Author Topic: Project: Mega2560 to L7 board test.  (Read 61961 times)

kn4ud

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 753
    • View Profile
Re: Project: Mega2560 to L7 board test.
« Reply #420 on: October 09, 2016, 08:35:43 AM »
I like the word CLOSER...  :) :D :
Allen

Lob0426

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1627
  • Let the Magic Blue Smoke out!
    • View Profile
    • Richards RasPi Web Server
Re: Project: Mega2560 to L7 board test.
« Reply #421 on: October 24, 2016, 07:50:36 PM »
Update:
The board has been working for a couple of weeks now. It only had basic function. No Limit/homing switches or other options.

Today I added in (2) 2 pin JST-XH connectors for my homing switches. Now I just have to add the other end of the connectors to one of my machines.

T2 Laser appears to work ok with the Mini-Mega. BCL is having some issues. I believe it is the new reporting format that is on grbl1.1c/d. Zax got a little ahead as T2 was the one not working before. He cured the Data wait problem and then found the change in reporting needed to be addressed. For BCL Ralph was having it too easy so grbl decided he got the next curve ball.

Everyone should Thank Ralph and Zax. They have continued to work on these issues as we try to move ahead to a microcontroller/s with more resources. They could have decided it is just too much work, like other software developers do, and then we would not be able to move to better more capable equipment. I do appreciate their work as they really are an integral part of having a chance to finish this project! I think I have the easy part of soldering up test equipment.

NOTE: The full size mega adapted to an L7 board works really well for both T2 Laser and BenCutLaser. Those of you that followed the earlier part of this project know that that also put up barriers to being complete. If you have the wires, Mega2560 board and a spare L7. You now can have the newest version of grbl 1.1d, with Laser Mode running on your Laser machine. Thank You @parnz for porting grbl 1.1 to the Mega2560. You should know that there is no gain in speed with a Mega board compared to a Nano running grbl 1.1c. And J-tech is sometimes a 1 to 3 seconds faster. But without a source J-tech did not lend itself to adding advanced features to our machines.

Just today I was playing with CoreXY settings. One of our members is scratch building a CoreXY type machine. Unfortunately I found that the Nano is too short of storage to operate with CoreXY enabled. This leaves you with the choice of making one of the CNC controllers on an Arduino Uno board if you are trying a self built CoreXY machine. I tried the Uno with CoreXY and it came in 50+ seconds behind the normal time for J-tech on a Nano or grbl 2.2c with Laser Mode on Nano. I believe this is due to the mixing that takes place between the A and B motors. I may try this on the Mega2560 boards and see if they can run faster. They have larger buffers and more RAM which may actually help with CoreXY. They are still only 16MHz.
Richard
A5 2.5W w/homing switches on L7
A5 2W diode self upgraded from stock 500mw
Printrbot Play. Ext bed, integrated tablet control.
Printrbot Simple Metal
USA

parnz

  • Newbie
  • *
  • Posts: 23
    • View Profile
Re: Project: Mega2560 to L7 board test.
« Reply #422 on: October 25, 2016, 01:18:27 AM »


I'm planned to create my custom controller board as well as my machine's structure. It's all DIY.
I want to add jog keys, e-stop, homming, weak laser mode switch and wireless communication.


My experiment:
-with grbl 1.1d + 328P UNO (clone) and FTDI chip: the stepper motor move like when you accelerate a car or motorcycle (speed gradual increase/decrease till reach target; I don't know how to explain please excuse my English
-with grbl 1.1d + mega2560 (clone) and CH340 chip: stepper motor move in steady speed.


But I just break my Rx0 serial pin of my mega2560 as used as a programmer to burn ESP8266 firmware.
Breaking somethings is a part of fun of building somethings.


ps. I'll write my own software as well. Coding is easy you can find the sample code on the internet. But the hard part is design the gui.
You need to think like you are user. The flow of using your software and how convention to use your software.
Why do I care? Because I plan to release it as an opensource.  ;)
If you can't write the code don't worry I'll write it for you but at lease you build it yourself (hint: Do It Yourself).


