ebay scum (fao: AXEL)

[Screaming_Rabbit]

whooahh!

I was at "The recording workshop" in Chillicothe/Ohio about 10 years ago... know what you mean  :-*

Greets, Roger

[TK.]

Xanboroon brings up a point that is really thought-provoking. Where are the boundaries of DIY? Are premade PCBs already against the spirit of the unwritten DIY laws, or is etching only a job for lunatic fringes, who would also create their own resistors/caps/semiconductors if they would be able to do this.

I think that the definition of DIY is a continuous process which will shift more and more into the direction of a macroscopic/modular approach in future. Some years ago each MIDIbox was totally handmade: built on breadboards, boxed in selfmade cases, decorated with selfmade panels... getting the parts was always a challenge, and bringing up the PIC was mostly only possible for experts.

Today we've reached a status where the majority of people prefer to buy premade PCBs, they buy premade panels, they buy premade housings. And a trend can already be regognized for ordering complete kits or special parts from single sources, although it costs some extra money (or not if the availability is bad). People are mostly like electrons... they prefer the least resistance

So, how will be the progress in the next years? When you watch closely to the latest microcontroller developments, you will notice that the handy DIP package slowly disappears from the market, new developments are mostly provided in a SMD or BGA package. There are a lot of charming solutions available in the meantime (e.g. ARM or TriCore based 32bit controllers, or huge FPGAs) which could bring us so much --- with the disadvantage of increased costs for equipment and a higher demand on solderings skills. The fact is that the demand for ready assembled modules comes once the industry has totally switched to SMD, and the distributors starting to discontinue the DIY friendly chips.

Open question: is the possibility to buy assembled modules good the the community or not. I'm not sure yet. Currently the restriction that only PCBs or kits are available limits the number of users - from my point of view this is favorable for the support effort as well as for the infrastuture (e.g. the costs for the webservers)

On the other hand it could bring new people into the community who are experts in other areas, like software customization/programming, testing, sound design, documentation, etc.

Maybe it would only shift the support effort to another direction. Less issues with a non-working PIC programmers, wrong connections, bad soldering joints, missing MIOS upload request, but more questions about software or customization problems - the propability to get an helpfull answer increases with the number of people who are working with the same stuff.

It's really a difficult decision...

Best Regards, Thorsten.

P.S.: reg the MBHP license - I don't like so much rules, they make the world too complicated

[Mickey_Sadler]

Hi All,

TK mentioned that the DIY friendly chips are going away to be replaced with SMDs. These can still be used by DIY folks. I just takes a bit of learning and practice and smaller tools to work with them. When I was 12 years old I got my first amateur radio license and built transmitters and receivers. That means I learned to solder at an early age. After technical school for electronics, I worked for a time for RCA in a plant where they built tube type radios, stereos, etc. Then I went to work for an aircraft manufacutrer, North American Aviation, working on aircraft electronics. I had to go to a class to learn how to solder. I had been soldering for around 10 years and thought it was an insult to have to go to a soldering class. The first day the instructor passed around a printed circuit board, everyone thought the board was just beautiful. Then he told us it was a reject board from the rocket division that was rejected for poor soldering! I learned a lot from that class, but the best thing I learned was that using good tools, good cleaning, and careful inspection (with a jewelers loupe if needed) good solder joints are a breeze whether they are on a 1/4 inch stud or a 200 pin connector that has the pins almost touching each other. SMD is just smaller, but still workable. Oh, I had to take that class, then pass an exam every year in order to stay employed - just like the aircraft welders had to be certified.

Don't be afraid of the SMDs, just use a smaller soldering tool and the proper adhesive to hold them in place while working. You will learn another very useful skill.

Later,

Mickey

[Sephult]

I have to agree with mickey.

Although when SMD is phased out, I am going to have to quit ;D

I hate BGA packages

I doubt that soldering short issues will decrease with premade boards using BGA or SMA packages. I work for a major electronics manufacturer and using smaller parts only makes it harder to find the shorts themselves, especially under BGAs. You wouldn't believe how many times they have to be reflowed! The community will probably find themselves filling out RMA forms for sending their premade boards back more than actually getting to work with them.

TK, does have an interesting word on how the support effort would change though. I also agree that too many rules makes everything too complicated.

