My Axoloti Plan

[strum]

Thank you for not being defensive about my suggestion.

It's up to you what you do of course but I would suggest working on breakout boards for the existing hardware and just leaving them as is, design a new board if you want but to break existing boards seems a bit futile to me.

I use a program called diptrace, I've looked at a few PCB design programs and it is by far the easiest to use I have found, it's free to use up to a certain amount of pins and very reasonably priced compared to other PCB design packages if you need the full program.

http://diptrace.com/

The first thing to build imho is a rack mounting panel or enclosure to mount the axo, some controls and a display in with some i/o jacks for gates/triggers and maybe the onboard ADCs and that's what I'll work on once I get the patch system I'm working on sorted out, I'll hopefully get around to working on some drawings by the end of this coming week and I'll post them.

I've made a panel already and have my Axo mounted on it. I have the jack sockets, encoders and display I just need to solder the headers on the Axo and wire it up.

I made a post with a link to the drawing of the panel and I think I might have posted some pics of it mounted in a rack case.

I'll draw up a range of panel mounting options and full and half rack enclosures and a standalone enclosure and post them hopefully in the next week or 2, you can take them and have them laser, plasma or water jet cut out of aluminium, plastic or wood or whatever you like if you like the concept.

[axoman]

Thanks for the link, although I'm already sorted in that respect. I just use Fritzing, nice and clean design, simple to use, yet still allows the design of custom components. They're across the water from me as they're in Germany, but I like the convenience of being able to design in Fritzing and have my designs be fabricated by them too. Really looking forward to that, seeing my own circuit board for the first time (hehehe). Although granted, it will be very basic stuff I design at first

Your suggestions are cool, and a lot of pople will be very pleased with them, but commercial developers still won't be because they're still using pre-arranged design. I'm lucky in that I know how to design, so your idea is no different to me than what Johannes is currently doing, where he's creating an Axoloti Control board with display and butons. I will buy one for sure, as will most Axoloti users I can imagine, but it can never get into a completed product I design cause it would be physically impossible to match it up to the physical aesthetical design of the products I have planned.

So to me, your idea is exactly the same as Johannes upcoming control board, it's a convenient way to help development that can also serve as a completed unit for non-commercial products.

Something you said though, reminded me about something I forgot to mention for the Axoloti Store. With an un-populated core available with breakout boards, two other obvious things to stock in the Axoiloti Store would be keybeds in various sizes and lengths, and to stock mod/pitch wheel units. You would be able to go to the Axo store and literally buy a complete synthesizer by adding the relevant selection of products to the cart and clicking the buy button.

Yeah I know ... I shall say no more but I just wanted to add that because I forgot it in the previous post.

But that would be true Axoloti, wouldn't it, to be able to go to the Axoloti store, the one stop shop to build whatever the hell you like (with both personal and commercial users in mind)

[strum]

I wasn't aware that Johannes was making a control board, I'll be interested to see how that looks but I've got a very clear idea already of the board I am going to build for myself anyway, if others like it then they are welcome to use the designs or use Johannes's

The more available options the better I think.

[axoman]

I detect some slight touchiness, but there is no need. I'm just pointing out that neither yours or Johannes projects are going to appeal to commercial develpers because both your products are 'pre-set' design. In other words, while you can configre what those boards control, you cannot configure the physical layout because everything is placed onto one board.

Use some genius and realise that both of you could make your projects far superior with one little change that will cost you absolutely nothing in production costs. All you need to do is design the board so that the display section, button section, and knob section can be seperated by cutting the board. Johannes already used this idea on the Axoloti Core for the MIDI section!

Do the same for the control boards and they'll appeal to a lot more users, and even small-scale commercial developers because you remove the 'pre-set' layout.

Axoman Industries - Thinking Harder, Getting Stronger!

[strum]

I detect some slight touchiness

That would just be your paranoia kid.

My designs will be entirely modular.

My first display, knob, button and jack socket system will not have a board at all, just point to point wiring into a header connector.

Why don't you start trying to make some of your ideas a reality.

[axoman]

That would just be your paranoia kid.

No, that would be fact, kid, as your new reply just confirmed.
You've got problems, real problems.

Why don't you start trying to make some of your ideas a reality.

No need for me to try, I already am, kid.

[strum]

We've all got problems kid, some more than others.

Perhaps you should be concerned more with your own problems and less with other people's

[axoman]

The only problem I have is getting by on practically zero - but I'm used to it.
You, on the other hand, have serious social interaction problems.

Get it sorted, kid.

[strum]

Lol, right, no worries, you have a nice day then.

Kid.

[axoman]

Here at Axoman Industries, every day is a wonderful day, kid. There's a playpen in the corner if you're bored, over there beside the visitor seating area.

[Ersatzplanet]

Wow, this thread degraded pretty quickly. The schematic of the AXO are available, just use what you want to use. As far as saying the multiple modular board concept "that will cost you absolutely nothing in production costs." is not exactly true. The added interconnections, whether they are wire or headers or whatever add cost. If you make it as one break apart PCB, you are wasting money making sections you may not use or sell. If you make it as separate boards, you have to buy quantities for cost savings and that means buying more and again having boards sitting around. It all depends on the quantity of boards being made. Most board houses charge by square inch and hole densities. If it is a non-commercial venture, or the board count is low, all of this matters little. There are many prototype houses out there that will happily make 4-5 boards for you fully populated and ready to test. I understand the concept of a modular design with separate PCBs that can be added to make a configured system that is custom made holds lots of appeal. Just look at the Arduino world with all its custom shields and add-ons. It is much easier to do if you never want to mount it to a front panel. If you do, then that has to be custom made for you configuration unless you want to go total modular (which is not a bad idea) where the separate control surfaces and interfaces are separate modules with their own front panels and interconnect behind the panel, but that is expensive too. I make Eurorack modules (Synthwerks, LLC) and we make a few modules like this - a main controller board and separate interface boards and configurations for it. The metalwork is unique but the PCBs are shared.

