PMA said:
Rectification is something happening as one of the possible consequences of the susceptibility of the circuit.
Could you answer the question:
Is a bjt front end more or less susceptible than a fet to near field magnetic field?
Is a bjt front end more or less susceptible to an electrical field than a jfet?
JPV
I tested about 20 different opamps for RFI sensitivity, in a high noise gain circuit. In our usual environment, we can easily find detection of TV row frequency or FM pilot. BJT opamps are MUCH more sensitive (in orders) than JFET opamps. Especially some BJT opamps are very sensitive, like new NS series.
Attachments
Rfi
Clear? Not at all. Quadratic (jfets) or exponential (bjt) both are nonlinear
. RFI susceptibility depends on much more things than only jfet/bjt's at the input stage. For example, on how the next stages and frequency compensation are configured. Moreover, a NPN/PNP input stage (read: CFB) can be made highly linear without any change of clipping (i.e. rectification due to RF), opposed to a LTP input stage (no matter whether equipped with bjt's or jfet's).
PMA said:
Clear? Not at all. Quadratic (jfets) or exponential (bjt) both are nonlinear
. RFI susceptibility depends on much more things than only jfet/bjt's at the input stage. For example, on how the next stages and frequency compensation are configured. Moreover, a NPN/PNP input stage (read: CFB) can be made highly linear without any change of clipping (i.e. rectification due to RF), opposed to a LTP input stage (no matter whether equipped with bjt's or jfet's).PMA said:
You do not answer the question
RFI is many thinks and coupling has many different paths.
When you make an EMI experiment and you draw generic conclusions you should know if the coupling was conducted noise via cable pick up or radiated coupling differential or common mode, magnetic or electric or plane wave.
Correct knowledge induces correct use and design.
Your statments are more like in the old age where rfi was considered black magic. Then came Don White, Morrisson and all the others and it is now a complex engineering branch difficult but with methods of analysis/design and carefull measurement procedures.
I am not saying that you are wrong, I am saying that correct use requires knowledge, therefore my questions.
If your circuit shows rectification in a certain environment with bjt's and less with jfet it shows to me that your circuit is susceptible to EMI whatever the font end. Perhaps in another environment and with the same circuit, your fet input will be more influenced or something else in the circuit.
Cheers,
JPV
PMA, isn't is ironic that you and I were debating the Blowtorch openings and their allowing RFI without a transmitter nearby, and these 'critics' who would throw away what we have learned over the years.
I must say that it, of course, depends on the environment where you live, IF this discussion will be important.
Once, over 40 years ago, I lived in the Santa Cruz mountains in an isolated cabin. My FM could only be picked up by a high gain antenna in a tree, aimed 50 miles away. I am sure that I got no significant RFI there, since home digital computers or fax had not been invented yet.
However when I moved this same audio equipment to another place (similar to where the Burning amp festival was located) I had RFI pickup no matter what I did with TUBE equipment. Why? I was line-of-sight across the SF bay of the Sutro TV tower which is the main TV tower for the entire SF bay area. It was so powerful, that I had to change all my cables to foil shielding to get 100% coverage, etc.
When we design, we prefer to design, 'reasonable worst case' to cover almost all situations. That saves embarrassment and problems later, yet we don't have to make it EMP proof, that may take an isolated environment that costs 10 times as much.
I don't mind spirited debate, but let us not do 'mindless' debate, quibbling over what book is relevant, etc. By the way, Ott's book is very good.
I must say that it, of course, depends on the environment where you live, IF this discussion will be important.
Once, over 40 years ago, I lived in the Santa Cruz mountains in an isolated cabin. My FM could only be picked up by a high gain antenna in a tree, aimed 50 miles away. I am sure that I got no significant RFI there, since home digital computers or fax had not been invented yet.
However when I moved this same audio equipment to another place (similar to where the Burning amp festival was located) I had RFI pickup no matter what I did with TUBE equipment. Why? I was line-of-sight across the SF bay of the Sutro TV tower which is the main TV tower for the entire SF bay area. It was so powerful, that I had to change all my cables to foil shielding to get 100% coverage, etc.
When we design, we prefer to design, 'reasonable worst case' to cover almost all situations. That saves embarrassment and problems later, yet we don't have to make it EMP proof, that may take an isolated environment that costs 10 times as much.
I don't mind spirited debate, but let us not do 'mindless' debate, quibbling over what book is relevant, etc. By the way, Ott's book is very good.
AndrewT said:
I wonder if these filters will be efficient in the Mhz range were they should be.
Don't you think that good shielding of the speaker cables with shield termination at the clean ground just at the entrance of the pcb would be more efficient.
JPV
JPV said:
Right on. It is refreshing to find around people with a little more than black magic beliefs or self promoting sales pitches.
I have measured myself dozens of opamps for noise, RFi, distortions, etc... and while I cannot claim any absolute quantitative results, the relative results are clearly showing no difference between JFET and bipolar input devices RFI behaviour. One notable exception: all bipolar micropower and CMOS opamps (some of them otherwise highly regarded, like the OPA365) which are definitely not good. But even with these, it takes some serios skills to lay out a topology that would make the effect audible.
