Question about unusual feedback arrangement

On Tue, 3 Mar 2009 22:30:11 +0000, Nemo wrote:

--- Here's the LTspice circuit list for the circuit you've described and its near equivalent:

Version 4 SHEET 1 1256 680 WIRE 64 80 0 80 WIRE 208 80 128 80 WIRE 352 80 272 80 WIRE 1008 80 928 80 WIRE 1152 80 1072 80 WIRE 0 144 0 80 WIRE 48 144 0 144 WIRE 208 144 128 144 WIRE 352 144 352 80 WIRE 352 144 288 144 WIRE 928 144 928 80 WIRE 1008 144 928 144 WIRE 1152 144 1152 80 WIRE 1152 144 1088 144 WIRE 176 272 176 256 WIRE 1072 272 1072 256 WIRE -144 288 -192 288 WIRE 0 288 0 144 WIRE 0 288 -64 288 WIRE 144 288 0 288 WIRE 752 288 704 288 WIRE 928 288 928 144 WIRE 928 288 832 288 WIRE 1040 288 928 288 WIRE 352 304 352 144 WIRE 352 304 208 304 WIRE 1152 304 1152 144 WIRE 1152 304 1104 304 WIRE -416 320 -416 288 WIRE -304 320 -304 288 WIRE -192 320 -192 288 WIRE 144 320 0 320 WIRE 704 320 704 288 WIRE 1040 320 896 320 WIRE 176 352 176 336 WIRE 1072 352 1072 336 WIRE -416 432 -416 400 WIRE -304 432 -304 400 WIRE -304 432 -416 432 WIRE -192 432 -192 400 WIRE -192 432 -304 432 WIRE 0 432 0 320 WIRE 0 432 -192 432 WIRE 704 432 704 400 WIRE 896 432 896 320 WIRE 896 432 704 432 WIRE -416 496 -416 432 WIRE 704 496 704 432 FLAG -416 496 0 FLAG -416 288 +12 FLAG 176 256 +12 FLAG -304 288 -12 FLAG 176 352 -12 FLAG 704 496 0 FLAG 1072 256 +12 FLAG 1072 352 -12 SYMBOL cap 128 64 R90 WINDOW 0 -39 32 VBottom 0 WINDOW 3 -35 32 VTop 0 SYMATTR InstName C1 SYMATTR Value 1.8e-12 SYMBOL cap 272 64 R90 WINDOW 0 -38 32 VBottom 0 WINDOW 3 -36 30 VTop 0 SYMATTR InstName 22e-12 SYMATTR Value 2.2e-12 SYMBOL res 144 128 R90 WINDOW 0 63 56 VBottom 0 WINDOW 3 66 58 VTop 0 SYMATTR InstName R1 SYMATTR Value 499k SYMBOL res 304 128 R90 WINDOW 0 67 55 VBottom 0 WINDOW 3 67 55 VTop 0 SYMATTR InstName R2 SYMATTR Value 499k SYMBOL Opamps\\LT1007 176 240 R0 SYMATTR InstName U1 SYMBOL res -48 272 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R3 SYMATTR Value 100k SYMBOL voltage -192 304 R0 WINDOW 3 20 128 Left 0 WINDOW 123 20 100 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value "" SYMATTR Value2 AC 1 SYMBOL voltage -304 416 R180 WINDOW 0 24 104 Left 0 WINDOW 3 24 16 Left 0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V2 SYMATTR Value 12 SYMBOL voltage -416 304 R0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V3 SYMATTR Value 12 SYMBOL cap 1072 64 R90 WINDOW 0 -39 32 VBottom 0 WINDOW 3 -35 32 VTop 0 SYMATTR InstName C2 SYMATTR Value 1.0e-12 SYMBOL res 1104 128 R90 WINDOW 0 63 56 VBottom 0 WINDOW 3 66 58 VTop 0 SYMATTR InstName R4 SYMATTR Value 1e6 SYMBOL Opamps\\LT1007 1072 240 R0 SYMATTR InstName U2 SYMBOL res 848 272 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R6 SYMATTR Value 100k SYMBOL voltage 704 304 R0 WINDOW 3 20 128 Left 0 WINDOW 123 20 100 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V4 SYMATTR Value "" SYMATTR Value2 AC 1 TEXT -392 456 Left 0 !.ac oct 256 10 10e6 TEXT 736 464 Left 0 !;ac oct 256 10 1e6

Running a frequency sweep I don't see much difference between them.

Even looking at the region where the reactances of the caps are close to the resistances of the resistors:

There doesn't seem to be anything wild going on.

JF

Perhaps he didn't have a 1 pF cap?

Or maybe the 0.1% resistors he used didn't go up to 1M?

Some spiral-cut thin-film resistors have about 0.3pF of parallel capacitance, and the original designer might have been designing around that. Few capacitor ranges go below 1.0pF, and the tolerance on the lowest capacitances tends to be high.

-- Bill Sloman, Nijmegen

YES. Mark

It sounds like he/she was trying to achieve a response similar to what is used for phonograph equalization (per RIAA standards).

See the section called Complete Active RIAA Filter on this website:

Bob

Thank you, lots of useful tips there.

I noticed that it was assumed he used thin film 499k resistor. However, they're just normal 1% metal film ones. Maybe there was a bag of them to hand when he started his design. I suspect he wasn't aware that thin film is lower noise (after all, no one knows every trick). Although with a target bandwidth of around 60kHz, maybe thin film has little effect even when the gain is so high (does it only give noticeable improvement at high freqs?)

Thanks again for the feedback, I feel I can tweak this circuit with more confidence now if we hit problems when it goes into production. All the side comments are really useful.