Tweet
Dear friends, when I was making a pulse detector, I found that the general front-end input is directly connected to the positive pole of the power supply. Because the coil is an inductive load, there is a lot of noise on the power supply, which cannot amplify small signals. When I want to use an isolated power supply to provide voltage to the chip, I still find the problem. Because an instrument cannot use two battery packs, it must be powered by a battery pack, which requires a DC-DC isolation module to complete, but the isolation module The noise is greater and unstable, so ask everyone, how to use a power supply for differential amplification, thank you, thank you ... a lot of thanks!
-
-
Have you studied the many PI detectors in this forum??
Common method is Battery Positive terminal to circuit "Ground" or better called "common'.
Then a Change Pump circuit running AT the TX Pulse rate produces a Voltage higher than "Common" for the op-amp's positive rails. The op-amp's negative rail is from a regulator (3-term linear) from the battery's negative terminal.
The key is running change-pump at the TX pulse rate. This way the switching noise is coherent to the RX sampling. -
Without seeing your circuit, I will make some general comments.Originally posted by liudengyuan View Post
The noise you are referring to can be produced from a number of sources but as it relates to DC-DC converters, they produce both coupled and switch mode radiated noise. The el'cheapo ones that flood the auction sites are notorious for bad circuit layout and design and are not optimised to reduce noise. If you are using such a converter, then it will need to be completely shielded to reduce the radiated noise. the conducted noise can be reduced by adding appropriate filtering to the outputs.
The noise from the converter is also related to the circuit current draw, generally, the switching noise increases with it so if your converter is operating near its maximum capacity, it won't help, so chose a converter with the load in mind.
Look at the data sheets of the modules to see where they operate the most efficient.
With reference to the coil switching and power supply in general.
The tank or reservoir capacitor should be located as close as practical to the TX mosfet and coil connector and the PCB or wiring kept as short and thick as possible.
The remainder of the power supply circuit should be decoupled with capacitors placed as close as practical to the front end amp and timing circuit ICs or MPU. Using a low ohm resistor (10-47R) in series with the supply pins to the amp will also help decouple the noise.
Finally, when laying out your design on PCB or prototyping circuit, keep the digital circuits as far away from the analog side as possible and keep the layout as small as you can get it.
This will help to decouple power supply and radiated noise.
As I say, these are very general comments, hope it helps.
Cheers
MdtodayComment
-
Because I want the op amp and the MCU not to share the ground with the coil, I directly take the two ends of the coil for differential input amplification, and use a single power supply to supply power, and use a voltage follower to simulateComment
-
So you are wanting an isolated supply for the front end amp or are you wanting to boost the voltage as Waltr has explained?Originally posted by liudengyuan View Post
I'm just a little confused on whether you are worried about supply noise or just asking on the circuit topology for the differential amp.
If you are looking for a differential front end design, take a look at Moodz work from a few years back, that may help.
I think rather than go back and forth it would be best to post up a schematic or at least a block diagram if possible.
cheers
MdtodayComment
-
-
Schematic for my target response tester. https://www.geotech1.com/forums/atta...9&d=1549206060Originally posted by liudengyuan View Post
Used 78L05's to generate +D(digital) and +A(analog) volts. Used an Op Amp to generate a ground half way between +A. Tx coil separate from Rx coil. Rx connected to difference amplifier across +A. Way I did it, not suggesting it is best way. Maybe one advantage, don't have a charge pump generating one of the supplies making noise.Comment
-
Creating a virtual ground is another good way to provide the pos/neq Voltages for op-amps.Originally posted by green View Post
A differential amp's biggest assets is isolation from Common Mode Voltages including noise.Mainly for isolation of differential amplification
I have a feeling something else is not correct. As suggested you need to post schematic of what you are doing.
Also, avoid any 'independent' switching power supply. If you need to create another Voltage then build a circuit that uses the same CLOCK, Synchronized, that is used for the TX pulse and RX sampling.
Even some of the simply VLF detectors use the TX oscillator to run a charge pump to create the negative Voltage for the op-amps.Comment
-
Why doesn't anyone use an instrumentation amplifier for preamplification? Will using an instrumentation amplifier reduce noise and high voltage interference?Comment
-
Probably the biggest problem I have is getting the IB Rx coil signal low enough at Tx turnoff. I used AD620's a lot for strain gage bridge amplification. Not enough BW for PI, but I see some that would be if input signal at turnoff is low enough.Originally posted by liudengyuan View Post
Scope pictures of Rx coil at Tx turn off. Have been looking at amplifier out, maybe looking at Rx with amplifier not connected will help.Comment
-
In all DIY pulse detectors, is there a common single-chip microcomputer that can replace lm556 ，4538,in delta pulses for timing, delay and detection, who can tell me, thank you very much!Comment
-
this turkish guy did sure the way replacing all CMOS chips on micro. this is PIM serija, DP2, DP3S.
i do not follow that he does. do not have interest. files shared by me have those files and you can see.Comment
-
Originally posted by liudengyuan View PostHave you read other threads in the forum? All of this has been discussed in detail. The Hammerhead II project shows the use of a PIC to replace the 555 & timing chips, and almost any PIC can do this. As for differential, look at the project forum called Differential Front-End PI. Also, here is a "standard" PI front-end, damping R and clamp diodes omitted for clarity. As you can see, it is actually differential, except for R1. As the coil voltage settles all you are left with is the thermal noise of R1. If the ground is noisy, it should common-mode out, assuming a decent CMRR.Originally posted by liudengyuan View Post
Comment
-
Why is the "standard" PI front end differential input?Originally posted by Carl-NC View PostHave you read other threads in the forum? All of this has been discussed in detail. The Hammerhead II project shows the use of a PIC to replace the 555 & timing chips, and almost any PIC can do this. As for differential, look at the project forum called Differential Front-End PI. Also, here is a "standard" PI front-end, damping R and clamp diodes omitted for clarity. As you can see, it is actually differential, except for R1. As the coil voltage settles all you are left with is the thermal noise of R1. If the ground is noisy, it should common-mode out, assuming a decent CMRR.
Comment
Comment