Moveable thermal anemometer design nulls battery discharge droop

All thermal anemometers work by inferring air pace from measurements of thermal impedance (Z) between a heated sensor and the encompassing air:

Z = T / P         (1)

The place P is the facility dissipated by the sensor and T is the temperature distinction between the sensor and ambient.

Wow the engineering world along with your distinctive design: Design Concepts Submission Information

There are two fundamental schemes for doing this.

1. Maintain P fixed and measure the ensuing temperature distinction T
2. Maintain T fixed and measure the facility P required to do it

An instance of the fixed energy sort will be present in “Nonlinearities of Darlington airflow sensor and VFC compensate one another”…

…and examples of the fixed temperature sort will be present in “Linearized moveable anemometer with thermostated Darlington pair”…

…and in Determine 1. 

Determine 1’s anemometer is uncommon as a result of it melds the sensor transistor right into a direct PFC (Energy to Frequency Converter) loop.

Determine 1 Fixed-temperature anemometer with direct power-to-frequency conversion.

To grasp how the Determine 1 circuit works, contemplate the case of zero airflow. You utilize ZERO trimmer R2 to set the quiescent base-bias currents for Q1 and ambient reference Q2. With the right adjustment, Q1’s temperature rise (~50°C) in nonetheless air, attributable to collector energy dissipation, reduces Q1’s VBE (by ~2 mV/°C) to equal or barely beneath Q2’s. The noninverting enter of comparator U1a is then barely much less constructive than the inverting enter. The output subsequently switches low, holding C1 discharged and resetting multivibrator U1b, whose output goes excessive.

This situation does two issues: It forces Fout = 0 and holds Q3 off.

Now let’s blow some air at Q1. The ensuing enhance in cooling tends to scale back Q1’s temperature, inflicting its Vbe to extend relative to that of Q2. This makes the comparability between U1a’s inputs reverse, releasing the reset on C1. C1 then fees by R9 and activates Q3, driving a t = 700-µsec pulse to Q1’s base by CALIBRATE trimmer R3.

The resultant pulse of collector present pressured in Q1 will be seen in Equation 2 (the place h_FE = Q1 present acquire and Rcal = R3 + R4):

I_C = h_FEI_B  = h_FEV/(Rcal),      (2)

This deposits a quantum of warmth on Q1’s junction:

t P = t I_CV = t I_B h_FEV  = t (V/Rcal)h_FEV = t h_FEV²/Rcal   (3)

which tends to return Q1 ‘s temperature to a worth heat sufficient to revive the unique zero-flow voltage stability with ambient-sensor Q2. Till Q1 achieves that temperature, U1 continues to oscillate, cycle Q3 on, and pump warmth into Q1.

Thus, a suggestions loop is established that acts to take care of a continuing temperature differential between Q1 and Q2. The common frequency showing at U1b’s output is subsequently proportional to the additional energy required to warmth Q1. The utmost output frequency for the circuit values in Determine 1 is 1 kHz. Applicable adjustment of R3 establishes virtually any desired full-scale move. Temperature monitoring between the Q1 and Q2 Vbe voltages gives good compensation for adjustments in ambient temperature.

The direct connection of Q1 to the facility rail ends in good effectivity (>90%) energy utilization, so whereas energy draw is (by definition!) depending on airflow, as proven in Determine 2, it’s usually modest: 200 to 350 mW.

Determine 2 Q1 energy draw versus air move is usually a modest 200 to 350 mW.

In truth, energy consumption is low sufficient that moveable battery operation, with an affordable multimeter for frequency readout, seemed engaging. A cheap stack of 4 AA alkaline batteries promised tens of hours of steady operation which might equate to lots of of air velocity readings. Nevertheless, as proven in Determine 3, direct battery energy of Determine 1 wouldn’t work very nicely, as a result of ±20 % roll-off of battery voltage throughout discharge. 

Determine 3 Typical AA cell discharge droop curves with an undesirable ±20 % roll-off of battery voltage throughout discharge, leading to a degradation of anemometer calibration accuracy.

The ensuing degradation of anemometer calibration accuracy could be excessive, particularly contemplating Equation 4:

t P = t I_CV = t I_B h_FEV  = t (V/Rcal)h_FEV = t h_FEV²/Rcal   (4)

that exhibits the square-law dependence of Q1 heating on provide voltage!

In the meantime, the seemingly apparent treatment of provide voltage regulation wouldn’t be very engaging both, as a result of ensuing influence on complexity, effectivity, and value. Luckily, Determine 4 exhibits an alternate easy, low cost, and environment friendly answer: base bias compensation.

Determine 4 Determine 1’s anemometer modified with U2, A1, and R11 – 14 to servo Q1 and Q2 bias currents to (largely) null the consequences of battery voltage droop.

Determine 5 exhibits the ensuing compensated energy curve (black) versus what would outcome with out it (crimson): higher than an order of magnitude enchancment!

Determine 5 Nulled (black) and uncompensated (crimson) Q1 heating versus battery voltage droop (5 ±1 volts).

 Nonetheless not excellent, however arguably adequate.

Stephen Woodward’s relationship with EDN’s DI column goes again fairly a great distance. Over 100 submissions have been accepted since his first contribution again in 1974.

 Associated Content material