choosing an oscilloscope

[Marks256]

What are some good oscilloscopes? I am thinking of buing one (ebay, of course). I don't think i will need anything over 100mhz. More is better, though.

 

[Marks256]

Ok, i guess 100MHz is quite a bit...

 

[Speakerguy]

I really like the Fluke 199C we have at work. Portability is nice. The small Tektronix are ok too, but I always find myself using the Fluke more often.

 

[Marks256]

Umm... maybe i should have mentioned my price range is a maximum of $150...

 

[Andy1845c]

I just got a Tektronix 2212 off ebay. Its only has a 60mhz bandwidth, but for me, right now, its fine. I can't remember now what I paid, but it was around 150, with 3 probes and manual.

About a month after I bought it, my dad came into a 100mhz HP scope for free figures, lol

What I would like to know, and maybe some of you can tell me, is what is the advantage of 100mhz over 60mhz or even 15mhz. What falls in that 60-100mhz range that is commonly measured? 100mhz is mighty fast, and to me is only usful in RF or computer processors. and 100mhz dosn't cover much RF past HF ham bands. And your not even talking 1 mhz there.

 

[kchriste]

Umm... maybe i should have mentioned my price range is a maximum of $150...

The Tecktronix 465 is a very good CRT dual channel scope. Make sure you get working probes (I say working because it is easy to blow a probe with RF/HV) and good scope probes can be pretty expensive on their own. You need an x10 and a x1 probe or a x1/10 combo and should have as many probes as your scope has channels. I don't think you'll get a LCD scope for under $150, but they are nice and compact.

What I would like to know, and maybe some of you can tell me, is what is the advantage of 100mhz over 60mhz or even 15mhz.

Remember that a 10Mhz squarewave contains harmonics up to 100Mhz and beyond. A 10Mhz squarewave will look like a sinewave on a 10Mhz bandwidth scope.

 

[Speakerguy]

If you are doing anything digital with a PIC at frequencies like 20MHz then you will want 100MHz or more. A square wave has as its component frequencies at X (the base frequency), then at 3X, 5X, 7X, etc (the harmonics). It takes several of the harmonics for it to look like a decent square wave. So you'll want 100MHz minimum for say a 20MHz signal.

 

[Nigel Goodwin]

If you are doing anything digital with a PIC at frequencies like 20MHz then you will want 100MHz or more. A square wave has as its component frequencies at X (the base frequency), then at 3X, 5X, 7X, etc (the harmonics). It takes several of the harmonics for it to look like a decent square wave. So you'll want 100MHz minimum for say a 20MHz signal.

This is a HIGHLY pessimistic point of view!, the ONLY thing running at 20MHz on a PIC is the clock crystal, and there's no need whatsoever to see how square it is, it either runs or it doesn't, and the waveshape doesn't matter. I'm not even sure if it's supposed to be a squarewave?.

While a 100MHz+ scope would be nice, it's hardly essential, and is only needed in an incredibly small amount of cases - probably 0.1% or less?, for normal hobby use. A 20MHz scope will do almost everything you ever need, and a 20MHz double beam scope would be FAR, FAR more useful than a 100MHz single beam (assuming you could find a 100MHz single beam?).

But ANY scope is a lot better than no scope, don't get carried away chasing high performance you might never need.

 

[Speakerguy]

*shrug* I'm frequently doing things that require the bandwidth. 200ns nstruction cycles ar 5Mhz base frequency, so I'd want at least 25MHz scope, and I might be looking for things beyond that at times (looking at transition times of switching devices, for example). Or looking at the HF noise spectrum of switchers, or trying to find out why and where in the heck some circuit is oscillating. At my old job they had real scopes - 8GHz Agilent's IIRC. What I'd really like is an MSO with bus decoding ability. We had a tektronix dealer here a few weeks ago demo one and it was awesome.

Just my opinion though. I admit I am an equipment snob with a tendency towards overkill.

And nah, the crystal output is a sinewave with some small harmonics on top. I got curious one day and measured it.

 

[Marks256]

Ok. I am really not sure what i need in a scope, so i am not sure what information to give. I am just going to use it on small digital circuits. Just looking at some waveforms. learn a bit about o-scopes.

So what would be the best bandwidth to get? 25MHz?


EDIT:

What does "double beam" or "dual Channel" mean? Does it mean that it can monitor 2 frequencys at a time? So, instead of just being one line on the display, there could be up to 2?

 

[3v0]

What does "double beam" or "dual Channel" mean? Does it mean that it can monitor 2 frequencys at a time? So, instead of just being one line on the display, there could be up to 2?

Double Beam and two channel mean the same thing. It means you can use two probes and display the waveform for each at the same time. Say you were looking at a ICSP interface, you could have the clock on one channel and the data on another. Many applications.

Do not overlook the HP/Agilent scopes. They are less sought after then tech scopes but rock solid just the same.

Get as fast or as slow a scope as you can afford. But as Nigel said be sure to get a 2 channel unit.

 

[Marks256]

Yep, i see the importance of a 2 channel scope. I am surfin' ebay right now.


EDIT: What is a Digitizing Oscilloscope?

