PSU Energy - AC - MCP (MCP39F511N)
What it does
This metric provider uses the MCP39F511N chip to read the power used by a device plugged into it. Normally this would be the device we are benchmarking.
We use the board implementation here: AMD00706
Spec highlights:
- 0,5% Accuracy on active power measurements in a 1:4000 dynamic range (~ 3.75 mW minimum resolution)
- 8 MHz internal processing clock (theoretical maximum of 12.5 us resolution)
- Power is averaged over multiples of 20 ms (Selectable from N={1,2,3,4})
- Internal accumulation of energy in register (currently not active in our setup. We use active power only atm.)
Implementation of the source code is mostly copied from https://github.com/osmhpi/pinpoint/blob/master/src/data_sources/mcp_com.c Credits to Sven Köhler and the OSM group from the HPI
Install
All compilations and linking will be done through the install script.
After that just plug it into your USB and please use Channel 1.
Classname
PsuEnergyAcMcpMachineProvider
Metric Name
psu_energy_ac_mcp_machine
Input Parameters
-i
: interval in milliseconds. By default the measurement interval is 100 ms.
./metric-provider-binary -i 100
Output
This metric provider prints to stdout a continuous stream of data every interval
milliseconds till it is stopped with
sigkill
or sigint
(Ctrl-c). The format of the data is as follows:
TIMESTAMP READING
Where:
TIMESTAMP
: Unix timestamp, in microsecondsREADING
: The value taken from sensor in the unit supplied or mW if no unit is specified.
Any errors are printed to stderr.
Remarks on the minimum average resolution on 20 ms
Although the chip has a very high internal sampling frequency it far higher than 20 ms it does not output measurement data from smaller time frames.
The effective retrieved data is always averaged over a minimum of one full grid power frequency cycle (typically 50 Hz / 20 ms). The reason for that is that the Active Power is defined as the integral over one full cycle
Technically you can also analyse only parts of a full-cycle. However it has no meaning for the Active Power as we want to analyze the consumption of the software and not small spikes in AC load that however might be negated with inductive or capactive induced phase-shifted surges later.