The power consumed by the transceiver can be important with certain applications. For example, when it's used with a solar or diesel powered BTS.
The power consumption of UmTRXv2 in dual channel mode is around 14-16W, depending on functionality, Tx output power and supply voltage.
How to measure power consumption¶
Current can be measured with the help of filters TI160808U601 and TI201209U121, which are present in all important power rails and have an internal resistance of approximately 0.3 and 0.1 Ohms respectively. The accuracy of measurements may not be as high as with resistors, but it is sufficient for a rough estimate.
The filters can be replaced with 1% resistors for higher accuracy and/or dynamic values.
Power consumption measurements¶
The tables below show the power consumption according to functionality, Pout and supply voltage.
Measurement conditions: Vcc=12.0V (power calculated from the DC current of Vcc supply), Ta=25 deg C, passive heatsink, active GPS antenna, Fout=942/942.2MHz, Pout=20/20 dBm modulated by tx_samples_from_file.exe with args: --rate 1083333 --gain 15, receiving by rx_samples_to_file.exe with args: --rate 1083333 --gain 15, mezzanine: UmSELv1-GSM900.
|brFunction||Power delta, W||+1 LMS,brPower delta, W||+2^nd^ LMS,brPower delta, W||brTotal, W|
|LMS Rx to file||-||2.40||2.40||10.32|
|LMS Tx from file||-||2.88||2.88||16.08|
Measurement conditions: Vcc=12.0V, Ta=25 deg C, passive heatsink, active GPS antenna, Fout=942/942.2MHz, Pout=20/20 dBm continuous ideal GMSK, TXVGA1=-10, TXVGA2=20, no mezzanines.
|brFunction||Power delta, W||1 LMS,brPower delta, W||+2^nd^ LMS,brPower delta, W||brTotal, W|
|LMS Rx Enable||-||2.16||2.16||9.72|
|LMS Tx Enable||-||1.56||1.56||12.84|
|LMS PA On||-||0.84||0.84||14.52|
Measurement conditions: Vcc=12.0V, Ta=25 deg C, passive heatsink, active GPS antenna, modulated by tx_samples_from_file.exe with args: --rate 1083333, receiving off, no mezzanines.
|--gain||Pout GSM900, dBm||Pout GSM1800, dBm||P, W|
Note: Spectrum becomes quite dirty at the gain 18 and above.
Measurement conditions: Vcc=12.0V, Ta=25 deg C, passive heatsink, active GPS antenna, continuous ideal GMSK, TXVGA1=-10, no mezzanines.
Note: Big difference between GSM900 and GSM1800 bands in this table is because of the matching issue in the original UmTRXv2. This is fixed now (see issue 41), new measurements are pending.
|TXVGA2||Pout GSM900, dBm||Pout GSM1800, dBm||P, W|
|1||2.4||--12.8- (see above)||13.2|
|5||6||--9- (see above)||13.38|
|10||11||--4- (see above)||13.44|
Measurement conditions: Ta=25 deg C, passive heatsink, active GPS antenna, Fout=942/942.2MHz, Pout=20/20 dBm continuous ideal GMSK, TXVGA1=-10, TXVGA2=20, no mezzanines.
|Supply voltage||Current, A||Power, W|
- Prototype 1, 03.01.2012: Powered from 8v and it takes ~ 0.5 A of static current (no fw loaded).
- Prototype 1, 14.01.2012: With FPGA bistream loaded (SRAM and Ethernet are enabled) in idle mode is approximately 6W:
- Current at 6.5V was 0.89A (5.785W)
- Current at 12V was 0.51A (6.12W)
- Prototype 1, 15.01.2012: Measurements :
L10-1 1.2 mv 0.3R LMS1 1.8v Analog L7-1 4.2 mv 0.3R LMS1 3.3v L9-1 1.3 mv 0.3R LMS1 1.8v Digital L8-1 1.8 mv 0.3R LMS1 2.5v L10-1 1.2 mv 0.3R LMS1 1.8v Analog L7-1 4.1 mv 0.3R LMS1 3.3v L9-1 1.3 mv 0.3R LMS1 1.8v Digital L8-1 1.9 mv 0.3R LMS1 2.5v L1 10.2 mv 0.3R 5V2 rail L2 2.8 mv 0.3R GPS L34 10.2 mv 0.08R 6.5v L40 0.4 mv 0.08R 3.3v FPGA L41 1.0 mv 0.08R 2.5v (FPGA+Clock+RF) L42 1.0 mv 0.08R 2.5v ETH L43 0.3 mv 0.08R 2.5v RAM L3 5.7 mv 0.3R 2.5v clock L47 13.2 mv 0.3R 5v RF L48 0.1 mv 0.3R 2.5v RF Total: 917mW (doesn't include 1.2V FPGA core). This is the power consumption of the system without the power conveters loss.
So you can see that from the main rails, the 6.5V one (which is the main smps rail for all subsequent linear supply) is the biggest current hog. And all that power goes mainly to the LMS and to the RF LNA. And in a smaller part to the GPS and oscillator (this is from a dummy firmware that just blinks leds, so FPGA loaded, but not doing much).
My multimeter is "only" 10.000 count and isn't good enough to measure the exact series resistance of each of these filters to get absolute values. You'd need a proper LCR to measure them I guess, but I don't have any.
- Prototype 1, 18.01.2012: measurements for idle mode ([Power_consumption_idle_18.01.2012.xls spreadsheet])
- Total consumption, measured at DC supply is 3.8W @ 8V.
- Each LMS consumes about 1 W in idle mode.
- For approx tests I suggest to use the calculated values of 0.025 Ohm for TI2012 and 0.14 Ohms for TI1608.
- The efficiency DC-DC converters is more then 85%.