mikep
Well-known member
Happy new year
I was trying a new transformer yesterday which has a slightly drooping LF response (roughly -.5 @100Hz, -1 @50Hz, -3 @20Hz) and realized that it may be possible to use a cap+resistor across the secondary to modify the LF response, as opposed to the more usual battle against HF peaking (not that you couldn't do both). in this case it was a 1:5 mic input device terminated with 50k. After lots of trial and error, I ended up with a .02uF cap in series with the termination R to remove its effect at low frequencies and it works pretty darn well (measures flatter and sounds subjectively better). I wonder now if this is something Ive had right under my nose and never noticed. I guess Ive always just assumed that any cap in a zobel network is in the pF range. I do remember seeing something slightly related in a schematic Steve Dove published. he had a reactive component in a line pad upstream of a mic transformer to modify the LF response.
Is there an error in doing this? there is one thing that bothers me a little bit. I know that in general, a transformer with drooping LF response (if Im not mistaken, due to not enough primary inductance) will do better with a heavier load. ie, lower termination resistance. but in this scenario, the loading at LF actually drops (in my case it goes to zero). and the reflected impedance is going to rise at LF (what the mic sees). but then there wasn't enough inductance to start with. At first I thought it was a band-aid, but maybe this is actually working against the root of the problem?
mike
I was trying a new transformer yesterday which has a slightly drooping LF response (roughly -.5 @100Hz, -1 @50Hz, -3 @20Hz) and realized that it may be possible to use a cap+resistor across the secondary to modify the LF response, as opposed to the more usual battle against HF peaking (not that you couldn't do both). in this case it was a 1:5 mic input device terminated with 50k. After lots of trial and error, I ended up with a .02uF cap in series with the termination R to remove its effect at low frequencies and it works pretty darn well (measures flatter and sounds subjectively better). I wonder now if this is something Ive had right under my nose and never noticed. I guess Ive always just assumed that any cap in a zobel network is in the pF range. I do remember seeing something slightly related in a schematic Steve Dove published. he had a reactive component in a line pad upstream of a mic transformer to modify the LF response.
Is there an error in doing this? there is one thing that bothers me a little bit. I know that in general, a transformer with drooping LF response (if Im not mistaken, due to not enough primary inductance) will do better with a heavier load. ie, lower termination resistance. but in this scenario, the loading at LF actually drops (in my case it goes to zero). and the reflected impedance is going to rise at LF (what the mic sees). but then there wasn't enough inductance to start with. At first I thought it was a band-aid, but maybe this is actually working against the root of the problem?
mike
