Parts for WP-34s: What would you like to see?



#2

As I mentioned in a recent post, after founding a company we quickly got around to building a CNC router. I'm sure I will have a chance to actually fabricate mock-ups of OpenRPN designs that I have waited years to actually handle soon enough. However, since what we have as a community is the WP-34s the majority of my effort will go towards enhancing that hardware. Interest in engraved labels seems solid, and as a result a custom engraving bit is undergoing design at the moment. I also need to get myself up and running with one of these machines to directly tinker with it. The far extreme, which I'm open to entertaining, is to move the guts into a new enclosure.

The machine is down at the moment awaiting new, more powerful, stepper motors. We have had excellent success thus far machining plastics. Try to keep suggestions focused on the 34s for now, replacement battery covers and other parts will come along later. I'm just looking for some other side projects to take on that will have my full interest.

Thanks!


#3

Delron seems to be a popular/durable material for a case or the keys.

Is carved Titanium a possiblity?


#4

Quote:
Delron seems to be a popular/durable material for a case or the keys.

Is carved Titanium a possiblity?


I don't know if Delrin is feasible, since durability is important, but it tends to be expensive.

Titanium? Unless you're going to use it for engraved keys (ouch, my fingers) I see no point. It would be very expensive as well.


#5

Delrin (sorry for the typo) is durable and can be machined very precisely.

Titanium is light weight - corrosion and scratch resistant.

So what would I like to see in a WP-34S case? The outer rim (grey) made of rubber for a secure grip and bump protection, the actual case machined out of a slab of titanium, and the back cover and keys machined out of delrin.

Copper might be an alternative metal to consider, although much heavier. It developes a beautiful patina over time and has antimicrobial properties.


#6

Delrin should not pose any difficulties. I already have working designs to fabricate an enclosure using deep drawn aluminum that I can rough cut and then mill to depth. This design includes a rubber "boot" to make it easy to grip, keep it from slipping on surfaces, and giving it a nice smooth overall appearance. Another idea I've batted around involves cutting aluminum extrusions to thickness as side walls and milling the rest from plastic... In this case the aluminum is still functional but part of this concept is cosmetic as well.

Titanium is just a no-go because even if we made enclosures from it no one would buy as soon as they saw the price tag. Realistically, to make such a part from billet would likely cost $200. We don't have ram EDM yet, but exotic hard to machine metals will be a better possibility once it is up and running in our shop. The place is turning into an engineers playground more every day it seems.

Copper is an interesting thought, and the aesthetic would be unique, but I'm not sure everyone wants a calculator that is designed to weather with age. I personally think it would be pretty cool, but I try not to consider myself too much when it comes making a marketable product.

Realistically, aluminum is by far the most likely candidate for viable metals. If anyone really wants to shell out for something tough I can have the case made out of stainless using DMLS and will need to polish it by hand to finish the part.

#7

Being as I am interested in this Hugh I thought I'd offer to help if needed, gratis. I have a tool & cutter grinder with all the attachments. And a lot of experience using one.


#8

Bruce,

This is a very generous offer... Especially considering I know of a few industrial surplus dealers who sell mill bits for the price of scrap (rarely more than a dollar) but that almost always require sharpening. You'll likely be hearing from me.

#9

Hi.

What about a custom case with matching features carved in keys' tops? Hence labels would only be needed for the keyboard faceplate.

Just an idea (and a wish) of mine.

Luiz (Brazil)

Edited: 2 Oct 2012, 3:08 p.m.

#10

I would go for a connector to the programming contacts. This would allow for a simple cable with either an FTDI 3.3V "cable" (the one with the chip built into the USB connector) or for Harald Pott's cable adapter. Many WP 34S users will not want to modify their units (crystal and all) but just flash the firmware and apply the overlay. Another possibility would be a simple cross over cable for unit to unit communication.


#11

Given that Gene is out of cables, this suggestion makes the best sense yet.


- Pauli


#12

I have about 40 left. If you need a cable, contact me... priority would be given for...

1) People who have not yet flashed a 34S

2) People who have but whose cable has broken

3) Then perhaps people who just want a spare.


But, contact me for a cable. Really... no I mean really this time...when these are gone they are gone.


