Beginners guide to building a hardware hacking lab | Hacker News Reader

Beginners guide to building a hardware hacking lab

voidstarsec.com

239 pointswrongbaud111 comments

I learned to solder while building a keyboard and flashing IOT devices earlier this year, and can't recommend Pine64's Pinecil Soldering Iron highly enough (just be sure to pick up a capable power-source while you're there).

> The Pinecil is a smart mini portable soldering iron with a 32-bit RISC-V SoC featuring a sleek design, auto standby and it heats up to an operating temperature in just 12 seconds when paired with a sufficiently powerful power supply unit. > -- https://www.pine64.org/pinecil/

Why do I need a soldering iron with processing power?

To let it turn off when you're not using it. It uses an inertial sensor.

That doesn't explain it, though. You don't need a microcontroller for that. I suspect the real answer is "to minimize parts count".

Is there a fixed-function circuit that can read a digital value out of a sensor and compare it with a threshold?

I'm not dounting it's possibly, but I'd be surprised if it's available without involving a Turing-complete computer.

Yes, there are tons of these sorts of circuits. It's something that is often needed, and these sorts of circuits have been used before ultra-cheap microcontrollers were a thing. A turing-complete computer is serious overkill for this sort of application.

I'm not sure if there is a single-chip solution (although I'd be a bit surprised if there isn't), but the fundamental circuit isn't that ambitious. Using a very low-end microcontroller is likely less expensive, though.

A PID for keeping temperature.

Your regular soldering station will also contain a microcontroller, same as a shocking amount of other tools. They can be incredibly cheap (a couple cents if you have enough volume), so even if you could do it with analog circuits they can be a reasonable solution. And of course this soldering iron comes with buttons and a display, so of course you use a microcontroller.

But I guess what you are getting at is that normally you don't brag about the architecture of your microcontroller. As you correctly suspect, that's done just for coolness points. Chances are an 8 bit AVR or an 16 bit PIC could have done the same thing, but on a soldering iron you don't care about the power draw of the microcontroller, and a 32bit RISC-V fits Pine64's brand better. It's the equivalent of "aircraft-grade aluminum" or "military-strength encryption".

Also the choice makes it easier to run Doom on it.

Regular soldering stations don't contain microcontrollers, they use a thermocouple to decide if the circuit should flow through the iron or not and some other smaller stuff to smooth out the temp gain / ensure the overshoot doesn't get too wild.

You don't. I can recommend my daily driver, a naturally aspirated charcoal furnace and several large blocks of copper.

Pinecil is surprisingly excellent. I bought it on a lark and was shocked at how fast it heats up and cools down. I'm talking a few seconds to get to temp. I don't do a lot of serious soldering but but I do I rarely bust out the Weller now that I have this little thing. I happily use it while plugged into my laptop charger.

I had basically the same experience. Got a Pinecil and built a keyboard. Love the thing, and would highly recommend it. Has also been great for modding and repairing Game Boys, Xbox 360s, GameCubes, etc. The default tip on the Pinecil v1 is my most-used, and very capable, but I've also gotten a lot of use from the finer tip cone as well as the chisel tip. I got the v2 iron as well, but still mostly use my v1. Still works great, and I guess I built up some trust with it that makes me reach for it first.

The microscope recommendations are a bit problematic, except for the mantis. When you can afford it, get yourself a binocular stereo microscope with sufficient viewing distance from the work piece. In contrast to USB microscopes, those will have no image delay, which can make hand-eye coordination difficult, allow for depth perception and have way superior image quality.

When buying a trinocular it should be a simul-focal one where the stereo microscope and the camera are focused and available all at the same time (no switching of the light path to the camera). Example for such a trinocular microscope: AmScope SM-4NTP. A bincular microscope with matching specs will be slightly cheaper.

Louis Rossmann explaining in more detail: https://www.youtube.com/watch?v=C_eQrbop-J4

That is definitely the segment that I have the least hand-on experience with. I've always used the mantis in my shop and not used much else for soldering. Feel free to toss on a PR with some reccomendations/updates!

