TRON Legacy Identity Disc (2023)

[zmaster]

Hello everyone,

I always wanted an Identity Disc and decided to make one. This is proving to be a journey... I figured I'd document some of my progress to help others make their own and/or avoid my mistakes. Don't take this as a tutorial, I'll make wrong choices and change my mind along the way...

Okay, let's start.
The Spin Master toys nowadays are expensive and, in my opinion, looked a bit too cheap. So, I started modelling the disc in Fusion360 with the idea of 3D printing it. By the time I was done with the external design, I head about the new Lightcycle Run discs. Those looked significantly better and I thought they would save me a lot of time. In particular I liked a few things:

• Built-in battery holders
• Clear rings (can light up in any colour, as opposed to the SM toys)
• Inner rings on both sides
• Supposedly bigger size than SM, closer to the "men" discs in the film

The first obstacle was opening the disc. The screws are covered by some plastic bits and can't be removed easily as you can't grip on them to pull them out and it turns out they're also glued. To remove them, I drilled a small hole and run an M3 screw in it, with the thread gripping in the hole I made. When the M3 screw reaches the bottom of the cavity (where the OEM screw is) it starts pushing out the plug. That breaks the glue and then it just comes out.

Here's the inside:

There are some "flaps" keeping the outer ring in place, but luckily they can be removed. The ring has some small lips than lock into the 2 disc sides, so it won't move.
The problem is there there's no clearance between the outer ring and the battery holders. As my goal was to reuse them, I used my dremel to make some space on the edges:

Unfortunately, there's not much space to be made as soon the battery terminals are reached. Obviously, some insulation must be added later...

Now, having a disc that lights up and you can swing around is nice, but I wanted something more "useful". So, I thought I could make some kind of wall-mount and use it as a light fixture in a corridor (not much light needed there and the current one there is too bright). For this reason, I started looking for RGBW led strips.

I initially bought a 5V addressable one, but it was heating up too much for continuous use closed inside the disc (it was getting to about 40°C / 100°F in free air).
So, I ordered a 24V COB RGBW led strip, which would be more efficient. This one doesn't heat much at all, however it's not addressable. Not a deal-breaker for me but also not ideal. The 24V is also a problem. But I'll talk more about the led strips later...

For the inner rings, I figured there's not enough space between the diffuser and the back of the slot to fit an led strip. And, even if there was, you can't bend an led strip in a curve. So, I figured the only practical solution was to remove the plastic from the side of the slot and shine light through there.

The result is pretty good, I think it looks better in real life than in pictures.

...and, that's it for now.

 

[zmaster]

Time for my first big failure. And I'm afraid it won't even be entertaining, but hopefully it will be helpful to someone who will want to attempt this, to avoid making my same mistakes.

As I mentioned, I was keen on taking advantage of the built-in battery holders. Alkaline batteries have a nominal voltage of 1.5V and NiMh 1.2V. There are 4 batteries, so that would mean 6V or 4.8V, which is a decent voltage to work with.
Unfortunately, the nominal voltage is misleading, in particular for alkaline batteries. Here's a discharge curve showing what I mean:

Source (great reference!)

These are the curves for a typical alkaline (red) and NiMh (blue) battery at 2A.
I have not set my mind on a specific led strip yet, but from the ones I have tested I know that they will need about 7-8W. That would be 1.33A at 6V and 2.66A at 3V. So I used the 2A curve as a ballpark reference.

In short, alkaline batteries are terrible at high currents. NiMh would be better, however I wanted it to work with any type.
So, I decided to design my own circuit to work down to 3V (=0.75V per battery, basically empty). Then, better batteries would give a better performance.

Before routing the PCB, I decided to do a real-life test on alkaline batteries. I'm glad I did.
I don't have professional equipment to do a constant-power discharge, so I did a 2A constant-current discharge too. Here's the result:

It turns out the was a very significant voltage drop of about 300mV (at 2A) with the battery holder I was using. The red line is the voltage measured at the load (an iMax B6 mini) and the blue line is the actual battery voltage.
I then did a quick test with the battery holders on the TRON disc and I got about 300mV of drop with those as well.

I did not anticipate such a big drop and this completely threw off my plans. As you can see from the graph above, my design would stop working after about 6 minutes with that alkaline battery (0.75V min). Actually, less than that, because at that low voltage the current draw would be closer to 3A.
Also, after 20 minutes of testing, the battery was hot enough that I couldn't hold my finger on it for more than a couple of seconds.

Things would be better with a NiMh battery. But the voltage drop on the battery holders would still be there and it'd be significant (basically wasting 25% of the power). I'm not sure how much it would heat up.
At this point, I felt like I was fighting too many issues to use these AA batteries, so I decided to scratch the idea and use a Lithium-Ion battery instead. I've got a few left over from a previous project and they're more suitable for high currents.