Have a good day to all!  ;D

Lob0426

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1627
  • Let the Magic Blue Smoke out!
    • View Profile
    • Richards RasPi Web Server
Re: Project: Mega2560 to L7 board test.
« Reply #423 on: October 25, 2016, 09:13:56 AM »
Cannot wait to see what you have planned.

You explained it alright. Laser Mode still has Acceleration/Deceleration in operation. Grbl was designed for CNC. So spindle movements were heavily regulated by accel/decel. Laser mode turns off the accel/decel for EVERY little speed or direction change. But accel/decel is still operating. So without laser mode the accel/decel is much more dramatic. If you put it in and out of laser mode, and use really low accel values, you can see the difference.


I was playing with the newest release of grbl. I am calling it 1.1d4. It seems to play much better with BenCutLaser than the prior 1.1 versions we are using now. They changed spindle speed formula again. And moved stuff around again so the code cannot just be imported over to your Mega version easily. I guess this will continue until they release 1.1 for the public.
Richard
A5 2.5W w/homing switches on L7
A5 2W diode self upgraded from stock 500mw
Printrbot Play. Ext bed, integrated tablet control.
Printrbot Simple Metal
USA

parnz

  • Newbie
  • *
  • Posts: 23
    • View Profile
Re: Project: Mega2560 to L7 board test.
« Reply #424 on: October 25, 2016, 07:34:18 PM »
Github is a powerful tools.  ;)
All changes they made from gnea/grbl are merged.  ;D

Lob0426

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1627
  • Let the Magic Blue Smoke out!
    • View Profile
    • Richards RasPi Web Server
Re: Project: Mega2560 to L7 board test.
« Reply #425 on: October 25, 2016, 08:16:31 PM »
Alright now that is just sneaky LOL. I will give it a try. Hopefully it will help me get the Mini-Mega up with both software. Then it will be worth looking into a an upgrade board. We are looking at a possible plug in adapter for the L7.

The prototype is hand soldered. It has the Mini-Mega and connectors for Homing switches. The picture below is old (Like two weeks).

What processor/microcontroller type are you looking at for a custom controller @parnz?

The Mega does have enough extra I/O for your plans. It's core is running at about top now with grbl. I tried a CoreXY build and that cost 30% processor for almost a third more run time on the job.
Richard
A5 2.5W w/homing switches on L7
A5 2W diode self upgraded from stock 500mw
Printrbot Play. Ext bed, integrated tablet control.
Printrbot Simple Metal
USA

parnz

  • Newbie
  • *
  • Posts: 23
    • View Profile
Re: Project: Mega2560 to L7 board test.
« Reply #426 on: October 25, 2016, 08:38:04 PM »
I'd like to build an controller shield that sit on top of mega 2560 like the RAMP board :)
It'll be made specifically to laser machine :)

Lob0426

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1627
  • Let the Magic Blue Smoke out!
    • View Profile
    • Richards RasPi Web Server
Re: Project: Mega2560 to L7 board test.
« Reply #427 on: October 25, 2016, 08:47:00 PM »
That would be the way to go. I saw some boards that might be good for a prototype on Amazon. They are on Ebay too.

https://www.amazon.com/dp/B00Q9YBQ04/ref=gl_it_dp_o_pC_nS_ttl?_encoding=UTF8&colid=1M2534R54DE69&coliid=I7JHKTFS69BVA

Then from there you could design a board to be made!
Richard
A5 2.5W w/homing switches on L7
A5 2W diode self upgraded from stock 500mw
Printrbot Play. Ext bed, integrated tablet control.
Printrbot Simple Metal
USA

ggallant571

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 2913
    • View Profile
Re: Project: Mega2560 to L7 board test.
« Reply #428 on: October 25, 2016, 09:04:31 PM »
I designed a 3 axis controller board that support both Nano and Mini. I/O pins for Mini were chosen to minimize trace so firmware will definitely require tweaking. Order was shipped on 10/17 from China via Hong Kong Post. It does not have tracking info so I am in the dark as to where they actually are. You are welcome to the design files.
SARCASM - Just one more service we offer here.

Lob0426

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1627
  • Let the Magic Blue Smoke out!
    • View Profile
    • Richards RasPi Web Server
Re: Project: Mega2560 to L7 board test.
« Reply #429 on: October 25, 2016, 09:32:58 PM »
We should be able to redo cpu_map for grbl to operate on that board too.

If I were to solder the Nano on the board and mount the Mini-Mega on header risers could both boards be stay on the board? You would only be able to plug one into USB at a time.
« Last Edit: October 25, 2016, 09:35:59 PM by Lob0426 »
Richard
A5 2.5W w/homing switches on L7
A5 2W diode self upgraded from stock 500mw
Printrbot Play. Ext bed, integrated tablet control.
Printrbot Simple Metal
USA

ggallant571

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 2913
    • View Profile
Re: Project: Mega2560 to L7 board test.
« Reply #430 on: October 25, 2016, 10:21:23 PM »
I suspect not as the 3.3V would be routed to both boards.
SARCASM - Just one more service we offer here.

Lob0426

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1627
  • Let the Magic Blue Smoke out!
    • View Profile
    • Richards RasPi Web Server
Re: Project: Mega2560 to L7 board test.
« Reply #431 on: October 26, 2016, 10:16:31 AM »
My thought is;
If you need to fix something on the board, could you add a diode to prevent 3v3 from feeding across? I was thinking that a Nano with Benbox could be soldered onto the board. Then people could use the original software and the newer stuff would be used on the Mini-Mega.
Richard
A5 2.5W w/homing switches on L7
A5 2W diode self upgraded from stock 500mw
Printrbot Play. Ext bed, integrated tablet control.
Printrbot Simple Metal
USA

ggallant571

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 2913
    • View Profile
Re: Project: Mega2560 to L7 board test.
« Reply #432 on: October 26, 2016, 03:47:37 PM »
First, I should have said 3.3V and 5V. Second, just speculation, is that applying a 5V signal to an I/O pin on the unpowered chip may be enough to power it.

I have always planned on sending you a blank board and they are cheap enough to play with.
SARCASM - Just one more service we offer here.

Lob0426

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1627
  • Let the Magic Blue Smoke out!
    • View Profile
    • Richards RasPi Web Server
Re: Project: Mega2560 to L7 board test.
« Reply #433 on: October 26, 2016, 05:09:48 PM »
The Nano and the Mini-Mega both have a regulator at Vin. External power if you are not connected USB is supposed to be applied at Vin.

Both boards 3v3 are out. I do not know if they are protected from IN. 5V the same. The trick would be to figure out how to set some diodes, so only the processor that is plugged into USB is powered. There is not a lot of room in there to set a diode at each boards connection to the 3v3 trace. Does anything on the controller board itself use 3v3? It is just brought out because it tis there? Look at your L6 schematics it does not look like it is connected. 5V is connected VDD.

We might find out. Solder a Nano on and then plug in a Mini. Power the Mini and see what lights up! Losing a Nano would be worth it. Losing a Mini would suck!
Richard
A5 2.5W w/homing switches on L7
A5 2W diode self upgraded from stock 500mw
Printrbot Play. Ext bed, integrated tablet control.
Printrbot Simple Metal
USA

parnz

  • Newbie
  • *
  • Posts: 23
    • View Profile
Re: Project: Mega2560 to L7 board test.
« Reply #434 on: October 26, 2016, 05:37:59 PM »
I assume all pin of both chip will be connected to each other right?
D1 of Nano connected to D1 of Mini Mega and so on except only Vcc pin.
If you power on the Mega and set pin D1 to hight then +5v will apply to D1 of the Nano and it will be fired even nano isnt powered on.


Sorry if I get you wrong.  :)