There was some silent grumbling in the community when we heard that the DL40 Module was elected to be executed but not really an outcry.
I don’t know whether it was sold in a quantity to be profitable, but I would beg for its live since it adds, as I think, the option to delegate asynchronous realtime features to the NETMF/Gadgeteer family that cannot be achieved in another way.
I don’t know wheter it is really a good idea to discontinue a module that is - as Brett says - „one of those opens the doors modules".
I don’t like seeing the FEZ Medusa Mini die before it could walk and before I could try it, but may be: To much work to keep it alive.
You can somewhat easily use the breakout modules with the DIP package of the LPC chip. This was targeted towards module makers and advanced users so sale was too low to keep. Of course we welcome anyone to make it and sell it.
@ Gus - Thank you for your answer. But it’s just the fascination of Gadgeteer to have interesting chips easy to use on a board with Gadgeteer Sockets. When I saw it, I just bought two of them. For 10 $ nobody will use the soldering iron. But let’s wait for the reaction of the community. If people like it too, they can give me a +1. If nobody wants it, nobody will make it.
@ gus is there going to be any dl40 available, or they are all gone?
What a shame that the sale was not sufficiently successful, but the idea of universal daisylink module to support new module development was excellent.
I noticed there were a few in the sale grab-bags, so there might have been a few.
Personally, I think it’s an awesome idea to have this module as well, but has it been successful? There are some great examples that have the heart of the DL40 (parallel CNC ) but there aren’t a lot of real-world use-cases documented / shared of how that can be used in real world situations - and I suspect that partially at least this would help open the eyes of people who might use this module as the starting point for their next big thing.
For example, how about as a display controller for a cheap module (HD44780 text module perhaps)?
Or what about a different form factor? Say if the 40-pin module was less breadboard friendly but say had a prototyping area, do you think that it’d have more appeal?
I agree with @ RoSchmi, it’s unlikely that people will prototype with the DIP version of this chip, it’s too easy to fall back to something else (duino based probably) rather than solder up something, which is why the prototyping area came to mind.
@ Gus - if it is not to much a business secret - just out of interest, from which quantity per year to you consider a cheap mudule like this to be profitable?
The problem with using the DL40 for that (specifically, but also in general) is that your host MCU can run the HD44780 faster and use less power doing it, while being cheaper, than adding the DL40 to run it.
If you REALLY need an external MCU to run your HD44780, then there are smaller, cheaper, lower-power options.
For that specific application, the DL40 doesn’t really shine in any area.
I thought it could especially be very useful were several sensors or buttons or displays or actuators have to be polled or controlled in a distance of several meters. Over the daisylink connection only 3 or 4 cables can be chained from one DL40 to the next
For long-distance stuff, a protocol like OneWire or RS-422/RS-485 (works over hundreds of meters) is much better than the I2C used by DaisyLink (works over a few meters).
DaisyLink is a solution looking for a problem, and it hasn’t found a problem yet.
@ godefroi - there is rarely a thing that cannot be made better in another way. But I think that if you use OneWire or RS-422/RS-485 protocol you need a processor at every distant station as well and under this aspect a set of DL40s could be a rather cheap solution. But you are right, there are only ideas, yet no working and needed applications.
For sensors there’s the DS2450 1-16 bit quad-channel ADC.
For digital I/O there’s the DS2408 8-channel programmable I/O.
There are also sensors that are 1-wire devices, ala DS18B20 (temp) and DS2438 (battery monitor/fuel gauge), DS2423 (pulse counter).
One could imagine driving an HD44780 using a DS2408 and a shift register. It wouldn’t be fast, but it’d work.
No microcontroller at all needed at the “edges”. DaisyLink, on the other hand, does require an MCU at the “edges”, because each node in the DL chain is a microcontroller.
If there’s a better way to build things, then why petition GHI to resurrect/sustain an unprofitable product so we can do them poorly?
O.k., knowledge and better arguments seem to be on your side. 7 pro-votes are are no majority. So may he rest in peace. Everybody here whishes the most profitable products for GHI and their customers.
thanks, I know all that - I was trying to find potential problems to solve, or show examples of problems that a DL40 makes sense… But seriously, there is still value in something like DL40 for this task - you only have to look at the number of I2C or SPI based add-on boards onto HD44780’s to see that it’s an often used ploy to minimize the number of pins; even though they typically just minimize the pin count, you could in fact abstract the interface to this away (you’d still need to implement the HD44780 protocol layer in the DL40IO C code).