TK Quote

"Today we've reached a status where the majority of people prefer to buy premade PCBs, they buy premade panels, they buy premade housings".

I think we will know when things have shifted in the wrong direction, as every MIDIBOX of the Week will be the same

      ;D

[smashtv]

Open question: is the possibility to buy assembled modules good the the community or not. I'm not sure yet. Currently the restriction that only PCBs or kits are available limits the number of users - from my point of view this is favorable for the support effort as well as for the infrastuture (e.g. the costs for the webservers)

I completely agree on the support effort point.  Some kind of built in throttle control keeps the community from growing too fast for it's own good.

On the infrastructure costs, it has always been my opinion that these should be paid from the only existing places that money changes hands in the name of the MIDIbox, Mike's shop and mine.

I don't know Mike's opinion on this though.

On the other hand it could bring new people into the community who are experts in other areas, like software customization/programming, testing, sound design, documentation, etc.

Maybe it would only shift the support effort to another direction. Less issues with a non-working PIC programmers, wrong connections, bad soldering joints, missing MIOS upload request, but more questions about software or customization problems - the propability to get an helpfull answer increases with the number of people who are working with the same stuff.

I can't count the number of times I have -wanted- to tell a customer, self-etcher, or one of Mike's customers to just send it to me for troubleshooting/repair so they could move on to MIDIbox fun.  

Of course I don't have the time to repair MIDIboxes for free or pay, and if I wanted another revenue stream from bench work I would just repair another few stacks of Galaga arcade boardsets for $75 per hour. ;D

One of my planned future projects is an online table of failure modes/fixes, but abusing these pretty modules to the point of failure is something I'm having trouble pushing myself to do.  It's going to be an expensive process also, burning up parts so I can tell you guys what part I intentionally fried on mine to reproduce -your- issue... ;)

The idea of never having to re-explain the tricks -some- people have to use to get a JDM working on their PC, or having to tell a fresh builder to check their solder joints a fifth time does appeal to me.

My email is a constant flood of tech questions, ranging from the most simple to fiercely complex.  TK I can only image (read:fear) the volume of questions you get in a day!

It would be stupid to assume that the best programmers in any field are also always able to construct the hardware they write for.  That in mind, the idea of looking past the hardware and welcoming a new class of end users who will hopefully add to the functionality and community is changing my way of thinking about DIY as it relates to the MB.

This all reminds me of the separate approaches towards consumer PC's.

Apple - closed hardware (not available for license to 3rd party manufacturers), no Mac clones, but for the gain of near perfect SW/HW integration.  No clones means no competition though, so the price stays high (relative to a PC clone) and less developers chose the platform.  Lack of a large developer market causes less developer tools for the Mac, etc.

PC - open hardware (available for license to 3rd party manufacturers), thousands of companies worldwide making PC's and more making component modules, at the cost of very poor integration between wintendo and the hardware it runs on.  Competition causes PC's to be affordable, exponentially more PC's than Macs are bought, so the developers naturally focus on the platform.  

Not intending to re-hash the Mac/PC holy war, but the history and outcome of these situations serve to show us the possible results of whatever path is chosen.  ;)

If the path of "hardware as a commodity" and focusing on the software side of developing the MB is taken, we must be careful to build in some type of protection against the Bill Gates types of the world, who will undoubtedly show up ready to pad his pockets via the hard work of others.  

I wonder how much time the Behringer people spent soaking up info on this site before taking "their" control surface designs to Uli?

P.S.: reg the MBHP license - I don't like so much rules, they make the world too complicated

Understood and appreciated!  :)

Best!

SmashTV

[moebius]

Well,

TK you have your sight tightly on the horizon, right? Microchip seems to still produces all of it's PICs in DIP cases. And from my knowledge - The First development kit for ARM7 core was released a few months ago by Oxford Semiconductors.

Miniatyrisation and integration provides us with more and more powerful devices. Yes, SMDs and SOICs, even some LQFP (or whatever manufacturer calls it) are still workable with good skills and using basic hobbyist equipment. Ball Grid Arrays are plain impossible. (But, Yes I wouldn't put a first timer to solder SMDs)

This makes "pure" DIYing (whatever it is?) very difficult - more powerful devices, less components, more complex multilayer PCB's, which you have to order from a prototype factory anyway.

