Help with wiring Makerbot v3.3 stepper drivers on RAMPS 1.4 board

I could really use some help sorting this out before I burn out more drivers. I'm trying to find out exactly how to hook up Makerbot v3.3 stepper drivers at 24v on a RAMPS 1.4 board. I know the board is designed for Pololus, but I'm trying to use the RAMPS board on a repstrap using NEMA. 23s. I also have successfully burned out a Gecko g251 and maybe even a second one. If anyone could take the time to tell me exactly which leads to hook up in the pololu slots, I would really appreciate it. If it helps, I have kept a log of my build here Blog
Richard

Edited 1 time(s). Last edit at 09/03/2012 10:43AM by RichardSalinas.

Pinouts for pololus can be found easy in an image at the bottom of each page there. I guess the image is this

You need to locate step dir and enable, and take them to the 3.3v drivers instead. To locate those you need the 3.3 source files or IC 3977 datasheet and physically follow the ic pins on the board. However i made you an image of the board and i added text for the pinouts of the idc connector. First top 2 pins are gnd. Ofc, besides step/dir/en interface, the 3.3 will need the power supply too.

Edit:
You should double check everything, including the above information. Good luck!

Edited 1 time(s). Last edit at 09/03/2012 08:55AM by NoobMan.

Thanks NoobMan:

If you could clarify a few more things please. Is the Reset connection ignored on the v3.3. On the pololu it's taken to sleep on the driver, so doesn't really go anywhere on the RAMPS board. This is where I find things confusing.

Also if you are familiar with the Gecko g251 driver, there isn't an Enable connection. Does the Logic Ground connection on the Gecko serve as the Enable? There is also a Reset connection too. Is it necessary to use?

If you can clear these points up for me, a lot would make more sense to me.

Edited 1 time(s). Last edit at 09/03/2012 09:25AM by RichardSalinas.

Specifically for movement purposes, step and dir are needed, where step is like a pwm signal giving the step rate, and dir when logic low moves to one direction (CW) and at logic high moves to the other direction (CCW-reverse). Besides this, there are other functions of the driver IC. Somewhat comparable from manufacturer to manufacturer, but should read datasheet of each IC for exact details of each.

Generally:

Enable is a function which disables the outputs of the IC. Rest of IC works, just its outputs are disabled.

Sleep is like a more generalised enable, where the entire IC (not just outputs) is put to sleep and thus consumes less power. Useful for low current situations. I dont think its used in reprap.

Reset is a part of the procedure, or a *state* of the ic initialisation. Sort of resets the ic to a home position, disables output, etc. Also partially works like a disable in this way, so the step signals are ignored until reset changes to normal state.


For A3977-3979 on the stepper driver3.3:

A3977 EN is active low (working) - high disables outputs. On 3.3 board it has an pullup, so outputs are off by default. So you need to actively use it: either use it as interface pin with uC, or otherwise to tie it to ground with a hard wire on the board itself. Reprap use this to disable the motors when not in use, to lower the general consumption. Stepper motors consume most when they are not moving, so using this makes sense.

A3977 Sleep is active low (disable chip) - when high enables chip. On 3.3 its tied up to high (5v) without even a pullup. So this is ignored by default. No action needed. I dont think its used in reprap.

A3977 Reset is active low (low puts it to reset state) - to be high in normal operation. On 3.3 board it has a pullup. No need to interface it. I dont think its used in reprap. Maybe makerbot interfaces it, since it had it there on the driver, but i am not familiar with their latest board and firmware. As for the pololu "being tied to sleep", its different perspective. Sort of speaking, the reset is not tied to sleep, better wording is both are tied to either 5v or gnd, depending on what case it is. Since its allegro chip i guess they are tied to 5V high.


For geko, or other manufacturers dunno: the functionality of each feature can be slightly different, e.g. one feature might work in reverse: enabled at high and disabled at low, whereas with allegro chips the same feature can be the other way around. Datasheet reading should clarify how each feature works.

Bottom line, you should only need STEP / DIR to control the steppers, but EN is also highly recommended on reasons of power economy and longer chip life (e.g. being used only when needed). So besides power efficiency which is a very good reason by itself, also the components with a certain lifetime, e.g. hour of use can benefit alot from the disable).

Reprap usually have the EN different per each motor. But one can hardwire the EN on extruder (tie to gnd). The X and Y at least can have the enable tied together. And the Z enable is the one which is mostly used, because Z only moves in between layers and it can be disabled while x and y moving because the Z rods are supposed to have enough friction to not let the position slip away. So Z enable should be different than x and y. But no need to bother with different enable setups, just as to figure its potential role. Coz if you want a different setup you would need to change pins the firmware, so no point to bother with it.

Basically, just interface the 3 pins: step / dir / en, and afaik you should be good to go.

