Do I need a scope?

[ksgeek]

Hi. Please excuse the total noob question...

I've got a security background (things like network and web application defense/hacking), but I'm fairly new to the world of hardware. I have a basic understanding of PCBs, components and ICs, and have a tiny bit of experience with logic analyzers. I'd like to set up a lab and I'm wondering if I'll be able to get by without a scope for a while.

Aside from normal soldering setup, I have a decent low-end multimeter and I'm planning to buy a Saleae 8-channel logic analyzer (not the pro version) and a universal device programmer (recommendations welcome--looking in ~$50-$75). I'm mainly interested in snarfing firmware and analyzing communications between components on embedded devices, and just learning more about electronics in general.

My question is, what types of things might I need a scope for, that I wouldn't be able to do with the things I described above? (Remember, I'm a newb, so the more you can explain it like I'm five, the better...)

Thanks!

[mjkuwp]

you don't need it until you need it.   you'll know.  I have scopes at work and I have one at home but I use the DMM way more often.  I also have a simpler, older and smaller Saleae and that is very nice and I use that fairly often.

I don't know what a "universal device programmer" is.

[Rbastler]

For a universal device programmer the TL866 could be the thing for you. If I understood correctly the meaning of universal device programmer.

[tablatronix]

ok so a scope is probably like a good network analyzer or wireshark, whereas without those you can get still the job done with manual labor, lots of small discreet inspections using cmd line utilities, and log file inspection, and diffing and hex viewing etc., but eventually you move past that and decide time is better spent learning and setting up better tools that do the same task in a few minutes.

personally I think you need to do one before the other, struggle is experience blah blah, but if you a have workbench and space, might as well goto a swap meet and see if you can pick up a $50 scope. Then decide if you really need a $400-$1000 DSO or MSO after using it for a bit. I am a lifetime noob and I am saving for a DSO because I have no space for any other option, I doubt I will even use it much but, can't build a house until you buy a hammer.

[james_s]

Most of the time you won't need a scope at all, but when you do need one, nothing else will do. The thing about a scope though is that you really have to learn how to use it properly before it will do much for you. It will show you what's there, assuming you are probing correctly, a topic in itself, but then you have to know how to interpret what you're seeing.

If you're a beginner and not sure if you need a scope or what it can do for you then I'd echo the advice above. Look for a cheap scope, I would suggest keeping it under $50 which is going to mean an older analog scope of modest bandwidth. That will give you a feel for what a scope can do and enable you to figure out if you need something better. If you get one that has XY mode you can use it to display "oscilloscope music" and other fun but not very useful stuff when you're not actually using it.

[ksgeek]

Thanks for the responses! You guys are awesome.

Maybe I'm not using the right term re the "universal device progammer." Basically looking for a cheap(ish) all-in-one programmer for eprom, ic , etc. Something where I could read contents of memory, program PIC, etc without having to buy multiple devices (maybe I'm pricing that way too low, but I can't afford to get into the $3-400+ range).

For the scope, are there specific (reasonably common) scenarios where I the Saleae wouldn't work, but a scope would? I've seen a scenario where data send/receive direction was determined by a slight variation in signal voltage (which a scope would see but logic analyzer wouldn't. How common are things like this, and what are some other examples?

[james_s]

No that's a reasonable term, or you could call it an EPROM programmer and everyone will know what that means, though many will do more than EPROMs now. I have a TL866 and it works reasonably well, my only gripe is the Chinglish in the software.

[julian1]

> I'm mainly interested in snarfing firmware and analyzing communications between components on embedded devices, and just learning more about electronics in general.

I found a $30 bus-pirate very good for connecting to random IC components (purchased or pulled from a PCB), and talking to them over SPI, I2C, figuring out protocols from datasheets and making them do stuff.

https://en.wikipedia.org/wiki/Bus_Pirate

In terms of a scope - it's really hard to know what's going on without one. A scope will tell you the basic information. Like is this op-amp oscillating or is there a valid clock signal on this IC pin.

[julian1]

I used the bus pirate for some time in the past. The next step to a more capable tool is the Saleae Logic Analyzer. More memory depth, better software ... you never look back! 
 

