Needing more than a spark test?

[homebrewed]

This is all getting good - except for the bent pin.
Getting all zeros when the input is shorted is, in concept, what is supposed to happen.
I practice, if all zeros happens, without at least some activity in the last two or three bits, that is cause to suspect something might not be all OK.
If all is well, you should sometimes see the LSB +/- 1 uncertainty.

A count of as low as 20 is more than 70dB down on a 2.5V reference, yet would be jangling the last five bits.
A count as low as (say) 5, (101), is 82dB down, and claimed SNR is 89dB, so a shorted input might give all zeros.
Except for the assembly mishap, I think you are doing great!

Re: The signal generator waveform. I am not sure about 27 bits. We only have 16.
1mV p-p out of a 2.5V reference is 2500 count, which requires 12 bits.
It can be tricky to deliver a signal generator signal in clean without also bringing in at least some unwanteds, and when it is reduced to only 1mV, they may start to show.

I still think you are acing this!

You are right on the money regarding my initial dismay at seeing all zeroes. I was expecting some (hopefully small) number of LSBs bouncing around. Still wondering if there's some sort of negative offset that is artificially lowering the noise floor. I'm going to do some testing with a plain old potentiometer to see how the ADC behaves once I increase the input enough to start seeing some action, bit-wise.

Regarding my comments regarding the apparent resolution issue, a little online searching revealed that my signal generator's output has a resolution of 14 bits. That explains it all right there. Oh, BTW, when I mentioned "27 bits" I was talking about LSB's not 2^27

I made a stab at using the polish wire trick but couldn't get the wire (a thin piece of nichrome wire) around the bent portion of the pin. But I haven't given up quite yet. The thought occurs to me that I may be able to get it behind both affected pins. But before I embark on that, I'm going to buy some solder flux to help prevent the otherwise-inevitable solder bridging.

The 1.5" diameter ABS rod I ordered for my "new & improved" focus ring should be arriving soon; and my additional 8 sources are claimed to be arriving today. So I have other stuff to do -- along with a laundry list of domestic things. The glazing on our greenhouse has severely degraded due to UV and at least one panel has just plain fallen apart (assisted by a tree branch) so it all needs to be replaced. This time with glazing that actually has UV protection, unlike the cheap stuff that came with the Harbor Freight greenhouse kit.

I have a little 4" battery-powered Makita saw that's perfect for cutting the panels to size. It is meant for cutting ceramic tile and cement board so the blade turns quite slowly, perfect for cutting plastic as well. I have a carbide-tipped 4" blade for it in addition to a diamond blade.. Turned around backwards, it cuts the stuff nicely

 

[graham-xrf]

You are right on the money regarding my initial dismay at seeing all zeroes. I was expecting some (hopefully small) number of LSBs bouncing around. Still wondering if there's some sort of negative offset that is artificially lowering the noise floor. I'm going to do some testing with a plain old potentiometer to see how the ADC behaves once I increase the input enough to start seeing some action, bit-wise.

I used smooth feeling potentiometer that had enough friction to stay put, and one of those ultra cheapo flea market LED 3-digit displays to get an approximate idea of what it was doing.

I made a stab at using the polish wire trick but couldn't get the wire (a thin piece of nichrome wire) around the bent portion of the pin. But I haven't given up quite yet. The thought occurs to me that I may be able to get it behind both affected pins. But before I embark on that, I'm going to buy some solder flux to help prevent the otherwise-inevitable solder bridging.

I got quite slick and practiced at it. I had a whole lot of 28-pin SOIC to remove. From Bruce @WobblyHand 's description, he has found that the little hot air gun is very effective. I still have to break out that kit and try it. I also harbor dark thoughts to try some of your methods. There is little oven, with the PID heater control method, and the temperature controlled eBay hotplate. I think one is supposed to pre-heat to some point before melt, then give it the beans until reflow happens, and then let take it off to cool. I have not yet invested in any of that stuff.

The 1.5" diameter ABS rod I ordered for my "new & improved" focus ring should be arriving soon; and my additional 8 sources are claimed to be arriving today. So I have other stuff to do -- along with a laundry list of domestic things. The glazing on our greenhouse has severely degraded due to UV and at least one panel has just plain fallen apart (assisted by a tree branch) so it all needs to be replaced. This time with glazing that actually has UV protection, unlike the cheap stuff that came with the Harbor Freight greenhouse kit.

Yep - I also have a daily fraction devoted to improving/replacing degraded stuff.

 

[WobblyHand]

My hot air gun set with 9 nozzles, and a spare heating element cost me $34.20 in April 2019. The Hot air gun clamp stand repair platform cost $26.73. Both ordered from AliExpress. Especially with the stand, it has made selective work very easy. Set up hot air nozzle adjust height, turn on heat, slide board under nozzle, remove part in 2 seconds. Remove part means either flick the part away, or simply pick it up with tweezers, because the solder on all the pads is liquid. It literally was that easy. You may have to wait a while if there is a ground pad under the part. If the pads are not overfilled with solder, putting on a new part is just alignment, and sliding the part under the nozzle. As you well know, it isn't always that easy, and alignment can be devilishly tricky. I have soldered 0603s by pressing down with a wooden stick and applying the hot air on old pads that haven't been cleaned of solder. That's handy if you just want to try a quick value change.

 

[homebrewed]

Here's a photo of my "ADC test setup". The signal generator is sitting on the seat of a nearby chair .


My jumper wires all are pretty long so I did a bit of surgery and shortened all the data lines. BTW the large green wire sticking out from underneath the ADC board is a jumper wire connecting the input power supply Ground pin to the digital ground. I did that before resolving the timing issues so I don't know if it helped or not. Also shown: a 10uf capacitor across the +5 & Ground pins. Also done prior to fixing timing problems, same caveat.

 

[graham-xrf]

Goodies!
So OK - it would seem that one Long Jun, starting from Songgang Avenue, Bao an District, in Shenzen has better access to a functional postal service to UK than folk in UK can enjoy when attempting to send a (very) small letter-style package to the USA! I notice that these sources have 0.8uCi of Am241, so a little less than the 0.9uCi of my other sources.

Amid much else going on, my wife brought me the couple of packages this morning (Yay)! Long Jun clearly was not thinking about plastic pollution much when he wrapped the whole thing in so much brown(ish) packing tape, it covered the whole polystyrene outer box, but now, we have these...

_ _

I will, of course, be trying something out pretty soon after lunch. If it glitters, vibrates, or does any kind of dance, I will let you all know.

 

[graham-xrf]

Radiation Detector-First tryout
The first thing to note is that I am going to have to get some more reliable information about what is contributing to the count, which means likely I have to use some lead. This gadget sees the background radiation, plus the stuff I have nearby, and keeps ticking up the cumulative dose past an alarm setting that multiplies the count rate to get dose/hour, or year, or something..

Also, it more or less freaks out quite a lot of the time, partly because of the alarm settings, which I have not messed with (yet), and, I guess, it may have something to do with the fact there are16 smoke detectors + two thorium mantles spread over the desk

What you get depends a bit on which part of the case the radioactives are closest to. That is not explored yet.
It can detect general ionizing x-rays, beta (electrons), and gamma. What it can see on the desk is all gamma.

One Am241 source, liberated from its ion chamber, delivered over 240 counts per minute, and looked like the count would be higher if I got closer, or gave it longer. It was about 75mm at the time, when the alarm freaked out.

Just putting the gadget on top of the pushed together pile of (10) sources as in the picture above started the count rocking up rapidly. After a few seconds, it freaked out again. I think the count had gone over 400. The sources were all still in their ion chambers, exactly as in the picture.
Some of the X-rays might have been from collisions with the steel and metals of the ion chambers.

All that happened on the desk was well contaminated , I mean "offset" by the contributions from the two Thorium mantles nearby. If I wave it at a lamp mantle, it freaks out more or less immediately. Only a few seconds, and the count is over 1000/minute without giving it as much a a minute to find out. At the moment, I can't do sensible stuff. I have to put the "extras" far out of the way first. Thorium mantles have now been temporarily stuffed down into the middle of the roll of lead that awaits a melting adventure, er.. I meant to say "procedure".



I think we can expect about 4Hz from a single source, and maybe 30Hz from a set of 8. I may be wrong about that, but I think that is a reasonable guess at minimum, of which only some smaller portion will work for us. I will try and do some more carefully set up counting soon. I want to get some lead sheet arranged so that this fooling about with count rates is somewhat less ad hoc.