There are several examples of this new generation DIY in several degrees:

Miss Parker (as seen somewhere on our forum) - http://www.axoris.be/index.html

Basicly a dedicated DIY DSP board for (Alesis DSP based) audio Fx. 2 sided PCB, doable in diy. Going to have "a dedicated software" for creating the new FX.

2181 EZ-kit lite based DSP - http://www.gweep.net/~shifty/death/

Instructions on turning a Analog Devices ADSP-2181 development kit into a DSP based Synth/ FX unit. Basic DSP SDK available from AD.

Chameleon - http://www.soundart-hot.com/english/index.htm

A commercial, professional quality open-ended DSP unit with own SDK to create new applications.

Basicly what you see is, DIY is moving to the software side. And the hardware is moving away from the hands of a DIY.

--

Remember what happened to synths after DX-7? Those old fashioned knobs disappeared - For nearly a decade! Why? So what's the future of DIY like? Hmm. It looks bright to me! You can have a power to create a synth, FX unit.. anything as powerful as build by a commercial company... You can have those chips already. (there isn't a law YET to prohibit selling chips to open source / DIY community - those com*unist basta.. now wait a minute, it didn't go like that nowadays,  te*orists!)

I'm hoping to see DIYish designs, yeah, maybe meaning semi-modular designs - With embedded uC/DSP , on a well engineered PCB, with lot's of I/O functionality. These would probably be running a kernel (RT/uCLinux, anyone) supporting HW devices. Boards from dedicated manufacturer, usable as core for hardware network firewall to... yes, DIY projects.  There are already linux ports for Motorolas Coldfire and Analog Devices Blackfin. Your favourite POSIX compliant platform in a chip, What are you waiting for? Oh.. those cheapish boards to do development with ;)

Maybe dedicated SDKs as well. Or in case of linux - many skilled people already having all the needed tools for developing for linux.

Bye, Moebius

p.s. T.K said:

P.S.: reg the MBHP license - I don't like so much rules, they make the world too complicated

Sorry: Do you want to people to try, or to exploit your design in a commercial manner? Or is it just allowed? What is commercial then? Ripping of that 150E from a friend and doing debugging at the forum?

Please, let us know.

Bye, Moebius

[jimhenry]

This discussion is getting interesting. Are things to the point where an SMD version of the core module and other "mature" modules could be designed? Would Mike and Smash be willing to take on making fully assembled and tested modules available? Is this a positive direction for the MIDI Box?

I suggest SMD because it might make more room available for connectors and I think it is less costly to have assembled than DIP.

Perhaps this would be less work for Mike and Smash because they would gradually get out of the kit business which has to be a headache.

Perhaps this would be less work for the MIDI Box community because we wouldn't have to try and remotely diagnose assembly errors.

The DIY part would be selecting and interconnecting modules, loading, configuring, and possibly modifying firmware, and designing and connecting the user interface. Maybe that is the most important part of the DIY experience?

[Ian_Hurlock]

Way hey there guys.

It would appear, TK, that you need to come up with something new to build so we have something else to yarn about  ;) ;) ;)

Anyway, TK, is that a field producing gate array ?

FPGA

All the best you cheeky boys :o :o

Ian

[Xanboroon]

Hey Moebius, Thanks for the publicity for Miss Parker, at least one guy of the uCapps forum seems to have interrest in our "almost DIY" DSP stompbox... :-)

Miss Parker is indeed a good exemple of the futur of what DIY can be.

All specs, schematic and layout done by our team. PCB realised by an external professional company because of the complexity. But all components soldered by someone (I means not by a machine...:-) ). Ok, I must admit to be honest that in the team we are three Electronical Engineers and that the guy who has designed the layout and soldered the components on the PCB made use of the equipment of the company where he works and he is quite good at this...

See the empty PCBs: http://www.axoris.be/MissParker/pics/PCB_empty.jpg

See the same PCB but with component: http://www.axoris.be/MissParker/pics/PCB_populated.jpg

Looks really profesional, no ? :-)

But again, like lot of people, I think SMD is not impossible to solder if you have correct tools (not all the package but some of them) ...

But it's true that DIY will evolve to someting much more "professional" where you will need to train you skills.

