Milkymist One RC1 Reports
Contents |
Initial Test on PCBA without any software
Below is shown that normal measured voltages and currents on RC1 boards without any software porting. This info is useful for diagnosis after EMS manufacturing or realization in studying/analyzing dissipation during standby mode which most parts are at quiescent or power shut down status.
Measurement of Voltage/Current during Standby/Quiescent Mode
Following data are measured from #1 & #4 pcba. The location(or reference) you can check them from RC1 Schematic. The method of measuring current is to taking apart related ferrite bead on dedicated power loop, soldering two wires on pads of ferrite bead and using Agilent 34401A digital multimeter.
| Location/Reference of Measured Voltage/Current | #1 PCBA | #4 PCBA | Spec. Reference | Datasheet
|
|---|---|---|---|---|
| L1, U1(3V3) | 15.6mA | 14mA | Typical Spec. 19mA | Page 3 of datasheet |
| L2, U1(AUDIO_AVDD) | 54mA | 51mA | Typical Spec. 53mA | Page 3 of datasheet |
| L4, U2(ETH_A3V3) | 4.7mA | 4.8mA | Not clear in datasheet | [1] |
| L5, U2(ETH_1V8) | 23.3mA | 23.5mA | Not clear in datasheet | [2] |
| L6, U2(ETH_PLL1V8) | 13.1mA | 13.1mA | Not clear in datasheet | [3] |
| L7, TP1(3V3) | 3.318V / 0.381A | 3.309V / 0.379A | [ ] | |
| L8, TP2(2V5) | 2.53V / 13.5mA | 2.538V / 13.5mA | [ ] | |
| L9, TP3(1V8) | 1.809V / 62.4mA | 1.823V / 61.1mA | [ ] | |
| L10, DCJACK(5V0) | 5.0V / 0.605A | 5.0V / 0.597A | [ ] | |
| L12, TP4(1V2) | 1.231V / 36.6mA | 1.228V / 33.5mA | 1V2 only used in fpga, Typical Spec. 15mA | Page 7 of datasheet |
| L13, U18(VAA) | 67.4mA | 67.6mA | Typical Spec. 67mA | Page 3 of datasheet |
| L17, U21(VIDEOIN_A3V3) | 170mA | 170mA | Typical Spec. 180mA(YPrPb) | Page 6 of datasheet |
| L18, U21(VIDEOIN_A1V8) | 11.3mA | 11.4mA | Typical Spec. 10.5mA | Page 6 of datasheet |
| L19, U21(VIDEO_AGND) | 99mA | 100.4mA | Reference | [ ] |
| TP5(AUDIO_VREF) | 2.232V | 2.227V | Typical Spec. 2.16V | Page 3 of datasheet |
| TP10(VD5) | 1.253V | 1.253V | Max. 1.226V | Page 3 of datasheet |
The original ods file is here.
Power Tree for Quiescent Mode
- Based on above information, a plot can be achieved as below:
- During current measurement, we got a strange phenomenon which is high current 120mA thus 6.7 times of typical quiescent 18mA definded in ds162. Later we got offical Xilinx feedbacks as below:
Q1. Accourding ds162.pdf, Iccintq typical quescent current of XC6S45-2 is 18mA. The current will rise up to 0.12A when only provide the Vccint for XC6S45-2. Is there any document to talking about this? or any answer to indicate this issue?
Ans. No document talking about this. What’s the purpose to provide Vccint only to Spartan-6? I cannot see any sense to do this.
Q2. Accourding ds162.pdf, table 4 "Typical Quiescent Supply Current", what this table measure condition? the issue is, my customer try to measure the quiescent qurrent, but the result are difference from ds162 table4. Is there any document to write it down? I need document to make customer well know what the difference, and make they to understand the result is acceptable
Ans. Testing condition: at nominal voltage, 25°C junction temperatures (Tj). Typical values are for blank configured devices with no output current loads, no active input pull-up resistors, all I/O pins are 3-state and floating. This is stated in Note 1, 2 of Table 4, DS162: http://www.xilinx.com/support/documentation/data_sheets/ds162.pdf Different quiescent values should be estimated with XPE: http://www.xilinx.com/ise/power_tools/license_spartan6.htm Besides, if the customer is using an ES chip to test the quiescent values, he should know: http://www.xilinx.com/support/answers/33641.htm
Q3. the issue is: only proivde Vccint for S6, and the Iccint rise up to 0.12A. at this condition, why s6 still work well??
Ans. FPGA cannot work if you only providing Vccint. Only after Vccint, Vccaux and Vcco2 receive their respective threshold will Spartan-6 begin to work.
Power Sequence on Supplies
Since Spartan-6 doesn't need a well-control power sequence on 1.2V, 2.5V and 3.3V. But here is still mentioned.
1.8V first 120us ramp up
3.3V first 2.92ms ramp up
3.3V second ramp up to stable level after 2.63ms of 1.8V
1.2V third 280us ramp up
1.2V third ramp up to stable level after 15.8ms of 3.3V
2.5V fourth 7.76ms ramp up
2.5V fourth ramp up to stable level after 19ms of 3.3V
According to p.6 table5 of [1], the power supply ramp time of VCCINTR(1.2V), VCCO2(3.3V) and VCCAUXR(2.5V) must be within 50ms. From above scopes gallery, Milkymist One meets Spartan-6's requirement of ramp time 50ms.
