RepRap Interface Standard

Thanks to Lars, we have a start now:

RepRap Interface Standard

Feel free to go ahead and find commonly used interfaces, describe them qith text, pictures, CAD models, whatever, tack a number onto them and bring them into the wiki. For example, hotend interfaces, both mechanical and electrical, are a good candidate. In case the page gets too long, it'll be split into one standard per page.

Why?

Well, it's simple. With an interface defined, developers no longer need to limit their development to a single printer. Instead they can work towards compliance with the standard and rely on a well working fit on every printer which complies, too.

Introduction of norms and standards in the early 20th century revolutionized the whole industry. Or could you these days imagine to have nuts and bolts which fit to one particular device, only? Likely you can't, you simply have M4, M5, M6 and M8 nuts and the comfort to know to have a replacement for about every device on the planet.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

uh oh, you said the "S" word.

I'm a "standardist" for exactly the reasons you cite, so i fully support the idea.

Seen your edits, @Buback, they go into exactly the right direction. Excellent! Also I think they should become a bit more abstracted.

For example, for the interface definition it doesn't matter wether the heating element is wound wire, a wirewound resistor or some kind of cartridge. What matters is that it's resistor-type (opposed to e.g. induction heating, which requires specific heating voltage frequencies), that it heats with 12V and a maximum of, let's say, 5A. Whatever the hotend developer puts into place, it can be heated with an electronics delivering 12V and 5A.

And then there are 24V electronics. Likely we need a distinction similar to "complies with RIS 12V" vs. "complies with RIS 24V". Similar to how not every bolt is M8, there are different sizes.

Good start!

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Just a thought:

What if we just put the heater and temperature sensor into the electronics module?

They would still be mounted in the print head. So a print head would not contain the heating element or a temperature sensor. These components would be part of the electronics module. Therefore we would not have to distinguish between 12V and 24V.

The print head then would only need to have a mounting option for a heater and a temperature sensor and something that allows the cables to run to the electronics.

This way even induction heating would be possible (If the size of the heating element did not change). Users would have greater flexibility to choose from the available electronics kits. And print head designers would not need to know how heating elements and Temperature sensors work.

Is there a reason to have the heater and temperature sensor as part of the print head, other then that they are mounted inside the print head?

The same could work for end stops. The definition of end stops could be that they need to fit into a box of a defined size. That they have not more than 3 cables of a certain size going to them. And a defined spot for screws to attach them to the printer.

End Stops could then be like "sub modules" of the electronics module. A electronics module would be compatible with several kinds of end stops(micro switch, optical, magnetic,..). But the printer bode would only need to provide the place and the holes for the screws.

Does that make sense?

Quote
JustAnotherOne
What if we just put the heater and temperature sensor into the electronics module?

Then we had zero electronics complying with the standard, because heaters are typcially not part of an electronics. Standards sum up and abstract what's already there.

And there's nothing wrong with having more than one standard for a specific kind of module. For example, smaller PC housings demand for Micro-ATX mainboards while full tower housings demand for ATX mobos. As both, ATX and Micro-ATX are well defined, a user knows what fits where.

Quote
JustAnotherOne
End Stops could then be like "sub modules" of the electronics module. A electronics module would be compatible with several kinds of end stops(micro switch, optical, magnetic,..). But the printer bode would only need to provide the place and the holes for the screws.

Much better :-) Yes, it works along these lines. One standard for main electronics, one for endstops. Then an electronics demands for endstops compliant with RIS xxx.

Standards for commodity PCs are public. They're a nice read on how things can be standardized: [formfactors.org]

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I don't think that we should cut the printer into pieces to create the modules. I think is is more about putting stuff where is has the most intensive interface.

If it is OK to have the requirement of space and holes for the screws for end stops then why not have a mounting location for a heater on the hot end?

It can still be a Electronics kit containing 3 end stops of type optical and x heaters of type y and x temperature sensors of type z. So heaters and temp. sensors would be sub modules.

The current situation is that most printers are seen as a whole and not as modules. So the heater is in the box. I would not think of these printers as incompatible. If someone has a printer like this and buys a print head module that does not come with a heater device he will just use the heater he has with the old one. If he builds the complete printer from modules he would not care if the heater came in the print head box or in the electronics box. He will have to make sure that the heater can be used with the print head and with the electronics.

So if the interface of the print head is just a hole and something to fix the heater, then all heaters would match. So there would be no reason to combine a certain heater with a certain print head. On the other side not all heaters match all electronics. So the heater should be either a a sub module or included in the electronics module, right?

Something that may be worth defining for future electronics is endstop connections. Specifically the order of the connections:

+------+
|signal|
+------+
|GND   |
+------+
|Power |
+------+

Currently there are variations in existing boards, but it would be useful for future developments.

I have updated the Wiki page.

I removed the suggestions for good heaters, and the description of temperature sensors. That is good information, but I do not think that it should be part of the standard.

@samp20 I added your suggestion to define the end stop pin-out. I put the Signal in the middle. This way a micro-switch end stop, that only needs two pins would be compatible if a 3 pin header would fit.

As I got no objections I created the entries for the electronics sub modules.

Does it make sense to define separate Mechanic modules for Cartesian and delta bots? Or should the standard be so abstract that they both match the requirements?