Edited 1 time(s). Last edit at 09/03/2012 10:23AM by NoobMan.

quick search or the geko drive apparently there is a feature named "disable":

"DISABLE PIN
TERMINAL 11 Disable
This terminal will force the winding currents to zero when shorted to ground (TERMINAL 12).
Shorting this input to ground (terminals 11 to 12) forces winding currents to zero and stops all output switching activity. The G251 will continue totalizing step and direction inputs if any are sent. The power supply current drops to less than 15mA. The motor will return to its original position when the disable input is released if no step pulses have been sent and the motor has not been moved more than 2 full steps."

On a3977 enable for normal operation is low, and for geko its high, e.g. when its low it disables outputs. To make it work would need to change the firmware accordingly, to have the EN inverted in operation (e.g. the other way around).

But i understand you want to change the pololu with 3.3, so dunno where the geko fits in.

Thank you very much. This is what I needed explained. I think I can go forward now with more confidence.
Richard

I initially decided to use the Gecko drivers ( I bought 2) , but with my various problems decided to try the v3.3's and purchased 4 ( they are 1/2 the price of a Gecko) hoping that they would serve my needs. My reasoning was if I couldn't get it to work , I could fall back on using entirely Makerbot electronics. I really don't want to do that.

Edited 1 time(s). Last edit at 09/03/2012 10:41AM by RichardSalinas.

Btw, your blog doest seem to load for me, is it up? And do you already have the a3977 drivers or you make them diy-way?

Cheers and good luck!

For some reason the link didn't set up. I fixed it, so it works now. I'm posting it here as well. Blog
I have 2 Gecko g251 drivers that are probably dead. My 3.3 drivers are in the mail. I'm also waiting on my arduinos since they are fried because of a dumb mistake I made with my power supply. Apparently when s**t happens it can really happen hard. I hope my luck returns.

Edited 1 time(s). Last edit at 09/03/2012 10:49AM by RichardSalinas.

The stepper drivers are perfectly interchangeable as they interface with uC via TTL levels. Doesnt matter if you use geko or 3.3 or w/e else as long as they are interfaced properly in our case with dir/step/enable. So this way you could use any steppers with any uC electronics.

The steppers are more linked to the motors themselives, in certain ways, but 3.3 drivers are quite adjustable to a range of motors. Other than that good practice is to use exact same drivers and motors at least for X and Y.

Followup reading for 3.3 a little bit here and more on mr Nophead's blog

As per mr Nophead suggestions there are 2 good mods: first to limit the 5v current, probably to say ~80mA, e.g. a resistor followed by a cap (like rc filter). Alternatively to use the smaller 78L05 but i saw you plan to use 24v so probably better to have the 7805 with a resistor because the power dissipation will be larger.
Secondary mod to increase the decoupling capacitance from +V near the ic, there is a 100 nf cap that can be changed to 220nf or more. I added another 220nf to the top side of the ic and felt a difference right away. I am not sure, but perhaps the latest 3.3 drivers already have bigger decoupling caps so this might not be needed in your case.

Edited 2 time(s). Last edit at 09/03/2012 11:25AM by NoobMan.

Once again thanks. I'll definitely look into the circuitry and figure out if the mods are beneficial. I do plan to keep the stepper drivers and motors the same for all axes. I plan to use the typical pololu and NEMA 17 for 1 (and later 2 extruders).

Once again thanks. I set up 1 axis on the RAMPS board with a v3.3 driver and it works great! You can see my setup here: Blog

Hi again
Just checked your blog, and congratulations both on blog and getting the motors moving.
Just a few of my opinions about pololus and a3977. There is an old forum discussion here about pololu 's current ratings, from which it can be seen that they are not *really* 2A in any case. Probably they are more in the interval of practical 0.7-1A. The 3.3 using a3977 on the other hand is quite robust. I am using a3977 also and at 2A with sinks they barely heat at all. So imho: 3.3 > pololus by far, and also 24v is way better. So i'd say thats a great start. Btw coz 3.3 has those leds showing the state of ouputs you can test them without motor and just look at the leds instead, thats one thing i like about it - gotta love those leds pattern.

I also like the feature of having the LEDs on the v.3.3. I wasn't quite aware of how nice of an indication they are until I ran my test.If you can, could you link the old forum post for the pololu's current rating discussion? And thanks for the positive feedback. This is an ambitious experiment on my part to see what I can make and if it will work for me. Maybe others can learn from my meandering.

Hmm, took a shot but cant find that thread atm, seems lost in the depths of repap.
Anyway, maybe i give pololus too little credit. Maybe they can get a little higher current with a good heatsink, but thats not easy to fix. In comparison, on the few a3977 boards i etched myself i left space for 2 holes and fixed ~4.5cm alu heatsinks with m3 screws, which is fairly easy and good. The contrast is as it looks like. I admit i could be biased here.