Divergence Meter Project

Links:
Design Considerations
Circuit Schematics
Parts List
Printed Circuit Boards
Warnings
Construction
Troubleshooting
Operating Instructions
Program Code
Photos and Videos
DS3232 Clock Option
B5441A Version
IN-18 Version
Project Status Updates

Introduction

A divergence meter (as seen in the anime Steins;Gate) is a device that will display your current world line number. Mine simulates the device shown in the anime and the visual novel.

This is the second divergence meter I've built. The first one looked nice, but its innards were based on a nixie clock kit circuit board, so it couldn't do fun things like display world line change animation. Also, it was built using techniques that would not make it easy for others to build one. The purpose of this second project was to build a divergence meter that others could build as well.

See my Design Considerations page.

This divergence meter uses IN-14 nixie tubes because they are readily available and inexpensive compared to other tubes. No nixie tube matches exactly the tubes depicted in the anime (IN-18 tubes, for example, don't have decimal points, so they can't display world line numbers). But my circuit should properly drive IN-18 tubes, so I have laid out boards for the IN-18 if anybody is interested (I have not actually had such boards fabricated as yet).

I did not design this meter to be cheap to build -- that would require multiplexing, and I preferred to use direct drive for the tubes in order to get greater brightness, and I wanted a circuit that would NOT have the dreaded "blue dot" problem with driving IN-18 tubes. I have a few nixie clocks, and I prefer the look of the directly-driven displays to the look of multiplexed displays.

Rather than try to design the high voltage circuitry needed to drive the nixie tubes, I chose to use a pre-made HV supply, the Taylor Electronics 1364 (http://www.tayloredge.com/), a nice device at $13.95 (and that price includes shipping).

The meter circuitry uses a PIC16F628A processor, programmed in assembly language. The device incorporates a Dallas DS1307 real-time clock chip that is used to keep the time and date, and has battery backup for the clock.

 

Video of the completed Divergence Meters:

   

Photos

Below is the assembled unit (the little screw heads on the corner edges have not yet been applied). The device uses an external 9-volt "wall wart" power supply, but it can also be run on an internal 9-volt battery. Testing with an alkaline 9V battery resulted in 42 minutes of runtime, so this would not be the normal power source.

Below is the assembled unit seen from the back. On the left in the internal battery switch. In the center is the power connector hole. On the right are two control buttons.

The next picture shows the assembled device unpowered for a sharper view of the components on the top layer. All of the components (except for the tubes) are dummies to match the look of the meter as seen in the anime. The little screw heads in the corners are in place here. (I didn't like the way some perf board holes were peeking out from the edge on the bottom, so I later filled those in with a little epoxy putty and repainted.)

 

Project Status Updates:

10-11-2018: Paul has completed his divergence meter. He used actual nixie driver chips for the dummy ICs on the case, just for a little added veracity. He used plastruct rod for the corner supports instead of hex standoffs – they take screws very well. He had case parts cut by Big Blue Saw, and he has two extra sets of the case metalwork left that he would like to sell - You can find details at http://www.nixies.us/2018/04/27/divergence-meter/
8-31-2018: HoLunW has developed a technique to invert the nipple tip of an IN-14 nixie tube, making the tube look more like the smooth-top nixie tubes shown in the Steins;Gate anime. He heats the nipple of an IN-14, and this allows the partial vacuum inside the tube to draw the softened tip inward. I imagine there is an art to getting this just right, but he's apparently gotten quite good at it.

At right is one of HoLunW's divergence meters built with inverted-tip tubes. He also uses spacers to make the 'tube sockets' more closely resemble the divergence meter as seen in the anime.

Note: I see that HoLun has listed completed divergence meters and case parts for sale on eBay and Etsy. Searching those places for "divergence meter holun" can find those items if you are interested.

HoLunW has also built divergence meters using IN-14 tubes that have a fine-mesh anode grid. These are apparently rare early models of the IN-14 tube (which I never even knew existed). At right you can see a close-up of two fine-grid tubes, one of which he has modified with his nipple inversion technique. Compare the anode grids in these tubes to the coarser-sized hexagonal grid anodes in the tubes on the divergence meter pictured above.

