Charge Amplifiers.

Charge amplifiers are used  with piezo electric transducers. Accelerometers are often piezo electric. No shadow over Analog Devices integrated accelerometers! They may be good for vehicle collision detectors but they are not nearly as sensitive as a regular piezo electric device.

I have picked up the vibrations from the 32 kHz crystal in a regular quartz wrist watch with a standard piezo electric accelerometer! (Endevco 2217E). Another time I attached it (with its magnetic adapter) to the frame of my massive 1963 Oldsmobile-98. A 2½ ton car. I connected the output from the charge amplifier to an audio amplifier and a loudspeaker. When the street was quiet from traffic noise I could hear the birds in the trees singing! Fortunately none of the birds did their usual "business" on the car, or it would have blown out the loudspeaker!

Due to the logarithmic characteristic of LogProbe, there is no gain setting on CHA-1. The dynamic range of 90 dB is likely to cover just about every need, and if it does not, we will cooperate with you, modifying the design.

One important use for a piezo electric accelerometer and the CHA-1 is monitoring of bearing noise. Increased noise from a roller or ball bearing is a prime indicator of an upcoming failure. Or lost precision in a machine.

Since even a perfectly good new bearing has some noise, magnitudes less than one towards the end of its life, the logarithmic response of LogProbe is ideal for the purpose. One can monitor the progress, or lack thereof, of the bearing throughout its life! With a linear indicator one would see nothing until just before the failure, and according to Murphy's Law, this will happen when nobody is watching! Besides, the DC output from LogProbe is easily connected to a meter or can trigger an alarm circuit.

Actually, ball- and roller-bearings for high precision machinery has to be 'run in' with a moderate load and LogProbe can of course be used for to monitor this process as well.

Onboard a ship, for example, a seized pressure bearing in bad weather can cost the life of crew and passengers and the loss of ship and cargo. LogProbe with CHA-1 and an accelerometer can easily be installed on the bearing pillow blocks along the propeller shaft with monitoring instruments in the machine room and on the bridge. They are self checking since they will show 0 until the engine is started (quiet ship) and showing normal readings thereafter if everything is all right. If the engine is not running the function of each installation can still be checked by tapping the pillow block with a screwdriver.

Installed on a milling machine or a lathe it will monitor the status of spindle bearings and give warning before tolerances of the machined parts become too wide due to worn bearings. Even tool wear may be monitored, since sharp tools work quieter than worn ones.

Even land vehicles are prone to the same potential problems from bearing failures. Bearings on trains, street cars and buses can be checked, either while in service or while in for service. An accelerometer mounted to a train track can discern if one wheel in a passing train is noisier than the others, indicating a possible wheel rim crack, flat spots, brake problem or a failed bearing.

An automotive shop can have great use of LogProbe in locating the source of noise. An accelerometer mounted to a rod allows for the monitoring of noise from various parts of a car engine. Say a bearing in a belt tensioner is bad. It is difficult to tell where the sound comes from. The alternator or water pump may be just as likely. Due to the logarithmic scale LogProbe is enough sensitive to pick up the faintest sounds, but not too sensitive for the loudest! A repair shop will soon learn what reading is typical for each device, so bad bearings can be spotted early. Of particular interest may be turbo compressors, where a bearing failure is likely sooner or later and where secondary expenses from a bearing failure may be considerable..

Motorcycle shops can put the bike on its stand or on a block, attach the accelerometer with a magnetic base to a fork or the frame, and spin the wheel by hand. A bad bearing will show direct. Especially important for the front wheel bearings where a failure can spell disaster.

Let us look at a bearing example:

The accelerometer is mounted on my bench grinder. It has no slip rings and pretty good bearings. It is reasonably well balanced. From left to right: it is spinning down and re-started. Then I went through the gain settings of the charge amplifier, one by one, as marked on the plot. Starting at 50 (x10 pC for full scale on some other instrument. I turned down the gain setting on the charge amplifier to 5 k and then up to 5, a 1000 x range, or 60 dB. Then I left it at the '20' setting and at the red arrow I switched off the power. The grinder coasts to a stop again over about 3 minutes. As the recording has dropped about half way I looked at the grinder. It is spinning at maybe one rev per second, and I cannot even hear it!

We have probably all heard about how one can listen to a rail road rail and hear an oncoming train for miles away. It sounds likely, the rails are very good conductors of sound. I decided to try the accelerometer - charge amplifier and LogProbe on a rail road! At La Conchita the rail road is visible for several miles. The 101 freeway runs nearby and Rincon Beach is just on the other side, to the west. I attached the accelerometer with a magnetic base to a triangular feature welded to the rail. Since I surely was going to register noise for at least a mile I needed a slow recorder speed, or the recorder was going to run out of paper before the train arrived. I selected the slowest, 50 seconds/inch. After a long wait, there was the train! I started the recorder. No signal! I could see the noise from the 101 when some heavier trucks passed. There was the train getting closer, ½ mile and still no signal! Something wrong? Then, suddenly, the recorder whipped up and down again as the train passed! The train was only a few hundred meters off when I got my recording!

I decided to wait for the next train,, but with the recording speed set to 5s/inch. When finally a train arrived I got the recording above. It was a 5 car Amtrack passenger train, the front car being the engine. The speed was about 100 km/h (60 mph) or 36 m/s. I have a 321 kB sound file from the event, but do not know how to make it downloadable from here. I can e-mail it to you if you want. Anyhow, You can see the "shoulders" of the recording above sloping at some 3 dB/s, so another 10 s later the sound is ~90 dB down from the peak. You can also see the signal going up and down by some 10 dB as each axel, or wheel, passes the accelerometer. This is quite some damping, 10 dB over only some 10 - 20 m. The railroad here had welded rails and new ties, recently upgraded crushed rocks. They are probably excellent dampers of vibrations.

For mail or orders, please contact us at