Oscilloscope
#1
Posted 30 March 2012 - 11:03 PM
I am getting to the stage in my projet where I need to start testing the output from my hardware and I am told I need an oscilloscope.
I know very little about scopes and wondered if anyone could suggest a scope that will do the job.
I don't want to spend loads, but I don't want something that won't give me a reliable reading.
The types of things I want to check are PWM signals from TLC5940s and clock signals from oscillators etc.
I have seen various types on ebay, some are digital usb ones with no screen and you see the output on your pc.
Others are small hand held ones and the last type are the typical bench top ones.
Here are a few examples:
Nano
Usb
Bench Top
Your opinions would be greatly welcomed, and ideally as cheap as possible, but obviously not to the point that they are useless. I have no idea of the accuracy etc that I need, but I want to check what is being output from some TLC5940s, the oscillator clock circuit Nevyn made to control the TLCs and also to see what the 74hc595 shift registers are outputing, plus anything else netduino related I made need to check.
Many thanks,
Andy
#2
Posted 30 March 2012 - 11:12 PM
However, if you want to look at a scope, why don't go for something that's will be more versatile for you along the way. So forget about nano (it's a toy) and probably USB too. The "Bench top" seems more like it even though pricing seems better than to expect in quality
As a rule of thumb you want a scope capable of measuring at least twice the frequency of the signal you want to measure (due to the Nyqvist folding theorem, according to which wheels appear to spin backwards on a 25Hz TV set, which they also tend to do).
Now, the "Bench top" has a bandwidth of "only" 25Mhz which means you can only measure signals slower than 12,5 Mhz at best. While 12,5Mhz can seem like much at this point, you will not be able to measure signals faster than that (.e.g. SPI or I2S).
I recently bought a RIGOL DS1052E 50Mhz scope at some 350$ which I am very pleased with. These scopes are on sale across the world right now and are great value for money. Highly recommended!
USB scopes are not much cheaper than regular ones and don't forget you'll have to boot that old PC up every time you want to take a look at a signal. Furthermore, the RIGOL mentioned has both USB device and host from which you can interface from a PC and save waveform screenshots directly onto a thumbdrive respectively.
Are you sure you want a scope, maybe you're looking for a logic analyzer? Then I would recommend something from Saleae, don't know if they have enough bandwidth though.
#3
Posted 31 March 2012 - 12:03 AM
I too have been looking at that owon model of scope as my old trusty tektronics scope literally blew up on me last week (something went pop, probably a cap and a lot of smoke came out) but it was an ebay bargain and i have easily got my money's worth out of the £50 i paid for it 10 years ago.
With regards to that Owon scope CPC/Farnell sell them and also offer a calibration service which certainly can offer some peice of mind when it comes to accuracy, although they are a bit more expensive than the Ebay sellers, you get the peice of mind for returns and warrenty then.
Depending on your budget and amount of time you have to look for a scope, you can snag some great bargains on ebay, if a CRO (Cathode Ray Oscilloscope) will suffice and you dont need a DSO (Digital Storage Oscilloscope) then you can get some great bits of kit for < £100. But it really depends on your needs, if all you require is to visulise PWM and Oscillator output then a CRO could well be a bargain for you. However if you also want to analyse bus traffic (spi, uart etc.) then a DSO would be more applicable as the storage feature allows you to scan back and forth the trace. Also if you want to capture waveforms a DSO with USB output should allow you to save the output as an image (or at the very least take a screen grab from the program).
As an alternative to a DSO for bus analysis you may find some thing like the IKA Scanlogic 2 to be adequet for your needs. I have just received mine as i needed a stop gap replacement for my scope and needed to be able analyse uart and spi traffic for the day job and I am very impressed with my initial experiments with it. They can be purchased for about £55 here.
Really it all depends on how much you would like to spend and what your plans for usage are, ultimatly you will not regret forking out for a DSO as there versitility pays for them selves in the long run.