And here is a divergence meter that HoLunW built using a set of fine-grid tubes. Very nice!

7-22-2018: There has been an increase in people building Divergence Meters since the start of Steins;Gate 0. Here are a few more successful reported builds (great work, guys!)...

First, we have Matthew's divergence meter. It works, but he hasn't finished the case yet.

Next is Rudolfo's divergence meter...also sans case.

And here's Max's divergence meter... complete with case!

Tommy not only built his divergence meter and case...

...he also got his divergence meter signed by Tatsuya Matsubara, one of the people who worked on the Steins;Gate show.

4-28-2018: People continue to contact me asking if I have printed circuit boards for sale... I do not. But if I know of people who have contacted me that have extra boards to sell (because that had several made), I put others in touch with them.

I do not currently know anybody who has boards to sell... but Paul Andrews has written to me with a request: He his interested in hearing from anyone who would like to pool resources to get certain parts made for building divergence meters. If you are interested, you can contact him through his blog page: http://www.nixies.us/2018/04/27/divergence-meter/.

 

4-2-2018: Aleksey has completed his divergence meter, shown on the right.

Instead of soldering his IN-14 nixie tubes directly to the tube board, Aleksey soldered socket pins to the tube board (others have also done this), and he soldered the leads of his nixie tubes to socket pins that he then arranged into a socket ring...and these socket pins are what plug into the socket pins on the tube board. I imagine it is easier to plug the tubes so equipped into the tube board that it would be to get all the lead wires into the socket pins on the tube board.

Of course, he spent extra time up front soldering the socket pins onto the leads and arranging them into the socket...but he may have saved time overall if that process was easier for him.

11-15-2017: I haven't updated this page in a while... so here are some new images of completed divergence meters:

On the right is Maurizio's completed divergence meter, which took about two months to finish. He documented his build in an imgur album: https://imgur.com/a/NKhRM

HoLunW made interesting variations of the divergence meter using 3D printed parts to make the cases (click on the image at the right to see a larger view of the parts). The 3D printed sub-structure includes the base, support posts, and battery compartment -- the aluminum case parts and top attach to the sub-structure with screws. He also printed a piece that attaches to the internal battery power switch so that when that switch is slid to the On position, the other end of the plastic piece slides in front of the external power connector (to prevent accidentally switching on the battery while the device is plugged into external power).

Olya completed her standard IN-14 divergence meter AND built a large IN-18 divergenge meter. Great work! She also wanted to thank "Nixie Keith" for supplying her with circuit boards and other parts. If you would also like to discuss getting circuit boards and certain other available parts from Keith, you can email him at nixiekeith@gmail.com.

Ian completed the divergence meter shown on the right. He used pin receptacles so that the IN-14 tubes with short leads could be plugged into the tube board (see the lower photo).

Close up detail of Ian's tube placement in pin sockets. The short leads are threaded through the plastic bases, then inserted into the pin receptacles. This is difficult to do, but he will be able to replace nixie tubes if needed without have to de-solder the old tube and solder in a new one.

5-23-2017: Matteo finished his divergence meter, shown on the right. He says that this divergence meter project was the first time he'd ever picked up a soldering iron, but my online tutorial made it a piece of cake. The divergence meter is not an easy beginner-level project, but Matteo is evidence that it CAN be done. Good work!

2-12-2017: I have completed construction of my IN-18 version of the divergence meter. I have updated the IN-18 version construction page and parts list.

Also shown here is the IN-18 version next to the IN-14 version for a size comparison. This things is huge.

2-9-2017: A couple more successful divergence meter builds have been reported. Mica sent in a picture back in December of the two meters he completed.

Also Olya sent in a picture of her working meter in February (still no case yet).

9-29-2016: Steven built his first divergence meter in 2013... and when he built his second one, he made a time-lapse video of the construction process. Great work, Steve!

 

9-21-2016: Keith built the divergence meter shown on the right, although he used a cigar box case for his first one instead of making a divergence meter-style case. He also has some extra printed circuit boards left, so if you are interested, you can let me (tomtitor@mindspring.com) know and I can put you in contact with him.

9-13-2016: Another successful divergence meter build: Maurizio Giulio has his device working! It doesn't have a case yet, but it does have Dr. Pepper and a metal Upa to keep it company.

