Welcome! Log In Create A New Profile

Advanced

RAMPS for Due!

Posted by bobc 
Re: RAMPS for Due!
January 10, 2014 08:00AM
Can some of you guys that are in the know update the ramps-fd wiki?

v1A is out there now.. so It would be nice to have some details on the wiki... especially the faults found and any work arounds etc.

Thanks
Re: RAMPS for Due!
January 10, 2014 06:50PM
updating the schema would be nice too smiling smiley
Re: RAMPS for Due!
January 11, 2014 06:44AM
and I am glad to say that mine doesnt have ambro's issues on the E0 dir and z min pins. smiling smiley waiting on another stepstick and my extruder motor so I can say for sure.

I am very happy with this. it seems to work perfectly on my udoo smiling smiley so thanks bobc smiling smiley
Re: RAMPS for Due!
January 11, 2014 09:00AM
That's good to hear and thanks!

I've been updating the layout to use the AREF signal for the thermistors, also a few other minor changes. I replaced the 1N4004 with a 10A rated package. I added a mosfet to control the LED, the Due pins can't source much current. I've also tried to beef up the copper for power traces and ground plane.

Looking through the list of issues, I think we have covered nearly all of them? I haven't found any larger resistor arrays that might be suitable, we could consider using 0805 instead, although that would take up more board space and might be difficult to fit.

So unless there are any more calls for changes, I think we should try to freeze the design and look at getting some more PCBs made up. The Chinese spring holiday is Jan 31 to Feb 6 this year.


What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license
Re: RAMPS for Due!
January 11, 2014 06:59PM
If you think we're there, then lets go for it!

I'll spend the next few days looking over things to see if I can see any gotchas.
Re: RAMPS for Due!
January 11, 2014 09:07PM
Ok, I have committed latest batch of improvements. I have discovered I am somewhat OCD when it comes to having things lined up, oh well smiling smiley

I did a quick test of reverse polarity using the v1 A design. Applying 12V reversed to the board, both PSUs I used provide power very briefly then cut out due to overload. The 10A fuse did not blow, and there appeared to be no damage to the 1N4004. On testing with a meter though, the diode has failed short circuit, this seems to be a common failure mode for silicon diodes. I should really get a scope on and see how long power is applied.

I don't think we could rely on the diode failing short circuit. That and a combination of slow acting PTC is possibly a source of problems in RAMPS. Hopefully the new design should be more robust.

I had a choice between a TO277 package and DO214AB. They are about the same size, TO277 might be harder to solder as it has a pad under the chip, but seems to have a wider choice of parts although there is not a lot in it. I decided to go with the TO277.

Edited 1 time(s). Last edit at 01/11/2014 09:07PM by bobc.


What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license
Re: RAMPS for Due!
January 11, 2014 11:15PM
Quote
bobc
Ok, I have committed latest batch of improvements. I have discovered I am somewhat OCD when it comes to having things lined up, oh well smiling smiley

NO COMMENT! winking smiley

I'm much the same.

Quote
bobc
I did a quick test of reverse polarity using the v1 A design. Applying 12V reversed to the board, both PSUs I used provide power very briefly then cut out due to overload. The 10A fuse did not blow, and there appeared to be no damage to the 1N4004. On testing with a meter though, the diode has failed short circuit, this seems to be a common failure mode for silicon diodes. I should really get a scope on and see how long power is applied.

I don't think we could rely on the diode failing short circuit. That and a combination of slow acting PTC is possibly a source of problems in RAMPS. Hopefully the new design should be more robust.

I had a choice between a TO277 package and DO214AB. They are about the same size, TO277 might be harder to solder as it has a pad under the chip, but seems to have a wider choice of parts although there is not a lot in it. I decided to go with the TO277.

Sounds fair re: RAMPS. All it takes is the diode to fail open circuit and everything gets >0.6V backwards across it.
Re: RAMPS for Due!
January 11, 2014 11:16PM
bobc - there is nothing wrong with having OCD when playing with this stuff. grinning smiley

Put me down for a v2 board when you get this one made. I haven't had the time to solder up the first board - too busy with work/refitting my printers/shapeoko/etc to make a go of it. sad smiley

Email or PM me with the particulars when the board goes to fab so I can get you some cash to offset your development costs.


- akhlut

Just remember - Iterate, Iterate, Iterate!