Is the Saleae effective as a source of generated/scripted SPI/I2C commands? I've never used one but know they're good for sniffing/debugging an existing protocol.

[daybyter]

This thread reminded me of this story:

https://www.reddit.com/r/TREZOR/comments/31z7hc/extracting_the_private_key_from_a_trezor_with_a/

[daybyter]

Nope...

"I'm mainly interested in snarfing firmware and analyzing communications between components on embedded devices,"

Seems like he needs a 6022be...   

[xani]

The general rule of thumb is if you have to ask if you need to, you probably do not need it.

Better ask "I need to do X and Y, what kind of test equipment I need to do/debug it?"

You might also look into one of those "all-in-one" PC-connected boards like Analog Discovery 2(https://www.digikey.com/en/product-highlight/d/digilent/analog-discovery-2?WT.srch=1) which just have variety of digital and analog inputs and outputs. Apparently there is also option to drive it from Python if you want to do something more complicated

[abraxa]

For the scope, are there specific (reasonably common) scenarios where I the Saleae wouldn't work, but a scope would? I've seen a scenario where data send/receive direction was determined by a slight variation in signal voltage (which a scope would see but logic analyzer wouldn't. How common are things like this, and what are some other examples?

There are indeed cases where a scope can discover things a LA can't and your scenario is one of them. However, what you describe is something hardware engineers resort to when there's really no other choice - e.g. for very high bandwidth connections. I don't think you'll encounter those unless you want to probe computers or networking equipment. Come to think of it, you haven't really mentioned what kind of device you'd like to examine. Knowing this would allow us to give you a better recommendation because for certain devices, even a Cypress FX2-based Saleae Logic is too slow.

Is the Saleae effective as a source of generated/scripted SPI/I2C commands?

There are commercial solutions for this kind of device but honestly, picking up an Arduino or an STM32 eval board works just as well (or better) if you have programming experience, which you apparently do. Tons cheaper, too, and it allows you to broaden your skill set since you'll then be dabbling in embedded development.

Also, since no one has mentioned it and you're on a budget, I'll just bring up sigrok. It's an open source software suite that comes with a GUI (PulseView) which is going to be a general-purpose analog/digital signal workbench. For now, it's mostly a logic analyzer and works best with Cypress FX2-based devices that can be had for as little as $10. Might be worth checking out.

[tggzzz]

For the scope, are there specific (reasonably common) scenarios where I the Saleae wouldn't work, but a scope would? I've seen a scenario where data send/receive direction was determined by a slight variation in signal voltage (which a scope would see but logic analyzer wouldn't. How common are things like this, and what are some other examples?

There are indeed cases where a scope can discover things a LA can't and your scenario is one of them. However, what you describe is something hardware engineers resort to when there's really no other choice - e.g. for very high bandwidth connections.

"Signal integrity" is a very important topic. If the receiver misinterprets its analogue signal input, then digital debugging is pointless. Note that a logic analyser's inputs may interpret the signal differently to the real receiver.

N.B. the signal's bandwidth and baud rate are completely irrelevant; the receiver doesn't "care" if the next transition occurs in 1ns, 1s, or 1hour! All that matters is the transition time, and modern jellybean logic can have transition times <1ns.

[abraxa]

tggzzz, I agree with you on all accounts. I was just under the impression that he was talking about signals that intentionally use more than two discrete values to convey meaning, and aside from purely analog values, I've only ever seen such configurations used for high-speed signals. I could be wrong, of course, and there may be low-speed applications for such and I just haven't encountered them yet.

[tggzzz]

tggzzz, I agree with you on all accounts. I was just under the impression that he was talking about signals that intentionally use more than two discrete values to convey meaning, and aside from purely analog values, I've only ever seen such configurations used for high-speed signals. I could be wrong, of course, and there may be low-speed applications for such and I just haven't encountered them yet.

I didn't follow the intention back through several posts, so you may be right in that respect.

