Where did this idea come from? The mosfet pinch voltage is largely irrelevant when it comes to RF immunity of a linear gain stage. That's because the mosfet transconductance in saturation (where any mosfet intended to work as a linear amplifier has its bias point, nothing to do with bipolar saturation) is proportional to the difference between the actual Vgs and the pinch voltage Vp gm=u*Cox*(W/L)*(Vgs-Vp). Otherwise said, when it comes to the gain (signal or RFI ingress), the absolute pinch voltage Vp value is irrelevant. A more refined analysis shows that low pinch voltage mosfets may have larger transconductance at the same bias current, so they are more susceptible to RFI, but the pinch voltage doesn't, by any means, define some sort of RFI immunity threshold, like inferred by Mr. Wayne.
A large pinch voltage would increase the RFI immunity of a mosfet switching stage, but not much that of a linear amplifier.
Last edited:
The idea may have come from actual testing and by large pair I was implying input capacitance having a large effect. I also spent 10 years in RF for GE and Bendix King although that was 30 years ago. It isn't the linear part it is when being forced nonlinear by signal overcoming bias.
The whole idea of avoiding bipolar transistors or op-amps to gain RFI immunity is backwards. Fix the product...
I suspect there are a lot of "high-end" audio components that would give you audible noise if a GSM phone is placed on top and it's searching for a network.
It's like complaining about backlight bleed in an LCD and then deciding to go back to CRT displays to get rid of it.
I suspect there are a lot of "high-end" audio components that would give you audible noise if a GSM phone is placed on top and it's searching for a network.
It's like complaining about backlight bleed in an LCD and then deciding to go back to CRT displays to get rid of it.
Are the mixing boards not problematic since they use sliders and such all over that just make big gaps in the chassis? Oh wait, they are all plastic 🙄 One might hope they have some shielding below?
They all have shielding.
While I used 48-input Yamaha mx400 with transformerless SS inputs/outputs, it was chirping with cellphones. When I switched to the console that I made from TOA-216 with input and output transformers, it is silent.
While I used 48-input Yamaha mx400 with transformerless SS inputs/outputs, it was chirping with cellphones. When I switched to the console that I made from TOA-216 with input and output transformers, it is silent.
Don't know how they are constructed generally, but it's certainly possible to filter all the inputs and bury traces between planes if needed.
I would expect pro gear to be better at this than "high end" commercial stuff. They might actually put their product through some compliance testing. Plus, pro gear is hopefully going to have all XLR connectors with pin 1 connected directly to chassis. Maybe even the Neutrik XLRs with the annular capacitors.
I was called to long Island to see if I could remove an RFI problem getting into music system. To make a long story short, a very high performance video isolation transformer I suggested wasn't used in system.... a wider BW model was used... They had ordered transformer and mfr suggested a 'better' model. I put in the one I spec'ed for them and no more RFI issues.
Don't allow much wider BW signals than necessary. [Amp may be wider by necessity but I/O filters for HF/RF needs to be included to limit BW to less than amp.] .
THx-RNMarsh
Don't allow much wider BW signals than necessary. [Amp may be wider by necessity but I/O filters for HF/RF needs to be included to limit BW to less than amp.] .
THx-RNMarsh
Last edited:
Ever now and then I have some stupid questions, like today.
I searched the net quite a lot but could not find a definitive answer, so maybe not so stupid.
Just how symmetrical are "complementary" NPN & PNP BJT's, assuming they have perfectly matched Vbe and hfe ?
Are they perfect mirror images of each other to 0.0001%, or only roughly symmetrical ?
We know that complementary JFETs are anything but fully symmetrical.
Thanks in advance,
Patrick
I searched the net quite a lot but could not find a definitive answer, so maybe not so stupid.
Just how symmetrical are "complementary" NPN & PNP BJT's, assuming they have perfectly matched Vbe and hfe ?
Are they perfect mirror images of each other to 0.0001%, or only roughly symmetrical ?
We know that complementary JFETs are anything but fully symmetrical.
Thanks in advance,
Patrick
They can not be. Because they have different number of doped layers, by design.
There are no electron of P type, nor holes of N type
There are no electron of P type, nor holes of N type
So how much distortion cancellation can be expected realistically in a NPN-PNP push-pull, at best ?
60dB or so ?
Patrick
60dB or so ?
Patrick
One thing to note, the Shure microridge and the JICO SAS are the same namiki profile. The sapphire rod will be a lot heavier than the microwall Be originally used and of course the resonant modes will be changed because of that. But the generator is still one of the better ones made so whilst it wont be as good in some respects it should still be one of the better cartridges available unless you believe in the cult of MC.
It's worse than you think. Samsung ran an advert last year with Rita Ora (no idea who she is, but the kids know of her) sitting in a studio, then putting her phone on the wireless charging cushion whilst she sings her song. Then she finished and checks facegram or whatever.
After a stalled start with UMA (calls over wifi) apple now support it and it's coming to more networks. Within a year or two that might become more viable as you only have 1 block of spectrum to worry about at 2.4GHz and you have contol over the signal levels. I had it 5 years ago and it was great. Drive home, hand over to wifi, great phone coverage in the house.
Meaning, of course, in the general case, that Vin is attenuated by a factor (1-beta). Ja?
Jan
The question now is: how dows the venerable V15III compare with the latest crop of higher quality MM cartridges, such as, for example, Orotfon 2MRED, which also has very high tracking at at a reasonable tracking wieght of 1 mV/cm at 1 kHz?
While i liked the V15III on my low mass Dual arm, I slightly pefer the Ortofon because of what i hear as a bit more even overall response (re: 47k // 10 pF), with a just 5 dB peak at 10 Hz (contributed by Dual's mechanical filters in the arm).
While i liked the V15III on my low mass Dual arm, I slightly pefer the Ortofon because of what i hear as a bit more even overall response (re: 47k // 10 pF), with a just 5 dB peak at 10 Hz (contributed by Dual's mechanical filters in the arm).
I'm sorry mr. Wayne, it doesn't hold water. You mentioned large pinch voltage, now you are shifting to a large input capacitance. It's true that the RFI trouble starts when the garbage drives the DUT into nonlinearities, but to have that happening, the DUT has to have some gain/transconductance.
The idea of using devices with a large capacitances as a RFI contamination countermeasure is, let's say, funny. Thanks for sharing your experience at GE and Bendix, at least it's not 40 years old.
The original sin is that the electron and holes carriers have different effective masses and hence the Boltzmann transport equations render different electron and holes mobilities. You can pair Vbe and beta until the cows comes home, the dynamic large signal behaviour of NPN and PNP devices will still be different.
Unless I'm missing something, a push-pull doesn't cancel distortion, a differential stage may.
DVV: this is something we are starting to explore over on the analog source forum. Slow progress as none of us have much time, but it's entertaining in that we all accept its a flawed medium and its just a bit of fun to pick apart long cherished beliefs. For example the 2M series may actually not be as good as the older SuperOM series 🙂.
But many heresies being committed, such as transconductance MM stages 🙂
But many heresies being committed, such as transconductance MM stages 🙂
- Status
- Not open for further replies.