[myhomelessmind.blogspot.com]
Re: RAMPS for Due!
January 11, 2014 11:32PM
I'll be getting some boards fabbed as well (well, 10 boards), which will be delivered to me in Aus. If anyone in Aus wants one, I'll probably have ~7 I can pass on (1 is already taken, 2 are for me).
Re: RAMPS for Due!
January 12, 2014 10:25AM
does the sainsmart due have any compatibility issues with this board?
Re: RAMPS for Due!
January 12, 2014 11:00AM
Quote
turutk
does the sainsmart due have any compatibility issues with this board?

I'm using a SainSmart with a board from Geeetech successfully. My SainSmart board itself is a little bit defective though. When powered on there's a high chance it won't start, and needs the reset button. Similarly, when done uploading firmware, it never resets, and always needs the button.

But otherwise it's working. I also have an original Arduino Due, with no reset issues.

Edited 1 time(s). Last edit at 01/12/2014 11:01AM by ambrop7.
Re: RAMPS for Due!
January 13, 2014 07:42PM
haha sounds like its going to be a busy February then! Polyphasic sleep does help but every year santa fails to bring me more hours in the day; maybe I am just too naughty smiling smiley .... But seriously thanks for all the hard work on this project! I am trying to work up my circuit design theory so hopefully I can contribute meaningful input down the track (lol just saw that pun).

I can certainly solder up and troubleshoot in the meantime, though with all the effort of revision and many eyes looking over the design I'm sure it will be minimal. So Cefier if im not the 1 already taken +1 grinning smiley

cheers
Re: RAMPS for Due!
January 16, 2014 04:38PM
Ok, I hope that the lack of issues on v2 means that no one has found any smiling smiley An ordering snafu means I haven't been able to prototype all the new changes, but I think they are low risk.

So I will aim to send gerbers this weekend, and I may get PCBs back before the Chinese New Year. (Year of the Horse btw)

Thanks to all for your input!


What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license
Re: RAMPS for Due!
January 16, 2014 05:14PM
I also want to try this board if possible. I will be getting an Arduino Due soon.
Re: RAMPS for Due!
January 16, 2014 06:06PM
I haven't had a chance to look at the board due to the Melbourne heatwave. Was 31 Deg C on Monday, 41+ Deg C days since then (Today is supposed to be 44), and most nights not getting below 25 Deg C. Temp is supposed to drop tomorrow, so I'll have a look at it then.

So I thought what the heck I'd look at it now.

Only thing that I can see is a non-issue as long as we document it:

The two Due mounting holes on the Power in/FET output side of the board have ground and a +VE voltage right near the hole (one per side). IMO this is a non issue (it's been that way for ages, and same on RAMPS) but we just need to document it somewhere that under no circumstances should you use metal stand-offs/screws, otherwise you'll get a short. With the other holes you're only shorting a signal (which is still not great, but not potentially fatal to your PSU/Fuse/PCB tracks/Arduino), so it's not a problem. I only mention it because I had someone locally do exactly that with a RAMPS board, and then couldn't figure out why it wasn't working and their PSU kept tripping. Fortunately their PSU tripped and shut off every time, rather than the PTC frying itself.

I'm happy with the board as it looks, and the ground/top fill now is much more even and consistent. Noticed the mods for the pads for the 100uF filter caps too - nice!

BTW: The thermistor inputs only have "0 1 2 3" next to them. Could perhaps we also place "Therm" next to them somewhere? I'm sure someone will get confused otherwise.

Edited 2 time(s). Last edit at 01/16/2014 06:29PM by Cefiar.
Re: RAMPS for Due!
January 17, 2014 10:28AM
Yeah, not weather for playing tennis smiling smiley

You raise a couple of good points, I will look at those.


What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license
Re: RAMPS for Due!
January 20, 2014 12:40AM
FWIW: I didn't get an order in for any boards yet, and Chinese New Year has basically put a halt on any new boards out of the place I order from. They won't be shipping anything out till after Feb 7th (more likely later, since I haven't ordered yet), so I won't be getting boards till after that time.

Sorry for the people who asked if they could get them from me.

Note: To the people who sent me email - are you expecting blank PCB's, or finished boards populated with components. I can probably ship you a blank PCB for nothing, but if you want a completed board, I'll have to pass on the cost of parts.
Re: RAMPS for Due!
January 20, 2014 11:30AM
Quote
Cefiar
FWIW: I didn't get an order in for any boards yet, and Chinese New Year has basically put a halt on any new boards out of the place I order from. They won't be shipping anything out till after Feb 7th (more likely later, since I haven't ordered yet), so I won't be getting boards till after that time.