A stereo microscope is a prominent fixture on my bench. It's on an arm so I can swing it into place. Can't recommend a brand, since mine was a dumpster dive.

I also like having a magnifying visor, so I don't have to move something to the scope. McMaster-Carr has a selection.

I have a stereo Amscope that I really love.

When you get up to buying meters, the main advantage of Flukes is that you can accidentally put 500 volts into them, and they survive it. When soldering, it's important to wear Jeans, or other heavy non-plastic full length pants.

I've never had solder come anywhere near my clothes when soldering electronics.

Yet :-)

I’ve seen more than once people tinning their solder irons by add some solder and flicking the solder onto the floor hard.

I did that for many years, and even hit my bare legs or feet a few times. But the solder cools really fast, and doing that doesn't cause a great deal of pain or damage. Just enough to remind you to be more careful.

These days, I flick the solder onto my soldering mat instead.

I’ve fried my 3D printer board once because of using a cheap meter. My good meter was out of reach so I grabbed a closer dirt cheap one to measure the output of the 24V bed. Instantly shorted my board and welded the leads to the screw. Long story short, don’t cheap out on multimeters.

What exactly happened there, and why do you think it wouldn't happen with a more expensive meter?

My fluke has a fuse on the ammeter.

Yes, unfused 10A range would have been my guess, but if device under test doesn't have short-circuit protection, it might as well die during the few tens of milliseconds that the fuse needs to heat up and blow, as well as a user error to be measuring voltage with switch in current measurement position.

I'm a beginner here, but I want to assume that if it was enough to weld it, the cheap meter wasn't the problem.

Sounds like you tried measuring voltage while your meter was set to measure amperage

Counterpoint: The IC at the heart of basically every cheap meter is a clone of the IC in an old Fluke DMM

https://www.pa3fwm.nl/technotes/tn27a-voltmeter.html

Honestly I would say cheap out on the DMM and splurge for a good oscilloscope, you get much farther with that.

But, I'm someone who doesn't do high voltage much, or really at all.

I would suspect the clamp diodes / protections would be external to the IC?

Right, I agree that if you're doing HV or need high accuracy, don't cheap on the DMM, otherwise I do think cheap DMMs are good enough for a lot of things and that being able to see the waveform is often more useful than a perfect static measurement

90% of the time when I'm using a multimeter, I'm just doing continuity checks anyway.

Yes, and while we're talking about probing, my favorite DMMs by far are the pen shaped DMMs because one hand holds both the DMM and the positive probe; small, no flat surface needed and always easy to read the screen. No more putting the DMM down on something and holding the probes on while craning my neck to read the DMM, pick up the DMM to move to another location, repeat, so tiring - I tried a pen DMM a few years ago on a whim and I'm always going to use a pen DMM now on.

I never tried the pen version. I actually have a very nice bench DMM that I picked up used for a good price. It's a bit of a "pearls before swine" piece of equipment in my hands, though. I've literally never needed the precision it has, or have done 4 wire measurements, or recorded a series of measurement over time, etc. But I enjoy knowing that I can if I ever need to.

That DMM gives me a similar convenience to what you describe, but I am almost never using a multimeter anywhere but on my bench -- that probably matters a lot.

so much this. Also, (1) if you are laying with 500V you should be careful. (2) cheap DMM will typically also survive (ask me how I know) (3) You can destroy 10 cheap DMM before is more rentable a durable Fluke

Something I've learned by looking at "clones" of my own products is that it's not a copy unless it copies my component sources, manufacturing, and quality standards.

It's also possible to cheap out on the meter while still getting a quality product. I like to look here for good reviews/tests of multimeters: https://lygte-info.dk/info/DMMReviews.html

I am a fan of the grey market Fluke 17B+. You get many of the features of the high end Flukes like the 87V without the high cost. They are my go to for non critical lab measurements.

You can put that into other quality meters as long as their CAT rating is sufficient and was verified by a reputable testing lab (Underwriter Labs, TÜV, etc.). Fluke isn't the only player in town with quality meters.

