... "unable to connect". I checked everything, and I mean everything, and no go. rats. Then it dawned on me, after reading some posts here about the difficulties negotiating the protocol with some of these chipsets at 115200, and I got to wondering if my loopback test had the chipset 'stuck' at 9600,8,N,1 so that the 115200 would never work. Consequently I opened another loopback check with minicom, only this time I set the chipset at 115200,8,N,1 and when finished I exited minicom without a reset. Viola!

Does one picture really say 10 thousand words, including thanks to everyone who helped me, and especially to those of you who are now famous, a BIG FAT thank you for the WP34s! ?




Cheers

[MarkHaysHarris777]

PS ... forgot to mention that my first successful flash attempt happened on my gnu/linux box running MySamba.EXE in wine. I had to create a sym link for the comm port in ~/.wine/dosdevices/ using the ln command from inside the dosdevices folder:

ln -sf /dev/ttyUSB0 ~/.wine/dosdevices/com1

... shows up as com1 -> /dev/ttyUSB0


... then just MySamba to use COM1.

I'm updating my flashbox now, cleaning things up a bit, creating a new PC-->FlashBox cable with DB9 connector, and preparing the new proto board to receive the bells and whistles next. When the crystals and caps arrive I'll give moding the 34s a shot. In the meantime, what a joy to play with the 34s. I'm appreciating its layout, and the new programming paradigms are very interesting.

Cheers

[walter b]

Congratulations and thankyou for the nice documentation.

(01-29-2015 05:57 AM)MarkHaysHarris777 Wrote:  Viola!

I guess you didn't want to call violets though that's what you wrote.

(01-29-2015 05:57 AM)MarkHaysHarris777 Wrote:  Does one picture really say 10 thousand words, including thanks to everyone who helped me, and especially to those of you who are now famous, a BIG FAT thank you for the WP34s! ?

Thanks for your very kind words. Though I don't think we can become famous by repurposing a little calc.

d:-)

[MarkHaysHarris777]

(01-29-2015 08:48 AM)walter b Wrote:  Thanks for your very kind words. Though I don't think we can become famous by repurposing a little calc.

You are welcome, and yes, you are famous... world famous actually (at least in engineering circles, universities everywhere among scientists and other math|tech nuts, all over the internet not limited but certainly including wiki (and MoHPC), and anyone (millions of us) who have ever used HP calculators and love them,... you are FAMOUS~! So there...

[MarkHaysHarris777]

Greetings,
I have reworked my proto board (FlashBox v1.0), cleaned up the wiring, and have begun the work to add flops, control lights, monitor lights, and cushioned cradle. The flash-box is based on the following schematic (if you find an error, please let me know and I'll fix it). The flash cable is actually very simple. With the exception of the FTDI PL2303HX there really isn't any electronics involved (unless you just want to) --- just wiring and two switches (ok, two resistors).



The following pics are the front, rear, and open views on the 'Flash Box Revisited'.





Last but certainly not least, credit where credit is due--- thank you Pauli and Walter, and all the rest of the team for bringing us repurposing and the WP34s! (the coolest community project on the net, bar none!)




Cheers gang,

[MarkHaysHarris777]

Greetings,
Well, I'm updating 'flash-box' with control circuits and lights. The first set of mods will be the reset ground shot, and erase strobe; the more important of which is the erase strobe. I am providing contact bounce conditioning to the two SPDT momentary 'push' switches so that the ground shot and strobe are clean, and so that the flop is 'holding' the strobe high instead of my finger on a dirty contact switch.
The LED on the reset circuit doubles as the unit power light and blanks when the reset push button is pressed. The LED on the erase circuit comes on when the strobe is toggled high and goes out when the strobe is released.
The control circuits and lights are based on the following schematic:



The following two pics are some angles on the Manhattan bread-boarding so far... monitoring lights and comparators coming soon.





