JR, your POV makes sense to me on the servo.
Servo and non-servo both look AC coupled. I'm failing to see the real point of it being there with all the AC coupling. Should I take it that the inverter path compounds DC offset of the U1A path, thus the servo?
Looks like the same light-weight output, just higher slew and bandwidth. But like the existing MLA7, just a 4558 driving the output and spec mentions 10K loads, I wouldn't be doing worse.mediatechnology wrote: ↑Tue Jan 15, 2019 1:04 pmWhatever works for you and can be made not to oscillate. I have some TLE207X somewhere. Can it drive low impedance loads or does it have a TL072 light-weight output?I suppose that also may relate to choices made around the op amp selection too? Figure 3 of the Valley TA-103 document (better HERE) mentions using 5532/5534 for driving lower Z lines, though I like the sound of TLE207x family fine and it appears to be fine for a direct sub in the Yamaha.
That was the starting thought!mediatechnology wrote: ↑Tue Jan 15, 2019 1:04 pmYes the inverter can be used to provide the anti-polarity output. In the two op amp topology of Self and Roberts common mode rejection is realized in the differential stage connected to the input transistor collectors. There's no need for a third differential amp to obtain common mode rejection like there is in the TransAmp, "Cohen," THAT151X, INA217 etc. (My hunch is that the two op amp lacks the symmetry of the three op amp which might affect HF CMRR but I'm not sure its worth it in most cases. The improvement you're making bringing NFB to the emitters is significant.)Is it unreasonable to use the inverter output as the rest of a balanced output? I assume the standard approach is to feed a differential amp with that pair of signals. I was previously thinking about cutting output connector traces to add impedance balancing to the outputs.
Looks like Self fig 14.11 to me, same thing I think. I'm still unclear on the gain pot here, looks like 2 that must be on the same shaft. Can't imagine two gain controls on any of the consoles he's referencing.mediatechnology wrote: ↑Tue Jan 15, 2019 1:04 pmYes I believe so. Need to look at it to see if we can call it "shared gain." (Had a look-see. No, it's not. It's a Ted Fletcher "SuperBal" gain stage.)The lower gain 'padless amplifier' without the servo in Self figures 13.5-13.6, I'm taking it that's a dual section pot used to control input and output stages simultaneously.
I've got a pad built into this thing already. I guess I can experiment with inducing common mode noise and compare real world. All my historical usage was LOUD, these days trending much quieter. I'll have to see what sort of gain range it falls into at first hack....autodidact and all.....mediatechnology wrote: ↑Tue Jan 15, 2019 1:04 pmI don't grok his pad man. IIRC its also on the dry side. I've always used a three R pad on the wet side.Sidebar: the comments about the 4 resistor pad in 13-4 are interesting, in that I feel I rarely see that used. Probably because I'm usually looking at transformer inputs where the 3 resistor does the job, yet it's a 3 in the Yamaha; easily changed.