Needing more than a spark test?

[homebrewed]

Umm, this is actually good.

As for my viewing pleasure? That's a stretch Thanks for the file. Very much appreciated.

I'm half thinking of modifying one of my ELS PCBs for this project. I have a blank PCB I can build. It has the display wired, and mounts for the display. Is it possible to buy one of your ADC PCB's? What do you want for one? Yes, I realize it hasn't even been fully tested, but it's a lot better starting point than I have right now. I have nothing built up to convert the stuff which appears to be riding on a 3V DC offset from U1.

I could wire up my PCB to your board. It would be a kludge, but I could then possibly get some data to look at.

I paid only $16 for 5 boards, including shipping (!). It would've cost a bit more but JLCPCB had a special deal going. I don't think it's worth worrying about $3 so just PM me your address and I'll put one in the mail. I mean, there still is some sweat-equity needed -- the opamp and ADC will cost you substantially more than ~$3! Oh yeah and let's not forget three SMT capacitors on the back side of the board that will need to be added. That will add maybe $.10 to the overall cost, too

I'm extending my offer to any other folks who are interested in the remaining boards. I'm sort of assuming that @rwm and @graham-xrf are interested as well so I'll hold onto at least 2 more until I hear back from them. For the record, it definitely is a prototype. Not designed as a plug-in "shield", that will come later.....if the results are deemed worthwhile.....

On that side of things I have been working on some driver code. I've got most of it fleshed out so I may be able to test the board fairly soon. The first test will be a simple one, just looking to see how many LSB's are dancing around w/o any signal input. FFT's of input signals will come later.

 

[graham-xrf]

Do we ever have to be determining a variable length of pulse?

Big pulses, small pulses, when they start out, there is a duration for when they have all expired. The decline from a higher amplitude does take slightly longer, but big or small, all is over by 15 to 18 uS. Nothing is contributed to the integration after a certain time.

The stretched version should have the same property, only slowed down.

My thought is that variable pulse duration detecting an "it's over" threshold, might introduce a factor that adds jitter to the integration count.
Is the method to clip the noise off the bottom?

 

[WobblyHand]

I paid only $16 for 5 boards, including shipping (!). It would've cost a bit more but JLCPCB had a special deal going. I don't think it's worth worrying about $3 so just PM me your address and I'll put one in the mail. I mean, there still is some sweat-equity needed -- the opamp and ADC will cost you substantially more than ~$3! Oh yeah and let's not forget three SMT capacitors on the back side of the board that will need to be added. That will add maybe $.10 to the overall cost, too

I'm extending my offer to any other folks who are interested in the remaining boards. I'm sort of assuming that @rwm and @graham-xrf are interested as well so I'll hold onto at least 2 more until I hear back from them. For the record, it definitely is a prototype. Not designed as a plug-in "shield", that will come later.....if the results are deemed worthwhile.....

On that side of things I have been working on some driver code. I've got most of it fleshed out so I may be able to test the board fairly soon. The first test will be a simple one, just looking to see how many LSB's are dancing around w/o any signal input. FFT's of input signals will come later.

PM sent.

 

[WobblyHand]

That's a pretty good deal for the boards. Similar to my ELS PCB. Pretty astounding really. Here's part of the rogues gallery on my desk.

The display will be mounted to the PCB. The Teensy will also be installed on the PCB. I probably won't install the buffer or transistors since they are used for the ELS. Then will cobble together the ADC board and this PCB, which in turn somehow gets connected to the detector. Anyways that's the idea.

Yeah, I know the parts are the bulk of the cost, but that's how it goes. Is there enough documentation (or silkscreen?) to assemble it easily? Have a BOM? Can I pull that info from your previous posts?

I'm trying to figure out how to get your code running. I think I have cobbled together most of the bits and pieces, but I haven't been able to compile it yet. I have XRF_MCA.ino (dropping the version) MCA.cpp, MCA.h, FastPeak.cpp and FastPeak.h. I didn't quite understand setting up the FastPeak library - I haven't rolled my own libraries yet. The incorrect installation of the lib could be part of the problem.

One of the things that is popping up is the newer version of Teensyduino 0.58.3 is quite unhappy with float_t. It's generating errors on that. Keeps saying float_t has not been declared. Could be that I'm still missing project files, however.

 

[graham-xrf]

I'm extending my offer to any other folks who are interested in the remaining boards. I'm sort of assuming that @rwm and @graham-xrf are interested as well so I'll hold onto at least 2 more until I hear back from them. For the record, it definitely is a prototype. Not designed as a plug-in "shield", that will come later.....if the results are deemed worthwhile.....

