Sandwizz - the next jumperless breadboard

[ledtester]

What I gleaned from the video:

- you upload a Kicad schematic / netlist to the breadboard (via USB)
- via a terminal session the breadboard will tell you where to place components
- it can verify placement of components
- the breadboard can provide some components itself - like an op-amp, function generator
- you can stack other modules to the breadboard, e.g. an Arduino Nano module

Update: There's a 3 minute overview video here:

https://youtube.com/watch?v=h3WQmGasE6k

Also, the built-in components include: op-amps, PGAs, IDACs, VDACs, comparators, ADCs, AMuxs and 74-series logic
 (according to https://youtu.be/nU-0VqIi4SA?t=175 ).

Another feature is the ability to stream audio from your computer over the USB connection into your circuit.

[Sensorcat]

Breadboards are notorious for causing problems with parasitics inherent to the board design, so it would be essential to provide information on parasitics and the resulting frequency limitations. But similar to so many startups (solar roadways!), zero technical information at present. This is strange especially together with the claim 'professionals' do/will use it.

As far as hobbyists are the target audience, my observation is that too much integration/automation means that people don't learn much and remain dependent. And when they hit the limitations of the product that becomes their world with respect to the hobby, it's over.

[mikeselectricstuff]

 

[SeanB]

Reminds me of the early electronic fault testing I had when in training. 3 hour test for finding 5 faults (yes they thought this could be difficult) and I sort of set an unofficial record by completing all 5 preset faults in 0.0minutes, as the software only counted in 6 second increments (decimal minutes), per fault, and the 8 relays in it could only realistically give you 8 faults, and you also had jumpers and test points, and could abuse the circuit to find those faults using a scope and multimeter, especially as we had a month to play around with it. Z80 based, and used RS422 serial to connect all the stations together on a 3 wire bus to a PC running the DOS based software. Yes did flip one over and removed the 4 screws, and did a quick reverse engineering to see that in most cases only 4 faults were possible per board that was plugged into the DB25 socket on the top panel.

[Benta]

Agree.

 +

[ledtester]

They announced that they'll be on Kickstarter starting June 4th:

https://www.microaware.com/sandwizz-on-kickstarter/

[MK14]

After watching some videos, and reading some stuff about this.  Especially the following video:


It seems, all they are doing is have a breadboard with perhaps around 60 separate 'columns' (interconnected sections of a breadboard).  Connect to an I/O line of a library card.  Which seems to be an Arduino, with each of its ports, connected to a separate column of the breadboard.

Which runs on the Arduino, software to check the correct component placement, create the signal generator signals (e.g. PWM outputs), and the various components, internal to their breadboard, are just features of an Arduino's I/O system.

E.g. A to D converters, CLB blocks (small configurable/programmable logic blocks, or even a small FPGA, on some boards, as VHDL seems to be mentioned, somewhere), if present in that MCU/Arduino version (or via software loops), it is not that clear, on their implementation.

I.e. Not really that innovative, as it is just connecting some version of the Arduino (they may offer different versions, later), with a pin of its I/O connected to one of the common (column) connections, on the breadboard.
By 'common' I mean, that if you plug a 40 pin DIP IC, into a standard breadboard.  Each pin, is connected to the other pins, in the same column, until you get to the separating middle section of the breadboard (A diagram would make this much easier to explain).

Then using software running on the Arduino, to 'create' the extra components.

I presume they plan on doing, analogue versions, which using banks of analogue matrix multiplexers, would cope with automatically wiring up analogue circuitry/components, on the breadboard.

As perhaps others in this thread have indicated.  The student, would probably be better off, learning the entire process or properly using a breadboard, along with wiring it up, getting the component placement right, and real life debugging, of why it doesn't work, e.g. a component is the wrong way round or in the wrong place, etc.

I'm not sure that the two inventors of this system, have properly thought this thing through.  As there are so many possible problems with this idea, such as the currents being too high, for analogue multiplexers, the time delays being too long, if software processes signals, possible extra electrical noise and too much stray capacitance/inductance, extra impedance from the multiplexers, and so on.

