Help me chose my first oscilloscope

[NEoX]

Hi! I'm new here on forum!

So, I'm a student of electrical engineering and I want to buy an oscilloscope to learn more about stuff and have in mind some projects to do. At the moment I'm working on project "supervision of pellet furnace" with ESP8266 and automatic refilling. For that project I build my own schematic and board, but it has some problems and I want to debug it.
So I need an oscilloscope, and also in next semester I have a course about oscilloscope and measure.

My budget is 450€ max!

1. I was looking at some models like Rigol ds1054z but i don't know if this scope is best buy for that money in 2018. (I will unlock to 100MHz if still can and all other things like I2C,RS232...)

2. My main interest is digital electronic/microcontrollers/serial comunications (I2C,UART,RS232....)

3. I want to buy a product that I can use for long, long time as a student and to do hobby stuff with it and maybe one day for work

Please suggest me some oscilloscopes that will last.

Thanks!

Sorry if this topic doesn't belong here.

 

[rjenkinsgb]

With that type of budget I'd go for a proper bench scope rather than any kind of USB/PC connected device.

The Siglent ones seem pretty good and you can get one from Siglent in Europe cheaper than they sell for on ebay.
eg. a dual channel 200MHz for 339 Euro.
https://www.siglenteu.com/digital-o...MI7ofkjeTp3AIVRbvtCh20NQWoEAAYASAAEgJ8LvD_BwE

You may be able to find cheaper than that from another European distributor.

Make sure you get some decent x10 scope probes rated for at least the scope input frequency, if it does not come with them.

Proper scope probes are absolutely vital for high frequency signals; in x10 mode they have a resistive divider in the probe itself that isolates the circuit from most capacitive loading.
Without that, waveforms will be affected and things can just stop working while a probe is connected.


Edit - just looked at the Rigol one you mention; it's not bad but personally, I'd go for higher frequency over number of channels - a lot of stuff runs at over 50MHz, even when just working with small MCUs etc.

 

[Nigel Goodwin]

Edit - just looked at the Rigol one you mention; it's not bad but personally, I'd go for higher frequency over number of channels - a lot of stuff runs at over 50MHz, even when just working with small MCUs etc.

I would disagree, it would be rare to require massive speed for micro-controllers, while speeds over 50MHz are common that's only the clock speed - anything you might want to scope approaches anything near that.

I've got a dual channel 50MHz Rigol, and it does everything I need or want, I've never even considered the risk of trying to upgrade it to 100MHz - as there's always a chance you could 'brick' it

 

[rjenkinsgb]

Horses for courses; I do a lot of embedded and industrial stuff and a 200MHz scope comes in handy at times. PICs are up to 200MHz internal clock and I've done PLA-based designs running at 160MHz.

The question related to a future-proof investment as well as current needs, which to me means probably faster with time - that is the major trend.

Also, part of my point was not paying extra for a four-channel over a two-channel.

Once you have a stable, locked display you can have one channel for reference and move the other around multiple points to compare things; you don't generally have to see all signals simultaneously to see what's happening.

 

[Nigel Goodwin]

Once you have a stable, locked display you can have one channel for reference and move the other around multiple points to compare things; you don't generally have to see all signals simultaneously to see what's happening.

No, that's why I have a dual channel scope.

 

[NEoX]

for the ISP, 3 channel come handy.. And Rigol that I mention can be unlocked by software key to 100MHz

 

[JimB]

And Rigol that I mention can be unlocked by software key to 100MHz

A piece of equipment which has been deliberately crippled by a (marketing department inspired?) software patch, does not inspire confidence as a tool to have around for the future.

Also, part of my point was not paying extra for a four-channel over a two-channel.

Once you have a stable, locked display you can have one channel for reference and move the other around multiple points to compare things; you don't generally have to see all signals simultaneously to see what's happening.

Only very rarely have I wished that I had more than two channels on the scope.

Another trick is to set up to use the external trigger on the scope and then the two available channels can be used to display other signals.

