TCSM2 interchassis cable » History » Version 4
falconia, 08/15/2024 08:22 PM
1 | 1 | falconia | h1. TCSM2 interchassis cable |
---|---|---|---|
2 | |||
3 | 4 | falconia | The smallest functional TCSM2 system consists of one [[TCSM2_TC1C_chassis|TC1C chassis]] and one [[TCSM2_ET1TC_chassis|ET1TC chassis]]; each chassis has its own backplane. A set of signals needs to be interconnected between the [[TCSM2_TRCO_card|TRCO card]] in the TC1C chassis and the [[TCSM2_ET_modules|ET module]] in the ET1TC chassis - hence a special cable is needed that connects the two backplanes. On the back of each chassis there is a DIN 41612 connector (size C, 96 pins total) with male pins - the interchassis cable attaches to these two connectors, S00 on TC1C side and V00 on ET1TC side. |
4 | 1 | falconia | |
5 | In a large system, each ET1TC chassis supports two TC1C chassis, hence whatever official cables Nokia once had (we haven't seen them) must have had a complex structure, with one half of V00 connector on ET1TC connected to one TC1C chassis and the other half connected to a different TC1C. However, if a hacker is putting together a minimal TCSM2 system in her lab and doesn't have Nokia's official cable, a simpler cable will need to be hand-crafted that connects the single TC1C to the single ET1TC. |
||
6 | |||
7 | 2 | falconia | h2. Homebrew cable project |
8 | |||
9 | 3 | falconia | At first glance, the problem seemed simple: because the male DIN 41612 connectors on the exterior back faces of TC1C and ET1TC backplanes are of "full C" size, as opposed to "1/2 C" DIN 41612 connector on [[TCSM2_ET_modules|ET2E]], the availability of mating connector parts (female DIN 41612, free-hanging inline for mounting on cables) is a little better; several options are available from Digi-Key _without_ cost-prohibitive MOQ. For our setup in Themyscira lab, we are currently using "Harting 09032966823":https://www.digikey.com/en/products/detail/harting/09032966823/3179778 for the connector body and "Harting 09030960501":https://www.digikey.com/en/products/detail/harting/09030960501/3179722 for the shell housing. |
10 | 2 | falconia | |
11 | However, there is one more obstacle: the connectors used by Nokia (on the exterior back faces of their backplanes) have strange plastic tabs on the inside, in the space where the body of the plug goes, as visible in "these photos":https://www.freecalypso.org/members/falcon/pictures/Nokia_TCSM2/ET1TC_backplane/ of ET1TC backplane. As a result, standard mating connectors like the Harting part linked above (cable-mounted plug, female contacts) cannot be inserted. It is absolutely not clear what kind of connectors Nokia used on their never-seen official cables, where to look for such connector parts, and even what they are (or were) called. |
||
12 | |||
13 | 3 | falconia | I (@falconia) thought about modding the plug connector, cutting or filing grooves into the plastic body so it would fit with Nokia's plastic tabs intact, but the task seemed daunting, particularly the precision that would be required. So I ended up agreeing to a bit of destructive surgery on the original Nokia parts: using a little flat screwdriver, I broke away the offending plastic tabs on TC1C S00 and ET1TC V00 backplane connectors, and now a standard DIN 41612 cable plug (or at least the Harting plugs we are using) inserts just fine. |
14 | 2 | falconia | |
15 | h2. Choice of wiring for the cable |
||
16 | |||
17 | Because of the way TC1C S00 and ET1TC V00 backplane connectors are wired, there are two viable options for wiring the homebrew interchassis cable: |
||
18 | |||
19 | # If the cable is wired one-to-one, such that the two ends of the constructed cable will be freely interchangeable (either end can go into TC1C or ET1TC), the single ET module will need to be inserted into ET1TC slot 04 (not 00), officially meant for second TC1C in a larger system. |
||
20 | # If the desire is to connect the single TC1C as "TC1C 0" to ET1TC, such that the single ET module will sit in slot 00, the cable would need to be wired diagonally, with distinct (not interchangeable) ends for TC1C and ET1TC. |
||
21 | 1 | falconia | |
22 | In either case, if only one ET module is to be used and not all 4, the number of connected wires in the interchassis cable can be reduced, excluding those that go only to higher-numbered ET slots. Please refer to the pinouts in "tcsm2.pdf":https://www.freecalypso.org/pub/GSM/Nokia_TCSM/tcsm2.pdf manual. |
||
23 | 3 | falconia | |
24 | h2. Interchassis cable in Themyscira lab: current status |
||
25 | |||
26 | The necessary cable was assembled on 2024-08-01, cable design by @falconia, soldering done by her day-job coworker: |
||
27 | |||
28 | https://www.freecalypso.org/members/falcon/pictures/Nokia_TCSM2/interchassis_cable.jpeg |
||
29 | |||
30 | 4 | falconia | Using the ET2E-C (C08763) module that arrived on 2024-08-12, with this module inserted into our ET1TC in slot 04 (our cable is wired for this option), our TRCO is successfully communicating with it: the built-in "Loop Test 1" is passing. This result is better than what we were getting with our initially-bought ET2E-T (C105507) module - thus it appears that our DIY interchassis cable is correct and working after all. |
31 | 3 | falconia | |
32 | 4 | falconia | The only interchassis signals that still remain untested are the clocks recovered from incoming E1 signals by the ET2E and fed to the [[TCSM2_timing|timing block]] in the TRCO. These signals will only be tested once we are able to connect the two E1 circuits between whichever ET2E module we use and our [[e1-t1-adapter:icE1usb]] - see #6464. |