#4
Posted 31 March 2012 - 01:33 AM
#5
Posted 31 March 2012 - 03:53 AM
#6
Posted 31 March 2012 - 07:50 AM
I second the comments on the Nano - forget it.I am getting to the stage in my projet where I need to start testing the output from my hardware and I am told I need an oscilloscope.
I know very little about scopes and wondered if anyone could suggest a scope that will do the job.
I don't want to spend loads, but I don't want something that won't give me a reliable reading.
Regarding the bench scope you mentioned, remember that it is coming from Hong Kong and if the couriers work out it's value then you will get stung for an extra 20% VAT on import to the UK.
Being in a similar position last year (i.e. not a big budget and needing a scope) I went down the USB route and bought a Piscoscope and a Saleae logic analyser.
I'm not an expert on this subject so all I can say is so far they have both served me well.
Regards,
Mark
To be or not to be = 0xFF
Blogging about Netduino, .NET, STM8S and STM32 and generally waffling on about life
Follow @nevynuk on Twitter
#8
Posted 31 March 2012 - 03:26 PM
#9
Posted 31 March 2012 - 04:48 PM
I'd to keep this shorter...
A periodic real waveform of any shape can be seen as composed of one or more sine waves, called harmonics.
See here: http://en.wikipedia..../Fourier_series
Analog bandwidth and sampling rate are not the same thing.
For example, you could see a scope targeted for 100MHz and 1GS/s =(samples/second). Upon the Nyqvist theorem, the maximum frequency should be 500MHz, but it's actually far away.
Any scope has an analog stage (input), and a displaying stage. While the analog stage is much the same for both digital and analog scopes, the displaying stage is completely different.
Analog bandwidth.
It's by far the most important, and typically expensive part of a good scope.
The analog stage *should be* an amplifier/attenuator that takes the signal on input (probe), and scales it for a certain factor. Theoretically, the shape of the input signal should be perfectly preserved. However, that's impossible.
Saying K the ideal scaling of this stage, the actual K' is mostly accurate up to a certain frequency. After that the decay is dramatic.
The frequency at which the K' = K / sqrt(2) is defined as the bandwidth.
When you are going to display the SPI running at 1 MHz (far away from the max), you could consider the shape of the waveform as a square wave. The steep is the edge, the higher are the frequencies needed for compose the waveform. Since a square wave should have "infinitely steep" edges, the frequencies involved are very high. For a 1MHz square wave of amplitude A, the next harmonic is the third (3 MHz) having amplitude A/3, the next is the fifth (5MHz) at A/5, and so away. The decay is not so fast.
Well, if you try to analyze the 1MHz SPI using a 10MHz bandwidth scope, then probably the wave form displayed will be pretty different from the actual one. That's because the analog stage of the scope "cuts" several important harmonics above 10MHz...and we've seen they're still important.
Sampling rate.
That refers to the digital scopes only.
Basically is the *maximum* sampling rate at which the ADC is able to sample the signal at the analog stage output. Of course, the higher it is, the better is the scope.
BTW, when the ADC is collecting the data, there should be some component able to store the samples. Bear in mind that this task has to be performed in *real-time*, thus is much faster than a normal PC. So, you might have a powerful scope capable of 1GS/s, but ...how deep is the cache?
If you cache offers some kBytes only, then clearly the period will be very short. For instance, collecting one sample each ns (1GS/s), a 4kB cache will be full in 4us. That could be a limitation, because you would collect many more samples: stop the refresh, and analyze them with ease.
Note: often you see something like "50GS/s equivalent". That's NOT meaning a super-ability to sample, nor any embedded wizard doing that.
Again, it's something related to the signal analysis. Have a look here: http://www.picotech....e-sampling.html
Hope it helps.
- mcinnes01 likes this
#10
Posted 01 April 2012 - 01:47 PM
#11
Posted 01 April 2012 - 03:02 PM
Here is my suggestion, but I'd like that someone else would give its opinion.
