... | ... | @@ -9,37 +9,40 @@ This page is a life document to decide on the features of the new WR switch. It |
|
|
## Front Panel
|
|
|
| Feature | Proposal | **To be discussed (TBD)** or Comment |
|
|
|
|-----------|-------------------------|-------------------------------|
|
|
|
| Data ports | Number: 18 <br /> Speed: 1 and/or 10 Gbps | **20 ports if space allows?** |
|
|
|
| Mgmt ports| 2 Ethernet (SFP cage + simple RJ45) <br /> USB Mini-B | **Add RS232 port as suggested?** |
|
|
|
| CLK IOs | Type: SMC <br />IN_1: 10 MHz <br />IN_2: 1 PPS <br /> OUT_1: 10 MHz<br /> OUT_2: 1 PPS <br /> OUT_3: 62.5 MHz | **SMA or other instead of SMC, differential, LVDS? [see](uploads/576f996e3ab0f2067d3d615ccad635c1/New_WRS_requirements_from_WR_Community.pdf)**<br />**Add input for calibration/TIC/timescale comparison or re-use existing?**|
|
|
|
| General LEDs| LED_1: Power <br /> LED_2: Status <br /> LED_3: Aux (localization) | **I've added locatiozation led, as suggested**|
|
|
|
| Data ports | Number: 18 <br /> Speed: 1 and/or 10 Gbps | |
|
|
|
| Mgmt ports| 2 Ethernet (SFP cage + simple RJ45) <br /> USB Mini-B | |
|
|
|
| CLK IOs | Type: SMA <br />IN_1: 10 MHz <br />IN_2: 1 PPS <br />IN_3: AUX <br /> OUT_1: 10 MHz<br /> OUT_2: 1 PPS <br /> OUT_3: 62.5 MHz <br /> OUT_4: AUX (abscal) | Check how SMA and AUX IN/OUTSMA can fit <br /> IN_3 AUX could be used in the future for internal TIC... |
|
|
|
| Port LEDs | LED_1: Link / WR mode / Calib <br /> LED_2: Synced / Activity | |
|
|
|
| Status display | Small/standard OLED | Replaces status LEDs|
|
|
|
|
|
|
*NOTE: There is no support for port speeds higher than 10 Gbps (coherent, 25Gbps and 100Gbps) - there is no real requirements for such feature and it would result in substantial cost.*
|
|
|
*NOTE: There is no support for port speeds higher than 10 Gbps (coherent, 25Gbps and 100Gbps) - there is no real requirements for such feature and it would result in substantial cost.*
|
|
|
*NOTE: The chosen FPGA allows for 20 ports, yet we consider 18 sufficient, usually people ask for less.*
|
|
|
|
|
|
|
|
|
## Back panel
|
|
|
| Feature | Proposal | **To be discussed (TBD)** or Comment |
|
|
|
|--------------|-------------------------|-------------------------------|
|
|
|
| Power | Power switch <br /> Power plug | It is easier to have it in the back for the case when optional <br />redundant/swappable power supply is implemented.<br /> **The location was controversial, we could study further**
|
|
|
| Debugging | USB Mini-B FPGA <br /> USB Mini-B CPU | **Not sure whether CPU output is needed/possible**|
|
|
|
| Power | Power switch <br /> Power plug | It is easier and more practical from design view-point, <br /> as well as saver to have it in the back |
|
|
|
| Debugging | USB Mini-B FPGA <br /> USB Mini-B CPU | |
|
|
|
|
|
|
*NOTE: There is no RS232 port in the back (as compared to the previous version of the WR Switch) - it was not really used*
|
|
|
|
|
|
## Power Supply
|
|
|
|
|
|
* Single power supply (not redundant, not swappable) in the basic version of the WR switch
|
|
|
* Standard connector (power/diagnostics) to allows WR switch versions with redundant/swappable power supply (e.g. with [Zippy M1U2-5650V4H](http://www.circotech.com/zippy-m1u2-5650v4h-650w-1u-redundant-power-supply-.html) )
|
|
|
* The basic version of the switch should have single power supply (not redundant, not swappable) and be prepared for redundant/swappable supply, two possibilities to explore
|
|
|
* Standard socket for redundant/swappable power supply already in place in the basic switch version (e.g. with [Zippy M1U2-5650V4H](http://www.circotech.com/zippy-m1u2-5650v4h-650w-1u-redundant-power-supply-.html) )
|
|
|
* Standard connector (power/diagnostics) to allows WR switch versions with redundant/swappable power supply and empty space in the WR switch enclosure
|
|
|
* Enough current for DWDM
|
|
|
|
|
|
*NOTE: There is no redundant/swappable power supply in basic switch version - it would be too expensive ([429$](http://www.circotech.com/zippy-m1u2-5650v4h-650w-1u-redundant-power-supply-.html) vs 80$) and it would induce vendor lock-in. However, the intention is to use a connector that is widely used with (redundant/swappable) power supplies (power and diagnostics) so that such a supply could be used without modification of the main PCB, if required*
|
|
|
|
|
|
## FANs
|
|
|
## FANs and airflow
|
|
|
|
|
|
* Standard FAN(s), not swappable, in the basic version of the WR switch
|
|
|
* Standard connector to allow swappable FANs
|
|
|
* Make a fan-less switch with FANs, i.e. include means that improved heat dissipation from the [Fan-less WR switch](https://www.ohwr.org/project/wrs-fl-hw/wikis/home) while installing FANs
|
|
|
* Provide swappable fans in the basic version of the WR switch (custom mechanics using standard fans to minimize costs)
|
|
|
* Good quality with speed measurement supported by HWm, Fan control in PCB
|
|
|
* Airflow - blow out of device to minimize airflow-induced jitter <br />**(TBD: blow to front/back - likely to the back since there will be no place in the front)**
|
|
|
* Airflow
|
|
|
* Initial idea: blow out air to the back (circulation from front to back) -further study needed
|
|
|
* Add air holes in the front, not to depend on the SFP cages
|
|
|
|
|
|
|
|
|
*NOTE1: There is no swappable FANs in basic switch version - it would be too expensive, increase price at 60 euro per module*
|
... | ... | |