I would expect the input capacitance to vary as a result of the enhancement or depletion zone changes caused by the e fields of the gate on the current distribution. And consequentially, I'd expect a "gain" modulation as well.
jn
jn
I haven't thought much about that. I was just thinking out loud about the physical mechanism.
I had a problem back in the 80's with emitter dip failures of some 100 volt very hot (edit:high frequency) transistors, they were punching through at high voltage because of the really thin base. Found them by comparison of Icbo and Ices. From then on, I consider depletion zone width issues for fets, bipolar, scr, and diodes when high slews or high v is involved. (edit: I had to modify the ATS to prevent voltage slew beyond the device design limit when testing BVces. To measure BVces, the machine would run the device active, then ramp the collector voltage up. Although the parameter is not an active one, the machine would run the device active to get there, and that would pop the device by punching the base.
jn
Last edited:
Moving on to more productive discussion: What would we use, in the case that we needed pretty good performance with either MM or MC cartridges, but still using Toshiba fets? I will give an example when I find it. The JC-80 phono stage is a good example.
Before answering any specific question: I would like everyone to look carefully at the Ciss graphs again. Ready? Ok, now NOTE THE SLOPE of the Ciss curve. It goes down with voltage, doesn't it? Does it ever go up? No. So the primary distortion mechanism is second harmonic. However, this is a push-pull design, so a certain percentage of the second harmonic generated will be cancelled, and there will probably be a 3'rd harmonic residue, that will 'come out of the woodwork' when the 2'nd harmonic is cancelled enough. However, since the N and P channel fets are NEVER the same capacitance for the same Gm, we don't expect the 2'nd to be nulled, just reduced. Therefore I would predict a mixture of 2'nd and 3'rd harmonic as the final residue of the subtraction process.
"(JD, you are aware of the aging effect of high dosages of medium wavelength ultraviolet on the epidermis?)"
A much better source of natural vitamin D than you can get any other way, I am all in with Jan on that one. Everyone is way to afraid of the sun, it is our friend if we are not stupid about it. (GOT MILK?)
A much better source of natural vitamin D than you can get any other way, I am all in with Jan on that one. Everyone is way to afraid of the sun, it is our friend if we are not stupid about it. (GOT MILK?)
Hi John,
Should I'd build an MC RIAA stage, I'd go balanced. (Probably not applicable for commercial products, though it was done by the late Allen Wright). It seems to me that in such a case (balanced), the cartridge will see half the capacitance of the total G-S junctions. Is that correct?
Should I'd build an MC RIAA stage, I'd go balanced. (Probably not applicable for commercial products, though it was done by the late Allen Wright). It seems to me that in such a case (balanced), the cartridge will see half the capacitance of the total G-S junctions. Is that correct?
In most cases, yes. There's also a 3dB noise penalty for topologies like Allen's.
According to what I've read, the medium-long wavelength UV is more potentially damaging because of penetrating deeper, while the shorter wavelengths, although provoking burns with overexposure, is the stuff that stimulates the vitamin D production.
It's also interesting to note that malignant melanomas, among the most dangerous of such, often occur in areas of the skin NOT exposed to light very much. A puzzle.
Back in the day when I worked in radiology, they were diagnosed by a kiddy mnemonic rhyme, and the flip of a coin : Dime size (dutch) was OK, a Quarter no good.
My various collegues also felt it obligatory to inform me, one by one, that tats were an increased carcinogenic risk.
(till it reached the stage of covering them up, looked more Pro too
)No more, dime-size and under can be deep-deep undercover "moles" , and tripping over a couple of cases of melanoma developing in tattooed tissue doesn't imply a high level risk.
There's Science for Ya.
(genetic memory does apply, in particular for those of African descent, the sun-shaded melanoma case)
Speaking of pro-deformation, a year after taking another life's turn at the traffic lights, I diagnosed myself with Hodgkins due to exposure to high levels of ionized radiation, considering my somewhat careless nature.
Turned out Toxoplasmosis (cat virus
)
Last edited:
And with that doubling of devices on both sides, we're back to the same capacitances --- however with a much closer match of variations in push pull of those, hence a better cancellation of even-order distortions. And as well, a likely better rejection of externally-induced noises.
The sister Discrete OpAmp thread went quickly to surface mount parts (sign of The Coming Apocalypse, but I digress). Makes me wonder if a 2013 MC phono design could include a phantom powered head amp up near the phono cartridge. Would remove the desire for differential inputs, but would need to work with existing two-coax signal cables (only).
Thanks,
Chris
Thanks,
Chris
Sure. Although even with more conductors the design chore is non-trivial if wideband feedback is involved. I did a dual channel charge preamp many years ago, with the front end in a cold chamber with the detector. The resulting overall design was hard to stabilize, and there was residual ringing in the step response that was evidently not deleterious, but it bothered me to see it. And I wasn't trying to do each channel with two wires!
An "open collector" single stage with just a FET and its source resistor would work, and even let shields be grounds for signal and power, like "mic power" for computers. No feedback possible, and not very fancy, but signal levels are low, so maybe enough.
But I really had in mind some arrangement with power on the shields and both genders of FETs - you know, something difficult and painful. That's the cowboy way.
Thanks,
Chris
- Status
- Not open for further replies.