LTSpice: CFP Mic Pre w/ Filtering

[Bo Deadly]

... so I built a THAT 1510 pre with switchable Cg. I used Nichicon UES "MUSE" bipolar 35V.

I noticed something very strange that does not show up in simulation. Wr-wr-wrong: see two posts down.

Consider these plots:

The first plot just shows a baseline. Cg is "large" and gain is measured to be 65.8 dB.

The second plot shows Cg of only 100u. The strange result is that gain in the passband is almost 6dB higher at 71.5 dB even though Rg was not changed.

The third plot just shows that distortion does not seem to be an issue when reducing Cg in the CFP mic pre (or at least with a 1510). In fact, it looks like distortion in the passband is better with that 3rd harmonic gone (second plot).

Note: 0 dBFS of MOTU Traveler = 8.6 Vpp = 11.9 dBu

 

[abbey road d enfer]

The first plot just shows a baseline. Cg is "large" and gain is measured to be 65.8 dB.

The second plot shows Cg of only 100u. The strange result is that gain in the passband is almost 6dB higher at 71.5 dB even though Rg was not changed.

Are you sure that the resolution of the FFT is sharp enough to actual sit the cursor at the very top of the peak? Or maybe the 717uF has larger ESR than the 100uF...?

The third plot just shows that distortion does not seem to be an issue when reducing Cg in the CFP mic pre (or at least with a 1510). In fact, it looks like distortion in the passband is better with that 3rd harmonic gone (second plot).

It looks like your 100uF is of much better quality than the 717uF.

 

[Bo Deadly]

It looks like your 100uF is of much better quality than the 717uF.

Whoops. False alarm. As Fonzie would say "I was wr-wr-wrong".

I was measuring two different amps (it's a dual pre). I don't have the low cut switches yet. So one is hardwired for low cut and the other isn't but the trimmer was turned down on one. If I max both trimmers the gain is similar: 72.7 dB and 71.5 dB. This is actually a little hotter than the gain eq for the 1510. Maybe Rg min is a little lower than I think (two 6.8 in parallel).

But at any rate, I think this mostly answers the original question. Using smaller caps in the CFP to implement low cut seems to work fine. At least the Nichicon UES 35V are working fine.