I was located a little higher up from the water on a hill in Tiburon. I didn't hear any TVI at the Burning Amp festival and I wonder why? Could it be shielding from the bridge itself? The location so near the water? Now that IS an interesting question.
john curl said:
I can't speak for others, but I used to design RF stuff so know how to lay out AF gear, on subconscious level.
In our TIASUR campus many students were amateur designers; but if some room did not have one they used stock amps, turntables, tape recorders. Who helped us to test for RFI, our fellow students running a collective ham station on the 9'th floor of our building with sophisticated forest of antennae on the roof. When they entered some contest nothing could help: lights in the building dimmed in tact of their speech, so high was the power. Demodulated SSB sounded like ducks and pigs.
Once I've found the way to silence them: I made a simple very dirty 3.5 MHz exciter frequency modulated by saw so it covered by harmonics all their bands: 3.5, 7, 14, 21, 28 MHz. I put the exciter into a slipper, connected a ground wire to a central heating battery, and antenna to my bed's iron frame...
They stopped transmitting... One mate come to our room, like for simple conversation, but his eyes apparently were searching for some nasty device around... When he gone I continued my listening session. 😀
Once I've found the way to silence them: I made a simple very dirty 3.5 MHz exciter frequency modulated by saw so it covered by harmonics all their bands: 3.5, 7, 14, 21, 28 MHz. I put the exciter into a slipper, connected a ground wire to a central heating battery, and antenna to my bed's iron frame...
They stopped transmitting... One mate come to our room, like for simple conversation, but his eyes apparently were searching for some nasty device around... When he gone I continued my listening session. 😀
Going back to the op-amp discussion wrt to RFI - you have to be carefull to draw a disctinction between different classes of op-amp. A micropower BJT input device is not a good choice for audio - it belongs in battery powered DC amplification applications (typically) where you can get away with heavy input filtering.
A lot of the newer audio amps run the input LTP or complimentry LTP at quite high currents and they use degeneration so they are less susceptible to input overload and RFI. Their is a tradeoff to be made in terms of intrinsic input reffered noise performance though.
Since op-amps do have high loop gains, layout to avoid noise pick-up is very important. One of the worst mistakes I have seen is putting the feedback resistors near a circuits output and then running a long trace all the way back to the inverting input. Well, if the OPA has 120dB OLG, you will probably get problems no matter what type of op-amp is used in a case like this. Lets also not forget about driving capacitive loads as well: isolate, isolate, isolate - easily done with a 50 Ohm resistor in series with the output.
A lot of the newer audio amps run the input LTP or complimentry LTP at quite high currents and they use degeneration so they are less susceptible to input overload and RFI. Their is a tradeoff to be made in terms of intrinsic input reffered noise performance though.
Since op-amps do have high loop gains, layout to avoid noise pick-up is very important. One of the worst mistakes I have seen is putting the feedback resistors near a circuits output and then running a long trace all the way back to the inverting input. Well, if the OPA has 120dB OLG, you will probably get problems no matter what type of op-amp is used in a case like this. Lets also not forget about driving capacitive loads as well: isolate, isolate, isolate - easily done with a 50 Ohm resistor in series with the output.
I will tell you this: Neither the Blowtorch preamp or the new Ayre preamp have significant RFI issues, and it cost us big dollars to get them that way. We both use FET input, and thick aluminum cases. However, it works! Just because we get good reviews isn't necessarily because of the RFI issue, but it seems a good thing to try to get right.
Many here might use a plastic box with a layer of conductive spray. That's OK, I made my first Vendetta preamp prototypes like that. It was all that I could afford at the time.
Now I make better shielded preamps, and there should be no argument about it. Why some people will not consider RFI protection or shielding important.
Many here might use a plastic box with a layer of conductive spray. That's OK, I made my first Vendetta preamp prototypes like that. It was all that I could afford at the time.
Now I make better shielded preamps, and there should be no argument about it. Why some people will not consider RFI protection or shielding important.
john curl said:I will tell you this: Neither the Blowtorch preamp or the new Ayre preamp have significant RFI issues, and it cost us big dollars to get them that way. We both use FET input, and thick aluminum cases. However, it works! Just because we get good reviews isn't necessarily because of the RFI issue, but it seems a good thing to try to get right.
Many here might use a plastic box with a layer of conductive spray. That's OK, I made my first Vendetta preamp prototypes like that. It was all that I could afford at the time.
Now I make better shielded preamps, and there should be no argument about it. Why some people will not consider RFI protection or shielding important.
Is thick aluminium case for shielding. I do not understand the necessity of thick because it is well known that the efficiency of the shielding is more affected by the seams and openings. If the material is a good conductor no need for thick because shielding is essentially a reflection phenomenon
One exception of course is for low frequency magnetic field.
JPV
We have explained this round and round, but people do not care. Thick aluminium or copper also shield against LF magnetic field, skin effect - eddy currents. Depends on thickness and frequency, of course. Thick conductive, non-magnetic material does not saturate.
Plastic with metal powder sprayed cover is almost useless, just better than nothing.
Plastic with metal powder sprayed cover is almost useless, just better than nothing.
John Curl said:
PMA said:
Don't you guys love this plural of majesty
Or is it "we" against "them" 😀- Status
- Not open for further replies.