Sorry for the people who asked if they could get them from me.

Note: To the people who sent me email - are you expecting blank PCB's, or finished boards populated with components. I can probably ship you a blank PCB for nothing, but if you want a completed board, I'll have to pass on the cost of parts.

...and I'll only be doing 1-2 of these fully populated boards for people - I simply don't have the time/money/patience/equipment to do more than about 2-3 boards total on a run.
Re: RAMPS for Due!
January 22, 2014 06:18AM
Hello and super compliments to everybody for this board!

I'm here to ask if is it possible to use this step driver with the RAMPS-FD: [www.massmind.org]
Here you can see it working with the RAMPS: [www.youtube.com]
My goal is to reach the 1/64th microstepping, I asked to James Newton how to cable it on a RUMBA/RAMPS and he told that I only need to cable Enable, Dir and Step; setting 1/64th micro with the jumpers on the stepdriver board (not on the RAMPS).
  1. Do you think it's possible?
  2. What are your suggestions on raching 1/64th with the RAMPS-FD?
  3. Is it true that 1/64th cannot be reached on classic Arduino (16Mhz) due to the low MCU's power calculation?
  4. Or is something related to the Firmware?

Another point, I need to control some servos and other stuff for a project that I've in mind:
  1. Is it possible to connect another Arduino trhu the SPI port of the RAMPS-FD?
  2. If so, can I use the second Arduino board, connected trhu the SPI, to control servos and other things during the printing process?

Thanks to everybody smiling smiley


---- "Let me make my move" ----
Re: RAMPS for Due!
January 22, 2014 06:51PM
Quote
Gordie
Do you think it's possible?

The step/dir/enable lines on RAMPS-FD are all 3.3V. You driver requires 5V. You will need to provide level translation to get your signals levels up to 5V.

The A498x and DRV8825 drivers both support 3.3V operation for the control side, so they work as-is.

Quote
Gordie
What are your suggestions on raching 1/64th with the RAMPS-FD?

Level shifting the signals to 5V if you want to use that driver, or find a driver that does 1/64th (or better) and supports 3.3V logic.

Quote
Gordie
Is it true that 1/64th cannot be reached on classic Arduino (16Mhz) due to the low MCU's power calculation?

No. the 1/64th has nothing to do with the power in the CPU. It's all about the expectations of speed that people have for a machine. If you want 1/64th steps but want it as fast as 1/16th, then the CPU needs to work at least 4 times as hard, as it needs to generate 4 times as many pulses to achieve the same distance over a set time.

At 16 Mhz, we're pushing the little Atmel MCU's near their performance limit, especially with all the other stuff going on in the background, to achieve the speeds people "seem" to expect. With an ARM, we should be able to do more. The Arm also can do some of the other more intensive code (that isn't directly related to steps) faster as well, leaving more time to do steps.

Also, the driver itself needs to be able to accept that many pulses in that same time frame. Many of the TB drivers seem to require the pulse to be "on" for a lot longer than the A498x's and DRV8825's, which means that you can't send anywhere near as many pulses.

Quote
Gordie
Or is something related to the Firmware?

One thing that will be related to firmware is that if you're using TB type drivers, they send to require a longer on-time. Some firmware out there supports this, some does not. Your mileage will vary.

Quote
Gordie
Another point, I need to control some servos and other stuff for a project that I've in mind:
  1. Is it possible to connect another Arduino trhu the SPI port of the RAMPS-FD?
  2. If so, can I use the second Arduino board, connected trhu the SPI, to control servos and other things during the printing process?

Thanks to everybody smiling smiley

You will need to write the code on both sides, but yes, it should be possible.

This method is how bobc plans to add support for more than 6 steppers (adding another board connected via SPI). Note that you need to provide some sort of "clock" to synchronise the signals if you care about everything working at the same time. This mainly would apply to steppers, but even with servos, it may be necessary (depending on what exactly you're doing).

That said, RAMPS-FD provides 4 Servo outputs (though they're only 3.3V), so if you don't need any more and the servos will operate from 3.3V, you won't need to take stuff off board.

Edited 1 time(s). Last edit at 01/22/2014 06:52PM by Cefiar.
Re: RAMPS for Due!
January 23, 2014 05:24AM
Thank you Cefiar your answers were very clear! I'm not a genius at electronics, I like to solder and build little circuits but I understand what you said.