[strum]

What DSP platform do you use for your eurorack modules out of interest?

[axoman]

Kinda hard to reply without sounding argumentative, but I'm not being. Nevertheless, I can assure you it's possible to make a modular board for exactly the same cost as a non-modular one. If tooling requires CNC for the fixing holes and components, then cutting a slot in the board at the same time costs nothing.

The only difference in cost is to the end user who wishes to actually split the board into modules. In that case, yes, there is extra cost, but it's just wires and some solder, and the only people who would need to do that, are those who need a specific layout. It's far better for that feature to be there than not, makes all the difference to the possbilities of finished product design due to flexibility of placement and arrangement.

Cutting the board during CNC is a no-brainer when it adds such massively improved flexibility at no cost. Or if you have a company being pedantic and want to charge a little extra, still, it's not as if it would be enough to change the end price in an substantial way. So even with little cost, it still makes a massive difference.

The other way is to sell the controller board fully-populated apart from the display unit, buttons, and knobs. Put those in a packet sent with the almost-finished board so that we can hand solder them to the correct points on the board, without actually soldering them to the board.

Regards the thread degrading quickly, I am not surprised in the least.

[axoman]

BTW, I suspect Discourse forum software to be designed by Google and the NSA, so please don't use it or join any forum that uses it.

[Ersatzplanet]

We make controllers mostly so no DSP is used in our products. We use some PIC controllers but that is all. I am letting the DSP stuff be built by others with more chops than me. So far the FCC has not looked at the Eurorack market much but someday they will and I don't want to have a bunch of product that all of a sudden has to have thousands of dollars in testing fees to sell. There are many Eurorack makers that have gone the Raspberry Pi and Arduino route, and a few with DSP's in them and I don't know which ones but imagine with the margins as tight as they are in Eurorack manufacturing, that they are the usual suspects in the lower pice ranges. Not many DSP farms out there in Eurorack.

[Ersatzplanet]

Have you priced out boards before? I am not trying to be argumentative but I commonly buy panalized PCBs and the houses typically do charge extra for routing odd shaped boards and I would love to use ones that don't. They chalk it up to programming time. It is typically a one time setup fee unless different bits are used for different slot sizes, then there is an added tool-change charge. Where do you get your boards done? I would love to find a vendor that doesn't charge extra for this. It is typically cheaper to have boards scored instead but then the limitation is straight lines and continuous ones at that. They also charge extra for anything other than a round hole for the same reason.

[axoman]

It sounds like they might be milking it, Ersatz!

No doubt they all have their different ways of charging, but as long as you design it to have the CNC use the same tool for cutting grooves as another feature on the board uses, then it shouldn't cost you extra. That's why I said, if they want to be pedantic about it, and charge extra, just swallow the extra cause whatever it is, it cannot be that much if something gets cut during the same tool run as other parts of the board. It can all be done in the same run using the same tool unless you need to have very specific tools to cut bevels etc. There's no reason the same tool used for cutting holes, cannot also be used for cutting either slots or a row of tiny holes (to make a snap-off).

Even if you chose to go for scoring, the price you would have to add to the finished product is peanuts compared to the value it has as a finished product. I'd pay twice the price for the flexibility than to not have it at all. I've not had any boards made yet, but I'll just use Fritzing's fabrication service for it when I do. Makes no difference to them where I put my holes etc.

[urklang]

Just wanted to chime in and echo what @Ersatzplanet was saying. This is open source stuff. No one is stopping you from realizing your own vision with it! No sense in raging at each other or at Johannes. Johannes is pursuing his own vision and of course it's never going to be perfect for every use case.

BTW, @strum, I'd suggest KiCad for PCB design. It's free and open source.

[strum]

Thanks urklang, I've tried KiCad and Fritzing but find diptrace just so much easier to use, I haven't done anything big enough with yet to have to pay for a license anyway, I'll deal with that when and if I have to.

You know what it's like I'm sure when you invest time learning to use a program and can do things easily in it switching to something else, even if better, can be a major setback in your workflow.

I think there is so much to get out of Axoloti as it is and I want to focus on that before even thinking about making anything other than addon control, display and i/o boards which will be pretty simple, along the lines of arduino shields.

[axoman]

No sense in raging at each other or at Johannes. Johannes is pursuing his own vision and of course it's never going to be perfect for every use case.

Are you trying to cause conflict or what?

No one was "raging at Johannes", nor was anyone suggesting he stops pursuing his vision. I'm pursuing my own vision regardless of what happens with this or any other project. As always, my suggestions are of benefit to the project, not just me. I'm well aware that creating a business and a product requires my own work. I'm going to the length of creating my own design knob and button caps, so believe me, I'm well aware of how much work is involved.

Suggestions are just that, suggestions.