For the Sound and Vibration Measurement Hat for Raspberry Pi road test, I'm reviewing Measurement Computing's IEPE Measurement DAQ HAT for Raspberry Pi. In this post: the Piezo Acceleration Sensor PCB Piezotronics 603C01.
I'll review the capabilities, use cases and mounting options.
The accelerator sensor is a single vector detector. It measures vibrations that are directed up its vertical axle. 100 mV/g is the output signal.
|In the next blog, where I'll calculate the effective acceleration of a device, I'll use this representation of the same signal: acceleration is 10.2 mV/(m/s²).|
It's intended for hazardous areas, physically and environmental hardened. Given its size and weight, it's intended for use on heavy motors, gearboxes, car parts, industrial machinery inside and outside. If you're looking for a lightweight device to specify vibration behaviour of small gizmo, this isn't what you'd select. It's more than 4 cm high, weighing 50+ g without cables or connector. Although it is very strong and can withstand abuse, there's one force you have to be careful with: shock. There's a ceramic precision component in there that can take a lot of vibration, but can get damaged when shocked/dropped. More on that when I review the magnetic mounting option.
The electrical connection is via a MIL 2 pin plug. This is a connection type suited for automotive, industrial, military environments. It handles moist, vibration and dust well.
With this family of MIL of connectors, you first make contact by pushing the male and female part together. Then there's a low speed (pitch? - the Dutch word is spoed, I may translate this wrong :) ) nut that you thread over the sensor, to firm the mechanical connection. This doesn't unscrew under vibration conditions in the range of this sensor.
Because of the low signal levels - and to maintain signal integrity, it's good to use a coaxial cable for the connection. The sensor, with its 50Ω impedance, expects it.
The PCB 603C01 is an IEPE type of sensor, and requires a constant current. This current is to be injected into the same cable that transfers the measurements. THat's easy, because the sensor signal is AC and the constant current a DC source. A cap has to remove the DC artifact from the coax cable payload before handing the signal over to an analyser. Power requirements for this sensor are a constant current between 2 and 20 mA. The DC voltage injected into the line should have compliancy between 18 and 28 VDC.
The sensor generates 100 mV/g. It properly transcribes up to 50 g (resolution: 350 µg), between 0.5 to 10000 Hz. Temperature range is between -54 to +121 °C. The temperature has impact on the measurements, and that's specified in the data sheet.
image: impact of temperature (in °F) on measurements. Source: product page
Mounting the Sensor
This could be a topic on its own. There are several documented and specified ways to mount the sensor, but your imagination is the limit. The way you mount the sensor on the device under test impacts measurements and ranges. The manual describes this, but I found clearer pictures on the internet. I'll use sources from acoem (it shows the magnetic mount I'm using) and dynamox.
I'm using configuration 4 (magnet) and 6 (bolted down). The gizmo for my roadtest has a stud coming out of the vibration table, and the sensor can be screwed on that. The connection is firm.
I also bought a Neodymium base that the sensor can be screwed onto. The magnet is crazy strong, and has a similar performance as bolting the sensor directly to the device under test (DUT), up to a certain frequency. When using the magnetic mount, it's best to first put the magnet on the DUT, then screw the sensor on it. Because the magnet is so strong, you can't control the force with which it attaches to the DUT (true!). The acceleration force can easily damage the ceramics inside the sensor.
Here's a performance comparison between different mounting options:
The firm installation with a screw/bolt has a linear result up to almost 10 kHz. My magnet mount works well up to 100 - 200 Hz.
Another watchout with the magnetic mount: it may interfere with the magnetism induced inside the motor you're testing. When I put it close to the fan of a Rigol PSU, the fan stops rotating completely.
In the next post, I'll analyse the data from tghe sensor as it's done in the mining industry. I got the algorythms from martinvalencia, and we tested if my LabVIEW designs could give the info they use in the Peruvian mining industry. Spoiler: yes.
|references (at the time of writing, the data sheets were off-line. Hope this is fixed when you click)|
|Piezo sensor product home page|
|Piezo Sensor datasheet|
|Piezo sensor specification sheet|