9-8-2016: Cori completed his divergence meter, shown at right. He modified the design slightly by using the pins out of a block of sockets (slightly smaller than the ones used for the connectors between the boards) so he can just plug the cut-short leads of the tubes into the pin sockets. This will make it easier to swap out tubes when they eventually die.

Some detail of the socket pins can be seen in the second photo. Buying separate pin sockets (as I did for my B5441A version) is expensive, but getting the pin sockets out of header strips is cheaper (but may require breaking the strip apart if the pin sockets can't be pushed out).

8-14-2016: Brandon made this time-lapse video of the construction of his divergence meter. This is fantastic because you can watch his meter go from bare circuit boards to operational status in under 15 minutes. If only it was that fast to actually build one in real life.

Thanks, Brandon!

3-19-2016: Frank was set on building a version of the Divergence Meter that used the large IN-18 tubes, so he and I went in together on an order for circuit boards. The IN-18 divergence meter is considerably larger that the IN-14 meter, with a width of about 12.4 inches (31.5 cm).
      It turned out that I had a few mistakes in the IN-18 version of the board I made back in 2012, but these were fixable with a jumper wire and a software patch.
      The decimal point (which the IN-18 nixie tube lacks) is simulated with a surface mount LED positioned in front of the appropriate tube...which worked out well. Frank insulated all of the tube socket pins with heatshrink tubing. My copy of the IN-18 DM is not yet completed (just a lack of time here).

3-18-2016: A few more people have sent me pictures of their completed divergence meters during the ten months since I last made an update post, so there they are.

Christopher, Tim, and Pierre each built IN-14 versions of the divergence meter. Christopher's DM is shown in clock mode, while Tim's DM and Pierre's DM (bottom) are displaying world line numbers. Congratulations, guys!

5-10-2015: Zach, another fan of divergence meters, successfully completed his divergence meter (shown at right). He re-created my circuit board designs using a program called DipTrace and uploaded them to OSH Park. He wasn't able to take my Express PCB files and open them with that program, so he went through pad by pad and recreated both the IN-14 nixie board and main board. Because his meter is working great, his circuit boards are probably fine (although I have not seen or tested them myself). All together, Zach paid $118 for three of each board (cheaper than ExpressPCB.com). If anyone would like to place an order for boards from the place Zach used, you can use the links he has shared below:

Main Board: 
https://oshpark.com/shared_projects/OzWVPrP7

IN-14 Board: 
https://oshpark.com/shared_projects/yB0Dg4q4

There is a difference on Zach's boards: He split the 9v battery input trace where my original board has the option of installing a diode. Zach's board has two pads on the bottom side of the board can be bridged together if you don't want to install the diode, and two through hole pads if you do. Zach reports that his version of the boards have silk screen markings, are ENIG coated, and are very high quality boards.

4-25-2015: A fellow divergence meter fan named Thobias paid someone to convert my divergence meter printed circuit board files (which are in ExpressPCB.com's proprietary format) into Gerber format files that he could use to order boards from other companies. He ordered some extra boards to sell, and I pointed recent inquiries about getting boards to him... but he has now sold out of his boards. Because he has sold out, he has granted permission for his files to be posted here.
    This is the link for downloading his files. Note: I do NOT know anything about how he used these files to place an order, so I can't answer any questions about them -- but the files may be useful to some of you for ordering boards from sources other than ExpressPCB.com (if you understand them).

5-16-2013: The last of the aluminum pieces for building the case have been sold. The only thing I have left are the pre-programmed PIC processor chips (since I can always get more chips and burn the program onto them).

11-30-2012: And the prototype circuit boards have been sold as well.

11-25-2012: The last of the full-up printed circuit boards (that have solder masks and silk screen printing) have been sold. All that remains are a few of the prototype boards that I ordered when I built my first prototype (these have no solder masks or silkscreen layer, and have some minor differences from the full-up boards). The prototype boards do not have the option of using the DS3232 clock chip (they can only accomodate the DS1307 clock chip). If you are interested in the prototype boards, let me know (they cost less than the full-up boards at $30 per set of two boards needed).