Hi Mark
Thank you for the kind offer. Given that I have the circuit of my own brewing in KiCAD, and that I won't be using a Teensy you can give one of those you are holding back to someone else interested.

I expect to be doing the same thing as you. The circuit and PCB build files will be available, and the prototype PCB will, like yours, be step-and-repeat copies. I am OK to send one to those who PM me. At the price you got yours, that is a great deal! From whoever supplied them, I wonder if they ship to UK?

My board design has Teensy as well as Raspberry Pi interface, options to include (or not) the 50/60Hz interference filter, a tracking option to use the (Glen Brisebois) ultra-low noise high bandwidth discrete FET input, (or not). There is tracking to mount the PIN diode directly on the board, as opposed to on daughter board end-on at right angles at the edge.

The circuit has links to use both AD7622, and AD7667 ADCs. The tracking underneath is identical. Also, there ground plane guarded stage test points taken to the edge. The whole thing is definitely a prototype board to experiment with.

 

[homebrewed]

Is there enough documentation (or silkscreen?) to assemble it easily? Have a BOM? Can I pull that info from your previous posts?

I will include the BOM and a hardcopy of the schematic. The board has silkscreen so it should be pretty straightforward, particularly since most of the small bits have already been soldered down.

 

[WobblyHand]

Hi Mark
Thank you for the kind offer. Given that I have the circuit of my own brewing in KiCAD, and that I won't be using a Teensy you can give one of those you are holding back to someone else interested.

I expect to be doing the same thing as you. The circuit and PCB build files will be available, and the prototype PCB will, like yours, be step-and-repeat copies. I am OK to send one to those who PM me. At the price you got yours, that is a great deal! From whoever supplied them, I wonder if they ship to UK?

My board design has Teensy as well as Raspberry Pi interface, options to include (or not) the 50/60Hz interference filter, a tracking option to use the (Glen Brisebois) ultra-low noise high bandwidth discrete FET input, (or not). There is tracking to mount the PIN diode directly on the board, as opposed to on daughter board end-on at right angles at the edge.

The circuit has links to use both AD7622, and AD7667 ADCs. The tracking underneath is identical. Also, there ground plane guarded stage test points taken to the edge. The whole thing is definitely a prototype board to experiment with.

JLCPCB is a very low cost manufacturer. I sent my PCB files to them and got a sweetheart deal for $7 for 5 boards. Must have been a new customer special. My first KiCAD board was through hole and I assembled it myself. As far as I know they ship anywhere in the world. That $7 included shipping! I don't think my next order will be that inexpensive, but I was happy with the quality of the boards. Every board that was populated just worked. But that was a digital board... Analog is different.

If you can share the design or a board, I'd be very interested. I like the idea of the FET front end. I went off to look up parts and found a lot of the FETs mentioned in the old app notes are unavailable. The new FETs don't seem to have many specs published, like for noise. It's like the original manufacturer sold the formula to a small manufacturer and these guys just make them, but don't seem to understand them like the originators did.

 

[homebrewed]

I'm trying to figure out how to get your code running. I think I have cobbled together most of the bits and pieces, but I haven't been able to compile it yet. I have XRF_MCA.ino (dropping the version) MCA.cpp, MCA.h, FastPeak.cpp and FastPeak.h. I didn't quite understand setting up the FastPeak library - I haven't rolled my own libraries yet. The incorrect installation of the lib could be part of the problem.

What kind of errors are you getting? The MCA and FastPeak libraries should look pretty similar, in terms of the directory structure. There's an examples folder (which can be empty), the cpp and header files, then a file called "keywords.txt". The latter is used to help the Arduino IDE highlight functions, much like it does for built-in functions. I've attached an example, generated for a boilerplate "test" library. Hope that helps.

 

[homebrewed]

It looks like I'm on the right track regarding improving my code's understanding when a pulse begins and ends. I've attached the latest result, and now I _am_ seeing a second peak next to that pesky aluminum background peak. The spectrum doesn't look exactly like the one I got when using the peak voltage, but the horizontal scale is different. between the two.

In the long run, it may turn out to be easier to identify the peak rather than when the pulse begins and ends, since they will be buried in the noise. Something to think about.

View attachment 435341

 

[WobblyHand]

It looks like I'm on the right track regarding improving my code's understanding when a pulse begins and ends. I've attached the latest result, and now I _am_ seeing a second peak next to that pesky aluminum background peak. The spectrum doesn't look exactly like the one I got when using the peak voltage, but the horizontal scale is different. between the two.

In the long run, it may turn out to be easier to identify the peak rather than when the pulse begins and ends, since they will be buried in the noise. Something to think about.

View attachment 435341

Attachment not visible. Get an Oops from HM.