The Picoscope 2204 is too limited (10MHz, many functions missing, etc) comparing to the price. The smallest decent is the 2207, but its cost is higher than a desktop, powerful scope. Of course you pay the miniaturization, but...money are money.
Same conclusion for the bitscope, other than I don't know what are you referring to.
The model of the link isn't bad, considered the price. Although is not highlighted, it's pretty easy to understand that's an Hantek.
So, I dug a bit into the official Hantek site, and I've see this one, that looks pretty nice.
http://www.hantek.ne...st.asp?unid=132
Here is an eBay.UK auction:
http://www.ebay.co.u...#ht_5956wt_1139
What I like is that it's a 60MHz (enough), it has 4 inputs, plus the external trigger (sometimes useful). Then a separate supply instead the USB, a function generator, and a deep enough cache for the samples. The case looks much robust, and they gives you a bag with all the kit. However, the price is about 270GBP...I'd say to feel better to choose a decent tool, instead a cheaper limited one.
Come one guys!...Anyone would add anything?
#12
Posted 01 April 2012 - 05:47 PM
To be or not to be = 0xFF
Blogging about Netduino, .NET, STM8S and STM32 and generally waffling on about life
Follow @nevynuk on Twitter
#13
Posted 06 June 2012 - 10:20 PM
While this model can still be found for sale I found that it has been replaced by the RIGOL DS1102E for the same price range as the older model (was $795.00 now $399.00) and has a 100mhz bandwidth. RIGOL DS1102E
#14
Posted 07 June 2012 - 09:36 AM
Just as an update, while looking for a recommendation on which Ocilloscope to buy I came across this thread, the RIGOL DS1052E was mentioned as a good scope (I am sure that it still is).
While this model can still be found for sale I found that it has been replaced by the RIGOL DS1102E for the same price range as the older model (was $795.00 now $399.00) and has a 100mhz bandwidth. RIGOL DS1102E
AFAIK you can turn any DS1052E into the DS1102E through a simple firmware hack details here although it looks like Rigol have seen sense and changed prices on the DS1102E to be about the same as an un-modded DS1052E. They are very nice cheap scopes, which I can recommend.
#15
Posted 07 June 2012 - 07:32 PM
Just as an update, while looking for a recommendation on which Ocilloscope to buy I came across this thread, the RIGOL DS1052E was mentioned as a good scope (I am sure that it still is).
While this model can still be found for sale I found that it has been replaced by the RIGOL DS1102E for the same price range as the older model (was $795.00 now $399.00) and has a 100mhz bandwidth. RIGOL DS1102E
The DS1102E doesn't replace the DS1052E. The DS1102E is a same board, chips, components, etc as the DS1052E, except it has a different label on it, and bit has been twiddled in the firmware to allow capture at 100 MHz. As @nakchak mentioned, with putty (or any other serial terminal program) and some very easy instructions on the net you can turn your DS1052E into a DS1102E.
-dan
#16
Posted 07 June 2012 - 08:11 PM
Thanks Dan & Nakchak,The DS1102E doesn't replace the DS1052E. The DS1102E is a same board, chips, components, etc as the DS1052E, except it has a different label on it, and bit has been twiddled in the firmware to allow capture at 100 MHz. As @nakchak mentioned, with putty (or any other serial terminal program) and some very easy instructions on the net you can turn your DS1052E into a DS1102E.
-dan
The main reason for the post was to hopefully prevent someone from paying too much for the DS1102. I found the DS1052E on Amazon and was about to purchase it when I did some more looking and found the DS1102E. The problem is that a lot of paces are selling the DS1102E for > $600.00 (Amazon & Ebay). I hate for someone (particularly me) to pay too much for anything.
Chuck
#17
Posted 13 June 2012 - 09:57 PM
#18
Posted 14 June 2012 - 12:07 AM
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users