10-26-2012: SOLD: Somebody asked to buy a set of three HV5622PG driver chips from me (since Mouser was out of stock, and I had three on hand...and I could order three more from Mouser as backordered replacements since I didn't need them right away). So I boxed up a set of the printed circuit boards, pre-programmed PIC processor chips, aluminum case pieces, and the three HV5622PG chips. But he has decided he doesn't want them after all. And Mouser informs me that my backorder has shipped.
    So, if anybody would like that box of parts instead, let me know (some of you have said you have problems locating the HV5622PG drivers locally). The price is $84.80 plus Priority Mail Shipping cost (either domestic or international)
    After that, I have four more sets of the circuit boards left (and a few of the prototype board sets available), and I have case pieces and processors for that many orders as well.

8-5-2012: Another report from a divergence meter builder! Chris's meter is shown at the right in its almost-completed state (still needs the tiny screws on the corners of the case). Nicely done. Once again we learned the importance of being very careful not to accidentally short the high voltage line to anything else when measuring and setting the power supply's output voltage (Chris zapped his PIC chip and needed a new one). The IN-14 tubes that Chris bought did not come with plastic bases, so he soldered them in place without bases, close to the board as shown (not exposing too much wire).
     Chris also had a problem with the clock display acting funny: It would freeze up, but then would start working fine after 20 minutes or so...until powered off and on again, when the strange behavior would reappear. The problem turned out to be a faulty clock backup battery, and everything worked fine after he replaced it with a good battery. So if your clock ever misbehaves, try a new battery.
     Chris also reported that his nixie tubes we displaying multiple digits in a couple cases, and had a hazy display to begin with, but these problems went away after the tubes were run for a while. I've seen this with a few IN-14 tubes as well...maybe some deposits get cooked off of the cathodes after the tubes are run for a while?

7-4-2012: All right...A report from somebody who completed his divergence meter! At the right you can see Justin's completed meter, which he used at a cosplay event. Good work! As a result of his building experience, I have updated the construction directions with a few extra notes. It's very important to be careful to avoid shorting the high-voltage output to any other lines, or you can blow parts. So carefully follow the directions when measuring the HV power supply output voltage, and make sure your test probe does not slip and cause a short.
     Also, I had some incorrect measurements listed in the pattern for drilling the holes in the back plate of the case. I have corrected the diagram...but you should check to see if the listed values match the spacing of your components, since some variation is possible!

5-19-2012: OK, no worries. I glued together another case using the slightly smaller aluminum parts, and everything still fits.

5-16-2012: Well, bugger. I got another order of the aluminum plates made (enough to match the total number of circuit boards I had made), but they were cut slightly under 1.25" tall (something like 1.235" tall). The shop only guarantees their cuts to plus or minus 1/16", so I can't get a refund because this is only about 1/64" undersized. Their cutting was perfect the first times I used this place, so I hadn't expected a problem. Anyway, still I have three sets of the 1.25" tall aluminum plates left for sale.
     Note that if your case pieces are slightly under 1.25" tall, you will need to shorten the height of the corner posts to match. If you are using the hex posts for the corner posts, you can use a metal file to file off a little bit from each end of the posts (the aluminum of the hex posts is soft, so the filing is pretty easy). Also, the slightly shorter cases mean that the fit will be very tight if an internal 9V battery is used, since there was only about 1/64" of clearance to begin with.
     Actually, I also have two sets of plates that I had cut 1.5" tall (from my first order when I was still experimenting with what size case to use), so if you fancy a somewhat taller divergence meter, I can sell you those -- you would need correspondingly taller hex corner posts (each made from two 761-2061-440-AL-7 pieces).

5-12-2012: I modified the software to version 1.05 today. This version supports both the DS1307 and the DS3232 Real Time Clock chips. See the DS3232 Option page.