Apart from that, communication systems will almost always have multi-level signalling for one of two major reasons:

• in order to get the max capacity out of a channel, there are many different types of modulation many of which involve phase and amplitude modulation (think constellation diagrams)
• noise, random and correlated, will be added to the signal by the channel

Start with 56kb/s modem signals, and work upwards

Good rule of thumb: analogue signals convey energy from here to there; digital signals convey information. You can't convey information without conveying energy

[ebastler]

I'm mainly interested in snarfing firmware and analyzing communications between components on embedded devices, and just learning more about electronics in general.

My question is, what types of things might I need a scope for, that I wouldn't be able to do with the things I described above?

Re-visiting the original question from a slightly different angle: I would argue that, especially as a beginner, there are many situations where you don't technically  need a scope, but it makes exploring electronics much more enjoyable.

Being able to visualize signals makes the inner workings of a circuit accessible and "tangible", which would otherwise be a black box. Starting with the simplest circuits -- say a bi-stable blinking circuit from two transistors -- it allows you to see what's really going on. You might be able to debug, adjust etc. a circuit with more basic tools, e.g. a multimeter. But looking at the output and the internal signals with a scope will teach you more about how the circuit works (or why it does not work  ).

I recall that, as a kid, I never really enjoyed electronics kits and experiment sets. Electronics was too much of a black box; it felt like you essentially had to trust the explanations in the instructions, and be satisfied with observing the external behaviour of your circuit. That only changed when I got access to an oscilloscope (quite a bit later).

[rstofer]

A logic analyzer is a great tool, particularly if it accepts a 'state clock' instead of a random clock generated internally.  But...

The scope will probably get 10x the amount of use of the logic analyzer even if the project is purely digital.  As stated above, if you want to see how things really work, you need a scope.  Can you work around it?  Sure!  But you're missing out on another form of learning - seeing.

Even the simple experiments like the charging of a capacitor are left as hand waving because there is no visible action.  Yes, you might slow things down and watch it with a V-O-M (or, for an ugly solution, a DMM) but it's not the same.  Just take a resistor in series with a capacitor and hit it with a square wave (eg from an Arduino) and watch the rise and fall curve on a scope.  Input to resistor, one side of capacitor to ground.  Probe the junction of the resistor and capacitor relative to ground.

Now, here's something fun:  You formed an integrator in the experiment above, turn it around and create a differentiator.  Run the signal into the capacitor in series with a resistor to ground and hit it with a square wave.  Again, look at the junction of the resistor and capacitor relative to ground.  You will see a positive going pulse at the leading edge of the square wave and a negative going pulse at the falling edge.

Even the most elementary circuits are a lot more fun with a signal source (even an Arduino) and a scope.

That new Siglent SDS 1000X-E 200 MHz 2 Channel scope is looking very attractive at $379.  That kind of scope will cover electronics for a very long time.  It has the bandwidth advantage over the other favorite, the Rigol DS1054Z.  It doesn't have as many channels as the Rigol so there is a tradeoff between bandwidth and channels versus equivalent $.

I don't see where you can go very far in electronics without a scope.  Maybe not day one, but soon...

[Electro Fan]

I'm mainly interested in snarfing firmware and analyzing communications between components on embedded devices, and just learning more about electronics in general.

My question is, what types of things might I need a scope for, that I wouldn't be able to do with the things I described above?

Re-visiting the original question from a slightly different angle: I would argue that, especially as a beginner, there are many situations where you don't technically  need a scope, but it makes exploring electronics much more enjoyable.

Being able to visualize signals makes the inner workings of a circuit accessible and "tangible", which would otherwise be a black box. Starting with the simplest circuits -- say a bi-stable blinking circuit from two transistors -- it allows you to see what's really going on. You might be able to debug, adjust etc. a circuit with more basic tools, e.g. a multimeter. But looking at the output and the internal signals with a scope will teach you more about how the circuit works (or why it does not work  ).

I recall that, as a kid, I never really enjoyed electronics kits and experiment sets. Electronics was too much of a black box; it felt like you essentially had to trust the explanations in the instructions, and be satisfied with observing the external behaviour of your circuit. That only changed when I got access to an oscilloscope (quite a bit later).

+1