Difference between revisions of "Sony E mount reverse engineer"
Hanskerensky (talk | contribs) m (Extended Hyperlink) |
|||
| Line 47: | Line 47: | ||
The value of this field is approximately 110*sqrt(focal length) at infinity focus! | The value of this field is approximately 110*sqrt(focal length) at infinity focus! | ||
| − | + | [http://elektrotanya.com/sony_nex-5r_5rk_5rl_5ry_ver1.0_level3.pdf/download.html SONY'S SERVICE MANUAL] | |
Revision as of 17:02, 30 June 2016
Sony E mount reverse engineer
This page contains data collected from public postings
http://www.dyxum.com/dforum/emount-electronic-protocol-reverse-engineering_topic119522.html
http://www.dpreview.com/forums/post/56133485 http://www.dpreview.com/forums/post/56587712 http://www.dpreviewer.net/forums/thread/3968755?page=2
E-mount protocol is UART-based, little Endian.
Pinout of the mount is as follows (These pins are from left to right looking at a lens from the rear):
1: LENS_GND - Lens motor GND
2: LENS_POWER - Lens motor power. Can be either 5.0v or unregulated Vbat (7.4v nominal) - This is somehow negotiated between body and lens
3: LOGIC_GND - Ground for lens logic circuitry
4: BODY_VD_LENS - As-of-yet unknown body->lens signalling. In nearly all cases, is normally high but pulses low at a very low duty cycle at 60 Hz (is it 50 in EU?)
5: LOGIC_VCC - Lens logic power. 3.15v All data lines described below are 3.15v logic high
6: LENS_CS_BODY - Handshaking line from lens to body. Always high when RXD is transferring data, normally low
7: RXD - Serial data from lens to body. Starts at 750 kbaud, usually negotiates up to 1.5 Mbaud during init. 8N1 lsb-first
8: TXD - Serial data from body to lens. Same speeds as above
9: BODY_CS_LENS - Handshaking line from body to lens. Always high when TXD is transferring data
10: LENS_XDETECT - (weakly) pulled high by the body. Grounded to LOGIC_GND by the lens, or sometimes via 680 ohm resistor (Viltrox EF-NEX II). Used to detect presence of a lens. Body will not output power unless this is grounded.
Lens ID [[1]]
The main focus position calculations (Packet 0x30 bytes 7/8 and bytes 25/26) appear to be reasonably correct/consistent with all native E-mount lenses I've looked at. These two values are always VERY close to each other and I don't know why there are two fields with almost the exact same data, but Techart seems to have gotten this one right. They are influenced by focal length but not in a linear fashion. They increase as the lens focuses closer (lens element moves away from sensor).
However, there's a third focus position field at bytes 31/32 The value of this field is approximately 110*sqrt(focal length) at infinity focus!
