EEVblog #692 - Digilent Analog Discovery Review

[jaxbird]

Just saw that Martin posted a quick view video of the Red Pitaya very recently:




Not much detail on performance, but confirms that the current software is very basic.

[ncoonrod14]

Quote from: ncoonrod14 on December 13, 2014, 10:44:00 AM
It looks like there's a catch to the educational pricing. In the terms of their pricing it mentions that the educational pricing is only available if the board is required for a class. Since I haven't tried to order one I guess you might be able to get around that but I'm not sure you could buy this unless your institution is listed as requiring their hardware 

Damn, didn't see that, yeah 

I can now confirm that you can order with just your school information. I entered my university, university email, and link to course descriptions of some of my labs in the space for the course that requires the Analog Discovery and mine is shipped. Great deal on a great product! 

[mtdoc]

Good to hear !

FWIW I was able to order one last year with my edX course info and a yahoo email address. No  .edu email needed.

[miguelvp]

I got mine from Microchipdirect { TDGL023 } for $219 USD.  They also have the BNC adapter board { TDGL024 } for $20. Ships direct to Canada.

Any extra fees? Shipping costs for example.

Why don't you try to add one of the 13 left to your cart?
https://www.microchipdirect.com/ProductSearch.aspx?keywords=TDGL023

Brokerage and such varies greatly and shippers that don't maintain a company presence inside Canada can't calculate it. Often the carrier will tag on the "extras".

Maybe, but now there are only 6 left, so they are selling at 2 per day average.

[LaurenceW]

If anyone in Europe fancies one, German distributer TRENZ Electronic has them, with educational (not US student) discount.  I stated that I am a Technical College lecturer (I am), gave details of the institution, and that was proof enough.  Ordered mine Monday morning, had it in my hands in the UK Wednesday evening! Shipping perfectly reasonable, and no d!cking about with customs delays/charges.

What's it like? - Dunno, Mother Christmas says I can't open it for another week! 

[plexus]

Hi All. I ordered from Microchip too. I am in Toronto Canada. I ordered on Monday and it arrived today!

Analog discovery $219US
BNC breakout $19.99
FedEx shipping $10.20
Tax $32.42
So all tolled: $325 CAD

FedEx left the package at my door as per my instructions. no other brokerage or other fees were incurred.

I am happy to report that it runs find on my Macbook Pro with Parallels 9!

Also, the connector port lettering aligns well to the pins on mine.

I just used the network analyzer to test the response of my stereo tube amps. yup! we have a problem. the left amp has a 15dB boost from 3kH to 30kHz. but, good to know! now I can troubleshoot...

[plexus]

Hi All. I ordered from Microchip too. I am in Toronto Canada. I ordered on Monday and it arrived today!
[snip]

Last year Newark Canada was selling them at U.S student pricing ($99.00), thats when I bought mine. I think its worth the money, I don't know if I would have bought it if it wasn't $99.00 bucks though. I think Newark sold there stock in a couple hours just from this forum.

I see now they don't even stock it in N.America and the price they want is over 4x what I paid last year.

http://canada.newark.com/digilent/410-244/design-kit-analog-discovery/dp/27W5499

yup. i couldnt find it for under $300 in canada just for the device. for me the package is worth $325 easily as I do not have an instrumentation lab and I only need such tools from time to time. so its perfect (for me). in fact for what you get it's unbelievable.

[tggzzz]

I've raised a question in a separate thread; if anyone has any pointers then please post them there.

The question is whether the Waveforms software running under Linux/Wine works with the hardware. I've noted how I've got the Waveforms working in demo mode, but obviously I can't take the next step.

https://www.eevblog.com/forum/testgear/digilent%27s-analog-discovery-does-it-run-in-wine/

Thanks

[plexus]

I've raised a question in a separate thread; if anyone has any pointers then please post them there.