So for the logic leveler should I use a circuit like that, right? [www.sparkfun.com]
Can I use the same circuit with the servos logic? You told me about the 4 outputs at 3.3v.

[Edit] I forgot to tell you, about the latency of the TB driver, in the comments of the YouTube video there's an explanation about using the capacitor to level the micro timing signals, also they talk about Marlin.

For the expansion board, you said what I had in mind; write the code into both of MCUs boards. I need to command also electromagnets that works with 12v, my idea is to use an Arduino board (connected thru SPI) with a relays shield to turn on/off magnets' states.
Thanks again.

Edited 3 time(s). Last edit at 01/23/2014 05:56AM by Gordie.
Re: RAMPS for Due!
January 23, 2014 05:26PM
Note: Those SparkFun converters you linked are uni-directional, so you may find that of the 4 channels, you only 2. For each stepper you'll need 3 signals (Step, Dir, Enable), and the signals all originate from the electronics (eg: RAMPS-FD) to the steppers (so they all need to be 3.3V in, 5V out).

There are bi-directional ones where you can use each signal in either direction, which will definitely be more useful (as you need less converter boards). That said, the Sparkfun ones are cheap, and while they don't do 5V->3.3V very well, you won't be using that part at all.

With regards to the servos, as long as the servo outputs are driven PWM or duty-cycle, then yes (basically means they're only using a digital signal).
Re: RAMPS for Due!
January 23, 2014 07:56PM
If we are talking about the THB6560AH, then according to the datasheet it has a logic threshold of 2V, so you shouldn't need a level converter BUT the circuit for that driver board [www.massmind.org] has pullups to 5V on the control lines. I'm not sure why, because the ENABLE is active high, and there are internal pull downs in the chip. Perhaps some control boards use open collector outputs? If you removed those pullup resistors, you would not need level conversion. Ah, that might be difficult they are in resistor networks sad smiley. That board is not designed to be 3.3V friendly!

Also the min step pulse is 2.3us, so that would be easy to do in firmware, some firmwares support this already.

I was planning to design some header boards to make it easier to connect external drivers, if I remember I'll do some next time I order some PCBs. It would be easy to do with piece of stripboard of course.

I'm not sure whether servos will require a level conversion or not, reminds me I should test that smiling smiley


What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license
Re: RAMPS for Due!
January 23, 2014 10:46PM
Could you use an ADuM3440 for level translation to external stepper drivers?


- akhlut

Just remember - Iterate, Iterate, Iterate!

[myhomelessmind.blogspot.com]
Re: RAMPS for Due!
January 24, 2014 04:40AM
Thanks for your answers mates. And I hope my questions aren't OT. If so I'll open a new thread.

bobc: Well I'm happy that my question gave you a reminder about the external connection of step drivers. Also the check about servos.
On the THB6560AH side, I've found out the original author of that board which luckly is in Europe (I'm italian) so I can buy easily. He sells that board in two versions "traditional" and "special", they have small differencies from the one I linked above.
Quote
Traditional version:
- Larger PCB (50 x 100mm).
- 4 current sense resistors to reduce their power and heat dissipation.
- Amplifiers on the output signal of the optocouplers, this will allow the drive to function properly with low current input signals.
- Seperate ground plane for signal and power.
- On board space for voltage regulator heatsink.
- On board space for THB6064 heatsink.
- Easy mounting with 4 fixation holes in each corner.
Special version:
- Smaller PCB (50 x 50mm) reduces cost.
- Driver chip is located outside the PCB for easy mounting on the heatsink.
- Fewer or combined components further reduce cost.
- 2 current sense resistors.
- Unlimited heatsink size for easy chip cooling.


Here is the site with schematics and more: [users.skynet.be]
Here is the schematic of Traditional, which I will buy: [users.skynet.be]
So, dear masters, which components do you suggest to change/remove to give it compatibilty to the RAMPS-FD? smiling smiley
I choose this Toshiba component because I was looking for a good Ampere driver, I think that this step driver is good for parallel wiring the step motors as it give up to 4A. Mine RepStrap has dual step motors system on the Z and X axes, so I can save one or two step driver outputs on the controller board. My plan was using the RUMBA or RAMPS and I didn't think about 3.3v when I saw the RAMPS-FD. That's the reason I asked my first question. Thanks for your good work on RAMPS-FD!