Also don't skimp on the meter, buy something decent. There is a lot to be said about the real life and noticeable differences in cheap and quality meters when testing or repairing electronics but in the realm of embedded device hacking I particularly found the following features handy:

- A high speed voltage bar at the bottom of the instrument which tells you if a voltage is indeed stable. You can detect the TX pin of serial lines with that. Cheap meters may have that bar but the update speed is way too slow to be useful

- A high speed and loud continuity tester. Many cheap meters (and some expensive as well) have a large delay between shorting out the leads and the buzzer going off. That's rather annoying when you can't make good contact long enough (also see below for test leads).

After working with a cheap meter for a while (Uni-T) I went for a Brymen BM869s and never looked back.

Also get yourself some high quality pointy, hard gold coated test leads. I personally can wholeheartedly recommend the Probemaster 8000 series.

Surprised not to see a recommendation for the Rigol DS1054 series, which is another good beginner scope and generally slightly cheaper than the Sigilent.

I can definitely add this, it's by no means a definitive list, and I'm hoping to get more insight to add to it over the years

And upgradeable, in more ways than one. :)

They just came out with a new model: Rigol DHO800/900

Dave's review: https://www.youtube.com/watch?v=S8jrpCoZyx8

Yes but can you trivially hack it to enable all the software locked features?

Yes, you can: https://www.eevblog.com/forum/testgear/hacking-the-rigol-dho...

That's amazing. Stay terrible rigol.

obsoleted by DHO800

As an aside, I'm not paid or associated or anything, but I did the hardware hacking class with this company and it was pretty good. If you're looking to get into the space it's not a bad place to start.

Its a good list, the only thing I would change is that I like a deeper workbench (30" is good 32" deep is even better) because I'll put a shelf in the back so that the test equipment can be vertical from the surface (and yet the probes can still reach the surface easily). For RF work, I've got a spectrum analyzer, vector signal generator, and an oscilloscope that are pretty much always there, and a network analyzer that I bring out when I'm trying to get debug RF networks.

And while modern (expensive) equipment is light and not so deep, if you're buying stuff used or surplus you'll end up with older gear that can be 18 or 19" deep and on a 24" deep bench that gets a bit crowded.

I finally shelled out for the Proline 32 inch deep benches and man do I love them. The overhead shelf and light are great, as is the French cleared rail.

This will be one of the signs that I've won the lottery, all my benches will be from Proline :-). The radio lab at work had them and I was quite impressed both with the fit and finish and the flexibility.

> ESD Protections

> The last thing that you want to happen is for you to accidentally destroy a device with static electricity, In order to avoid this, it is always a good idea to get an ESD wrist strap or an ESD protective mat.

Work is a different story, and maybe it's because I've only ever done electronics work on the relatively humid east coast, but I've personally never have taken special ESD precautions during my hobby time (such as grounding my desk at home or wearing a strap), yet have never discharged static electricity through any one of my personal circuits.

Am I doing it wrong?

>Am I doing it wrong?

If you're not killing anything, then probably not. It's not a bad thing to be conscientious of, but if static just isn't a problem for you and you aren't handling super sensitive components, stressing about it isn't terribly productive.

Only if you want to sell it.

Stuff has gotten a lot more robust, except for specialized things such as laser diodes. Also, once an IC is soldered onto a circuit board, it's not invincible, but less prone to damage because the capacitance on the board limits the voltage spike.

The only time I ever damaged IC's was when I had my office chair on one of those plastic mats that lets you roll your chair around on a carpeted floor. That was a static electricity generator.

If you're regularly getting static zaps during dry winter weather, on things like doorknobs, it's a sign that you need to be more careful with your electronics.

> Stuff has gotten a lot more robust

It has, yes.

But even back when things weren't as robust, I never got into the habit of using a wrist strap. Just remembering to touch something grounded before poking at things seems to have done the trick. I don't think I ever fried a device through ESD.

