Digilent Analog Discovery 3 announced (June 14 release)

[mwb1100]

Digilent Analog Discovery 3 announced (June 14 release).  No pricing details.

  - https://digilent.com/shop/analog-discovery-3/

Seems like there's not much over the AD2.  What am I missing?

Maybe there will be a fire sale on AD2?

[egonotto]

Hello,

125 MS/s like the Red Pitaya.

Best regards
egonotto

[xtech]

Does WaveForms for AD is better than for DD? I've used in the past WaveForms with Digital Discovery and this was really a bad experience. Finally I've sold it because in my opinion it was barely usable. Using behemots like Tek TLA715 was more convenient than WaveForms. Or maybe simplicity of Saleae made me so lazy

[switchabl]

They claim a "larger-than-ever buffer size", unfortunately I couldn't find an actual number. The max. 16kpts per channel is the biggest gripe I have with the old one.

[ataradov]

Does not looks like a significant improvement. I assume it is just a continuation with marginal spec bump, probably inspired by the need to update components. Availability of parts has shifted drastically over the years, so they may not be able to source stuff anymore.

WaveForms is the same as for DD. For AD I would call it workable, not ideal, but not completely useless. Not enough to sell mine. It is is a useful thing to have in some cases.

But I would not use DD, there are far better logic analyzers out there.

[Anthocyanina]

They claim a "larger-than-ever buffer size", unfortunately I couldn't find an actual number. The max. 16kpts per channel is the biggest gripe I have with the old one.

i was hoping for something along the lines of 1Mpoints, but no, it's only 32k per channel max with the AD3. got that from here https://twitter.com/ATaylorFPGA/status/1663932993742815234

Does WaveForms for AD is better than for DD? I've used in the past WaveForms with Digital Discovery and this was really a bad experience. Finally I've sold it because in my opinion it was barely usable. Using behemots like Tek TLA715 was more convenient than WaveForms. Or maybe simplicity of Saleae made me so lazy

i think the waveforms software is great for the AD2, for things where i don't need to look at high frequency signals or too low level signals i end up using the AD2 over any of my other scopes, and the beta versions are released very regularly and they implement changes suggested by the users very often as well.

the AD2 price has been updated to 299$ on their website.

[alm]

But I would not use DD, there are far better logic analyzers out there.

I use mine mainly for the pattern generator with programmable output levels, but then I got it cheap used. I wouldn't buy it just for the logic analyzer indeed.

[J-R]

Resellers have the pricing for the AD3 listed:
$428.72 - https://www.digikey.com/en/products/detail/digilent-inc/410-415/19235661
$415.34 - https://www.arrow.com/en/products/410-415/digilent

Like others, I had some hesitation with buying the Digital Discovery, especially given its age, but recently couldn't resist snagging the $280USD kit on ebay, basically new in box, for about $125USD delivered.  I think I mostly bought the AD2 and DD for the educational value.  I have more capable devices sitting around.

[croma641]

From 99€ (AD1 introductory price) to 230€, to more or less 350€ for the AD 2..... I hope that a new open source or similar will be available..

[switchabl]

The bottom of my black AD1 says "Sponsored by Analog Devices, Xilinx". There is no way you could sell something like that for anything close to 100€ if you actually have to pay regular prices for all the parts. The ADALM2000 is slightly cheaper but it is only 12 bits and the software is not nearly as good (open source though).

[chickenHeadKnob]

The specifications page is irritating.
It has a USB "C" connector. That's Great. They don't actually say if it is USB 2 hi-speed or  USB 3 , 3.1 or whatever.

[Weston]

Ouch, that pricing. The AD2 was creeping up in pricing over time too.

I think they are cool devices, even if the WaveForms software has some rough spots. I have been having students use the AD2 in some feedback and controls labs I designed and it works pretty well. Due to the differential input channels its actually a very good impedance analyzer, you can measure mOhm range impedance up to the MHz range.

I wonder what the edu discount will be, education is their biggest market. They sold a whole lot of AD2s during covid when all the students had to be remote.

At a $400+ price point the AD3 (or the AD2) does not seem like a very good value for the money. Once you throw in the added cost of the BNC adapter board and oscilloscope probes the AD3 costs more than entry level DSO from Rigol/Siglent with a built in waveform generator and is not too much cheaper than a entry/edu level scope from Tek/Agilent.