I also believe that anyway, everywhere in the world there are a lot of people with so much imagination that solutions for DIY hobbyist will anyway pop-up whatever how the technology will evolve.

Modular approach with premade small modules that you can interconnect with a lot of possibilities are indeed a good solution and a step to ease the DIY for people that don't have acces to expensive tools or are not expert in soldering SMD components...

Xanboroon

PS: BTW, SmashTV, you can see that I'm not again premade PCBs... For instance the Miss Parker PCB is something that I'm really not enough competent to realise it myself...

[TK.]

Moebius Sorry: Do you want to people to try, or to exploit your design in a commercial manner? Or is it just allowed? What is commercial then? Ripping of that 150E from a friend and doing debugging at the forum?

Of course, I don't want that users get ripped off, but independent from the terms in any license, people will always find a way to bypass the rules. The circuits of MBHP are small and common, they can be easily modified to create something which is not covered by the license. It's only a question of knowledge and effort.

So, how can people be protected from such "bad guys"? We have to keep an eye on public platforms like forums, ebay, etc... and we can warn the potential customers if possible. The guys won't get support from us - thats all we can do. On the other hand: if somebody is able to sell assembled MBHP PCBs for a reasonable price, if he works together with Mike and/or SmashTV (in order to keep up the existing infrastucture), when he spends the same effort in customer support like SmashTV, then a collaboration is possible. As I mentioned before: in some years we possibly rely on such a service. And a inexpensive alternative solution is a good protection against overpriced ripped designs.

Ian: It would appear, TK, that you need to come up with something new to build so we have something else to yarn about  

Anyway, TK, is that a field producing gate array ?

FPGA

Yes, it's a field programmable gate array. Don't be afraid of being triggered with such a new wanna-have in the next 2..3 years - I think that the current platform is sufficient for my own plans until then. But once I feel frustrated enough about the limitations of a 8-bit microcontroller, and once I see that the use of state-of-the-art technology for non-electronic experts is feasible, I will switch.

A FPGA has the advantage that it is a very generic device, which allows to do a lot of very different things with the same hardware. It's so flexible, that very dedicated units can be implemented without much effort, which are doing tasks with best execution time (nanosecond accurate), e.g. functions for scanning buttons/encoders/pots, serving LED matrices, etc... but also complex waveform generators (-> synthesizers), audio processing units, sequencer engines or just serial interfaces (you want to have 256 MIDI In/Out Ports? no problem ;-) ). Even a customized CPU could be programmed into the array, but due to the implementation effort and number of required gate functions/memory, it's mostly easier to use a seperate microcontroller connected to the FPGA. The current disadvantage is, that development environments for FPGAs are very expensive, but this could change in the next years once FPGAs get more populated. My dream solution is a FPGA with integrated CPU, clocked with 100 MHz minimum, low-cost programming interface and free available verilog or VHDL compiler/simulation/synthesis tools - let's see... we are young, we have time :)

Best Regards, Thorsten.

[smashtv]

Too much to reply to!  But I'll try without creating another novel.;)

Smash, where are you located? I currently live in Montana myself, but I am from Ohio, so I know precisely what you mean...

Oklahoma, dead center of redneck land.  Don't get the wrong idea, not everyone here is stupid, just mostly the loud ones.  And there seem to be a lot of loud ones.....;)

I have heard some legendary stories of bored exploits of the angry gun-toting meat eaters in Montana..... :o

(there isn't a law YET to prohibit selling chips to open source / DIY community - those com*unist basta.. now wait a minute, it didn't go like that nowadays,  te*orists!)

hehehe we'll know the fun is almost over if we see Harman buying up chip plants.  You scared me there for a minute, I thought someone had changed the US's "enemy of the week" and nobody told me! ;)

Are things to the point where an SMD version of the core module and other "mature" modules could be designed?

Yes but this would be an expensive process.

(I use the word expensive here, but the thought is more than testing/prototype/setup costs, the largest cost is the amount of time needed to create new documentation)

Would Mike and Smash be willing to take on making fully assembled and tested modules available?

I can't answer for Mike on this, but yes.  

There are -many- variables to consider here though:

Is the community ready to support people with no electronics knowledge as they wire these pre-built modules together?