I sometimes ground my chair through a 1-meg resistor when I'm feeling zaps during dry winter months. Also, run some copper tape along front edge of workbench, through 1-meg to ground, as a place to conveniently touch.

My bout with damaged chips was when I was programming early EPROM microcontrollers, and a zapped chip bore a strong resemblance to a firmware bug, so it was a frustrating time until I discovered the actual cause.

Nah. Modern chips generally have great ESD protections. If you aren't creating a spark, it's probably not gonna kill it.

The only time I need ESD straps is if I'm working with genuine old school metal gate 4000 series CMOS chips or some whacky, super sensitive input that ESD circuitry is omitted from because the leakage will mess it up (high end RF and ADC spring to mind).

> yet have never discharged static electricity through any one of my personal circuits

How would you know unless you took precautions? Static electricity doesn't necessarily make itself obvious by arcing, yet the voltage can be high enough to damage sensitive components.

Also damage sometimes isn't obvious in the sense that that a component or circuit is killed outright. Instability, reduced lifetime and increased susceptibility to future ESD events are other known consequences of electrostatic discharges.

For more information: https://www.youtube.com/watch?v=JkECJlA2Fg4

Excellent! I could quibble over details of this or that brand but that would be silly. The only category I would add is a ventilation system, especially if you have a 3D printer but even soldering and de-soldering can generate a lot of noxious fumes. Of course it depends on the environment your lab is in but assuming it's a basement or some closed in area, having a good exhaust system can be a godsend.

Great point! I will add a small paragraph on fume extraction this week

As a beginner, I've been a big fan of the Digilent Analog Discovery. This guide lists it under "logic analyzers", but it's also a passable oscilloscope if you don't need more than 30MHz bandwidth. Because the target market is undergraduates learning electrical engineering, it's designed to be inexpensive, doesn't take up much space, is easy to get started with, etc.

As a fellow beginner, I have always been a fan of Salee https://www.saleae.com/] stuff, for much the same reasons you describe.

The problem with Saleae IMO is that their current price point is way out of reach for beginners. The software is wonderful (as is the hardware), but it's hard to justify as a newcomer to the hobby at that price point

Eek. I have not looked at their prices in too long! You are absolutely right. My memory suggested Saleae based on prices like this: https://www.adafruit.com/product/378 ($USD 149. Very clearly, no longer the case!

Appreciate the pointing out the staleness.

I think Saleae's made way in the low-end market to Chinese Saleae clones that have similar software and hardware to their original models – at similar prices, too.

Lots of talk about soldering irons but none on lead. If you do use lead solder please treat the area as contaminated, don't touch your face, and wash your hands afterwards. Lead poisoning is a quiet and long term thing. A little bit of care upfront will prevent a sad end for your loved ones in decades.

The earliest lead-free solders were definitely not beginner-friendly. Things have improved. I prefer an alloy called Kester K100, which is lower melting and more solder-like. Others might chime in on alloys that work for beginners. Also worth considering are solder pastes that you can apply to the joint and then heat up with an iron. And flux pen.

solder paste is awesome. At a FabLab I worked at we got pretty proficient at grinding 2-layer circuit boards with a CNC machine (1/64" bit), cutting solder masks with a craft vinyl sticker cutter, applying solder paste over the mask and then baking in a toaster oven. Great process for making 10 of something.

I still haven't found a lead-free solder that I am willing to use on a hobbyist level. I'm sticking with the leaded variety. People get paranoid about the lead, but it is a very manageable risk.

Never verified but I was always told the fumes from lead free are more noxious than the leaded counterparts, maybe because of the higher temp requirements. Should be using a fume extractor regardless but I’ve always been more conscious of it when working with lead free.

Fumes from leaded solder, rosin, flux, are on the order of 100x more harmful than lead free

Can you provide citation? Leadfree also requires rosin and flux, no?

Is the lead even in the fumes? The temp it evaporates at is much higher.

I understand "lead-free" fumes are actually worse, but you should avoid all fumes regardless.