#13

Since I'm finally getting involved in this I'll be contacting you for a cable. Given that quantities are limited, if anyone has a cable that is broken but otherwise has an intact connector it could be useful for mold production or reverse engineering.

From what I can see this will be a matter of sourcing correctly dimensioned pogo pins, cutting a bunch of retaining clips (if they cannot be found pre-fabricated) and finally making a mold. I have access to a product that is normally available only in large quantities for the filter fabrication market, it's a two part polyurethane and is exceptionally durable. If anyone has any thoughts feel free to chime in.

In the long term a solution really needs to be found that will provide WP-34s software suitable hardware to implement everything that could reasonably be desired. This is likely better suited to another thread, but look at projects such as Raspberry Pi and Arduino. Aside from power consumption, all of the necessary hardware is ready to go. In fact, even making a derivative of Raspberry Pi using SRAM or some other low power memory solution will be easy to fit into a calculator sized enclosure and still appeal to other applications where being able to run on minimal power and remain wireless is critical. It was awesome of HP to give users a couple of products to properly hack... But unless the 30b lasts as long as the 12c it will eventually go away and the software will need a new home.


#14

I documented my homemade connector here. While it works fine as-is it would benefit from being embedded in plastic.


#15

Katie; i don't know how i missed that the first pass.

every time i look at one of your kludges i feel like i have nine thumbs and a toe on my hands.

#16

Ah, great to hear this! So there is a chance again to use a real hardware 34s. I just sent an email to your hotmail address which, I hope, is still working. I remember you once mentioned you did not receive the one or other message. ;-)

Dieter


#17

Where abouts in Germany are you Dieter? Maybe I can help you out with flashing if you don't have a cable yet.

I'm near Dortmund.


#18

Ah, fine - I'm just 40 minutes North. ;-) At the moment I will see if I can get a cable from Gene, since - as Walter always says - the 34s is a "moving target" that will require frequent updates. :-)

But thank you for your kind offer. I think I'll get back to you if it turns out I'm too... er... "intellectually challenged" to get my first 30b/34s to work. ;-)

Dieter


#19

The 34S has pretty much stabalised. We're not planning on adding any new features, just bug fixes from here.


- Pauli


#20

I see. So the cable still is required for the bugfixes. ;-)

BTW - I just noticed the Lambert W code I posted on September 23 is included in the current 34s version (lambert_w.wp34s version 3277). But it doesn't seem to be used for the latest emulator, since that version does not return 33 valid digits for x close to -1/e. Could you check some results of the latest version to make sure there is no error in the code?

At the same time the Lambert W code may be improved in that it requires one local register less - the content of R.02 may be kept on the stack. Simply change the following lines:

...		...
51 LamW0_common:: LocR 002
... ...
67 Num eE
68 [times]
69 FILL
70 STO+ X
71 [sqrt]
72 FS?C .00
73 +/-
74 RCL L
75 _INT 003
76 /
77 -
78 // Newton iteration for W+1
79 LamW0_wp1_newton:: ENTER // and x close to -1/e
80 +/-
81 e[^x]-1
82 RCL[times] .00
83 Num eE
84 [times]
85 RCL- Y
86 RCL+ Z
87 RCL/ Y
88 RCL+ Y
89 FS?C .02
90 JMP LamW0_converged
91 CNVG? 06 // test if absolute error < 1E-32
92 SF .02
X[<->]Y // or simply
DROP // [<->]XZZZ
93 DSE .01
94 JMP LamW0_wp1_newton
95 LamW0_converged:: DEC X
96 JMP LamW0_finish
BTW, this code is used both for W0 as well as W-1. Finally, the current lines 56 and 57 may also be removed - a common threshold of -0,35 for both branches should be fine.

Dieter


Edited: 7 Oct 2012, 10:46 a.m.


#21

The code should be the same as you posted. I'll try to check it over again to be sure.

The local registers aren't an issue -- they aren't using the user's RAM at this point -- instead they live in volatile RAM after the xIN command. Step count reduction becomes more important that minimising local register use.

I'll look into the common threshold as well.


- Pauli

#22

I finally got around to comparing the two versions of the lambertW function.

I don't see any difference between the September 23rd code and what is in current builds. I've defined constants for 1/e rather than playing with exponents etc but that shouldn't change things. Likewise switching to local registers and flags.