[Anthocyanina]

The specifications page is irritating.
It has a USB "C" connector. That's Great. They don't actually say if it is USB 2 hi-speed or  USB 3 , 3.1 or whatever.

i asked them for more information, and they replied, with regards to USB speed, it will still be USB 2.0

sadly, the AD3 seems like a mere incremental improvement over the AD2. They could have gone USB3 and leveraged USB C's power management thing and made the aux power connector unnecessary, but they didn't do either of those things. Another thing they mentioned was that it would still have the same 2 voltage ranges as the AD2, so the only improvements seem to be a faster sample rate, double the sample memory (and single channel can use 64k of memory), slightly higher waveforms per second, and a new case, which i'm not to fond of as i rather enjoy the AD2's translucent case since i like seeing the insides of it.

link to their post partially answering my questions: https://forum.digilent.com/topic/25916-analog-discovery-3-information/

i just noticed a curious thing, unlike the AD2, the AD3 doesn't have XILINX's logo at the bottom, nor does it have the ADI logo nor the TI logo as recent AD2s do. how curious! does that mean the end of their partnership? at least as far as providing ICs for the AD devices, hmmm!

[Howardlong]

The specifications page is irritating.
It has a USB "C" connector. That's Great. They don't actually say if it is USB 2 hi-speed or  USB 3 , 3.1 or whatever.

I'd imagine it'll be 480bps hi speed based on what's not said.

Interesting they didn't go with Power Delivery, but chose an DC power jack instead, presumably based on the additional BOM cost. 

I'm sure I'll pick one up in time (I already have 2 x AD2, 1 x AD1 and 1 x DD) as the form factor & software are excellent for both field work and on-the-go hardware development, but the improvements over the AD2 are so marginal I have little incentive to dive in right away.

The potential game changers for me would have been: substantially deeper memory, increased sampling rate to 250 if not 500MSa/s, and power supplies to +/-15V.

[artag]

Power Delivery is fairly widely available on chargers but is it well supported on PCs ? It certainly won't be supported from a USB 3 or 3.1 port which is often the source of a USB C cable.

[jasonRF]

I have never used any version of an AD, but am interested in seeing the specs of the lock-in amplifier they advertise for this new version.   

[Digilent]

Hi! Digilent here. Thanks for discussing our humble new product! We wanted to drop in to clear up pricing questions: it will retail on our site for $379 (cheaper than AD2), and available to academia for $279.

We've updated the page with this information, since you all aren't the first to want it!

[egonotto]

Hello,

It would be fine if there were more than two levels of sensitivity.

Best regards
egonotto

[2N3055]

Hello,

It would be fine if there were more than two levels of sensitivity.

Best regards
egonotto

Agree..

It still has toy front end, max of only 32kS of memory.  It should be max 200-250 USD retail and 100-120 USD academic.
It is a didactic toy.. No doubt useful but too expensive for what it is...
$379 buys you professional entry level standalone scope these days..

[egonotto]

Hello,

They (AD, AD2) have an input divider that reduces the input voltage to a factor of 22/104 = 0.211 in the most sensitive area. This protects the input, but increases the noise. I would prefer a bigger factor.

Best regards
egonotto

[jasonRF]

At a $400+ price point the AD3 (or the AD2) does not seem like a very good value for the money. Once you throw in the added cost of the BNC adapter board and oscilloscope probes the AD3 costs more than entry level DSO from Rigol/Siglent with a built in waveform generator and is not too much cheaper than a entry/edu level scope from Tek/Agilent.

In the US the AD2 often sold for $419 including the BNC adapter and two probes, and it looks like the AD3 will be about the same.  Can you tell me which Rigol/Siglent models have built-in waveform generators and cost less?

In any case, I think the AD2/AD3 compete more against other USB scopes than against traditional bench gear.  And in that market I think the AD3 pricing isn't bad considering it is in the category of 'nice' USB scopes that work as advertised and have reasonable software with a lot of features.  The closest 'nice' device in terms of performance that I am aware of is the picoscope 2205a MSO which sells for $539.  It has 2 analog and 16 digital channels, nominally 25 MHz bandwidth, 48 kSamples shared memory, and can use all channels simultaneously at 250 MS/s (500 MS/s with only 1 analog or 8 digital channels).  It does have a proper front-end with 4 mV/div sensitivity, and with an 8-bit ADC that makes it twice as sensitive as the AD.  But the single-channel 1 MHz function generator (+/-2V, 600 Ohm output impedance) is way wimpier than the AD, it doesn't have true differential inputs if/when you want them, no pattern generator, and no power supply.

Do I wish the AD3 was cheaper?  Yes!  But I think it is priced competitively. 

jason

[J-R]

Actually, the AD2 frequently could be bought direct from Digilent with $200 off if you buy a development board, of which a common choice was about $100.  So that would be $299 for the base kit.  (It currently is being listed for $299 as well.)  Then they also have offered 10% off coupons for signing up to their newsletter.  Finally, they offer student pricing, which you really just need access to a school/university e-mail address to access that.  So add it all up and the AD2 was quite a bit less.

