2.5amp for nema 17's is ok, but I'm surprised your running nema 23's that low. You need 4.5 even for the 1.5nm versions really.

 

Beg your pardon they are rated at 3a. 2nm and above at 4.5a

Or do you mean 2.5a per channel, I won't lie, this is where I get a bit confused, I know nema 23's 2.5nm unipolar are 3a per phase. 4.6a parallel

I'd crank those pots right up and see if it makes a difference

 

Hey, Jeff,
I am starting a build and want to use the TinyG as well. Noticed you use a 24V Power Supply (I assume since the TinyG max is 30V).
Any problem driving the 425oz steppers with the TinyG. Any overheating? What did you do for cooling? Any issue with the slaved steppers driving the gantry? Do they ever get out of sync? Do you still love the TinyG? or would you recommend another
Board? While I am an Electrical Engineer, I am new at CNC.

 

Just started sourcing parts for the build, but the Canadian dollar is killing me at .78 cents!
Thanks for the awesome Aluminum Plates from SlyClockWerkz! Will the Xpro driver work properly with a nema 17 400 step, with x,y using the nema 23 200 step or should I just stick with the 200 step nema 17? Waiting for the back-order items to be available openbuildspartstore.com before ordering the rest. Unless there is a Canadian dealer for parts?

 

Me.
www.fuzztech.ca

 

Big order your way

 

Ha! After all the nice cutting and carving of the last days, it was prone to having trouble today. I was missing steps on the X axis while pocketing a job. After hours of testing and observing the machines behavior I came to the following theory.

While running in the Y direction, the X was long enough idle for the half flow power to kick in. So, when the X axis received instructions to move into the wood, it didn't have full power in time to make that move and got "stopped" in its way, hence missing steps. After returning to the XY zero I could see about 6mm was missing due to a couple of Y runs up and down.

As a check I run a full profile cut of the same shape (3 passes at DOC 5mm) for 2 times and had no issue at all! The machine returned to the XY zero spot on.

Anybody have an idea of how to address this issue?

 

Well, another driver gave trouble right after this one, so I'm done with the crap. Going to try some others I ordered a couple of months ago. Anyway, I would stay away from these stepper drivers: http://www.ebay.com/itm/TB6600-4-5A...Motor-Driver-Board-New-Version-/281151742275?

They are sold by many different traders on eBay and elsewhere. All the same. Total crap.

 

It may depend on the particular driver chip being used but I am not certain that it happens that way.
From the standby condition it is usual for the driver to return to full power at the instant of the first step pulse where current is increased before the motor is moved.
If you are losing steps due to lack of torque the remedy would be to increase the drive current to the motor but you should also perhaps check that your Velocity and Acceleration settings are not pitched too high.

Tweakie.

 

I think those drivers have an extremely bad reputation so replacement is the way forward.

Tweakie.

 

The settings on the drivers was good for Amps (2.7) and steps. Also got the confirmation from Planet CNC's Andrej that the switching speeds of optocouplers used in these drivers are to slow. So, I'll sell them to a Thai that wants to build his own CNC (and if you see what that means, you will understand that these crappy drivers will be the crown on his work!). LOL (a lot!!!).

The other ones are from Haoyu and seem to be better. We'll see. If that doesn't work I'll go for some Leadshines or maybe even Gecko. Tomorrow no cutting (only cables and wires) and getting these drivers operational.

Thanks for the input, @Tweakie.

 

That makes me sad. That is what I bought for my machine. I don't have it together yet to test.

 

@Nick W , it depends what you wanna do with your machine right after it is operational. If you want to tweak it, play around to get an idea about operating it etc. and gain experience with CNC, use them until they break and then replace them. If you want to go a.s.a.p. into cutting more serious stuff without getting stopped (like me) by this kind of trouble, go straight for better drivers.

Most of us want to go cheaper than the professionals because of budget reasons. Totally understandable, especially given the fact that there are offers for cheap drivers all over the place. Makes one wonder why buy expensive drivers if cheap ones can do the work too? It's due to the relative inexperience with these machines that one can think like that and disregard the opinions of the more experienced and seasoned CNC machinists. Listening to them will cost you more money, but as it turns out, probably less than spending x times $15-20 for drivers that break down all the time. I bought a batch of 4 and than again a batch of 6 of these crappy drivers. To have some spare in case of. So, now I have spend $170 on a bunch of drivers that turn out to be useless on the little bit longer run. Add to that 6 x $ 16 for the other drivers ($ 96) and I'm already at $300 including post and packaging. I reckon for that money you'll have a Gecko 4 axis driver set.