Cheers

[John Galt]

That's a lot of work for a couple LEDs... :-)

I like electronics as much as anyone else, but the Erase status LED is accomplished easily enough by adding it to the Erase input. If you're using the Prolific device for 5V it has more than enough current to drive the LED. I agree with you in that it's inconvenient to have to hold a finger on a button, so I just used a toggle (maintained contact) switch. No flop required.

Reset is more difficult because it's active low, but only momentary contact is required. So for it, use a double pole momentary contact switch instead, with one of its poles "ground" and the other 5V (or 3V if you determined it has enough current for your LED). No electronics required.

Anyway, nice work.

[walter b]

Well, these young engineers have passed simple switches for long, they think in electronics instead.

d;-)

[MarkHaysHarris777]

Greetings,
Today I'm adding Tx Rx activity lighting facilitated by the TI LM339N quad open collector comparator chip, and two yellow LEDs which I chose to mount in the upper right-hand corner of the flash-box just to the right of the cradle. I chose to monitor both lines (Tx, Rx), and each line uses 1/4 LM339. The theory is really simple, each comparator is given a reference voltage from a resistor based voltage divider (on the inverting input of the amp) and the non inverting input is then 'compared' to the reference voltage. If the line (active state low) drops below 2.5v, then the led glows. The current limiting resistors were chosen deliberately high (even though it means dimmer LEDs) to keep the current drain lower.

The Tx Rx LED circuits are based on the following schematic:



In addition to the LM339N (and Tx Rx LEDs) I also added the power switch (DPDT 5.1v & 3.4v) and insulated the connections:



The following two pics give alternate elevations of the new components. I had to move the 3.4v distribution lines to the back side of the Manhattan board.



In this final pic of the LM339 from the right end you can see the mounting for the Tx Rx LEDs, and the 100 mf cap which I decided today to place on the 3.4v line from the FTDI /



Cheers

[MarkHaysHarris777]

Some of you may have noticed that the dandy brilliant LM339N from TI has four open collector comparators and that two of them are hanging around uselessly wondering what their place is in the scheme of things (like many of us). What shall we do with the other two op amps?
Glad you asked.
Using pretty much the same schematic (for the Tx Rx monitors) amp (3) will monitor the reset pin voltage (J32), and amp (4) will monitor the 3.4 Vcc line voltage. If the reset pin (J32) from the 34s goes high (this is a signal that the 34s is powered on) amp (3) will signal this by turning on Q out and lighting our LED (this is going to be our visual indicator, since the screen on the 34s looks 'dead'); when we initiate the reset ground shot, amp (3) Q out will turn off the visual indicator (probably a green LED, haven't decided yet). Amp (4) will monitor the 3.4 Vcc line voltage and turn on a red LED should the voltage drop below 2.8v, or maybe 2.7v, volts. At any rate, this indicator is a visual warning (normally off) that indicates potential problem with the flash, or perhaps even a warning not to proceed with the flash.
Coming soon.

PS I am remaking my Katie clips (sigh). I chose the wrong material. The aluminum is soft enough to work with, hard enough to hold, but apparently too brittle for the long haul; one clip (tip) broke off. Fortunately because of the large pins, the other clip held fine and the transfer completed. But, I now need to repair the Katie clips. I'm going for brass this time (same design) only this time I'm going to reinforce the tips of the clip with a ball of solder on the lower inside. (I'll send pic when done). This should prevent the clips from 'breaking' or bending under pressure of the pogo pins; the real problem, again, is that the pogo pins even with pointed tips need to be held securely (3 mm travel at least) BUT without popping loose, or breaking. Well, I've already been back to the drawing board, now its back to the work bench.

Cheers,

[MarkHaysHarris777]

I'm experimenting with a new Katie clip; this round is a heavier gauge brass sheet (a little harder to work with, but will be absolutely reliable. I was able to produce the tip without visible metal fatigue but I decided to reinforce the tip with solder just the same. Also got the idea to bevel the edge of the tip with side cutters which produces a little sharper gripping edge for contact with the plastic ledge around the pogo pads.




cheers