We use a DS90C387-DS90CF388 link for many years to make a display link with a cable of 15m. On the design we set the pre-emphasis level according to our 15m cable.
Now we want to use the same board-pair with different shorter cables (1m, 5m, 10m). With a shorter cable you can expect overshoot at the receiver side (too much pre-emphasis).
Is this allowed or will the receiver be overstressed?
Glad to hear you like the device and have been running 15 meters for years. That is a good length and it sounds like you also have a good cable.
You are correct, for the shorter cables, you ideally want to turn down the pre-emphasis level to obtain the optimual signal quality. I am not too concerned with the over voltage effect to the RX input, it is pretty robust and allows for large VID and rail to rail levels. I suspect at 1meter you might not need it enabled at all, and if you are happy at 15 meter use a setting for 1/3 and 2/3 for the 5 and 10 meters. Ideally use a high bandwidth differential probe (eg TEK P6330) at the RX input to view the resulting signal quality and dial it in for the best eye and minimum overshoot.
Please also make sure your shorter cables are of the same quality as your 15m one. Usually pair-to-pair skew is the limitation on this typeof interface and skew should be minimized.
DPS APPS / SVA
In reply to John Goldie:
Thanks for the reply.
To explain the situation more in detail:
We would like to use the same board (with pre-emphasis resistor setting based on the 15m cable: 100Ohm, Vcc (100%) pre-emphasis) to drive the 1-15m cable-range. So when we use this board with a short cable (e.g. 1m) there will be for sure an overshoot at the receiver side. According to your replay this is allowed. Can you confirm this?
Thanks in advance,
Kind Regards, Stijn.
In reply to Stijn Depraetere:
When you observe the waveform (check also single-endedly and differentially) - if the signal does not go below Ground or exceed VCC, then it is acceptable and not expected to do harm. As noted over-equalization of a link will normally cause excess jitter and reduced sampling margin. If however your settings, media, and resulting signal quality have acceptable margin for your application, then again you should be fine.
If the signals exceed the rails, you are generating extra noise sources and also turning on the ESD protection paths and this scenario is not recommended.
I have seen some cables provide additional signals (for example two pins pulled to a combination of levels) that report back to the host what cable it is, then the proper settings can be set or switched in. I understand your desire to use one set source board and this is not an option at this time. Just sharing how others have self-configured for short, med and long cables to a single host board.
DPS APPS / SVA / www.ti.com
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with respect to these materials. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.