Oscilloscopes

[Anonymous No.398350]

I've recently taken to amateur electronics as a hobby and I think I want to get an oscilloscope. I've been thinking about this for two primary reasons: I think it would be an effective learning tool and I think it would give me a better understanding of signal processing, which also rather interests me.

I've taken a look at craigslist in my area and most of the listings are either $100-$200 "not sure if works" purely analog antiques or $9,000 mysteries.

Ideally, I think I'd like something digital instead of analog for the built-in readouts, but is seems there are a lot of options out there. I've also looked up a few USB/audio jack setups you connect to your computer, but a lot of those seemed to have a lot less functional range than the real deal (or warnings like "exceeding 5V may damage your computer hardware".)

Does anyone have any experience with oscilloscopes that could be had for, say, less than $400 and could give some suggestions? caveats? tips? What are the pros/cons of analog vs. digital scopes?

Thanks in advance.

[Anonymous No.398355]

I have a Korean CRT one from the 80s. Its OK for basic stuff. A few weeks back I picked up another one that had been put out for hard rubbish. It did not work but I may be able to fix it. The junk pile one is a BWD845 and I downloaded a full service manual no worries. Have never used a digital one so can't comment there.

There are a lot of oldfags diseased estates being sold off so you never know you luck.

[Anonymous No.398360]

Re the computer idea. I have used a sound card as a storage cro but being designed for audio are AC coupled only and if you want to use it for RF useless. You can voltage divided of course for higher levels.

[Anonymous No.398400]

Years ago I bought an analog 'scope (25MHz) from a pawn shop for $85. It works well enough for amateur electronics projects, I can't really imagine it failing to fill my needs though I would feel more comfortable if it was, say, 100MHz.
I haven't used a digital 'scope in a while but I was kind of put off by the screen resolution. I'm sure they're better now but I still find analog 'scopes more "charming"; getting the trace focus set right, that sort of thing.

[Anonymous No.398612]

Don't you have some good old Metrix ?

You don't find em at every corner but I've seen many under 100€ here. (France)
The same models I used in high school

[Anonymous No.398625]

>>398400

Your 25MHz analog thing quite likely displays something even at 100MHz input. "Something" is often enough to deduce if things are working.
On the other hand, 100MHz scope will still have probe capacitance, which quite often shits on your measurements above 25MHz. Well, unless you have fet probes.

IMO one of the nicest things in digital scopes is that you can use a PC to play with the recorded data. Screenshots are easy to take, too.
Screen resolution tends to be high enough to show scope's own noise, so I haven't seen it as a problem. More pixels looks better, though.

[Anonymous No.398643]

OP here, thanks for the replies.

>>398400
What are the sort of situations where having that additional bandwidth is useful? Besides RF. I'm correct in thinking that the rated frequency is what signal frequencies it's able to detect (theoretically) and not something like a sample rate?

>>398612
Looking up Metrix locally either comes up with nothing or stuff that's $600+; a bit too steep for me.


I actually got looking a little closer, and it turns out the one in my original pic seems to be in working condition (pic related.)

It's 20Mhz, comes with 2 new traces and seller is asking $100. Is it worth a shot?

[Anonymous No.398654]

Unless you can find a scope locally that's in good working order, I'd just get a Rigol.

Analog scopes are a bit harder to use, but the old hands swear by their ability to pick up "rare" excursions, like noise that only comes once in a while. Not likely to be an issue for you.

The rated frequency is the analog bandwidth. For a digital scope, the sampling rate is generally much higher so you don't get sampling artifacts.

[Anonymous No.398655]

>>398643

If it's a digital scope, manufacturer tells you both the maximum sample rate and the analog bandwidth. If only one frequency is given and the seller isn't trying to scam you, it's the analog bandwidth.

The required bandwidth depends on your needs. 25MHz limit is easy to reach if you want your logic signals to look logic signals. For example, 10MHz square wave has strong 30MHz (and to limited extent 50MHz, 70MHz and so on) content and will look rounded on 25MHz scope.
10MHz is quite normal clock for cheap 8b microcontrollers.

[398655 No.398666]

>>398655

Oh, and this wasn't meant to say that you can't use a 25MHz scope to look at 10MHz square wave. It just won't look perfect, that's all.

