Model 805M1 TO-5 3-Wire Accelerometer

The model 832 and 834 datasheets show the operating temperature from -40°C to +125°C. The lower limit of -40°C is not low enough; we need to measure down to -55°C. What is your recommendation for meeting this requirement?

We tested the bias at -55°C. Test result of 832-0500 DC bias change with temperature is shown below, DC Bias changes about 0.5% at -55ºC compared to 25ºC:

Total current supply is 4.1uA at -55ºC which is still within specification. However, for continuous usage down to temperature of -55°C, the model 832M1 and 834M1 are recommended.

Do you happen to have any more detail for the assembling the model 832 and 834 to a board? Our board assembly department is a little concerned about manually soldering this part. It says on your data sheet that the accelerometer can’t go through solder reflow at high temperature and that manual soldering is recommended. I was hoping for some more clarification on this statement.

The reason for the caution is the potential risk of sensitivity shift in the output after reflow soldering. The units will survive the reflow soldering process. We caution against this process since we have seen a 1-2% sensitivity drop after reflow soldering. For our reflow profile we have a peak temperature of +250°C since we use non-leaded solder for RoHS compliance. A lower reflow profile may result in negligible sensitivity shift. If you can use leaded solder such Sn63 or Sn62 (183°C and 179°C eutectic respectively) then the peak reflow temp should not have to exceed +210°C (60 second max). This should then allow reflow soldering.

We generally conformal coat our circuit boards to protect the circuitry, would there be any concern with conformal coating (810M1, 820M1, 832M1, 834M1)?

No, there are no concerns with conformal coating. The seismic mass system and electronics are all hermetically sealed under the cover.

Can we bake the circuit boards after conformal coating (810M1, 820M1, 832M1, 834M1)?
Yes. There will be no problem with an over-night bake at +93°C on the model 832M1. We bake the units for 24hours at +121°C during manufacture.

Just a clarification, at 0g output, is the accelerometer output Supply Voltage/2? So that when we have a negative acceleration we approach 0 but not negative?
Yes, you are correct. The output will swing nominally +/-1.25V about the bias voltage. For a +/-100g range accelerometer with 3.3V excitation (bias at 1.65V), the output would be nominally be 0.4V to 2.9V.

Can TE Connectivity provide a higher temp version of model 832M1 and 834M1?
Yes, we can make high temperature version which operates from -40°C to +150°C, but the current consumption will be 60uA. The model numbers are 832HT and 834HT.

 
If I use structural epoxy around the perimeter to reinforce the vibration sensor onto the circuit board (after soldering) will this affect the vibration response of the sensor? Is there a reinforcement technique you would recommend?
No, this will not affect the response of the sensor and in fact it is recommended to reinforce the sensor attachment after soldering. Typically we recommend the customer use a low viscosity cyanoacrylate adhesive (such as Loctite 4501) and allow the epoxy to wick underneath accelerometer to fill the gap to the circuit board.

What mounting techniques and materials are recommended to achieve the best high frequency response for the board mountable accelerometers (810M1, 820M1, 832M1, 834M1)?
To achieve the best frequency response, we recommend mounting the accelerometer directly to the structure to be measured. An adhesive can be used to secure the accelerometer. Take precautions not to short the output pads underneath the circuit board. Good frequency response can also be achieved by mounting the accelerometer onto a ceramic or hybrid circuit board. FR4 boards should be avoided for applications requiring wide bandwidth measurements since the FR4 material can impart a resonance to your measuring system. If attaching wires to the output pads then these need to be properly secured/anchored at regular intervals to minimize cable motion that can add noise and resonances to the output signal.

What is the material composition of the plating on the solder pads of model 832 and 834 accelerometers?
The circuit board traces are Titanium-Tungsten plated with Nickel and Gold. 50micro-inches min of Au (99.9% pure gold per MIL-G-45204, Type III, grade A) over 50-350micro-inches of Ni (per AMS-QQ-N-290, Class I.

What is the recommended value for the blocking capacitor to use in the excitation circuit of the IEPE model 805 and 808 accelerometers?
A capacitor value of 10µF is recommended.

Can the model 805 and 808 models be mounted directly to the measurement surface with epoxy?
The outer casing on the model 805 and 808 series accelerometers is connected to circuit ground. If the mounting surface is non-conductive, then there will be no issues. However, if the mounting surface is conductive, then care must be taken to ensure there are no ground loops in your installation. It is advised to use the optional isolating mounting case illustrated below for such installations to avoid any ground loop troubles.

Your datasheet for the model 832 and 834 series accelerometers indicate an excitation voltage range of 3.3 to 5.5Vdc. Can the accelerometers be used with a lower excitation voltage?
Our engineers have confirmed that these accelerometers can be used with a minimum excitation voltage of 2.7Vdc. We specified 3.3V in our datasheet to give us some margin. We also confirmed that we had previously performed a signal warm-up test on our accelerometers. The signal converged to 98% of its final value at 30msec. There was no overshoot. It was typical of a single-pole response characteristic that was determined by its filtering …95% (lapse of three time constants).

In reference to questions above, how does the lower excitation voltage affect the full scale measurement range?
Although the model 832 and 834 series accelerometers are designed to be operated by 3.3Vdc battery power for optimum performance, the accelerometers can also be powered by excitation voltages (ExcV) ranging from 2.7 to 5.5Vdc. However, excitation voltages other than 3.3Vdc will affect the full scale range of the accelerometer since the bias voltage is a function of excitation voltage.
The following formula can be used to calculate the full scale range of the accelerometer when using different excitation voltages other than 3.3Vdc.
Full scale range (g) = [ExcV – 0.3V - (ExcV / 2)] / Sensitivity (V/g)
Example; a model 832-0200 with z-axis sensitivity of 6.41mV/g and 2.8Vdc excitation
Full scale range = [2.8V – 0.3V - (2.8V / 2)] / .00641V/g = 172g