The question is whether the Waveforms software running under Linux/Wine works with the hardware. I've noted how I've got the Waveforms working in demo mode, but obviously I can't take the next step.

https://www.eevblog.com/forum/testgear/digilent%27s-analog-discovery-does-it-run-in-wine/

Thanks

I can't give you a definitive answer. but, I wondered the same thing about running Waveform on windoze on my mac with parallels. Since parallels works with other meters I have that have USB (eg. data collective multimeter and dual channel thermometer work fine connected to the mac with the software running under parallels). I was confident it would work. Luckily, it does!

So if you system can connect to other similar devices and work then I would say that it is likely it would work. you may want to buy one from a place that will take returns and perhaps let them know you want to test it first. or just call Digilent and ask?

Let us know if you get it working.

[c4757p]

It almost certainly does not work under WINE.

[tggzzz]

It almost certainly does not work under WINE.

What leads you to that conclusion? Have you tried it, or something similar?

What would be the expected symptoms (and preferably cause) of the failure?

[tggzzz]

It almost certainly does not work under WINE.

There's an informative post by attila in the thread
https://www.eevblog.com/forum/testgear/digilent%27s-analog-discovery-does-it-run-in-wine/
Someone with the same name on the Digilent forum is a Digilent employee

[mtdoc]

It sounds like there's a cross platform version of the software in Beta testing.  That's good news!

[plexus]

Not sure if this was posted before: http://www.jensign.com/Discovery/About/index.html

[c4757p]

It almost certainly does not work under WINE.

There's an informative post by attila in the thread
https://www.eevblog.com/forum/testgear/digilent%27s-analog-discovery-does-it-run-in-wine/
Someone with the same name on the Digilent forum is a Digilent employee

Interesting, thanks. The USB bits seemed a bit hopeless when looking at the Analog Discovery earlier this year for use with a summer Circuits course - we ended up having the non-Windows students run it in a VM. Perhaps it's changed. I'll have to take a look at that.

[c4757p]

Sorry, this was back in May/June! I don't remember. I remember farting around moving DLLs about trying to make it not crash.

[amyk]

At least their name won't be confused with Agilent anymore...

Digisight is already taken, but Sigsight is still available

[tggzzz]

I presume even though I don't know the formula that from those two values you could calculate the frequency for -3dB attenuation.

In general no, but in practice plausible presumptions could be made.

I have heard Dave talk about -3dB being the point at which the bandwidth of an oscilloscope is determined. Are there also other lesser factors which are used to determine the bandwidth?

Digilent must have had some reason to be so apparently conservative in their claimed BW.

3dB is the standard for scopes and just about all equipment, but it is only part of a specification. Other parts of a specification that can be important (depending on the application) include passband ripple, rolloff rate and phase response. A scope should be benign in all those respects, but be very wary of digital scope sampling rate, since that can hide more than it illuminates.

ISTR reading somewhere that it was felt that the target markegt (beginners) would complain if they felt they has been mislead by finding the amplitude of a 5MHz signal was too low. Personally I don't agree with that, since they will need to understand what equipment's spec does and doesn't mean.

[mtdoc]

ISTR reading somewhere that it was felt that the target markegt (beginners) would complain if they felt they has been mislead by finding the amplitude of a 5MHz signal was too low. Personally I don't agree with that, since they will need to understand what equipment's spec does and doesn't mean.

That is correct.  On page 16 of the technical manual it show plots of the frequency response and states:

"You can see from the plots, that this circuit exceeded the requirements for 5MHz of bandwidth, and the -3dB
point is more than 20MHz. However, since many students who will be using the Analog Discovery don’t
understand the concept of “-3dB” is the “bandwidth” of an instrument, and that a 1V input signal with -3dB
applied will measure 0.707V, it was felt from a marketing standpoint to specify the bandwidth of the analog
inputs as less than -0.5dB as the “bandwidth”. This ensures that when connecting a 10MHz signal on a
traditional instrument (with much higher bandwidth), and the Analog Discovery, the measurements will be
very similar, and lead to less confusion."