Oh well :-(


- Pauli


#23

I tried the current emulator version and noticed that arguments very close to -1/e return results with about 30 digits accuracy. On the other hand, my original version, run as a regular user program, achieves about 33 digits accuracy. My first thought was a possible error in the 1/e constant, but I could not find any wrong digits in the file containing the compiled constants.

Something must be different. Maybe you can enter the Sept 23 version as a user program and see what results you will get:

DBLON -1 [e^x] +/- RSD 32
=> Wp = -0,99999 99999 99999 93132 74117 92325 7455
=> Wm = -1,00000 00000 00000 06867 25882 07674 258
These are the exact results. The 34s Lambert W function should return the same result with 33 valid digits. I cannot tell the actual result since my current version now is slightly more exact than before: Wp is +2 ULP off and Wm is exact.

Dieter

#24

The PCBs have been delayed and currently I am on holiday. So I won't have the boards ready for another few weeks I guess.
But so far there has not been that much interest in the DIY cable. There has been more interest by far in the USB and IR boards that can be installed inside the calculator
I have no clue as to why that is though. Maybe people who are prepared to make their own cable, prefer to go one step further and modify the calculator. Or maybe everybody who is interested in the WP34s already has a cable.

The point is: I wouldn't put too much effort into making a professional connector. At least not without evaluating how much demand there would be for such a connector.

Katies solution works well and isn't that difficult to make. (Other people might have a different view on that of course.)
Thanks again for the parts you sent, Katie! - I haven't got round to making my own connector yet though.


#25

We've written about the original programming cable draining the batteries often enough ;-) So every reasonable user will prefer USB power through an additional board within the WP 34s probably.

#26

I'd be interested in the board for making a cable when you get more. I searched online but I could not find anywhere to buy it. I'm presently working on making a cable along the lines suggested in Katie's earlier post.

Hopefully we can collectively sort out the best practices for cable making and induce some retailer to offer all the components together as a kit.

#27

Engraved keys is, for me, the most wanted !


#28

Cool. Aside from requiring complete disassembly and needing to fabricate a fixture to hold the keys in the correct orientation, engraving keys is one of the easiest things possible. If we get really crazy and use some of the smallest mill bits on the market it may be possible to pull off with the machine intact.

#29

Brushed aluminum backing would be nice, rounded towards the front....
Yes, I've just committed sacrilege and described an Ipod Touch form factor. Sorry if that offends anyone!


#30

Had calculators advanced as high end electronic devices into modern times, I'm not sure that you're far from the mark. Apple follows good principles of design, and if similar construction methods were used to an iPhone or Ipod with good tactile keys I doubt anyone could complain within reason.


#31

I would gladly pay $149 for a WP-34s with a better casing and keyboard (no stickies), xtal+ir and printed manual.


Possibly Related Threads...
Thread Author Replies Views Last Post
  [WP-34S] Unfortunate key damage with update to V3 :( svisvanatha 5 611 12-10-2013, 11:37 PM
Last Post: Les Bell
  WP-34S (Emulator Program Load/Save) Barry Mead 1 339 12-09-2013, 05:29 PM
Last Post: Marcus von Cube, Germany
  DIY HP 30b WP 34s serial flash/programming cable Richard Wahl 2 494 12-04-2013, 11:14 AM
Last Post: Barry Mead
  WP 34S/43 ?'s Richard Berler 3 454 11-10-2013, 02:27 AM
Last Post: Walter B
  My FrankenCulator (wp-34s) FORTIN Pascal 4 456 11-09-2013, 06:18 PM
Last Post: FORTIN Pascal
  WP 34S Owner's Handbook Walter B 5 624 11-09-2013, 05:34 PM
Last Post: Harald
  wp 34s overlay and programming. FORTIN Pascal 6 583 11-08-2013, 01:28 PM
Last Post: Nick_S
  m.dy in display of WP-34S Harold A Climer 3 360 11-05-2013, 11:28 AM
Last Post: Andrew Nikitin
  WP 34s : An old problem coming back Miguel Toro 2 340 11-05-2013, 03:00 AM
Last Post: Marcus von Cube, Germany
  [WP 34s] Pressure Conversion Factors Timothy Roche 8 679 11-04-2013, 07:17 PM
Last Post: Dave Shaffer (Arizona)

Forum Jump: