Hacker News Reader

Comment by tzs | Hacker News Reader

The parts list includes 9 resistors, 8 capacitors, a photodiode, a dual op amp chip, a switch, and a BNC connector.

The op amp chip is in an 8 pin DIP. The capacitors, the photodiode, and 8 of the 9 resistors are all in packages with leads. The switch and BNC connector can easily have leads soldered to them

Then there is R3, a 40 M resistor that the parts list calls for getting as in SMD. In fact, the parts overview link specifically calls for a black SMD.

As far as I can see, R3 is the only thing that needs a PCB. Everything else would work fine with a solderless breadboard, or any other common non-PCB hobbyist electronics construction method.

Is there some reason R3 has to be black and has to be an SMD resistor for this thing to work? Looking at the schematic it appears to just be part of the feedback loop to control the gain of the op amp in the first stage amplifier, which should work no matter how one decides to get 40 M of resistance.

View on HN

Hey, project owner here. Glad you all enjoy this! :-) I've never built my particle detectors on a breadboard. It is extremely likely that they will oscillate or frustrate in another way. You have to keep in mind that the amplification is enormously large, per particle only a few thousands of charges are being generated and turned into a still tiny voltage. The lower 10M-feedback electron-detector version works if _soldered_ on a prototyping/veroboard. Those with regular hole patterns - if the parts are put really close together.

Please don't be put off by the circuit board requirement, I have listed it on kitspace such that it is really easy and cheap to get one: https://kitspace.org/boards/github.com/ozel/diy_particle_det... Even if you have never ordered a PCB, it should be straight forward using kitspace as a proxy to get it right.

For beginners, I would propose soldering the electron-detector first (on the plus side, it has 4 times more the sensitivity) and if that works swap few parts and upgrade to the alpha-spectrometer variant since that is a bit more tricky to operate and get running. I've commented on the through-hole/SMD choice and similar questions in this twitter thread https://twitter.com/0zelot/status/1260931205676990466.

In short, I choose leaded components over SMD where possible such that it is easy to solder. But analog signal integrity and low noise vs. signal require a circuit board.

Presumably this would be suitable for soldering Dead Bug or no circuit board style - component to component.

even the alpha-spectrometer might work if soldered wisely on a regular prototyping grid. in any case, thoroughly cleaning after soldering with alcohol will help. but considering the really cheap and fast PCB productions nowadays, the hassle is most likely not worth it.

The short answer is that you can put it on a breadboard, but it will require attention be paid to the first op-amp circuit.

The first op-amp is used as a charge amplifier. A high energy charged particle will deposit a charge Q on un-grounded node of the reverse biased diode. The voltage on the output of the amplifier is V_out = -Q/C_feedback + V_bias, where v_bias is made with the voltage divider on the non-inverting input. The resistor R3 is only used to adjust the frequency response (and I assume stability) of the amplifier. It will low pass the signal at ~800 Hz.

The problem and the reason they want you to use small low capacitance components is that in order to get a large charge to voltage conversion factor, C_feedback is small (5 pf). If you build the circuit on a breadboard and place the op-amp output and input on consecutive lanes then parasitic capacitance can easily add another ~10 pF to that and drop your amplification.

To get around this, you'd have to lift up that leg of the capacitor and solder your connections to it instead. This will avoid stray capacitance from lowering your gain.

There seem to be builds using a through hole resistor for R3:

https://github.com/ozel/DIY_particle_detector/blob/master/ha...

The opamp appears to be configured as a transimpediance amplifier. 40M is a lot of gain. Probably keeping traces short and the resistor away from the board helps. That’s probably why they mount the resistor vertically when they use a through hole in the pic above (my guess anyway). Using a SMD keeps the traces short, which might be why they use it in the current version.

Another option could be just to air wire the resistor. Sometimes people bend up the legs on the opamp and wire the resistor/input directly to the IC pin.

I think you might have issues on a breadboard. Air wiring the whole thing would probably work. Or using protoboard, trying to keep the traces short.

> There seem to be builds using a through hole resistor for R3

That's the "electron-detector" version, which uses a 10M feedback resistor. The alpha-spectrometer version is the one that uses the 40M SMD resistor.

I don’t think that makes much of a difference to the part selection (it’s the same board). If you look at the schematics they say you can use either through hole or SMD. SMD is preferred:

https://github.com/ozel/DIY_particle_detector/blob/master/ha...

Reply on news.ycombinator.com