[Nominal Animal]

i asked them for more information, and they replied, with regards to USB speed, it will still be USB 2.0

Sad.  I was hoping it would be USB 3, so it could stream the ADC data at 14 bits, 125 MS/s (2 Gbit/s + USB overhead), continuously.

Many/most Windows machines probably can't handle that kind of USB bandwidth (in an userspace application), but who cares?

[David Aurora]

At a $400+ price point the AD3 (or the AD2) does not seem like a very good value for the money. Once you throw in the added cost of the BNC adapter board and oscilloscope probes the AD3 costs more than entry level DSO from Rigol/Siglent with a built in waveform generator and is not too much cheaper than a entry/edu level scope from Tek/Agilent.

In the US the AD2 often sold for $419 including the BNC adapter and two probes, and it looks like the AD3 will be about the same.  Can you tell me which Rigol/Siglent models have built-in waveform generators and cost less?

In any case, I think the AD2/AD3 compete more against other USB scopes than against traditional bench gear.  And in that market I think the AD3 pricing isn't bad considering it is in the category of 'nice' USB scopes that work as advertised and have reasonable software with a lot of features.  The closest 'nice' device in terms of performance that I am aware of is the picoscope 2205a MSO which sells for $539.  It has 2 analog and 16 digital channels, nominally 25 MHz bandwidth, 48 kSamples shared memory, and can use all channels simultaneously at 250 MS/s (500 MS/s with only 1 analog or 8 digital channels).  It does have a proper front-end with 4 mV/div sensitivity, and with an 8-bit ADC that makes it twice as sensitive as the AD.  But the single-channel 1 MHz function generator (+/-2V, 600 Ohm output impedance) is way wimpier than the AD, it doesn't have true differential inputs if/when you want them, no pattern generator, and no power supply.

Do I wish the AD3 was cheaper?  Yes!  But I think it is priced competitively. 

jason

The issue with the Picoscope alternative has always been the half assed software. Diligent absolutely crushes them on that. I don't know why both companies are so locked into only working on half of the system each. Picoscope T&M 7 could have been a game changer, but after the eternity it took for a "stable" release it wasn't even stable. And at the other end- Diligent could have simply banged on a couple BNCs, done something about the input ranges and made the AD3 an absolute cracker... but they didn't.

I really don't know what either of them are thinking with a lot of this stuff.

[Anthocyanina]

At a $400+ price point the AD3 (or the AD2) does not seem like a very good value for the money. Once you throw in the added cost of the BNC adapter board and oscilloscope probes the AD3 costs more than entry level DSO from Rigol/Siglent with a built in waveform generator and is not too much cheaper than a entry/edu level scope from Tek/Agilent.

In the US the AD2 often sold for $419 including the BNC adapter and two probes, and it looks like the AD3 will be about the same.  Can you tell me which Rigol/Siglent models have built-in waveform generators and cost less?

In any case, I think the AD2/AD3 compete more against other USB scopes than against traditional bench gear.  And in that market I think the AD3 pricing isn't bad considering it is in the category of 'nice' USB scopes that work as advertised and have reasonable software with a lot of features.  The closest 'nice' device in terms of performance that I am aware of is the picoscope 2205a MSO which sells for $539.  It has 2 analog and 16 digital channels, nominally 25 MHz bandwidth, 48 kSamples shared memory, and can use all channels simultaneously at 250 MS/s (500 MS/s with only 1 analog or 8 digital channels).  It does have a proper front-end with 4 mV/div sensitivity, and with an 8-bit ADC that makes it twice as sensitive as the AD.  But the single-channel 1 MHz function generator (+/-2V, 600 Ohm output impedance) is way wimpier than the AD, it doesn't have true differential inputs if/when you want them, no pattern generator, and no power supply.

Do I wish the AD3 was cheaper?  Yes!  But I think it is priced competitively. 

jason

The issue with the Picoscope alternative has always been the half assed software. Diligent absolutely crushes them on that. I don't know why both companies are so locked into only working on half of the system each. Picoscope T&M 7 could have been a game changer, but after the eternity it took for a "stable" release it wasn't even stable. And at the other end- Diligent could have simply banged on a couple BNCs, done something about the input ranges and made the AD3 an absolute cracker... but they didn't.

I really don't know what either of them are thinking with a lot of this stuff.

i don't know about Picoscope, but yeah, on Digilent's side, i wouldn't say BNCs would be super important, particularly at the bandwidth offered by the AD3, but even one extra voltage range, a 1vpp full scale range would have made it much more appealing. too late to call this one an AD2.5, lose some money, spend a few months updating the front end to support an extra voltage range, then release a 3 voltage range AD3 at the 400$ the AD2 used to cost lately? of course it's too late  but one can dream