TLDR:
Assuming 8W of power draw and a minimum voltage of 3V to make things work:
- it's not feasible to use AA alkaline batteries
- it would be possible to use AA NiMh (or AA Lithium) batteries, but with significant power loss on the battery holders and a question mark on battery temperature for prolonged usage

 

[zmaster]

I finished designing the electronics and now I'm waiting to receive the prototypes.
In the meantime I can finish off the mechanical stuff.

After removing the side plastic of the inner rings, there's nothing connecting the inner part and the outer part of the disc, except for a bit where the OEM PCB is, so they're all floppy. I made some 3D-printed brackets to bridge the 2 parts back together, while leaving a gap for the led strip.
Turns out I made them too high and I could no longer fully close the disc. I had printed them in ABS and glued them with a cyanoacrylate gel (basically, superglue, but I chose the gel type so it would fill small gaps). The bond was so strong that, after chopping most of them away, I had to dremel off the remaining glued part.

Before making them again, I decided to think about fitting the Li-ion battery.
Without making big modifications, the only suitable place for the 18650 cell is where the OEM speaker is. Because, even removing the OEM battery holders, the "floor" in that area is higher due to the battery cover, so there's less space available.
I decided to remove the speaker stuff since I'm not planning to have sound effects.

I then designed some 3D-printed supports to hold the battery in place. This wasn't as easy as I thought. The geometry is not trivial and space is tight. After several iterations I got it how I wanted and made the brackets too.

This is the end result:

The battery holders also act as brackets for one side of the disc and have a top side that can be attached with screws to clamp the battery.
The second screw is just to avoid the top side rotating.

I will share the files of the 3D printed parts when I'm done and I'm sure there are no more issues with them... there always seems to be something in the way, hopefully this is the final design.

If you want to design your own, my advice is:

1. The part must lay below the posts along the inner ring, otherwise it will hit the posts on the other side of the disc.
2. Always try to put the 2 sides back together after glueing any part. Because removing them after they're glued is not fun.
3. If you want symmetry, be aware that most mounting screws are not a good reference to use. Instead, look carefully and plan in advance where you want to place everything.

 

[ismaelram21]

The screws that are glued in, is that the same spot with 4 plastic circles on top of the disc? Will I have to drill it out first, then use m3 screw?

 

[zmaster]

The screws that are glued in, is that the same spot with 4 plastic circles on top of the disc? Will I have to drill it out first, then use m3 screw?

There are no glued screws, it's the 4 screw covers (the "4 plastic circles") that are glued.
You basically have to drill them out, either completely or enough to put something in to pull them out.

This is what I did. The M3 screw once reached the bottom (existing screw) pushed the cap out.
Obviously this only works if the hole is tight enough for the M3 screw to grip on it.

 

[zmaster]

Soooo... it's finally done. I had a lot of trouble with my electronic board, particularly with the switching regulator. I'll talk about it in my next post, but for now here are some pics and video of the final result.

 

[MattMunson]

This was an amazing read and the results are phenomenal. Great work!

What was the final verdict on the size of the new disk? Was it indeed bigger? Better? Mo' Accurate?

 

[Kylash]

Any chance you were able to mess with the SD card slot? I tried a few cards in it and occationally got the original chips to read on a reader but not long enough to see whats on them or try messing with other sound files.

 

[zmaster]

This was an amazing read and the results are phenomenal. Great work!

What was the final verdict on the size of the new disk? Was it indeed bigger? Better? Mo' Accurate?

I never had a Spin Master disc, so I can't compare, but Kylash (funny enough, next one to comment) made a comparison in this YouTube video.

Any chance you were able to mess with the SD card slot? I tried a few cards in it and occationally got the original chips to read on a reader but not long enough to see whats on them or try messing with other sound files.

I don't live in the US, I bought just the disc on eBay, so I don't have any card to play with. Looking at one of your videos, I'm not even sure they're SD cards. They look more like MMC, but even those don't seem to have the contacts quite in the same place.
Edit: I take it back, the connector has contacts that seem to match an SD card.
Why are you interested in them exactly? Replace the sounds?

 

[Kylash]

Yeah, I believe thats what they use at Disneyland to make the custom sounds for people that they record, and change the light colors. Theres some data on the cards that they use for that. But i couldnt get them to read.

 

[zmaster]

Here are more technical details about the build. There would be a lot to say, I'll try to keep it short but informative.