There isn't lead in the fumes. The risk of lead exposure is all about touching the solder.

Yup. That's what I understand too.

Similar to lead paint, lead exposure is from lead dust primarily. Soldering produces lead dust, and it is relatively difficult to prevent further contamination

Lead free rosin is less dangerous than leaded rosin, check the SDS data sheets

>I understand "lead-free" fumes are actually worse, but you should avoid all fumes regardless.

Read the SDS for any rosin available leaded and lead free. Lead free will have P260, leaded will have P261, a significantly higher risk. And P302 vs P301, and other additional risks.

I agree with the conclusion: always use a fume extractor with flux regardless of the type. They're a cheap insurance policy. For what you pay to insure your car for two weeks you could insure your lungs for all soldering.

"you can wash your hands but you can't wash your lungs"

I've also heard the folk wisdom that the leadfree fumes are worse for you, targeted instead at automated high-volume RoHS compliant factories than the hobbyist hunched over his protoboard but wouldn't know how to find out if one if more harmful than another.

We have The Woz on video with solder in his mouth and he is pretty old. This is just fear mongering BS.

Lead solder didn't bother me at first (I had read all the arguments about vapor pressure, solubility of lead, people's lived experiences, etc. Plus, my initial experiences with lead-free were terrible). What convinced me was when I realized how much fine solder dust I was making every time I cleaned my soldering iron tip. More than once I've had the dust from the brass sponge spill out onto my desk and leave a big grey mark.

I've been using some low-temp lead-free solder from ChipQuik in both paste and wire form, and it works perfectly fine for prototyping. It doesn't seem as strong as other solder types, but the low melting point (even lower than leaded solder) means I don't have to stress the chips as much with temperature swings (great for rework), and I can keep my iron cooler, which reduces tip oxidation.

Jim Williams' workshop: https://www.flickr.com/photos/mightyohm/sets/721576294448252...

Ha ha, that looks like mine.

For electronics work a video microscope is the way to go:

https://www.aliexpress.us/item/3256803915366071.html

I have the 180x, though I mostly use it zoomed out. Perhaps some would prefer the 120x for more working distance. Be sure to get a kit with the light.

https://www.aliexpress.com/item/2251832167751376.html

(I'm told the stand can be found on ebay with lower shipping cost).

For oscilloscopes, Check out the new Rigol DHO800/900 series. They're super compact and USB powered, but basically have the 5000 series features and are 12-bit. The lower bandpass ones are fairly inexpensive and they're hackable to higher bandpasses (craftily, instead of eliminating hackability Rigol has made features like the logic analyzer only exist in higher end models).

I'd like to see some inexpensive options for sidechannel analysis, I've tried varrious things with SDRs monitoring shunt resistors and really had any interesting signals blasted out by LDO noise and USB noise.

Back when I worked on libsecp256k1 I really wanted to setup a CI rig that ran blind side channel analysis on a commit by commit basis, e.g. by looking at the cross correlation of aligned traces with different secret material, but I wasn't able to get something working.

A few notes:

- For soldering stations I think used Hakko's are very worth considering. The recommended KSGER unit is known to have design issues on older iterations such as ground issues. It can be a bit of a mixed bag whereas the high end Hakkos are a very safe bet so if you can find one at a good price I would recommend that over a KSGER-like.

- Fume extractors might be a waste of money from what I hear. I know people who do a ton of soldering that don't use them at all and I took part in a soldering course about a year ago where they said that there are a lot of regulations on what can go into the solder and the flux core so they are pretty harmless under normal circumstances. I use a kitchen fan when cooking on the stove to avoid breathing all the VOCs that get suspended in the air but when I solder I just avoid inhaling when the evaporated flux plume comes.

Fume extractors are nice for not getting the smoke all up in your face at minimum. Any filtering they may do is a bonus.

The CH341A requires a somewhat large warning on it, given there are still many floating around that'll happily push 5V data lines still when supposedly configured for 3.3V

Later revisions supposedly fixed this, but the reviews on the Amazon link seem to indicate it may be of older stock.