I really do want to check out my red-tip TIG welding filler rods. They are 2% thorium. Maybe one can make circular "grooves" in the lead control shield around the diode, and bend some TIG rod to pack into them. Perhaps secure with some hot melt glue. I have to check out thorium. I don't think it has photons the same quality as Am241.

 

[graham-xrf]

Maybe Thorium is not so useful. I am still hunting out a good spectrum. I really thought it should be.
I notice that quite a lot out there display on something called "Spettro".
There is enough alike in Theremino MCA (looking at the colours) and the Ref buttons to make one think they had some common origins.

 

[WobblyHand]

My lost package, which was my original order from Digikey showed up today. It seems I have duplicates of everything. Eventually Digikey will want to charge me for it, so I have to decide what I want to send back. Might send back a couple of ADC's, as I have two now (actually 4 in my possession) but I am on the fence on the LTC amplifiers. It's good to have stock for future projects, and perhaps a repair, but, having to pay another $100, I'm not so happy about. Decisions, decisions.

Curiously, it has all the parts I ordered including the "back ordered" ones - but the notification I got at shipment, was the LTC6268-10's were not in stock.

Have to make a better work set up for electronics, with more room. Soldering individual bits is certainly awkward. Probably rearrange the desk tomorrow and see if it works out better. Need to make the microscope more convenient - can't do the little stuff the way I used to.

 

[graham-xrf]

Perhaps Thorium?
It seems one needs about 4 sheets of 2.5mm thick Plexiglass between the lamp mantles and the the detector to block the beta particles, otherwise it produces a background count that swamps most of the peaks.


Check out the counts from the daughter products on the way to becoming lead atoms.
There is one at about 80KeV, a big one at 212KeV, and some others a bit over 300KeV, more at about 590KeV, and another at 90KeV.
These could let us see Tungsten, Gold, Platinum, Mercury, Lead, and then a few others that we surely are unlikely to be interested in!
It would also give us returns from lead shielding that maybe we also don't want.

I found this stuff at --> https://www.randomuseless.info/spectra/results/mantles/index.html

 

[graham-xrf]

Variable gain amplifiers question..
@homebrewed :
Hi Mark
On your signal conditioning board, you had a choice of three gains, 1X, 10X, 100X, done by tapping into the outputs between opamp stages, though I am not exactly sure how the P2 device does it, and the output is clamped not to exceed 3.3V+0.65, nor go below -0.65V.

Do we know how much gain we need, or that we still want to switch gains in the amplifier chain?
If the dynamic range of the ADC is enough (about 105dB), and around 2V is expected for the biggest energy we expect, then scaling the X-axis can be done in software. This can be anything from setting the X-axis range, or set to portions of interest, to having log scaling, octave or decade. I see (in videos) this is readily done in MCA displays of XRF videos on data already collected.

I can imagine one might want to "zoom in", perhaps collecting counts from a range of lower energy returns, while ignoring pulses from higher peaks known to be there, just letting them clip on the rail, and not get plotted anyway.

In my design, there is the first stage TIA, followed by op-amp stages for post-TIA gain, one of which is a place for an optional 50Hz/60Hz interference filter stage. There is also the final ADC driver which has gain=1. This is where I decide whether there is a software switched variable gain addition.

It's the interface?
In my circuit, the power for the analogue signals section and ADC is isolated entirely, using a cheap low power high frequency switcher with transformer (MAX253), followed by a 2.5V ultra low noise LDO regulator (ADM7160AUJZ2.5R7). There is some passive filtering and noise control as well.

To compete with a photomultiplier tube method, gain of the entire arrangement is enormous, quite hard to keep stable, and tricky over noise, bias, and offsets. Any arrangement to switch gains, or select tapped signal routes, in my case, would need opto-isolated digital control, because the alternative would bypass the attempt to prevent connection of the computer digital 0V to the clean 0VA.

Hence the question. Are we intending having the ability to alter at least some of the gain range by hardware?
Perhaps yes, while prototyping, until we know how much gain we need?
I had thought, given you have a whole signal conditioning board, and some experience with it, you might have had a first-order guess.

It's OK if we just don't know yet. I was just poking around looking at opto-isolated stuff commonly used for things like RS485 and whether or not a ready-made software switchable variable gain amplifier might be good.