LED strips​

I've tried 3 different LED strips:

• 24V COB RGBW, non-addressable
• 12V COB RGB, addressable in groups of 5cm (about 4")
• 5V RGBW, addressable each led

They roughly used the same power. Higher voltages are more efficient (= less heat).
Based on my testing, the type you want to use depends on what you're looking for. I suggest:

• If you want spinning/loading effects, use 5V individually-addressable LED strips.
  The 12V/24V strips are addressable in groups. I tried to make a spinning effect with my 12V one, with several different strategies (spin just 1 group, 2 opposite groups, 4 groups, fading groups in between, etc). Nothing really worked well.
• If you want static, 12V or 24V non-addressable LED strips are better IMHO.

Whichever you choose, I recommend COB strips, otherwise individual LEDs will really stand out in the outer ring (they're still visible on a COB strip, but they're very close, so I'm okay with that).

Another thing to consider is the thickness. The strip I used is 1.8 mm thick and it barely fits without fully removing the OEM battery holders (it's COB, the silicone in the center compresses a bit, unlike classic WSxxx strips).
My other COB strip is 3.2 mm and the non-COB one is 2.2 mm.
All of these measurements are with adhesive still on. I installed mine without it, so we're probably talking about 1.5 mm.
So, you may need to take off those AA battery holders completely.

Control board​

I designed my own circuit board. I don't know if there are off-the-shelf boards that can be used, I did search a bit without luck, but maybe there are specialised stores for props. But I also liked the idea of making my own, so that's what I decided to do.

Unfortunately, while the result is functional, I made several mistakes. I was planning to share my design, but it's just not good enough as I had to use some tricks to workaround some design choices.
So, I will instead share my "lessons learnt", to avoid others making my same errors.

WARNING: this is very technical, boring stuff ahead

General​

• Li-ion batteries have a minimum voltage, when they're empty, around 3V. You need to account for 200-300mV of drop between wires, battery protections, MOSFETs, etc. So, IMHO you either need to power your logic with an LDO from Vin to 3V or boost the voltage and use a linear regulator from there.
  If you decide to power your logic from 3V, be aware that some (most?) MCUs require 3.3V to make the USB communication work (ie. the Raspberry Pi Pico). So, that may affect your choice.
• I highly advise to add a battery protector. They're cheap and can be found on eBay/Amazon, though I had to go to Aliexpress to pick one with the specific overcurrent rating I wanted.
• Coming from EAGLE, KiCad is nice and mature

Switching voltage regulator​

• If you want to use a 24V LED strip, I recommend using two Li-ion batteries. I used only 1 and the problem is that the duty cycle of the boost controller will be about 90%. That's at the limit of what they can do, it will work for a steady state, but it has very little room for correction on a load transient, making it poorly responsive.
• Talking about load transients... if you drive the LEDs with a PWM, that's exactly what it will cause.
  Be aware that the Texas Instruments POWER Designer tool only determines the size of the output capacitor based on the desired output ripple voltage. THIS IS A PROBLEM. If the LEDs turn fully off, the best the controller can do is stop charging the inductor. But it's already charged and it will dump all its energy on the capacitor. That can cause an overshoot of several volts if the cap is too small.
  So, you need to consider what will happen on 0%-100% and 100%-0% load transients.
• Keep in mind that ceramic caps derate a lot when they have a DC bias. And, while the TI tool takes that into account, not all caps derate the same way. I eventually figured my output capacitors would have an effective capacity of just 10% the rated value!
• I suggest to use fast PWMs, otherwise you'll need huge caps that probably won't even fit.
• And finally there's the fun part. Compensating the control loop. This is a bit of black magic to me, but you'll probably find tools from the manufacturer to calculate the bode plots of the transfer function and check for the stability criteria. Ok, I'm making it sound too scary... it's not THAT bad, but don't make my mistake of going 3V->24V. Otherwise, as mentioned above, you're putting yourself in a corner.
• When you design the battery overcurrent protection, you may think that the maximum input current will be ILAVG + ∆IL/2, but that's only for the steady state. It will be higher during load transients. So, look at the IC maximum switching current instead.

TLDR: designing and tuning switching power supplies is hard.

Button and power switch​

• I re-used the OEM power switch. Since I don't know its rated current, I used it to control a power MOSFET instead of making it bear the input current. I suggest you do the same. I advise picking one with a very low RDS, I used an AON7423.
  The current leak on the gate is so low that it will not discharge the battery (I estimated it'd take more than 100 years).
• I also re-used the OEM rubbery push-button, because I liked the feeling. It works great!
  You just have to draw the "2 nested forks" footprint on the PCB and use gold-plating to avoid oxidation (I used ENIG... good enough).