5-11-2012: Fixed a bug in the error handling routine. This routine displays "666" if there is a problem communicating with the DS1307 real time clock chip. A missing or dead clock backup battery can cause this error. BUT, the error display was not working upon power up because the high voltage power supply had not been enabled yet! This has been fixed. But, in versions of the software prior to 1.04 (including all processor chips shipped out before today), a missing or dead clock backup battery results in a blank tube display upon power up. To see if a dead CR2032 backup battery is causing a blank display, you can replace the backup battery with a wad of folded up aluminum foil in the backup battery holder, which shorts the battery contacts together. The divergence meter will run this way, but won't remember time settings if power is removed, of course. See the Troubleshooting section.
     Version 1.04 also adds a version number display feature. When you press Button 2 to get to the Settings Menus (from Date mode), the version number of the software will be displayed until you release Button 2. So most of the time, you'll bip right past it on your way to the Hours Setting, but if you want to read the version number, just hold down Button 2 as long as you like on the way in to the Settings. Versions of the software prior to 1.04 will not display a version number.

5-9-2012: Updated the Construction page. Also, a few people apparently tried to send me email around May 1 that never got through. If you sent me email and did not get a response, please send it again. Thanks.

5-8-2012: I built another copy of the IN-14 Divergence Meter using the new production printed circuit boards (just to be sure that everything was fine with them)...and so that I could eventually try replacing the DS1307 clock chip with a higher accuracy DS3232 clock chip. The DS3232 will surface mount onto the back of the board. I put the DS1307 in a socket (normally I solder it right to the board) so that I can easily remove it and add some jumper wires when the time comes to switch chips. Changing to the DS3232 will require some software modifications, so it's not just as simple as soldering on the DS3232 instead of the DS1307.
     Oh...and since I made the extra divergence meter, I'll probably sell it on eBay at some point. I'll send email to all of you who asked me about getting a pre-built divergence meter, so that you have a chance to bid on it.

5-1-2012: I now have the printed circuit boards to sell. I will also sell pre-programmed processor chips. I also have some of the aluminum pieces for making the case. See the Parts I Have For Sale section of the Parts page for details.

4-20-2012: Version 1.03 of the program code and hex file uploaded (fixed error in Year-setting routine).

4-13-2012: I have placed an order for circuit boards. These will be the full-up boards: double-sided with plated-through holes and solder masks on both sides, plus silk screen layer on the top side. Once the boards arrive, I will list them and some other items here that you can buy from me for building a divergence meter. The new boards are very similar to the board patterns already posted here except for the following changes:
      I tweaked the tube board a little by routing the high voltage trace and anode resistors further in from the edge of the board to decrease the shock danger while handling the device if it's powered up when it's not in its enclosure. I added some vias so that all traces run to the tube leads on component side of the board (opposite the side the tubes mount on) because I've been told that in the eventual laity that you ever need to replace a bad tube, the hole through-plating can accidentally get removed during the desoldering. With all of the traces running to the pads on the side opposite the tubes, there will be no problem even if the through-plating is removed from the lead holes.
     The main board is basically what was posted already. It has the two resistors I needed that were not on my prototype board, and also has a place where a DS3232 clock chip could be surface mounted on the backside of the board (since this is more accurate than the DS1307 clock chip...but it will require software changes).

3-30-2012: I updated the software to version 1.02 so that the device powers up in clock mode (instead of divergence meter mode). The reason for the change: After a power outage, you would not want the device to come back on (when power is restored) in divergence meter mode and have it display the Steins;Gate world line number for an indefinite length of time, because it is bad for the nixie tubes to display a static number for long periods of time.

3-22-2012: I completed the case for the B5441A version of the Divergence Meter. I used hex posts for the interior corner posts this time, and you can see details of that construction (versus using hardwood corner posts). Also, fixed a bug regarding the restoring the 12/24-hour clock setting from the value stored in EEPROM after power-up.

3-17-2012: Built a divergence meter version for Burroughs B5441A tubes.

3-7-2012: The video of the completed divergence meter has been posted to YouTube. It occurred to me that I didn't have to wait until this damn cold got better so that I could do narration for the video without coughing...I just needed to add descriptive text. So now I'm done.

3-5-2012: The Parts Lists page is done. Let me know if I missed any parts. The final 1.00 version of my program code is now posted. Let me know if you find any bugs.

3-4-2012: Arrggh... I'm sick with a nasty cold this weekend, so I wasn't able to shoot the video showing all the features of the device, but I'll get to it later. I did make a page about the various design choices I made while designing my divergence meter. Also, there was a bug in the time adjustment routine, so I fixed that and I'm running the test again (it seems to be working, chopping off a second after my user-set interval to make up for my DS1307 running a little fast).