Will there be any market protection for me as a fabricator? (In other words, after I invest another $4k in prebuilt modules, Will I be forced to watch them rot on the shelf when someone who puts an unrealistic value on their time and can match or beat my price?)

Is this a positive direction for the MIDI Box?

No idea at this point.  I do like the idea of the focus shifting over to MIOS dev rather than hardware troubleshooting!

Perhaps this would be less work for Mike and Smash because they would gradually get out of the kit business which has to be a headache.

Personally I would feel wrong about -not- offering a kit for the current hardware.

The kit thing is a major headache, I spent about 4 hours last night splitting out and packaging 30 CORE kits and 25 DIN kits.  Not complaining though, I knew there was much more to it than buying parts when I started offering the kits.

Hey Moebius, Thanks for the publicity for Miss Parker, at least one guy of the uCapps forum seems to have interrest in our "almost DIY" DSP stompbox... :-)

I'm very interested in MP, You have had my attention since the first announcement here.

Hopefully one of these days I'll have time to build one join the fun!

PS: BTW, SmashTV, you can see that I'm not again premade PCBs... For instance the Miss Parker PCB is something that I'm really not enough competent to realise it myself...

Like I said before, a PCB is a commodity part, just custom made.  

Some of the PCBs I design here way outstrip my fabrication capabilities, so I have them made just like everyone else.  

To limit myself to what I can prototype in house would not be wise.

On the other hand: if somebody is able to sell assembled MBHP PCBs for a reasonable price, if he works together with Mike and/or SmashTV (in order to keep up the existing infrastucture), when he spends the same effort in customer support like SmashTV, then a collaboration is possible. As I mentioned before: in some years we possibly rely on such a service. And a inexpensive alternative solution is a good protection against overpriced ripped designs.

Cool.  Let me know when it's time for me to build some test fixtures.

I think that the current platform is sufficient for my own plans until then. But once I feel frustrated enough about the limitations of a 8-bit microcontroller, and once I see that the use of state-of-the-art technology for non-electronic experts is feasible, I will switch.

Very cool!  My one and only feature request for the first generation of FPGA hardware running MIOS is that you include the Pac Man Core. ;D

Best!

SmashTV

[Jack]

I bought it here:

http://www.fpga4fun.com/

My dream solution is a FPGA with integrated CPU, clocked with 100 MHz minimum, ...

Are you speaking from NIOS (not MIOS)  softcore processor? Or a real integrated cpu?

PS:I started last month with CPLDs and FPGAs prototyping and programming. It's really hard to understanding, programming and prototyping FPGA's!

Sorry, I don't want to start a new topic.

Jack

[Duggle]

hi Jack,

great link.

Its seems FPGA for DIY really is here!

This technology really fills a gap where h/w interfacing

requirements and speed proclude uControllers.

Also the softcore processor is really interesting. It seems the device necessary to support it may be a bit expensive at the moment. There are some really affordable starter kits there.

cheers

[TK.]

Are you speaking from NIOS (not MIOS)  softcore processor? Or a real integrated cpu?

no, I'm not speaking about softcores, they mostly allocate too much macro cells, or the performance is not good enough.

Best Regards, Thorsten.

[pay_c]

To come to a totally other point...

All this stuff made me think of an (really WEIRD) thing:

Let´s say somebody made an old Midibox (like that MBPlus or so) and doesn´t need it anymore. He is allowed to sell it on Ebay or so, right? He has to claim that there will be no official support and it´s based on GUI, but anyhow you could sell your stuff, right?

*NO*, I won´t! Naaaaah and I also don´t think many people would (it´s some thing *YOU* made not some company). But this thing just keeps in my mind...

Thnx in advance!  :)

[AndrewMartens]

no, I'm not speaking about softcores, they mostly allocate too much macro cells, or the performance is not good enough.

Best Regards, Thorsten.

I think it might be interesting (down the road) to look at one of those Pluto or Pluto-II boards and see how it might interface to a MBHP Core module - one can still use a PIC running MIOS as a real microcontroller interface, and then load a synth-only core onto the FPGA.  That would be one way to maximize the power of the FPGA without wasting macro cells.

The Dragon seems interesting, but at nearly 3x the cost of the Pluto-II, it's a little on the pricey side.