Li-ion battery charger​

• Charging from the USB, if you want to do it properly, you can only pull 500mA without doing complex stuff to negotiate more power, which is a can of worms. That translates to slow charge (ie. 6h for my 3000mAh battery).
  If you're only going to use wall chargers, you can pull more... at your risk.
• In general, without doing very complicated stuff like USB-C Power Delivery, you won't have enough power to charge the battery AND run the LEDs. So, you'll have to enable charging only when the disc is off.
• I used a linear charge IC that I saw on several other designs, the MCP7383x. It's very simple to use but it sucks. It heats up a lot and using big copper pours on the PCB is almost useless due to the very poor thermal condoctivity of the package. When the battery is low (biggest voltage delta = most heat to dissipate), the thing gets burning hot in a few SECONDS and it's essentially operating in thermal limit instead of current limit.
  I advise to use a switching charge IC instead.

Dimension checking​

• If you have a 3D printer, I suggest printing a 3D model of the board to check fitting. But, unlike me, you should also model the height for the mating connectors and wires... otherwise when you get to close the disc... it won't. And you'll have to bodge things up to make them fit.

Okay, enough rambling. I hope I wrote down everything I learnt doing this.

 

[zmaster]

Small update: I made a wall mount for the disc.

Initially I was going to make something like the back mount on the film, holding the disc with magnets. Then, due to limits of my 3D printer, I thought about making something similar to the holder in the Lightcycle Run Backpack, with spring-loaded bits. Eventually, I told myself I was overthinking it and a simple bracket grabbing on the inner ring slot would work just fine.

So, here's the disc mounted above my TRON Legacy "Frome".
By the way, I don't recommend the Frome because the canvas is very reflective, ruining the dark colours. It could be fine if light comes from the side though.

 

[zmaster]

STL files​

Here are the STL files for the 3D printed parts.

Clip	Clip to support the plastic around the inner ring after cutting a slot to let the light through.
BattClip*	Battery holder.
PcbIsolator	A thin sheet of plastic to put under the PCB to isolate it from the LED strips for the inner ring.
WallMount	The wall mounting bracket.

 

[devsfan1830]

STL files​

Here are the STL files for the 3D printed parts.

Clip	Clip to support the plastic around the inner ring after cutting a slot to let the light through.
BattClip*	Battery holder.
PcbIsolator	A thin sheet of plastic to put under the PCB to isolate it from the LED strips for the inner ring.
WallMount	The wall mounting bracket.

If ya haven't already, might I suggest uploading at least the wall mount to Printables.com and/or Thingiverse.com? I had been looking for something to mount my disc and there is literally NOTHING on either for this. I found yours by complete accident lol.

 

[zmaster]

If ya haven't already, might I suggest uploading at least the wall mount to Printables.com and/or Thingiverse.com? I had been looking for something to mount my disc and there is literally NOTHING on either for this. I found yours by complete accident lol.

Okay, I suppose that's of sufficient general interest, as opposed to the other parts. Thanks for the suggestion.
I just uploaded the wall mount to Printables. Tried Thingiverse too but it keeps giving me 404 errors and behaving erratically.

 

[devsfan1830]

Okay, I suppose that's of sufficient general interest, as opposed to the other parts. Thanks for the suggestion.
I just uploaded the wall mount to Printables. Tried Thingiverse too but it keeps giving me 404 errors and behaving erratically.

Sounds par for the course for Thingiverse lol. Thanks!

 

[Sundowner]

Okay, I suppose that's of sufficient general interest, as opposed to the other parts. Thanks for the suggestion.
I just uploaded the wall mount to Printables. Tried Thingiverse too but it keeps giving me 404 errors and behaving erratically.

Haha I made the switch to printables for my designs for the same reason.

Your disc looks great! I scored a new 2023 Park Disc for $70! The Parks disc is larger and in my opinion better of the two. So the SM will do for my Daft Punk costume build.

 

[devsfan1830]

Okay, I suppose that's of sufficient general interest, as opposed to the other parts. Thanks for the suggestion.
I just uploaded the wall mount to Printables. Tried Thingiverse too but it keeps giving me 404 errors and behaving erratically.

Working on a little something based on your wall mount. I plan to upload it to Printables as a remix when I finish it. Waiting on a piece of cut acrylic for it.

https://imgur.com/a/48wbuVt

 

[zmaster]

Working on a little something based on your wall mount. I plan to upload it to Printables as a remix when I finish it. Waiting on a piece of cut acrylic for it.

https://imgur.com/a/48wbuVt

Looks good

II would consider adding some weight on the front of the base for extra stability (metal? sand?). I did something vaguely similar once and it was very stable in one direction but would flip fairly easily backwards, because the weight was not centred in the base.

Also, It's up to you, but no need to upload as a remix. You did most of the design yourself

 

[Sundowner]

I've been trying to print the mount (thank you by the way) but I can't save the gcode after slicing. I keep getting an error but it could be on my end