Max input voltage for the QA401 is spec’d at 26dBV with the attenuator enabled or 6dBV without. I understand this is the voltage across the +/- inputs. What is the max allowed voltage on inputs relative to the analyzer ground?
Hi @gvl, the input opamps on the QA401 are supplied from +/- 6.1V rails, with a series 180 ohm resistor. Generally, you want to limit any excursions beyond the rails to limit at 10 mA, which is what the OPA1612 can handle. So, I = (Vin-6.5)/180, and if I = 10 mA, then Vin = 8.3V. So, if you apply a sine with a peak of 8.3V, at the sine tip you’ll hit the 10 mA. Now, the OPA1612 is rated for 10 mA continuously, so an RMS would be more appropriate here, which means a larger input could be handled. Additionally, there are steering diodes (BAS7004) that will shunt voltages beyond 6.5V + Vf to the rails. These diodes should have a lower Vf than the protection diodes inside the OPA1612. But the OPA1612 diodes aren’t really characterized such that they have published Vf versus I curves.
Finally, there’s an input C that is 33uF/50V polarized. Now, a lot of folks forget that if you apply 10V DC to a cap, that 10V will momentarily show up on the output. So, the cap helps for steady state, but doesn’t solve the transient problem.
In short, you can apply up +6V to each input all day long. The negative excursion magnitude will be tied to the 33uF input cap to tolerate the negative DC. As long as current levels are low (which they are because of the high input Z), the caps don’t seem to care that much. But just know than negative DC over will probably degrade the caps eventually.
How far you go beyond 6V depends on the ability of the external circuit to limit current. If it’s capped at 10 mA, you can do that all day long too.
And with the atten on, the problem gets much better. The 50V cap rating becomes your upper bound.
1 Like
Thanks Matt, do I understand it correctly and it should then be safe to measure bridged amps up to 100Vpp with the attenuator enabled and a 2-resistor amp load (R3 shorted, actual resistance values TBD to ensure safe voltage across the +/- inptus)?
Hi @gvl, yes, the circuit as you show is safe to apply with atten enabled. For a bridged amp, the upper leg will be very small, and the lower leg will be very large. The internal FDA will aim to sum, and it should correctly (I think?) give the resultant magnitude. But the CMRR will not be good because the noise in one leg will be very different from the noise in the other leg. Some push/pull aka bridged aka balanced signals really want that CMRR to help with noise and distortion. And so, if you try the asymmetric setup and it’s not as expected, it’s possible the imbalance is the cause.
Since 1 ohm is easy and cheap, I think it would be better to try with two 1 ohm resistors AND THEN short one if you are curious and see how the measurement changes. Some amps probably won’t care at all, and some amps (maybe filterless class D???) might care a lot. I wish I could say I knew the answer for sure, but I don’t 
Thanks Matt, good point about CMMR, I didn’t think of it.