---- "Let me make my move" ----
Re: RAMPS for Due!
January 24, 2014 10:27AM
I am interested in powering servos from the RAMPS-FD, but I did not take the 3.3V logic level into accout. So I am wondering if it is a good idea to run them directly from the RAMPS-FD. I am looking forward to your results with servos!

I read once that it is advised to have a separate power supply for the servos (I think they can draw a lot of current in a very short moment). Moreover they can be driven with 6V (some high-voltage servos also with 7.4V) for more power and faster response, but 5V should also be fine (for most scenarios). Do you think making a small extension-board (quite similar idea to the EXT-4 or EXT-6 board from bobc) with a level-converter for the pwm-signal and a separate power supply (maybe just a line regulator with some large caps) is necessary to get the best performance out of the servos?
Re: RAMPS for Due!
January 24, 2014 11:04AM
On servos, I did some research before and it appears most servos will accept a PWM voltage of 3-5V, so that should be ok, although I didn't find an official standard, it seems that 3.3V compatibility is quite common.

However, even small hobby servos might take 500mA, which is already getting to the limit of the 5V supply from the Arduino. I did make a mistake on the V1 board by providing the servo power from the main 5V rail. RAMPS has a separate 5V rail for the servos, which I didn't appreciate. Bigger servos run off higher voltages and use several amps, so you couldn't run those off a shield anyway, an extension board is a good idea.

In RAMPS-FD v2 the 5V can be supplied externally. On both versions 5V is only used on board for the endstops and is provided for external connectors (e.g. SD card etc).

On V1, there are a couple of options:

1. Cut the track to the Arduino 5V on the bottom side of the board between JP101 and the shield OR cut the 5V pin off!
Then provide external 5V power on one of the connectors AU3 or AUX4.
You can then plug straight in to servo headers.
Remember that without 5V power the endstops won't work properly.

2. Connect the servo to PWM signal and ground on the RAMPS-FD header, and provide power separately to the servos.


What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license
Re: RAMPS for Due!
January 24, 2014 12:04PM
Quote
Gordie
Here is the site with schematics and more: [users.skynet.be]
Here is the schematic of Traditional, which I will buy: [users.skynet.be]
So, dear masters, which components do you suggest to change/remove to give it compatibilty to the RAMPS-FD? smiling smiley
I choose this Toshiba component because I was looking for a good Ampere driver, I think that this step driver is good for parallel wiring the step motors as it give up to 4A. Mine RepStrap has dual step motors system on the Z and X axes, so I can save one or two step driver outputs on the controller board. My plan was using the RUMBA or RAMPS and I didn't think about 3.3v when I saw the RAMPS-FD. That's the reason I asked my first question. Thanks for your good work on RAMPS-FD!

Ok, that is useful info. I will order one of the boards for testing.

One issue that I nearly mentioned but didn't was the question of opto-couplers. It seems on the skynet boards, they run off 5V which is provided externally. The optos would sink about 12mA each at 5V, or about 7.5ma at 3.3V which is a bit on the high side for the Due.

The optos also add a small extra delay of 1-2us, which may need to be accounted for.

So I think there is a definite need for a header module with level converter and pulse stretcher. I would like to use the standard pMinMo 10 pin connector, but it takes up a lot of space, so might be tricky. A possibility is to use the 4 pin motor connector for enable/step/dir/gnd, and hope that no one will try to plug an external driver into a Pololu motor output.

Here is a pic showing the concept anyway.



The two pin header is need to get 5V to the header board if the external driver does not provide 5V (seems to be an optional feature).


What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license
Re: RAMPS for Due!
January 25, 2014 04:30AM
bobc: Yep, great!
Here you can find a video with the Pololu to PMinMO Adapter usable on the RAMPS and compatible boards like RUMBA: [www.youtube.com]
I understand what you had say, I think this step driver can be very useful for the dual motor system and, of course, for reaching the 1/64th microstepping.
Do you have any blog or place where you take any updates on your RepRap's experiments?
Re: RAMPS for Due!
January 25, 2014 09:07AM
@bobc: Ok, then I am going to try option 2 when it comes to testing servos.

@Gordie: I heard once that using a dual motor setup for one axis and microstepping is not a good idea, but I am not sure if this is also true for a dual motor + dual driver setup. Read more about this issue here.
Sorry, only registered users may post in this forum.

Click here to login