[ledtester]

The Kickstarter has launched:

https://www.kickstarter.com/projects/sandwizz/the-sandwizz-breadboard-concept/pledge/new?clicked_reward=false

Some of the rewards:

$97 gets you one Sandwizz breadboard
$130 gets you a Sandwizz breadboard and an Arduino Nano library card (need to supply your own Nano though)

[coppercone2]

don't know, I can see a breadboard as mental exerciser


maybe you should leave the thinking machine out of this one and spend some time practicing. Its like weight lifting, for your mind

[Evi]

All this buzz is simple Cypress PSoC5 and nothing more.
And of coarse PSoC Creator is much more powerfull than this parody.

[temperance]

Interesting product. But I have a few questions.

Can I connect for example ten of those together and have them communicate with each other in a self configuring network in order to emulate a multi layer printed circuit board but then in the horizontal plane? Also, does it detect if I make stupid mistake and rewire it automatically?

[PlainName]

I.e. Not really that innovative, as it is just ...

Yes, quite, which is why they have been many, many versions of it available since forever!

A lot of innovation is blindingly obvious once someone else has done the hard work. Personally, I wouldn't touch this (mostly because I don't do breadboards) but I think it's a clever product.

[Evi]

Quote from: temperance on Today at 03:23:45

Interesting product. But I have a few questions.

Can I connect for example ten of those together and have them communicate with each other in a self configuring network in order to emulate a multi layer printed circuit board but then in the horizontal plane? Also, does it detect if I make stupid mistake and rewire it automatically?

>
It is very problematically, but you can buy this for $20 and have all the fun>https://www.infineon.com/cms/en/product/evaluation-boards/cy8ckit-059/

[shabaz]

I think it's a mess of a product. They could have made a type of product with relay connections to (say) some simple built-in test equipment (i.e., operating a bit like a multi-channel data acquisition device integrated into a breadboard), and that might have been a more interesting thing. I don't think I'd buy this product as-is at a quarter of the price.

They've sucked out all the fun from using a breadboard for quickly trying out things, and physically seeing the connections. Plus, their method will be unattractive to beginners who won't know how to use KiCad or to console into the device to use the interface. Plus, if you're not a beginner, then the electrical limitations will soon be a significant hindrance.

Then, piling on top of that, the price is probably out-of-step for the type of product it is; an Analog discovery with any student discount, plus a reasonable breadboard or two (so that you don't need to immediately disassemble a project to create a new one) would be far more useful.

I think they will have a very small set of users, and the device will lay on a shelf unused after a short while.

[Evi]

https://www.electronicdesign.com/technologies/analog/article/55088230/electronic-design-sandwizz-schematic-configurable-breadboard-now-on-kickstarter

Each Sandwizz-breadboarded CY8C5868AXI-LP032 “SoC” can provide up to four op amps, four comparators, two programmable gain amplifiers, three digital-to-analog converters (DACs), three Vrefs, and three 4-input analog multiplexers on the analog side .

[MK14]

Part of the advantage, of (real) breadboarding (solderless) something, is that you can quickly experiment, optimize and fine-tune the design, to make it work, just right.

Such as, rapidly changing some resistor values, by physically changing the values, until the output voltage meets the requirements, the supply current is reasonable and the occasional oscillations (instability), is reduced enough for the early Alpha-prototype, to get its first PCB made up.

I.e. components, such as resistors, can be changed on the fly in seconds.  Allowing rapid development, debugging, optimization (at least initially, on very early, well before per-production units, i.e. such as proof of concept etc).

Having to mess around with a computer/Ki-Cad, re-compiling maybe (I don't know how its workflow works) and uploading the new settings to the breadboard, sounds rather fiddly, annoying and puts possibly significant delays into the development path.

On the other hand, some changes, especially for components or component values you don't have readily to hand, and if there are many changes, for the current development iteration, may actually be quicker and easier (theoretically), overall.

[coppercone2]

it means you need more kits, I imagine the big problem is inductors. Capacitors are still fairly ideal in parallel, but inductors and transformers are trouble. I think a higher E series inductor kit would ease things up. A transformer kit is hard to imagine since their kind of like the ASIC of passive components


Air coil inductor, Iron core inductor, and ferrite inductor E24 series kits. That is something wrap your head around.

Capacitor kits for foil, ceramic and electrolytic are fairly easy to put together.

For inductors and ferrites you need adapters though, its hard to get alot of them in TH form now.