I wanted to determine a ballpark number for the power efficiency of op-amps as amplifiers, so I decided to make a simple LTSpice simulation. Having no particular op-amp preference, I chose the first one: LT1001.

Diagram and simulation results attached. (I didn't see a max-current spec on the op-amp datasheet, but I'm assuming it can source 4-8mA.) "Ix(u1:6)" is the output pin. From the LT1001 datasheet, Vsupply+ and Vsupply- are "Ix(u1:7)", "Ix(u1:4)". The current draw at the power supply pins does not match that produced at the output; in fact the current from the supplies is fixed. This probably explains why the simulation indicates that the op-amp sources 16mW (rather than dissipates) when the load is 1k; the op-amp supposedly dissipates 8mW when the load is 3k. (I'm getting these numbers by hovering the mouse over the op-amp symbol, but the numbers are good about summing to zero so they seem somewhat sensible.)

Is this just a failure/limitation of the op-amp modeling, or am I doing something wrong? Assuming it is just the Spice model, could you offer suggestions (e.g. alternate model) for how to make it work?



(with 1k load resistor)
--- Operating Point ---

V(railhigh): 10 voltage
V(raillow): -10 voltage
V(input): -2 voltage
V(n001): 1.22653e-005 voltage
V(output): 6.00005 voltage
I(Rload): 0.00600005 device_current
I(Rfeedback): -0.00200001 device_current
I(Rin): -0.00200001 device_current
I(Vinput): 0.00200001 device_current
I(Vlow): 0.0015999 device_current
I(Vhigh): -0.00159987 device_current
Ix(u1:3): 4.8918e-010 subckt_current
Ix(u1:2): 8.87816e-010 subckt_current
Ix(u1:7): 0.00159987 subckt_current
Ix(u1:4): -0.0015999 subckt_current
Ix(u1:6): -0.00800006 subckt_current


(with 3k load resistor)
--- Operating Point ---

V(railhigh): 10 voltage
V(raillow): -10 voltage
V(input): -2 voltage
V(n001): 1.25712e-005 voltage
V(output): 6.00005 voltage
I(Rload): 0.00200002 device_current
I(Rfeedback): -0.00200001 device_current
I(Rin): -0.00200001 device_current
I(Vinput): 0.00200001 device_current
I(Vlow): 0.0015999 device_current
I(Vhigh): -0.00159987 device_current
Ix(u1:3): 4.89177e-010 subckt_current
Ix(u1:2): 8.8782e-010 subckt_current
Ix(u1:7): 0.00159987 subckt_current
Ix(u1:4): -0.0015999 subckt_current
Ix(u1:6): -0.00400003 subckt_current

 

It's the model for the opamp which is limited. This will be the case for most opamp models you could get. Opamp spice models focusses more on accurate input and output simulation than power simulation.

If you want to calculate power dissipation in the opamp, it's better to just look at the information provided in the datasheet. Although it might not give you the information to precisely calculate the losses, there is sufficient info for calculating the worst-case losses.

 

Thanks for the info!