This can be fixed with fairly simple modifications luckily.

It's very straightforward to modify to run at 3.3V. There's a few good guides online. It's one cut trace and a short wire tack.

I'd suggest adding a few boxes of common resistors/capacitores/leds and a simple power supply.

I agree. For initial stock, anyway, those cheap large variety packs are awesome. They'll also train people to check values and not assume that the markings on the parts accurately reflect what they are.

> a simple power supply.

Old computer PSUs can often be had for free, and a great first project for someone at a hardware lab would be to build a PSU breakout box to make it safe and convenient to use as a bench supply.

During the pandemic, I bought a Chinese JBC clone that takes real JBC tips. Temperature calibration was possible and it heated up fast enough, but not like a real JBC station. For 15% of the price, it's a steal and close enough.

Also got a 100 Mhz oscilloscope with paid options but hackable to 200 MHz, WiFi (COTS USB), and a 16-channel IIRC logic analyzer.

Picked up an Agilent PSU with a bad fan. That was an easy fix. Will have to calibrate it but it's already pretty close.

Then I spent too much money on a Fluke 289 DMM. The price went up significantly now, I wouldn't recommend it now.

In general, buy used, repair, and recycle.

For RF low cost tooling I suggest Evilcrow RF v2 (https://github.com/joelsernamoreno/EvilCrowRF-V2) instead of Yard Stick One.

This is a related question but how do mere mortals or scrappy hardware startups afford to work on devices with USB4? Outright purchases of USB4 capable scopes are outrageous. Rentals aren't cheap either.

The price on the "MisVision Trinocular Microscope" (it's "MinsVision" on Aliexpress) is a copy and paste error, I guess? I could only find devices like this for around 370 dollars/euros.

Hah, I'm surprised he mentioned Lauterbach at all. It's ruthlessly powerful, but not remotely hacker / user friendly.

Lauterbach is much beloved in many hacker / reversing communities, largely for its power and wide compatibility.

Not an easy thing to start with, I'll certainly grant you that.

What happened to jtagulator? I would have thought this would have been included as a go-to JTAG tool.

I decided to leave that out mainly because it has been discontinued, but I could add it as a reference!

The Glasgow Interface Explorer [0] is starting to ship out now as well. I got mine a couple weeks back. It’s a neat option for a multi-purpose debug tool.

[0] - https://www.crowdsupply.com/1bitsquared/glasgow

For SWEs what is a good way to get an intro into hardware?

I think the best approach is to start taking things apart that interest you, and learn along the way. For example - on my blog I use things like arcade cabinets and home routers to introduce some hardware reversing concepts:

https://wrongbaud.github.io

There is also nothing wrong with getting some of the arduino starter kits on amazon and using those to learn how to interact with various peripherals, etc.

Adafruit tutorials and Neopixels can be fun with a very low barrier to entry. Get into sensors and networking from there. The RP2040 by Raspberry Pi is a great chip to start learning with micropython.

Local hackerspace, just show up.

I second this.

Also, cheap electronics kits can be a great way to get your sea legs, especially if you take the time to work out why the circuits are designed as they are.

Erh, is the scrolling of that website weird to everyone?

I think you don't need all that, but you can start simple with less, like a simple Arduino kit.

Also you can get some Tektronix or similar if you get used. I know I would love some old used HP or Agilent bench multimeters.

I believe it is a bug affecting IOS devices in Git-Wiki. https://github.com/Drassil/git-wiki-theme

Reader mode is saving me on iOS

Is anyone else having a terrible time navigating this on mobile?

I can’t scroll up or down properly at all, the page is weirdly scaled to mobile with text being cut off on both left and right, and there is an odd modal up top that doesn’t work properly as it resizes itself constantly.

Works fine in reader mode.

on chrime mobile: desktop mode is also helping out

Also has an unnecessary 12 megabyte gif halfway down, presumably as a middle finger to people using a data plan.

(oh and another 7mb gif below with extremely poor optimization)

Reply on news.ycombinator.com