I agree with you that students should learn what instrument specifications really mean.

On the other hand, the reality is that without the opotional  BNC adapter board and a x10 oscilloscope probe, the usable bandwidth will be below the actual -3dB point.

[c4757p]

Students should definitely learn what the specifications mean, but unfortunately, you really can't shove too many things at them at once before things start going over their heads. In the Electronics course I've been TA'ing, we tried to keep a focus on common engineering practices like decibels and the -3dB point all the way through, but that's only a recent change - the guy who taught it last year probably didn't use the word "decibel" once.

It tends to be important to limit the d(surprise)/dt to keep their attention, so I do like that they're keeping to a novice-friendly style of specification. If we used this in our class, we'd have no problem looking up or testing the actual -3dB point on our own - in fact, it'd make a good lab exercise to help learn the equipment

[tggzzz]

It tends to be important to limit the d(surprise)/dt to keep their attention, so I do like that they're keeping to a novice-friendly style of specification. If we used this in our class, we'd have no problem looking up or testing the actual -3dB point on our own - in fact, it'd make a good lab exercise to help learn the equipment

That's a justifyable position, but I can't help remembering the a distinguished professor (Eric Laithwaite, an early proponent of MagLev trains and Linear Motors)...

On principle in the final year exams he set for undergraduates, he would include three questons:

• one question could be answered by everyone that had attended and understood his lectures
• one could be answered by people that had explored beyond his course material
• one couldn't be answered in the time available

and he expected that an engineer could determine which was which and avoid the impossible question. If you couldn't, then you deserved not to get good marks.

But nowadays students demand to be spoon-fed and not exposed to the real world.

[c4757p]

But nowadays students demand to be spoon-fed and not exposed to the real world.

Meh. Some definitely do. I've seen students who, faced with a relatively challenging lab exercise, put their hand up and essentially ask for it to be designed for them. That's not what I'm talking about, though. Even for the ones who want to learn, you have to plan how to deliver information to them so as not to overwhelm them. Unfortunately, a few people think that if you put any effort into teaching methods beyond just bashing students over the head with information and expecting them to thirstily drink it up, you're spoonfeeding them, and some people think that if a student can be overwhelmed or doesn't just teach himself everything, he's demanding to be spoonfed.

[tggzzz]

But nowadays students demand to be spoon-fed and not exposed to the real world.

Meh. Some definitely do. I've seen students who, faced with a relatively challenging lab exercise, put their hand up and essentially ask for it to be designed for them. That's not what I'm talking about, though. Even for the ones who want to learn, you have to plan how to deliver information to them so as not to overwhelm them. Unfortunately, a few people think that if you put any effort into teaching methods beyond just bashing students over the head with information and expecting them to thirstily drink it up, you're spoonfeeding them, and some people think that if a student can be overwhelmed or doesn't just teach himself everything, he's demanding to be spoonfed.

All of which is valid. The comment you quoted was over-the-top - but the rest was valid.

I'm becoming a grumpy old man.

[c4757p]

but the rest was valid.

Yup, it was.

I'm becoming a grumpy old man.

Easy solution - you should have been grumpy your whole life, like me! Then you only become old.

There is another thread where generalisations about specific or small sample sizes were on topic. The "hobbyists are useless" thread. This thread has remained useful can we please keep it relevant for future reference?

That one's full of unpleasant people and comments, though!

Okay, I'm done here, sorry. You're right, getting offtopic.

[jaxbird]

ISTR reading somewhere that it was felt that the target markegt (beginners) would complain if they felt they has been mislead by finding the amplitude of a 5MHz signal was too low. Personally I don't agree with that, since they will need to understand what equipment's spec does and doesn't mean.

With high Z the bandwidth of this device is about 20MHz (-3dB): (note the Digilent software does not permit exceeding 10MHz, but using the API you can extend that to 20MHz+)




However with a standard 50R load the -3 dB bandwidth (function gen) is pretty much exactly 5MHz: