3. TC Modding
After testing that all the parts would work properly when put together, it was time to get down to actually putting the eSATA port into the TC. This required a little more disassembly. The power supply unit (PSU) was simply lifted up out of its position and moved out of the way to begin with as shown in Figure 3-1. The same figure shows a highlighted area in the upper right corner, which is where I wanted to put the eSATA connector. There were plenty of other spots to choose from along the sides or even the top, but I wanted to make this as neat as possible and keep all the ports in the back. It seemed like there was just enough horizontal space to squeeze that eSATA port in between the Kensington security lock and the corner of the TC. Vertically, it looked like the port would fit under the bundle of antennae wires in that corner, which could help secure the connector in place as well.
Of course, in order to start drilling and dremmeling, the logic board had to be removed to eliminate the possibility of metal shavings getting caught in the circuitry. So the three stand-off posts (Figure 3-1) were removed and the PSU was disconnected from the logic board and set aside.
Figure 3-1. PSU moved out of the way. Target area for eSATA port highlighted. Note three posts to be removed.
In order for the eSATA port to fit, the rubber molding had to be removed (Figure 3-2) which left the connector structurally weak. A little hot glue was applied around the junction where the cable meets the connector to reinforce it (Figure 3-3).
Figure 3-2. eSATA cable before and after removing rubber molding.
Figure 3-3. eSATA connector reinforced with hot glue.
Figure 3-4 shows the eSATA port lined up in the desired location of the TC. Note that it will block one of the posts that were removed. This is not a big deal since there is plenty of other support structure built into the TC.
Figure 3-4. eSATA port lined up in desired location.
Four more small screws were removed (Figure 3-5), three that held the logic board down and one that held the antennae wire bundle to the metal support frame. Once these were removed, a small piece of support came out (Figure 3-6). Two metal tabs (Figure 3-6) were bent back slightly to release the logic board and one last connection for the status light was unplugged (Figure 3-7) before the board could be lifted out.
Figure 3-5. Four small screws marked for removal.
Figure 3-6. Four screws and support removed. Note the metal tabs holding the logic board in place.
Figure 3-7. Disconnected status light connector.
Once the board was lifted out, the antennae connections to the Airport card were accessible (Figure 3-8). These were carefully unplugged (Figure 3-9) with needle nose pliers after making a note of the color and order of the connections (G, W, R, B) while minding the thermal grease everywhere. The logic board was then set aside.
Figure 3-8. Logic board removed, but antennae still connected. Thermal grease everywhere.
Figure 3-9. Antennae disconnected. Remember order: G, W, R, B (left to right in this photo).
To find the best position for the port, the external eSATA cable was plugged into the eSATA connector and lined up on the edge of the TC (Figure 3-10). Then, two pilot holes were drilled (Figure 3-11) as a starting point for dremmeling out the porthole. The alignment was double checked with the eSATA connector (Figure 3-12).
Figure 3-10. Used connected eSATA cable to best judge position.
Figure 3-11. Drilled pilot holes.
Figure 3-12. Checked eSATA plug against pilot holes.
After a little work with the dremmel, a hole was made through the plastic and it was time to start working on getting through the metal frame (Figure 3-13). Before working through the metal, some duct tape was placed behind it to catch all the shavings so they didn't get all over the thermal grease (Figure 3-14).
Figure 3-13. Initial porthole, through plastic only thus far.
Figure 3-14. Duct tape used behind the metal to catch shavings while dremmeling.
After several iterations of dremmeling and checking to see if the connection could be made through the porthole, it was done (Figure 3-15). Note that I purposely left side tabs and some metal for the eSATA connector port to catch on (Figure 3-16). Also, the antennae wires were moved out of the way and the little screw mount was bent up. Later, this was used to help hold the eSATA connector in place.
Figure 3-15. Port hole after several iterations of dremmeling.
Figure 3-16. Some metal tabs and edge remain for female eSATA connector to latch on to.
Figure 3-17 shows a better view of the small gap between the plastic and the metal that was later used to hold the eSATA connector in place. The gap was made a little wider just by prying it open using a Phillips head screwdriver. This allowed the metal tab on the long edge of the eSATA connector to fit in nicely (Figure 3-18).
Figure 3-17. The metal tab on the female eSATA connector should fit in the gap between the metal and plastic.
Figure 3-18. eSATA connector tab in the gap.
Once the eSATA connector was in place, a few test plugs and unplugs were carried out to make sure that the support structure would hold (Figure 3-19). Then the internal eSATA connector was removed so that the logic board could be put back in place after clearing out any stray bits of plastic and metal.
Figure 3-19. Tested that the male eSATA plug would connect properly to the female port.
The antennae were reattached before putting the logic board back in. The order of the wiring may not be important, but I reconnected them in the same way just to be on the safe side (Figure 3-9). It might make a difference since the four antennae have different sizes (see Figure 3-6, around the edges of the TC), which may be important to the 2.5 vs. 5.0 GHz frequencies. The tabs that held the logic board in place (Figure 3-6) had to be bent a little by prying with a Phillips head screwdriver in order to re-secure the board in place (Figure 3-20).
Figure 3-20. Prying the tab a little with a screwdriver secures the logic board in place.
Once the logic board was back in place, three screws were used to secure it and the additional support structure (Figure 3-21). The forth screw that held the antennae wire bundle in place was purposely left out for now. The status light connector was re-attached. Two of the three metal support posts were put back, leaving out the one by the newly created eSATA port. The eSATA connector will be in the way of this post, so it was just left out entirely. There is plenty more support structure built in to the TC, so the effect of leaving this post out on the structural integrity isn't anything to worry about.