3-2-2012: I have finished the programming of all the features I initially intended to include in the device (in both divergence meter mode and clock mode). The last thing I did was program the time adjustment routine that will add or subtract one second from the clock registers after a user-specified number of hours. I was rather disappointed with the performance of the DS1307 Real Time Clock, since mine was gaining one second in about 11 hours. Anyway...testing the time adjustment routine will take a couple days. I have uploaded my first draft of the operating instructions.

2-22-2012: Added routine that animates through World Line numbers seen in the anime and visual novel as you press the button. I'm now adding the option of manually entering any world line number to be displayed on the tubes.
     Also, I noticed that if I touch the ICSP VPP pin while the device is working, it can cause it to reset... so I added a 10K resistor to the back of the main board to pull the RA5/MCLR high to prevent this glitch (it probably would never occur when the unit is in its case, but better to remove the possibility). I didn't have a hole on that line, so I soldered one lead of the resistor to the trace. That resistor is not yet shown in the schematic or board artwork.

2-21-2012: Basic routines are done for setting the time, date, 12-24 hour format, and clock brightness. It's nice that the thing is now displaying the proper time!

2-19-2012: I continue to work on the programming to make the device into a nice clock. There's lots of tedious stuff (I'm writing the user interface routines for changing the clock settings now), and it isn't as exciting as working on the Divergence Meter's world line displays (which I'd like to get back to so that I can add more stuff there). But the clock stuff must be worked on as well, since I expect the device will spend most of its useful life displaying the time.

2-17-2012: This week I got the software to interface with the DS1307 Real-time Clock chip, and I've been doing the programming to make the divergence meter be a clock. Here's a fun tip: If you don't put the CR2032 backup battery in its holder, the DS1307 will not respond reliably (but it does respond intermittently, just to confuse you into wasting hours trying to figure out why you can't talk to it). It also responds reliably if you just stick a wad of folded up aluminum foil into the battery holder to short the battery line to ground...but it will forget the time when you power off with no backup battery, of course.

2-10-2012: I have worked out the software interfacing with the high-voltage serial-to-parallel driver chips that I use to display numbers on the nixie tubes. I wrote a routine to randomly generate and display world line numbers that looks like the animated display you see in the visual novel when world lines change (after studying the animation frame by frame to get the timing correct). The bright pulse you see when the world line animation comes to a halt is accomplished by normally running the display at less than maximum brightness so I can ramp it up for the pulse. To make a nixie tube display dimmer, you can't just feed it less voltage -- you have to very rapidly switch the tubes on and off (faster than humans can detect the flickering) and adjust what fraction of the time the tubes are blanked. The advantage of the directly-driven tubes in my design is that they can be brighter than a multiplexed display, so I can make the pulse bright.

 

Construction Details

The construction photos I took during the assembly of three divergence meters can be seen on the Construction page. Study all of the construction page images and text, and if you have any questions, you can email me at tomtitor@mindspring.com.

 

Circuit schematics

Circuit schematics for the main board and the tube board are below (click to embiggen). This version of the main circuit board incorporates two resistors that were added to the prototype. Note that the tubes are labeled T1-T8 on this schematic, but they were changed to T0-T7 on my final boards (because that is how they are refereed to in the software).

   

 

Warning: HIGH VOLTAGE!

This device uses 170 Volts to drive the nixie tubes, so there is a danger of shock. If you kill yourself, don't go blaming me, because you have been warned. Once the circuitry is safely enclosed, and you make sure that nothing can contact the high voltage lines, it should be safe to handle, of course. This device does not feature the bare wires below the nixie tubes seen in the original divergence meter in the anime (bad design there, Okabe), because the IN-14 tube leads are inside their plastgetting ic bases, which are in contact with the top of the case.