JimB

 

[ronsimpson]

I have too much junk........
I have one box that tracks 64 digital signals at one time. (1 or 0 only) It can not measure voltages.
I have many scopes that only measure analog. 2,3,4 channels
I have a scope that does 2 analog and 16 digital channels. (I think 8 is just fine)
The point is that each box does some jobs and not other jobs. The analog + digital box is a compromise.

My slowest scope is 100mhz. That is good for many things. (50mhz might work for you)
I also have a 200mhz and 500mhz.
Scope companies often use the same PCB on many projects. It is some what common to have a 500mhz scope that can remember 100,000 samples. OR 100mhz scope that can remember 500,000 samples. OR a 2/4 channel scope that has 500,000/250,000 samples. They use the same memory board. Memory = Money

Some scopes can do math! I want a scope that can do this: (FFT)

In my case that is more box(s).

in next semester I have a course about oscilloscope and measure.

Go to the measurement lab at school and play with the scopes. Ask questions. Really use them.
---edited---
I used to work for a scope company. Some times they have a educational discount. Ask your school.

 

[DerStrom8]

I have had the 2-channel Rigol (DS1052e) for several years now and it does most of what I need it to do. Granted I don't really deal with RF but it sounds like you probably aren't either.

I actually just acquired a DS1102e (the 100MHz 2-channel version of the Rigol) and am trying to sell it for $275 + shipping (241€ + shipping) but it's used and only has one probe:

https://www.ebay.com/itm/263876336655

 

[RichMarkley]

Rigol sells good scopes, but the major manufacturers are also now selling lower-priced scopes. Take a look at the RTC1000 from Rohde & Schwarz. This starts at €835, and it has professional grade specs, such as a sampling rate up to 2 Gsamples/s, 2 Msample memory depth, a bandwidth up to 300 MHz, and a waveform update rate of 10,000 waveforms per second. You should be able to use this scope for a “long, long time"

 

[gophert]

PICs are up to 200MHz internal clock and I've done PLA-based designs running at 160MHz.

Those would make one hell of a fast Pellet Fueled Furnace Controller.

 

[dknguyen]

For digital communications, a logic analyzer is much more useful than an oscilloscope.

I agree with Nigel. If you're just doing MCUs, 50-60Mhz is enough even if the MCU has a clock speed of 80+ MHz. Rarely do you need to inspect such a high speed signal that carefully in an MCU application. The fastest thing in the system is probably the clock and all you really need to do most of the time is confirm that it's there, not discern it's exact waveform, and most serial communications (certainly all the common ones) run much slower than the clock anyways so you don't need the bandwidth for that either.

It's different if you're doing RF, or building PLLs, but you didn't say that's what you were working on. If you're building SMPS, you *might* need that bandwidth.

That extra bandwidth costs a lot and the money would be better spent on a little logic analyzer or something else to help you with your MCU work. I'd much rather have 4 channels and 60MHz bandwidth than 2 channels and 200MHz bandwidth (or even an imaginary scope with 3 channels and 200MHz bandwidth).

 

[gophert]

For digital communications, a logic analyzer is much more useful than an oscilloscope.

I agree with Nigel. If you're just doing MCUs, 50-60Mhz is enough even if the MCU is 80+ MHz because rarely do you need to inspect such a high speed signal that carefully in an MCU application. That extra bandwidth costs a lot and the money would be better spent on a little logic analyzer or something else to help you with your MCU work.

I'd much rather have 4 channels and 60MHz bandwidth than 2 channels and 200MHz bandwidth.

Note that a PIC running at 160MHz requires four clock cycles per instruction so that is eight instructions to make a port pin go high, then low to make a complete output cycle. The scope would need to handle 160MHz/8 = 20MHz

 

[dknguyen]

Note that a PIC running at 160MHz requires four clock cycles per instruction so that is eight instructions to make a port pin go high, then low to make a complete output cycle. The scope would need to handle 160MHz/8 = 20MHz

Only if you use a scope to look at (rather than something like a logic analyzer)it and only if you actually need to discern it's specific waveshape rather than just the presence of a pulse. In my experience, the vast majority of the time you only need to know that it's there (I assume your example is bitbanging). 60MHz should be enough to discern the presence of a 20MHz pulse train. If your resources are limited, the extra money is better spent elsewhere.

