AMX Schematics & Layout

[Olly]

I've been having another look at the TX circuitry and am particularly impressed with the use of a CUK converter for the -ve voltage generation.
This is a much cleaner solution than a charge pump and capable of higher currents with proper regulation, lower ripple and much less noise.

[Carl-NC]

The project is suspended until I have time to work on it.

[lxs17757414132]

AMXHave you started working yet? I'm looking forward to it

[KingJL]

This is certainly interesting and worth looking at down the road, but my first priority is to get the fundamentals working. For that I'd rather not introduce any extra elements that might affect the target responses or the tilt. Also, my PCBs are on the way.

I understand completely!

[Carl-NC]

This is certainly interesting and worth looking at down the road, but my first priority is to get the fundamentals working. For that I'd rather not introduce any extra elements that might affect the target responses or the tilt. Also, my PCBs are on the way.

[KingJL]

Here is V1.1. Changes are:

ADC footprint corrected
Polarity of C51 corrected
Added R56 (0Ω) to the output of IC7b for DAC option

Thanks to everyone for the feedback!

[ATTACH]n411747[/ATTACH]
[ATTACH]n411746[/ATTACH]

Carl,
I once worked with a very successful project manager who would say "In every project, there comes a time that we need to shoot the engineers and produce a product!"
That being said, I have been playing around with the addition of a noise cancelling loop to the first stage of the differential FE. I believe there are significant improvements overall noise performance of the FE with these loop(s). Also there is significant improvement of the suppression of the power mains interference (50, 60, & 400 Hz) area.

Providing screen shots of both noise loop and non-noise loop circuits for comparison:
AMX FE non-noise cancel circuit and simulation outputs (same as published in AMX V1.1):



AMX FE with noise cancel circuit and simulation outputs:




If interested, I can provide sim(s).

[Carl-NC]

I made a small PCB that holds 2 single 8-pin opamps and plugs into a dual opamp footprint, specifically to substitute a dual with 2 singles.

[ivconic]

Yes, I will draw all the components and then if component like resistor is not needed; simple wire jumper will be put.
Or omit and left with unused pin holes. As the schematic dictates.
Good and at the same time bad thing is that LME is dual opamp, so I can't try there any single one, once I draw the connections for dual type.
That's bad, and good aspect of that is because I can use one LME for both input opamps and by that simplify the wirings and supply.
Same apllies for other two opamps.
Of course; through-hole pcb will not be in the same "league" with SMD 4 layer board, but hopefully will be good enough to test the concept and exercise the code side for dev board.

[Carl-NC]

Thanks for the kind words.
I never mind answering questions and, yes, I prefer to do so on the forums rather than pm/email.
Ivica, now you see that the options are very similar and can easily be accomplished by kludging in resistors as needed. Kludging is easier to do in a thru-hole board so for SMT I prefer to explicitly include the extra components.

[ivconic]

I am/was not defending Carl... He is a big boy. I was really just expressing my appreciation of his effort.

I absolutely agree!
Also my full respect goes to Tony Tinkerer, Paul Moodz, Willy Bayot, you KingJL, Olly and other participants who are present here and trying to contribute this very interesting project!
From someone who wasn't the least bit interested in PI detectors; you have turned me into an "over-interested d(m)onkey" who is dying to see it live and get it working properly!
Because the initial criteria are very ambitious and because I have not seen such a machine yet.

[KingJL]

[FONT=Comic Sans MS]... and KingJL there is no need for you to "defend" Carl from me...

I am/was not defending Carl... He is a big boy. I was really just expressing my appreciation of his effort.

[ivconic]

Olly and KingJL there is no need for you to "defend" Carl from me, because I consider him a very good friend and hence my freedom when I address him.
And believe me, I am addressing these and similar questions and "inquiries" to Carl only and only for educational reasons.
And I'm a "slow learner" and often have need for further explainations.
I could have contacted him via email and I'm 101% sure I would have gotten the same answers.
But I deliberately do it on the forum, publicly, to leave a trace, for the next generations of enthusiasts and those who want to learn something.
I think that Carl understands my behavior very well and that I don't need to make any special excuses.
And of course; my main motive is to learn something new but also something old that I didn't fully understand.
..
Thanks Carl, clear and understandable as always!
So now I get it. All three options are equally good, it seems to me.
Choice will be conditioned with opamp&coil specs, in short.
And now I am tempted to draw pcb with all the option included!!!

And to switch between the options with jumpers.

[Carl-NC]

Can you (if not asking much) separate both options in two separate schematics and explain in short both options, pros and cons for each one.

Here is the voltage mode preamp, based on Tony's original schematic:


You set the gain with resistors and it has more stable gain vs coil resistance.

Here is the same schematic but in current mode:


The only difference is the input resistors = 0 and coil resistance determines gain. Will have lower thermal noise.
Here is the single-ended version operating in voltage mode:


This should have slightly lower noise than the differential voltage mode above. Replace the input resistors with 0 ohms to make it a single-ended current mode preamp with even lower noise.

[KingJL]

I have re-created the AMX V1.1 in Kicad V7.0 (NO EASY TASK), if anyone is interested.

New version!!

Changes:

Changed attributes of small filled zones on top copper layer from "Thermal Reliefs" to "Thermal Reliefs for PTH". I usually leave all filled areas to "Thermal Reliefs" to prevent "tombstoning" (which I have encountered... but usually because the filled areas were larger). I decided to keep this board as identical to Carl's gerbers as possible.

Removed the trace that bounded the GND fence filled area at bottom left of top copper layer. This was an artifact caused when exporting the original gerber to the PCB that I missed in my cleanup.

Added the the attribute of 1 mm board edge clearance for copper layers to match the original gerbers.

Added 1 mm fiducials to each corner (front and back) to aid those who have/use "pick-n-place" equipment (Brian, this is for you!).

I will be removing the link to the previous version.

[KingJL]

... respectfully I think that Carl has done an amazing job with the experimental platform...

I totally agree!