In industrial applications, equipment can often suffer harsh conditions. While some of these are more obvious (e.g. temperature extremes, vibration, liquid and dust), others are less obvious such as power issues. In this part, I run some automated tests to try and determine just how much power the Harting MICA consumes and how it behaves when faced with different voltages.
Voltage Tolerance & Power Consumption
To understand the tolerance of the Harting MICA to various voltages and determine its power consumption, I first completed all other tests as this test has the potential to be destructive. On-paper, the unit claims to be able to run from 12V or 24V (±5%) via the I/O port, or from 802.3af 48V Power-over-Ethernet (PoE). As I have no compatible PoE equipment at this time, I did not test PoE capabilities.
For testing, I decided to push the limits somewhat. Power was supplied from a Rohde & Schwarz HMP4040.04 Programmable Power Supply previously reviewed. The power supply was commanded to step from 0V to 29.900V inclusive in 0.100V steps, as this covers even the worst-case voltage that may be expected from a 24V (nominal) solar system with sealed-lead acid batteries on maximum cycle charge voltage of 29V. At each voltage step, the supply first turns off the output for 1s, then turns on the supply at the requested voltage, waits 45s for boot-up to complete and then reads 128 voltage and current pairs from the supply at 0.3s intervals, recording the averaged voltage and current and computing the power from this value. This was done with the MICA containing a Transcend 16GB microSD card and with the Bosch CISS attached, running the CISS Gateway, MQTT Broker, NodeRED, NAS and Debian containers and an active LAN link to represent a reasonable workload.
The requirement to turn off the power supply output and turn it back on at each voltage step was because it was found that the MICA will not boot if the voltage input is slowly ramped up and instead sticks on either a red light or a blinking-green LED. This is not unusual to encounter equipment that behaves like this, although it could pose a problem if on a remote site. The running of the test took a number of hours and resulted in a total of 300 steps and around 240 unexpected power removals and reboots. I can report that the MICA suffered no ill effects from having been rebooted so many times unexpectedly.
From the testing, it seems that the MICA is able to operate over a much wider range than the paper specifications may allude to. The paper specification ranges are highlighted in yellow – perhaps the voltage specifications are listed as such as the output from the GPIO pins follows the input voltages, thus to ensure the GPIO voltages are “correct” for interfacing to other equipment, the input voltages must be kept within the on-paper ranges. However, if this is not an issue in the application, it seems that the MICA was able to operate reliably from about 6.5V through to almost 30V, although perhaps with additional stress to the components.
It is clear that the power curve approximates a constant power of about 2.8 to 3.0W under my heavy load. Changes in the power curve are predominantly due to efficiency differences of the internal switching converter which is used to derive the voltage rails that power the components. As a result, it seems the lower voltage operation below about 6.5V is unreliable as the converter appears to hit a 500mA maximum current limit (as listed in the specifications). The peak converter efficiency seems to be about 13-14V, with the power dissipation increasing as the voltage increases above 26V despite reducing current, suggesting it would be unwise to stretch the limits too far. In all, no damage was incurred during this test despite pushing beyond on-paper specifications.
Supply Dip Tolerance
Another potential hazard in power supplies is significant power dips as may happen in automotive power buses for example. To simulate this, I used the Arbitrary Waveform feature of the HMP4040.04 to produce two different waveforms to study the behaviour of the MICA.
In the first waveform, 24V is supplied for 5s, followed by a dip to 12V for 1s, repeating continuously. All of the voltages in this case are within the operating window, but a sudden near-step-change of 12V is not particularly “friendly”. Leaving the MICA to run for an hour, it seems rather unperturbed by this.
Stepping up the difficulty, I dropped to a lower voltage. In this case, power is delivered at 12V for 10s, dipping to just 6V for 1s and repeating continuously. We know from prior testing that operation at 6V is somewhat borderline, however, it seems that the MICA may well have enough capacitance within it to ride-out such disruptions as it did not falter at all.
Of course, it would be possible to generate even harsher waveforms that make the MICA fail, but that is beside the point. The fact that it was able to handle even this much is noteworthy in itself.
Conclusion
The Harting MICA claims to be able to run from 12V or 24V (±5%) via the I/O port. To understand its voltage tolerance and power consumption, I ran an experiment using a Rohde & Schwarz HMP4040.04 Programmable Power Supply to step from 0V to 29.900V inclusive in 0.100V steps, pushing beyond the ratings of the MICA as this covers even the worst-case voltage that may be expected from 24V sealed-lead acid batteries on maximum cycle charge voltage of 29V.
It was found that the MICA will not boot if the voltage input is slowly ramped up and instead sticks on either a red light or a blinking-green LED. This is not unusual to encounter equipment that behaves like this, although it could pose a problem if on a remote site. The running of the test took a number of hours and resulted in a total of 300 steps and around 240 unexpected power removals and reboots. I can report that the MICA suffered no ill effects from having been rebooted so many times unexpectedly.
From the results, it seems that the MICA was able to operate reliably from about 6.5V through to almost 30V, although perhaps with additional stress to the components and with GPIO voltages that follow the input voltage. The power curve approximates a constant power of about 2.8 to 3.0W with changes in the power curve predominantly due to efficiency of the internal switching converter. In all, no damage was incurred during this test despite pushing beyond on-paper specifications.
Testing voltage tolerance with cyclical waveforms that included 24V (5s) to 12V (1s) and 12V (10s) to 6V (1s) were tolerated by the Harting MICA with aplomb, operating without any interruption or anomalies. This suggests that there is sufficient capacitance within the MICA to ride through even significant dips in the power bus, making it well suited for harsh industrial applications.
---
This post is a part of the Harting MICA CISS Complete IIoT Starter Kit RoadTest