 

[gophert]

l

Only if you use a scope to look at (rather than something like a logic analyzer)it and only if you actually need to discern it's specific waveshape rather than just the presence of a pulse. In my experience, the vast majority of the time you only need to know that it's there (I assume your example is bitbanging). 60MHz should be enough to discern the presence of a 20MHz pulse train. If your resources are limited, the extra money is better spent elsewhere.

Yes, it would need to handle a 20MHz signal, it would need to be something more than 20MHz to do that.

 

[KeepItSimpleStupid]

FWIW:

Interesting views. I have a 3 ch analog Kikisui and a 2ch digital(Tek) I want to fix one of my low bandwidth battery powered CRT scopes. When I was working, the battery powered scope was preferred. process control. I hated the Phillips scope. We did have an old Tek GHz scope. Portability is sometimes important.

If I were doing uP embedded designs I'd be tempted to get a scope that :
1. can do an FFT.
2. Can do with help, real power calculatons. e.g. 3V supply, low freq clock and or real power calcs at power line frequencies. Help means something else to ge currrent.
3. Can do pseudo-differential measurements.
4. Protocol like i2C analysis.

All of those tasks would require compromizes. Its a bunch of incompatable stuff.

 

[dknguyen]

l


Yes, it would need to handle a 20MHz signal, it would need to be something more than 20MHz to do that.

Oh, I see now. You were just trying to point out that if the OP wants to observe a 20MHz square wave he needs something at least 3x higher bandwidth. 3x is just barely enough but sufficient to know that something square-ish is there. Five would be preferable since you can see the 5th harmonic of the square wave, but might be wildly more expensive for his budget.

I also find isolated channels super super important. Or at least, when not stricly needed they are extremely convenient. But that rules out most scopes, even much more expensive ones, unfortunately. I don't know why that isn't standard.

 

[gophert]

Oh, I see now. You were just trying to point out that if the OP wants to observe a 20MHz square wave he needs something at least 3x higher bandwidth. 3x is just barely enough but sufficient to know that something square-ish is there. Five would be preferable since you can see the 5th harmonic of the square wave, but might be wildly more expensive for his budget.

I also find isolated channels super super important. Or at least, when not stricly needed they are extremely convenient. But that rules out most scopes, even much more expensive ones, unfortunately. I don't know why that isn't standard.

Also, coupled with my first post on this thread, that the need for a 160MHz microcontroller is extremely rare and, in the case of the OPs first project, humorous.

PS, I started with a 15MHz analog scope. I upgraded to a 100MHz Hitachi Analog scope (2Ch) because it was being sold for $50 with two NOS probes at a swap meet. My old 15MHz scope is still good enough for most things I do.
.

 

[dknguyen]

Also, coupled with my first post on this thread, that the need for a 160MHz microcontroller is extremely rare and, in the case of the OPs first project, humorous.

PS, I started with a 15MHz analog scope. I upgraded to a 100MHz Hitachi Analog scope (2Ch) because it was being sold for $50 with two NOS probes at a swap meet. My old 15MHz scope is still good enough for most things I do.
.

uggghhhhhhh 15MHz so low lol. That said, I would rather have a 15MHz, 4-channel, channel isolated scope over a regular 60MHz one. Bandwidth is real low on my list of scope priorities for MCU work.

 

[Nigel Goodwin]

I also find isolated channels super super important. Or at least, when not stricly needed they are extremely convenient. But that rules out most scopes, even much more expensive ones, unfortunately. I don't know why that isn't standard.

Fairly obviously it's because it's expensive to add, and only very rarely required.