As mentioned above, when testing the naked device (outside of its enclosure), DON'T touch any of the board traces, especially the high voltage supply, the HV line on the 8-pin J1 connector, and the high voltage trace along the back edge of the tube board. The HV pin is pin #7 (numbered from right-to-left in the image below). It's safest to first plug the DC connector into socket J2, and then set down the board, and ONLY THEN plug the DC power supply into the wall outlet. Disconnect in the reverse order. The charges on the high voltage capacitors drain away to a safe level in less than 10 seconds after the device has been disconnected from the power. If I need to access the buttons while testing the naked boards, I lay the device on a non-conducting support (I stack up a few books as needed) with the buttons hanging out over the edge so I can get to them without touching near any of the high voltage traces.

Be especially careful when handling the boards if the 9-volt battery is attached, because you could easily have the 9-volt's power switch in the wrong position (or bump it into the wrong position while handling the boards) and power the device. I recommend connecting the 9-volt battery ONLY once you have the device mounted in its enclosure.

Also, if you have a 9-volt battery installed in the device, DO NOT turn on the power switch from the 9-volt while the device is plugged into the external power adapter, because this would be a Bad Thing as the external power supply tries to run current backwards into the battery. I don't have a diode in there to prevent back flow of current since I didn't want the voltage drop. This device is primarily for use with an external power supply, and the battery is only for brief mobile uses... You should take the 9-volt battery out when you don't intend to use it, and you won't have a problem. (Update: The new printed circuit boards have the option of adding a diode to protect the 9V battery from be being charged if the battery switch is left on while the power adapter is connected. This requires cutting a trace on the board, and then soldering in a diode such as the 1N5817 Schottky Diode to the provided pads. This will cause a voltage drop, and therefore the battery will not last a long before the voltage to the processor drops below its brownout level.)

 

Printed Circuit Boards

Artwork and files for the circuit boards can be found on the PCB page. I designed the boards using the free software from expressPCB.com, but you probably don't want to just go there and order some because they are quite expensive if you just order a small number of them. If you are interested in getting some boards, you can contact me (see the Parts page for a list of components I have for sale). And if you are interested in circuit boards for other nixie tubes (such as the IN-18), we can discuss that.

 

Parts Lists

The Parts Lists page is now done. Let me know if I missed listing some part.

 

Program Code

I did the programming for the Divergence Meter in assembly using Microchip's free MPASM application. Here is the Version-1.05 program code that you can download (it's a plain text file). Don't laugh at my code... this is only the second PIC project I have built, so I'm new to this. If you spot any bugs, please let me know.

Here is the Version-1.05 hex file that can be used to program the PIC16F628A processor (it's also a plain text file). I transferred the hex code file onto the processor using a PICkit 2 Programmer, which attaches to the 6-pin J3 connector of the main board (pin #1 is marked by a small triangle on the board and a similar white triangle on the PICkit device).

 

Operating Instructions

The operating instructions can be found on a separate page.

 

B5441A Version

I did make a version of the divergence meter that uses Burroughs B5441A nixie tubes. The Burroughs B5441A (and corresponding National N5441A) tube has a smooth top (no nipple) like the tubes shown in the anime, and contain decimal points (unlike IN-18 tubes). This makes them a wonderful choice for making a divergence meter. The problem, however, is that it is very hard to find these tubes. I managed to snag a total of 21 used B5441A tubes on eBay after searching rather diligently for four months. This version required making a new tube board and some changes to the software, but it uses the same case and main circuit board as the IN-14 version. If you need the version of the software for the B5441A meter, let me know.

You can see pictures on the B5441A version page.

 

IN-18 Version

An IN-18 version of my divergence meter design has been successfuly built. You can find out more information about this version on the IN-18 version page.

 

 

NOT included:

Things my divergence meter does NOT do:

No cross-fading of digits in Clock Mode. Sure, it could be programmed, but I didn't want to bother.

No automatic Daylight Saving Time corrections. Sure, it could be programmed, but what with all the different DST rules around the world, I didn't want to bother.

Can't display a decimal point in a tube at the same time a digit is being displayed. This is because I have the current limited to the decimal points (to their rated amounts), and if a digit is also being displayed in the tube, the decimal point gets barely enough current to show up. Two decimal points in the same tube (with no digit displaying) will display fine.

 

 

Dedication

This project is dedicated to all the Science Anons on 4chan.org/a/

El Psy Congroo.

 

If you have questions, send me email:

--"Tom Titor" of /a/
tomtitor@mindspring.com