Close

Mini DisplayPort - The Headache

A project log for D-DAQ

automotive parameter & performance monitor & logger

michael-obrienMichael O'Brien 06/07/2014 at 02:430 Comments

Two days ago I noted that the mDP connectors are reversed, and I forgot to confirm their appropriate pinout before including them in on the project. I've already placed an order through a different supplier, ironically located near my place of work, for the correct orientation receptacles and today is when I set apart some time to figure out who named the right pinout. I knew this would be a bit of a headache, but hopefully not too much so. With the help of some "helping hands" I popped on two receptacles on the ends of the cable and switched my DMM to conductivity mode. Before I started mapping out the pinout, I looked back at the spec for DisplayPort to see what was supposed to happen inside the cable because I had read up on it being a cross over cable. I had 2 sources to reference: Wikipedia/Apple and displayport.net.

I mapped out both sources on a spreadsheet, used VLOOKUP to pull the names of signals so I didn't have to do this part by hand. My first reference is the footnote at the bottom of the pinout table on wikipedia stating:

This is the pinout for the source-side connector; the sink-side connector pinout will have lanes 0–3 reversed in order, i.e. lane 3 will be on pin 3 and 5 while lane 0 will be on pin 10 and 12.

My next reference, a little more detailed, was this graphic that I've seen dozens of times:

What the quote leaves out is what is shown in the graphic. The differential pairs are not treated as a single signal, but as individual conductors, thus the positive and negative lines are swapped. Next bit is the fact that the graphic is wrong. Pin 1 is defined as a ground in the spec. Wikipedia's first reference is Apple's 2008 DPF defining the specification, pin layout, and mechanical aspects of a cable and receptacle. Well, page 2 of that PDF has the proper pin layout, which corrects the above graphic when considering Mini DisplayPort. If you ever see if again there are 3 bits of info that are accurate: DP_PWR is a NC pin, i.e. the cable doesn't have a conductor for it, the AUX channels and Config channels do not change pins between source and sink.

Well, after mapping everything out and then flipping the pins, I came across 2 problems with using the reversed receptacle: the HDP line is connected to ground and the 14V line is connected to ground. I'll post later than the latter doesn't concern me with the Proto 1 board, but at the same time, it's a safety problem and also would make the board short out as soon as a cable is plugged into it. On the flip side, out of the 14 conductors available in the cable, I'm only using 12! I can simply reroute these 2 traces to the respectively unused pins and then I can use either of the receptacles in case there is a supply issue later down the road. Yes, it changes the pinout of the signals coming out of the other end, but at the same time, I'm using an adapter board to have the cable plug into the display board at a right angle. Just need 2 versions of that now ;).

Discussions