[Anonymous No.398693]

more bandwidth = more resolution

The cheapest digital Rigol DS1052E are about 300 bucks.

I found a 10Mhz two channel Japanese thing with a manual
on my Craigslist. Ideally, I would have gotten a Tektronix 50Mhz, or 100Mhz model, but Ebay is as sketchy as fuck for these machines.

[Anonymous No.398701]

>>398350

decades of experience with scopes, professionally, back to the 1970's.

DO NOT buy any of the older Tek or HP scopes without thoroughly operating each and every single knob and dial. they were/are exceedingly high quality instruments; the problem is that the switches used to generate all the functions become intermittent, and are basically unrepairable. NOT standard parts.

more or less: a 'scope with a CRT is OLD.

DSOs are the way to go. There are now decent chinese-made digital scopes for < $500. Sparkfun sells one or two of them for example.

for general shop use, audio, digital stuff: two channels, 60 - 100MHz bandwidth, a real trigger (most do these days) external trigger input. color etc is not really necessary.

USB scopes are not much good for anything beyond audio frequencies --- last i looked, and this could change.

the hard part of DSOs is the gigasample/sec DtoAs, and storage for that data.

If you want to see, reasonably accurately, the rise and fall if one mS pulses from an arduino, say, you want 100MHz.

I own a Tek TDS1012, bougt it new in 1997. i lurv it.

[Anonymous No.398706]

>>398625
> probe capacitance

this is true. to actually see stuff less than... 100nS? you'll need to know about capacitive loading and all that rot, and ways to deal with it.

>>398654

this is true -- the first scopes i used professionally were the old Tek 500 series, the boat anchors. sloooow! 10MHz BW, some of them. the best of that era in many ways was the 465, 100MHz, analog, and great trigger system. YES you could see shit on that scope that no way my little DSO will do without great trickery. it's because the analog scopes are... electrical analogies, and the phosphor "remembers" a glitch that the DSOs need to capture, simulate, display. that layer of abstraction fucks you up.

still, those were techniques for dealign with gear that was incapable of capture and analysis that even my lowly TDS1012 is a no-brainer at. It's not really much of a "loss", it's all gain.

shit that was "high speed" in 1980 is slower than what's in toys, today. here, newer is better.

[Anonymous No.398723]

Got a 100MHz Gould scope for free from university. If you're a student keep your eyes open, they often give old stuff away.

[Anonymous No.399059]

I've done a bit more research and I think I'm leaning toward the owon sds7102.

It seems the firmware was a little sketchy when they first came out but has gone through a number of revisions since then (mid 2011)

The first google result for it (at least for me) returns a 60+ page blog/forum post about the model.

Also
>that big screen

Anyone worked with one of these personally?

How annoying were the hard/crappy buttons or the "slipperiness" of the base?

Anyway, thanks again for all the help.

[Radiofag No.399081]

The Tektronix 453 is one of the greatest oscilloscopes ever made.

It is easily calibrated and maintained by an amateur.

[Anonymous No.399106]

>>399081
Don't you generally need test equipment to calibrate test equipment?

[Radiofag No.399107]

>>399106
With a modern digital multimeter and your computer's soundcard to act as a signal generator, you can accomplish a lot.

[Anonymous No.399108]

>>399107
Ah, makes sense.

[Anonymous No.399142]

Try eBay rather than craigslist. You don't need to be limited to your local market with something as small and easy to ship as a scope. I remember seeing some nice used Tektronix digital ones for $150-250.

[Anonymous No.399370]

>>399106
>>399107
You can only calibrate to the accuracy of your reference. Cheap DMMs and on-board soundcards are not very good references.

[Anonymous No.399385]

most scope work doesn't really need super accurate calibration, in my experience.

[Anonymous No.399390]

>>398350
I think I paid $200 for a 20-year-old analog Tektronix maybe 15 years ago. It is 30 Mhz I think. It is the same ones I used in electronics class in high school (or very close to them).

There is a place in St Louis (USA) that deals with this type of electronic device repair/resale, as well as test equipment calibration. They can repair them and the guy tested the stuff he had for sale.

--------

The little digital oscilloscope-cards that you hook up to your PC are generally pretty crappy. They sound too good to be true, and they are.