I think it would be good if the team of OB would put more emphasis on going for quality parts when it concerns the electronics. Of course it will drive up the price of the total project, but the result will pay itself of and lead to more satisfied builders with less problems. It can also prevent builders from going astray and have to learn the hard and costly way what is good and what not.

Ah well, I'm of now to replace the drivers for the new ones. Wish me luck.

 

which drivers are you going to use now?

 

These ones http://www.hotmcu.com/tb6600-1-axis...l?cPath=7_30&zenid=7h0gfih59sfgr61lvjsj9g9kn5 since I already had them lying around for months. They seem to be reasonably good, for the price. Possibly less trouble but I don't expect them to be completely trouble free. If so, I'm a lucky machinist.

 

TB6600 stands for Toshiba 6600 (I believe). The stepper drivers are based upon that chip, an older version is the TB6560. The ones I took out are some no-brand Chinese drivers of which you can find the picture and details in the link to eBay I posted earlier.

The ones I'm using now are the Haoyu, also Chinese but of better quality apparently.

The problem with the previous ones is, that they are simply to slow to process the instructions and start to loose steps when you wanna go a bit faster. And that already happens with speeds above 800mm/min!

 

Maybe I'm lucky, but I have a 4 axis TB6560 based board for over a year and I'm happily cutting over 4000mm/min, no problem yet

 

Anyone have a view on the maximum torque motors that a stock OX can actually make use of?

The GT3 belt must limit it?

 

Im running close to 400oz steppers on my OX and cutting aluminum and unhardened tool steel. Just have to get your speed and feed right.

 

What speed/feed do you use for aluminum?

 

What voltage and current are you putting through them?

 

Running 5 amps through the steppers. I think the im using 15k rpm and 450mm/m.

 

Don't you get the annoying slow tick when step jogging?

They Both use pull up resistors to get the max current out of them, for very short periods, not enough for nema 23's. Avoid toshiba chips. You want mosfet micro stepping drivers.

 

Got any example (video or so) of that slow tick? Do you have a suggestion for mosfet micro stepping drivers? Thanks in advance.

 

I'm afraid not, but as an example a step jog of 0.1000 mm would take 3 tics taking around 3 seconds.

I have now changed to these:

http://www.cnc4you.co.uk/Stepper-Motor-Driver-4.5A,-50V-CNC-Microstepping-CW5045

It now takes one and is instantaneous, this may well be partly due to kernel speed being 100khz instead of 25, but doesn't explain why I was losing steps with the toshiba chips. Can only put it down to being underpowered. (Pots were all the way up at 4.5a and running cool with fans and heatsinks.) also this caused chattering at any speed, I have spoken to various cnc part suppliers who have confirmed that the issue seems its the chips themselves that restrict the circuitry to the current. And they are all rated on peak current not average, the type of sign wave you get from these is a triangular shape so suffer fade. hope this helps you save money from these badly specified drivers. Jon

 

Hi guys,

As I'm tuning in my OX I was just thinking about my step/mm settings on my GRBL controller. I'm wondering if using the "stock" OX setup is using the same number I came up with.. By no means are my numbers are perfect, but for me its been a place to start..

So, i'm using the CNC xPRO controller V2 with 0.9G:
$100=53.333 (x, step/mm) 1/16 STEP
$101=53.333 (y, step/mm) 1/16 STEP
$102=355.56 (z, step/mm) 1/16 STEP

 

Okay for nema 17 and possibly 1.5nm 23's only

 

I looked at those drivers some time ago. But basically they're the same as other Toshiba based drivers. I guess these guys got some of the troubles fixed and charge the double prices for them. For that price I would go Leadshines for just a little bit more.

 

I don't think it's the Toshiba chips that cause the problem but the switching speeds of optocouplers used in these drivers are to slow, hence causing delays in the signal and missing steps. I guess the chips are ok, but the Chinese fumblers went cheap on the rest with these kind of bad results.

 

No..Not so much the case. When i say i lost steps with mine it was only on the nema 23's the z ran perfectly with its nema 17. I've since given these drivers to my dad who is using them for 3d printing with no problems what so ever. And if it were the optocouplers these board would have alot worse rep than they have.

Although you may be right that it could well be that its not the chip.. Its could be the h bridge that lets them down..

Just found this article on it, well worth the read:

http://www.ruggedcircuits.com/the-motor-driver-myth

Conclusion of which is if an L298-based motor driver i.e toshiba type, states 3a max current. what your actually getting is 0.8a through each channel. this is the problem. mosfet drv type drivers are the only type thats going to power 23's correctly.