 

[Nigel Goodwin]

Presumably it's a digital scope - this converts the signal with an ADC, and stores it in memory, so you can freeze and store one shot events.

As for the type of scope you need, consider it as you would a car!.

Speakerguy79's point is that you MUST have at least a Porsche 911 Turbo, because you can't do 170 miles per hour otherwise.

My point is you spend more time going to the shops, or sat in traffic, than on the race track!

 

[Marks256]

What kind of o-scope do you have, Nigel? If you don't mind me asking.

 

[Speakerguy]

A basic dual channel 200MHz digital scope is a Z28 Camaro at best, you can get those for 1-1.5k from a number of companies. I'm just one of those buy it once, buy it for life types.

That Tek MSO4000 I saw a few weeks ago starts qualifying as supercar territory. The memory storage and bus decoding was just awesome. It would make debugging I2C systems a joke.

Also, definitely get two channels. That is certainly more important than speed.

 

[BeeBop]

I agree on two channels; phase angle measurement is important, and you can only do that with two channels. I would add that for a first scope, you might want to consider an analog, for two reasons:

They are less in demand = you can get one for less dollars, usually.

They will force you to learn how to use the cursors, rather than just read out the values from the Measure Menu. The Math Menu can also make one lazy. If you can use an analog scope, you will have no problem using a digital, but if you can use a digital, there is no guarantee you will be able to use an analog o-scope. If we are onto the car analogy, then it is like learning a stick shift first, rather than an automatic.

 

[sam2]

There is a good tutorial on ebay about buying an oscilloscope there.
It is writen by a guy who repairs them.
You will have to look around for it, its writen by one of the guys that bids on alot of them.

I bought a tek 465B,100MHz, off ebay and it works great.

I did alot of research before I bought it.

I only looked at ones that were claimed to be in good working order and gave a money back guarantee.

There are some great deals( and some bad ones) on ebay.
sam

 

[kjennejohn]

To answer some questions and clarify a few points:
-"Dual channel" means you can hook up two probes and track two signals at once. The principle application of this is to use an enabling signal, such as /CS (chip select) on a RAM chip, to act as the trigger so you can see the relationship of the data and address lines (among others) to this central signal.
-"Dual beam" means you have two guns painting two traces on the display at the same time. This is prefered for two-signal work. Most lesser two channel 'scopes have one gun, so either display one channel, then the other; or do what's called "chopping". Chopping is where you show one signal's trace for a small period of the total sweep, switch to the other to show a small portion of that, and keep doing this switching throughout the sweep period of one full display. The problem with this is you run the chance of missing some key event because something ends up getting cut out because the chop is looking at the other signal at the time.
This is where modern digital scopes with LCDs shine. Both signal are read and digitized, then both are displayed on the LCD at the same time.
-"Digitizing" simply means the incoming signals are measured on a timed basis, as binary values, then stored in onboard RAM for later (near instant) display. The better scopes allow you to take one reading only (sort of a "freeze frame"), then study it at your leisure or download it to a computer for later, deeper analysis.

The vast majority of the scopes you find on Ebay are servicable analog scopes. These typically have two inputs, 20MHz to 100MHz, but one gun, so supply chopping controls, with all the limitations mentioned above. The early DPOs (Digital Phosphor Oscilloscopes) will give you "freeze frame" ability, but the storage is you and a camera, as the results only stay on screen for a short time. And they aren't cheap, easily more than the $150 you mentioned.

There are pros and cons to analog and digital scopes:
Analog will still do 98.5% of what you want to do in electronics work, IF you are only interested in frequencies, levels and rise/fall times. Which is what most electronic troubleshooting is about. It excels at showing minor discrepancies in a continuous signal because you see the irregular, infrequent events as dimmer signals amongst the strong, regular signals. You can catch ringing and overshoot problems easier, for instance. But they suck at catching one-shot events in the middle of a lot of traffic, or in long periods of no activity at all (seems like?). But, a modern digital scope with extended triggering capabilities will catch these rascals easily.

If you must catch infrequent events reliably, with expanded triggering capabilities, or store traces and download them to a PC, then digital is the only way to go. Even units running less then $400 will do this. I saw an ad in an email recently (search these forums, or Google, using "OWON", as I believe this has been covered here before) of a two channel scope that has 25MHz bandwidth with 100MS/s (megasamples/second, the more the better) with must-have features comparable to my $2400 Tektronix 4-channel color LCD scope. Of course the manual is Chinlish in the worst way (but usable) and isn't very informative about using the USB port, but it can download displays to a PC for printing. It has a color LCD; variable phosphor (nice!); and a limited, but usable, built-in math package. It has a measurement bar on the screen's right that allows you to track five values on-the-fly. One ad even says they throw in a free battery, which suggests portability, but I couldn't find anything about installing the battery in the manual. Beware!
Amazingly, it has more memory per channel than my Tek!

I hope this answered more then it confused.
kenjj

 

[BeeBop]

kenjj,
Thanks, that was really worth reading!

 

[Nigel Goodwin]

What kind of o-scope do you have, Nigel? If you don't mind me asking.

I've got two, a GW 20MHz double beam, and an Hitachi 30MHz double beam.

At work I use a Grundig 50MHz double beam.