There are full-digital oscilloscopes of course, but they cost $$$$$. $30,000 is not out of line for an upper-end one.

Also if you want to work with vacuum tube circuits a lot, I was told that you would do better with an all-tube oscilloscope and all-tube-voltmeter. Because of the internal impedance they can measure vacuum-tube voltage ranges directly rather than using resistance, so they end up doing a bit better than typical devices that are transistor/resistance based.

[Anonymous No.399396]

>>399390
Shoutout St. Lou!

What kind of practical tasks (specifically) could would a function generator help with? Like say you want to build a "x," how would you utilize the function generator?

[Anonymous No.399401]

>>399370
>>399385

A typical scope Y amplifier isn't that accurate to begin with. For example, my relatively new scope has a specified Y accuracy of 1.5%. Even quite cheap DMMs are much better than that.

I wouldn't trust on sound card's amplitude accuracy, but there's no reason to believe the timing accuracy is shit. Depending on the reconstruction filter, it might be a good idea to use calibration frequencies which are generated by integer divisors (11025Hz, for example), though.

[Anonymous No.399408]

>>399401
Audio codec clocks drift quite a bit, in particular with temperature. If you're doing audio work you'll know the importance of synchronizing to a common master clock.

[Anonymous No.399535]

Watch this video,
He recommends a rigol ds1052e

youtu.be/R_PbjbRaO2E

[Anonymous No.399542]

>>399396
>What kind of practical tasks (specifically) could would a function generator help with?
I got the oscilloscope AND a function generator because I had an interest in building a tube audio amp myself (something I still haven't gotten around to doing :\ )
For audio amps, you need a clean signal generator and an o-scope to see what the amp is doing.

Computer soundcards are often not clean--the 'sine' waves they put out are notchy and not smooth, like from an actual analog source. Not in a way you can hear, but it can be seen easily on the right equipment. For testing audio circuits that different matters.

[Anonymous No.399546]

I bought an tektronix 2230 100MHz digital storage scope for 100 euros and I swear by it.

a quick look on ebay shows some nice tektronix scopes for sale actually, so I'd definitely recommend checking that out

[Anonymous No.399549]

>>399535
>
I got this, its pretty okay for the money!

[Anonymous No.399550]

>>399549

Did you do the 100mhz mod

[Anonymous No.399551]

http://www.youtube.com/watch?v=g89oAZ9Hj_8

[Anonymous No.399553]

>>399550
No, I got no need for it, I work mostly below 1 MHz, I do Tesla coils and stuff

[Anonymous No.399556]

>>399551

dat mitchglob

[Anonymous No.399564]

>>399550
I don't think you can mod new DS1052Es anymore (or at least it's more difficult), Rigol fixed the hole last year. And while they're still solid scopes, they're not really the best bang for the buck in that price class anymore.

[Anonymous No.399575]

>>399408

Typical analog scope's timebase accuracy (and nonlinearity) is measured in percents. Even shitty sound card's clock is much better than that.

[Anonymous No.399590]

>>399575

true, but anyone who has used a scope knows that other things are more important.

a scope gives you a picture of what is going on.

if you haven't used one, i excuse your lack of knowledge.

[Anonymous No.399608]

Calibrator fag here....

Test equipment is a must same for temp and humidity requirements. But I would suppose in the end it really would be up to you the user to determine if you really need your o-scope to be professionally calibrated.

Pic related Fluke's 5820 scope calibrator.

[Anonymous No.399678]

>>399608
>$40,000 used

the fuck?

[Anonymous No.399697]

>>399678
Welcome to the big leagues.

[Anonymous No.399718]

>>399697

pls explain

[Anonymous No.399736]

>>399697

bs m8 even if i trash my oscilloscope every time it needed calibrating and got a new one it wud still be cheaper

[Anonymous No.399739]

>>399736
Do you have a 2GHz scope?

[Anonymous No.399743]

>>399739

y would I have a 2ghz scope

[Anonymous No.399747]

>>399743
Because the 5820 calibrates up to 2 GHz, that's why it's so expensive.

[Anonymous No.399803]

>>399747

i can buy 40,000 klondike bars with that money

wat do you need 2ghz for