Depends what you’re doing. When doing failure analysis I use our Tektronix TDS5054B in conjunction with a DMM (Keithley 2001 or something similar). The scope, obviously, is invaluable when I need to “look” at waveforms and look at rise time, modulation, phase shift, etc. It also has some pretty cool calculation capabilities. But there are some things the DMM does better:
The A/D resolution of any scope is pretty bad, and hence you can’t make precision voltage measurements with it. A DMM has much better A/D resolution and thus much lower measurement uncertainty.
Your standard scope probe has an impedance of 10 MΩ. A modern DMM has an impedance of around 1 GΩ, at least on the lower ranges.
Unless you use a differential probe (expensive) or do a channel subtraction (crappy resolution), a scope is a single-ended measurement. A DMM is always a differential measurement.
Yeah, this is a big one. I’m normally a software dude but I’ve been doing some power analysis stuff at work recently. The Aligent 34970A is a good DMM with 6.5 digits of precision, but it doesn’t have enough time resolution for my purposes (need microsecond samples). So I’m using a Tek TDS3054 'scope instead… but because I’m trying to measure current with a shunt resistor setup, I need differential measurement. Two sets of probes wasn’t accurate enough, so I got a differential probe… which I promptly almost melted because it was only rated at +1v maximum instead of the +12v I needed.
Ultimately got a proper TCP202 current probe and all is happy…
Yea, doing power measurements using a digital scope can be tricky. Have to be careful with common mode voltages. And while a clamp-around current transducer might work, it will give you all kinds of error you’re not aware of. Four reasons: 1) Limited frequency response. 2) Different propagation time vs. your voltage probe. 3) Zero drift. 4) Unless it’s a Hall-Effect type, it can’t measure the DC component of the signal.
So while I’ve used a scope to do what you’re doing, I normally use a dedicated power analyzer when I need to measure power. We have a couple of old Voltech PMiv 4-channel power analyzers on hand. These things are great. Very simply to use (you don’t have to worry about isolation), very good accuracy specs, and they give you a ton of info (real power, apparent power, etc.).
Can you tell me more about this? Also, this is OT, but I prefer messing around at home with my breadboards than going into the labs, just because I don’t have to wear pants and I can make as many pots of coffee as I like without dipping unduly into one of the professor’s private stash of fancy coffee.
What’s your signal generator? And do you use one of the console DMM, like HP makes, or a handheld thing to get a rough estimate?
I was aware of some of those issues but it’s good to see a compact list in one place. The spec sheet for the probe claims 50 MHz bandwidth, so I hope that 1) isn’t an issue. The drift and DC components aren’t a problem for my purposes–I am mainly concerned about relative differences in current. I have a device which, when running a test program, should have a very clear pattern of current draw that repeats every 16.7 ms. In this case, current is a proxy for overall performance, and I see that occasionally, current draw is maybe 15% lower than baseline. This is enough evidence by itself that something is very wrong with the power management software in the stack. The device has onboard sensors, but the time resolution isn’t nearly enough (like 100 ms).
Looks pretty good! 50 kHz sampling rate is a little low for my purposes, but would be sufficient. I might have to mention these units to our lab guys…
I’m cheap, so at home I use a midrange handheld DMM (the one I linked to above) and an old-ass CRT oscilloscope. Also a bunch of shitty $5 DMMs that I don’t trust for much more than continuity and “is there any voltage at all” testing.
At work, I had been using a console DMM (the Aligent I mentioned), but it doesn’t have enough time resolution. So instead I used a Tek LCD oscilloscope and a clamp-on current probe. It is not perfect, as Crafter Man mentioned, but for my purposes the flaws aren’t too relevant.
To the extent that I still do hobby stuff at home, I’m pretty much all-digital. Not much call for a general purpose signal generator, so I don’t have one of those. These days I pretty much play around with Arduino, and external components hooked up via serial or I2C. I’ve gotten pretty good at getting these to work without much trouble, so I don’t even have a logic analyzer of my own. However, I did recently play with one of these:
I’m not the lead, but I’m on the core team. The Logic device worked great for diagnosing all the serial peripherals we have to contend with. Fun stuff.
BTW, I can go into a lot more detail here, but I’m not sure what kind of baseline you’re looking for. For instance, I can give a short overview of shunt measurement vs. clamp-on probes if you’re unfamiliar with the concept. Just trying to get a feel for your starting point.
Well, OK. In my defense I wasn’t trying to hijack the thread – I’m getting a second BS in EE with double major in CS and I’m somewhat frustrated with simulation tools like PSpice. So, let’s say I want to work on economizing the number of flip-flops needed to do something simple like a clock using Common cathode 7-segment display at my apartment, I’m just trying to fiddle around with stuff for fun and to try to multiply my knowledge into designing things that are economical and work in ways I can predict.
I’m not even one year into EE – I just know two terms of digital design and the usual linear course (but I forgot a good bit of linear – KCL and KVL and asynchronous clocks and J-K flipflops is about as much jargon as I can spill. I do know how to operate an o-scope, and all the bog-standard stuff, but that’s just the basics).
So, no, I don;'t have a big project I’m working on (except some pet projects that don’t fit into the regular stuff and are probably beyond me, like designing a mic-preamp that works with a vocoder, and building it all from the ground up including sourcing out low-noise transformers and DAC…you see, I don’t really know what I’m talking about, but I’m just trying to screw around with stuff since I have the basic nuts-and-bolts theory on my own time).
It’s fair to call me a beginner with the average 1st-year labs+theory of linear and digital circuits/design. I’m accomplished enough in my erstwhile discipline in academic research as an expert in phenomenology and technical philosophy to know when I don’t know even half enough to have the correct questions.
As I mentioned above, keep in mind that a current probe, especially one with active electronics, *will *have a different propagation time vs. your voltage (scope) probe. And it will be frequency-dependent. This could cause all kinds of grief, because (ideally) your RMS calculations assume the instantaneous voltage and current signals are sampled simultaneously. A power analyzer takes care of this issue. I would buy a couple used ones.
I have a bunch of scopes around here. Every time our company moved they threw out old scopes and I picked a few of them out of the trash. You can get used tektronix scopes fairly cheaply, and they are well built and work well, and they do seem to last forever.
The only scope I have that I bought new was a DSO Quad. It’s tiny and quirky, but it actually works fairly well once you figure out how it works. The user interface is a bit unintuitive (like I said, it’s quirky). Not sure I’d recommend it because the USB interface went a bit wonky on it. Now when I try to save pictures I see them in the directory but I can’t download them to my laptop. and I can’t delete the images that are saved on it. But for a cheap storage scope with decent bandwidth and its really cheap and small and did I mention it’s cheap? It’s also so small it fits in your shirt pocket. I still use it quite a bit even though the USB is futzed on it.
BTW, BSEE with 24 years experience here. If you ever get stuck or want advice or just want to bounce ideas around there’s a bunch of us around here. Feel free to ask questions.
Didn’t see the additional posts until now. The DSO Quad looks kind of neat – I guess when you consider the price of your average textbook, $200 or so doesn’t start to seem so bad! Hope I didn’t hijack the OPs thread too much – but I didn’t want to not respond with my thanks to everyone for having so much to say about an interesting topic.