I tried to hack my Ultegra Di2 but it didn't work. I tried the aftermarket battery in the seat post wired into the the wire for the battery that goes to the junction box. It seems the new CANbus protocol won't easily allow it. I dremeled the battery mount to try an get the electronics out to put it in line w/ the battery to make it work, but the little electronic boards are glued in and encased pretty securely so it wasn't possible. So what next? Admit defeat and use a stock battery mount and battery? Please don't let me suffer like this:) Somebody help a fellow WW out!!

How about just keeping the entire battery clamp mechanism without attempting to dremel out the circuit board (cut off the cage mounts etc just keep it as small as possible but still functional), then solder a couple of wires from the battery holder terminals and lead them to a custom 7.4v source. Pop the battery clamp down the seat tube with some extensions to the junction box, and up to your own custom battery.

_________________
Di2x2

What about that Icarus Lights dude. He thought his ultegra seatpost batteries would be plug & play. Maybe email him, google Icarus Lights.

I just ordered another battery mount. I can plug it into the internal junction box and just leave it in front of the BB (in the down tube). I'll do what you suggested and solder wires to the battery mount. I'll just make those wires long and have my mini tomiya connector at the end to plug into my tenergy battery.

I e-mailed the folks at icarus lights twice and haven't received a reply yet. They state on their website that they will soon have an Ultegra Di2 battery. That would be easier. Maybe they will reply next week and we'll see what's up.

The other thought I had is about a satellite shifter. Since there are 2 plug ins on the shifter, the extra one can be used for a satellite shifter. I wonder if it too needs the CANbus protocol or if the satellite shifter just sends a simple up or down shift signal and then the shifter adds the CANbus protocol?

I really appreciate all your guy's input. It's very helpul

Interested to see how this works out for you. I'm waiting for a di2 frame, tossing up between ultegra and dura. Seatpost battery and satellite shifts are a priority so please update with your progress. Good to know the ultegra at least have ports for satellites.

I'll let you if the battery mount soldered to connectors for tenergy battery work as soon as I do it. I should get another battery mount sometime this week. I'll get an extra Di2 wire to have the connector to solder it to an after market satellite switch too. Should know if they work next week:)

I've not worked with any of this Di2 stuff to know fully what's going on, but I can state that there's nothing in the CANbus interface that should have anything to prevent you from hooking up a different battery. 7.4V is 7.4V. Unless there's some kind of authentication chip being used to prevent you from using aftermarket batteries, if you hook up the right terminals it should work.

I think there is a chip. I connected everything well and soldered the connections fine.

So I just saw on Calfee's facebook page that they use basically the same battery for Dura ace and Ultegra Di2. The Ultegra Di2 however needs their Ultegra Di2 "module". So there you have it. That module is similar to the battery mount from shimano in that there is a chip of some sort that identifies itself as the battery or communicates w/ the FD (brain) via CANbus protocol. So we need a computer guru to figure out how to make a chip like Calfee did so we can do our own hacks. For now, I'll just wire my battery to the new battery mount I have coming. BTW, that stock battery mount alone felt very light:)

I ran into the same issue trying to use some very small batteries that will slip inside the frame near the bottom bracket of my Tarmac SL4. I connected my batteries through the battery mount (with alligator clips) and this does work, verifying that the ID chip is in the battery mount.

The mount itself is 24.5 grams without the easily removed hardware. It is much too big to go inside my frame and is glued together - I don't think it can be taken apart without cutting it up. It may be possible to remove the circuit board by destroying the plastic housing.

Besides the cludgy stock mount and saving battery weight (mine are 21 grams for the two) I would like to be able to play around with additional shifter switches. I connected an oscilloscope and the signals are evident on top of the battery voltage. I can't tell if it is actually a CAN signal - it says so in various places on the internet but it is difficult to tell if one person made this up and got copied. If so it would not be very difficult to hack.

Chiva, please keep us posted on dismantling the battery mount!

I don't know if this image will work.

Can you post a picture of the scope capture chipg? You said the signals were evident, can you describe how?

CAN is actually a high level protocol specification that makes minimal demands on the underlying signaling. Basically it just requires some kind of open-collector type arrangement like i2c etc for collision detection on a single or pair of shared bus lines. There are various standards for the physical layer, none of which this implementation could be following. It's something homegrown by shimano to be sure.

One more thing; in some parts shimano has sealed them by filling the plastic housing with epoxy. This is a common technique also called potting in the electronics world. It nicely waterproofs things and protects them from physical shock. But it means you may not be able to meaningfully deconstruct this mount. Keep that in mind as you attempt to disassemble. If you find a wad of plastic you've reached the center of the tootsie pop. Going further will likely just destroy what's encased inside.

I just see a bunch of small, evenly spaced, voltage spikes and than a a bigger drop that I presume is the motor coming on. I did not trigger it very well. I took the signal off of one of the extra connectors in the shift levers. I am assuming the signal is the same everywhere as there are only two wires. Pedaling with one hand, shifting with the other, trying to watch the scope.

I believe the cpu is in the front derailleur but have not verified this.

Unfortunately I am having too much fun riding the bike to do much right now. I may get some more data Wednesday.

Do you know what the level of the voltage spikes was?

A half volt or less as I remember. I did not try to resolve it well and may have been clipping a lot off if it is a fast signal.