Figure 3-21. Logic board back in place, additional support structure replaced, three screws put back and status light reconnected.
Next, the eSATA connector was put into place under the bundle of antennae wires (Figure 3-22). This was difficult, but eventually I got it wiggled into place making sure the long-edge tab of the eSATA connector fit into the gap (Figure 3-17) and the short edge tabs of the connector went behind the metal tabs left in the frame (Figure 3-16).
Figure 3-22. eSATA connector fits very snugly under the antennae wires.
Once the eSATA connector was in place, the metal tabs of the TC frame were pushed back and the screw standoff was bent back in place to secure the connector (Figure 3-23).
Figure 3-23. Bent small metal tab (left) and screw-standoff to hold eSATA connector in place.
The smallest tab was pushed back all the way using needle nose pliers and another test plug/unplug was done to confirm the connection would still work (Figure 3-24). After this step, the antennae bundle was screwed back in place, though the screw wouldn't go in quite all the way any more. Lastly, the other long-edge eSATA connector tab was bent back, flush against the metal shielding to get it out of the way of the bottom cover of the TC.
Figure 3-24. With male eSATA plug connected, pushed small metal tab all the way back to hold female connector in place.
At this point, I tried to put the PSU back in and found that the SATA cable was in the way of power board (Figure 3-25). To get around this, the metal bracket, the insulation and shielding of the PSU were trimmed as minimally as possible (Figure 3-26 and Figure 3-27). I was really hoping that this wouldn't affect the PSU performance or cause some electromagnetic interference of the PSU with the SATA data cable. Neither of those turned out to be a problem later on though.
Figure 3-25. SATA cable in the way of the PSU.
Figure 3-26. Minimal trimming of bracket, insulation, and shielding.
Figure 3-27. Trimmed back insulation and shielding exposed PSU internals.
After trimming, the PSU was carefully guided into place, outer end first while routing the SATA cable around it, then setting the other end in place (Figure 3-28). This was not easy and took a couple attempts, but seemed pretty secure once it was done (Figure 3-29).
Figure 3-28. SATA cable being routed around the PSU.
Figure 3-29. PSU back in place with SATA cable routed around it.
Before closing up the TC, I decided to make two more slight alterations. These are completely optional, but I figured it wouldn't take much effort to potentially add a little better cooling to the PSU, since this is a major area of TC failures. First, the temperature sensor was secured to the metal bracket next to the PSU (Figure 3-30). It seemed reasonable that if the PSU starts to get hot, so would this bracket.
Figure 3-30. Relocated temperature sensor closer to PSU.
Second, the cooling fan was flipped over and rotated 90 degrees so that it would blow directly on the PSU (Figure 3-31). Originally the fan was held in place with rubber rivets, but these were not reusable. Flipping the fan over allowed me to use short screws to hold it in place (Figure 3-32). The ones I chose were a bit too large in diameter and I cracked the plastic, though they still held the fan tightly in place. Flipping the fan was also necessary since just rotating it would not have left enough slack on the connector to reattach it to the logic board.
Figure 3-31. Flipped and rotated fan 90 degrees to blow on the PSU.
140540965Figure 3-32. Used screws to hold fan in place.
At this point I thought I was ready to reassemble the TC and test it out to make sure it was still working, but when I went to put the bottom cover back on I realized that I had totally overlooked the fact that the SATA connection to the logic board stood straight up and the bottom cover would not go on (Figure 3-33). While a SATA to eSATA cable with a 90 degree SATA connector exists, I didn't want to have to take out the one I just put in, so I ordered a SATA couple to connect the existing SATA cable with a 90 degree angle in the TC to the SATA/eSATA cable I just put in. Once this was in place, there was plenty of room inside the TC for the extra cabling (Figure 3-34).
Figure 3-33. Reconnected RAID box. Note that internal SATA connection sticks straight up.
Figure 3-34. SATA coupler in place.
The bottom cover was set back in place, but the corner where the eSATA port was installed didn't quite fit right. Some of the metal trim around the corner edge of the bottom cover was trimmed with clippers to make it sit in place properly (Figure 3-35). Also, since one support post was purposely left out earlier, only nine screws were used to reattach the bottom cover.
Figure 3-35. Bottom cover replaced. Highlighted corner of bottom cover needed to be trimmed a bit to sit in place properly. Also, one less screw used since one support post was left out.
The last step in the reassembly was to reattach the rubber bottom after cutting holes in it so that the screws holding the fan in its new orientation didn't make bumps that would cause the TC to sit unevenly (Figure 3-36). No additional adhesive was required to make the rubber bottom stick in place. The fully reassembled TC with eSATA port can be seen in Figure 3-37.
Figure 3-36. Cut holes in the bottom of the rubber base before reattaching it.
Figure 3-37. TC with eSATA completely reassembled.
Everything was reconnected (though the RAID box was still not quite finished), powered up again and put through the paces. The TC functioned properly and all volumes from the RAID box (except of course the "APswap") were mountable. Finally, Time Machine was turned back on and the backup completed without a hitch (Figure 3-38). I also tested that restoring from Time Machine would also work and that went fine as well (Figure 3-39).
Figure 3-38. TC and TM still worked after all that modification.
Figure 3-39. Tested that the backups really worked as well.
During the time it took to complete the TC modding, the eSATA bracket I ordered arrived, so it was time to move on to finishing the RAID box.
Continue to Part 4: RAID Modding and Final Assembly
[Part 1: TC Take Apart and Testing] [Part 2: RAID setup and testing] [Part 3: TC Modding] [Part 4: RAID Modding and Final Assembly]