@JustAnotherOne, I have a few arguments to put forward against your suggestion:

1. I am not aware of anyone using the VCC, Signal, GND standard. We would end up with 3 configurations in use instead of the current two: Signal, GND, VCC and Signal, VCC, GND
2. The Signal, GND, VCC can still work with mechanical endstops since they connect between Signal and GND. Endstops that use Signal and VCC will still need their own pulldown resistor, and thus a GND connection, since not many control boards have their own.
3. The current two configurations don't cause damage if the wrong type mechanical endstop is used. However if someone tried plugging a Signal, VCC, GND mechanical endstop into a VCC, Signal, GND header then VCC and GND will be shorted.

Your solution does have one good point in that it can connect a mechanical endstop to either VCC or GND since they are both adjacent to Signal.

Quote
JustAnotherOne
If someone has a printer like this and buys a print head module that does not come with a heater device he will just use the heater he has with the old one.

If every printer uses different holes for mounting the heater, it won't work. Making sure such tasks work out are the entire point of a standard.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Quote
samp20
@JustAnotherOne, I have a few arguments to put forward against your suggestion:
[...]

In general it's a good idea to look at what's already there. If a standard deviates from a de-facto standard, there should be very good reasons. Because, let's face it, it's very unlikely any of the electronics designers will put more than a few minutes of effort into being RIS compatible until the standard is really widespread.

Three of the most common electronics (RAMPS, Sanguinololu, Gen7) come with the same pinout, so I wrote down this: [reprap.org]

In this case, and IIRC, there was even substantial discussion about the ideal pinout, because it's different from the former Gen3 one. The goal was to avoid destroying the electronics by accidently inserting a plug the wrong way, which is given with GND in the middle.

Nevertheless I added a direction latch (what's the exact name in english?) against wrong insertion. Allowing wrong insertion is not exactly user friendly.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

@samp20, Traumflug:

Your solution with GND in the middle is the better choice. I wasn't aware of the discussion.

@Traumflug:
The Idea with the heater is that all print heads use the same hole for mounting the heater. Are there any reasons for incompatible heater mounting solutions on the print head?

First, thanks for your editing, @JustAnotherOne. Even if we don't agree on everything on the first spot, discussing various views is the best way to find good results.

Quote
JustAnotherOne
The Idea with the heater is that all print heads use the same hole for mounting the heater. Are there any reasons for incompatible heater mounting solutions on the print head?

Perhaps we should leave the decision of the exact heater type to the hotend manufacturer. What matters is just the connector towards the electronics and the mount towards the extruder. For best results one should glue the heating resistor/cartridge into the heater block anyways, so it isn't exchangeable.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

I agree that discussing is what this standard needs. And I'm happy to discuss all the points. The aim should be to find the best solution, that is probably not my solution.

I never heard of gluing in the heater. Is probably interesting to find the best glue for that. It must be able to deal with the heat and be a good heat conductor, at least better than air.

What I know is that it makes sense to have something that presses the heater to the side of the hole to increase the contact surface.

If the heater is part of the hot end then the standard must also define the min and max resistance of the heater. The min resistance value is needed so that the electronics will not be short circuited by the heater. The max resistance is needed for the minimum Voltage to drive the heater. It also helps detecting if the heater is broken.

I think it makes sense to have the heater as part of the electronics(for the reasons mentioned). But If there are reasons (as you mentioned) then I don't care. So if it would be beneficial to have the heater in the print heat then update the Wiki accordingly.

I think it is good if a standard contains the reasons for its definitions. So it would be good to have them in the Wiki, too.

Quote
JustAnotherOne
I think it is good if a standard contains the reasons for its definitions. So it would be good to have them in the Wiki, too.

Good suggestion to add justification notes to the standards.

Hot end grove standards:
I don't know why J-head, and E3D grove mount designs deviate, I would like to see only one supported by Reprap.

Edited 1 time(s). Last edit at 04/16/2014 05:13PM by A2.

Quote
JustAnotherOne
I never heard of gluing in the heater. Is probably interesting to find the best glue for that.

Answered this just a few days ago in the german forum: [forums.reprap.org]

Quote
JustAnotherOne
If the heater is part of the hot end then the standard must also define the min and max resistance of the heater.

Electronics side defines a given voltage and maximum current and the heater should be able to deal with this envelope. I'm just a bit struggling wether definition of the electronics side can go into RIS 1 or wether it needs it's own RIS.

Now, endstops are defined in RIS 2 and it looks reasonable to refer to this definition in RIS 1. Could be the same for heaters. Or whole extruders?

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Quote
Traumflug

Quote
JustAnotherOne
If the heater is part of the hot end then the standard must also define the min and max resistance of the heater.

Electronics side defines a given voltage and maximum current and the heater should be able to deal with this envelope. I'm just a bit struggling wether definition of the electronics side can go into RIS 1 or wether it needs it's own RIS.

Now, endstops are defined in RIS 2 and it looks reasonable to refer to this definition in RIS 1. Could be the same for heaters. Or whole extruders?

I think we need a definition of this. I did not create the RIS2. I think about the number as being the version. So having the end stop as RIS2 makes no sense in that context.

If the Number is the Module level (0=complete printer; 1=module, 2=sub module) then you are right that the whole print head would be a RIS1 and a heater as part of the print head would be a RIS2. The ugly part here is that then you have an interface from the Electronics to the heater. So that would be a interface between a module and a sub module of another module.

I don't see a benefit in using the number to signal if module or sub module. Do I miss something? If not then I would suggest to use the number as version number. So everything would be RIS1 until we do a incompatible change.

End stops have also a close connection to the mechanic and only a loose connection to the electronics. So here again it might make sense to put the end stops as sub modules of the mechanic. And again it would create a ugly module to submodule of other module interface.

So maybe we should forget about sub modules at all and have the heater and the end stops be normal modules. What does everybody think about that? We are here for discussion, right?