[Connor]

I doesn't fully modify my board yet, I've just received the ICs.
But, until now, I just removed the 10k resistors (disabling the feature which reduce current to the stepper motors when the motors were idle) and the speed was raised from 750mmpm to 1000mmpm without loosing steps. I'm using 3Nm steppers with 3A/ph, powered at 24v and TR16x4 screws.

I've made some testing at 1200-1300mmpm but some steps are still loosed.
I'll do all remaining mod's steps in the next days...

The hissing noise was also reduced, not completely, but still reduced.

So, you can test it just removing these 10k resistors. If no difference will occur just undo.

It worked for me.

Thanks

Looking at my board.. I don't have a 10k near pin 3.. I have a 4.7k that looks like the one everyone is removing and placing somewhere else on the board. Is that the correct resistor?

Also, I would just rather by-pass the funky opto isolator stuff instead of trying to add in another 7414.. I don't think there is a need to add another one..

[Connor]

Hi Conor,

Audible frequency varies person to person, usually lower ages vs olders. Change the oscilator capacitor for TB6560 to a lower value i.e. 100pf and produced frequency will be out of ear range.

I don't have any surface mount 100pf's in stock.. I have some through hole ones I probably could do.. only issue is they're 20% tolerance.. I don't think that's good enough.

[Connor]

I removed the 4.7k resistors from the enable.. .. I also noticed that my Dip switches May have been set wrong.. But, it's hard to tell.. According to the documentation, I had it set a 50% current... But, I swore I it was faster then when I had it set at the 75% setting.. I never ran it at 100% for very long because I didn't see any marked improvement it..

When I removed the 4.7k resistors, I DID notice a little bit of a performance increase in the speed. (before changing the power current settings).. and now.. at 100% it's even better.. Can't wait to get my 100pf caps and try them to see if it helps with the high pitched noise.

[interzis]

at least at my board, 4.7k is marked on the PCB. the resistors are 10k (marked 104).
yes, I also work with 100% current setting, I need 3A for my steppers.

I'm glad your board work better after these mods. keep going, I'll do the rest of mods soon

good luck

[Connor]

at least at my board, 4.7k is marked on the PCB. the resistors are 10k (marked 104).
yes, I also work with 100% current setting, I need 3A for my steppers.

I'm glad your board work better after these mods. keep going, I'll do the rest of mods soon

good luck

Mine is marked 4.7. I didn't look at the resistor itself.. or measure it for that matter.. My surface mount 100pf caps should be here this coming Monday. We'll see how that does.

[Connor]

My 100pf caps arrived yesterday and I installed them. I'm happy to report that my wife can now be in the same room as my machine and isn't complaining about the super high pitch sound (that I couldn't even hear) In addition, I noticed another increase in speed/torque. I took it one step further.. I went from 1/2 stepping to 1/8 and In EMC, you can specify the Step Time and Step Space (Time to hold the On and Off for the steps) I had it at 5000nano seconds.. I took it to 1000 and it work even better, at least on the X axis, I latter discovered they Y had a bit of chatter, like maybe it was dropping a few steps or something.. so, I took it to 1250 and both axis are very happy. I'm now able to do to 300IPM Rapids without it stalling. This is all with the cap change out and removal of the 4.7k (or 10k, which ever it is) . I've yet to fix the issue with the Opto Coupler for the clock.. and I'm wondering if I need to. I'm also a bit confused as to why the Y did fine with the 1000ns steps and the X had issues until I went to 1250ns.

Thanks, Connor

[john-100]

Hi Conner ,

one possible explanation for the difference between the X and Y axis could be the wide variation in oscillator frequency form one IC to another

for a 330 pf capacitor the frequency range is 60 Khz to 200 Khz
the typical frequency being 130Khz

John

[Connor]

Hi Conner ,

one possible explanation for the difference between the X and Y axis could be the wide variation in oscillator frequency form one IC to another

for a 330 pf capacitor the frequency range is 60 Khz to 200 Khz
the typical frequency being 130Khz

John

I'll swap the axis around and see if it follows the motor, or stays with the driver chip.. All axis have been upgraded to the 100pf cap.

I'm a bit confused with everyone messing with the CLK part of the circuit, cutting out the Opto or adding in a extra 7414.. What's the deal there? I'm also trying to figure out which modification swaps the GND and Supply on the enable opto.. I want to check that and see if it's messed up on my board or not. One thing I found that was strange, is each chip has it's own enable, EMC, uses just 1 output for ENABLE.. So, I had strange things happen when I was configuring everything for the first time.. I finally figured it out.

[doorknob]

I'm a bit confused with everyone messing with the CLK part of the circuit, cutting out the Opto or adding in a extra 7414.. What's the deal there? I'm also trying to figure out which modification swaps the GND and Supply on the enable opto.. I want to check that and see if it's messed up on my board or not.

My interpretation of post #1 is that the OP found that the step pulse did not have a sharp trailing edge, rather it displayed some timing jitter (see the oscilloscope traces), which was cleaned up by adding in the extra inverter (and inverting the signal polarity in the Mach3 setup).

There also have been reports that the optos on the board are relatively "slow", which could possibly lead to missed steps, and some people had reported that signals were not making it through the opto (possibly due to having weak LPT port output drivers). Also, the optos seem to be superfluous, inasmuch as the board layout shares the ground between the board's inputs and outputs, so it's not particularly well isolated even with the optos in place.

As for the swap of ground and supply on the enable opto, the OP acknowledged in post #19 that his board does not actually have the wiring error that he originally thought was there, and so he posted a revised version 3 of the schematic minus that modification.

See http://www.cnczone.com/forums/821195-post19.html and also see http://www.cnczone.com/forums/attach...4&d=1283911876

[john-100]

H Conner ,

back in page 1 of this post there is two issues people have been trying to correct

on early versions of the board
the changes to the clock circuit are attempts to reduce the switching time so you can
step the motor faster
( bypassing the opto isolator or decreasing the value of the pull up resistor etc )

adding the second 7414 schmitt trigger inverter is to improve the noise immunity

in new versions of the board you may not need to make all the mods

John

[Connor]

Well then, I would say that I'm done then. Changing out the CAP and removing the funky idle circuit looks to have increased the performance as much as I can expect to get out of this board. 300IMP isn't anything to sneeze at, and the Super high pitched noise is gone. I'm still not sure why I can't get my Z speed up.. I think the stepper is just too small for much more than what it's at.. But, that's okay it works good enough for what I'm doing.

[bluewave65]

I made an SMD eagle PCB for generally better look instead of universal board

Dave

[GrootWitBaas]

From my understanding, the opto is the wrong way wired, I have swapped the two legs and this had some improvement for me on the general performance, and my LED's are working correctly. The biggest improvement was obtained with the oscillator caps changed to 100pF. So for me all changes done in a short list is:
1. Change caps (100pF is what I used)
2. Add 7414 (as per post #1)
3. Swap opto's 'Out' and '+5v' around (also per post #1 see picture below)
4. Add Separate 7812 for fan
5. Add proper ground to Main Heat sink.

I don't believe all changes is a must, but while I was at it I did all the above.

@bluewave65 nice PCB board, maybe you can also add the bits for a 4ch board ?

Groot

[interzis]

re-read the discussion, the optos are OK, you don't have to swap their pins

[sonic1954]

I got my speed on the z axis up by doing a fresh configuration file from the one that comes with mach3.

[GrootWitBaas]

re-read the discussion, the optos are OK, you don't have to swap their pins

Not sure where this was said, but I based the decision to do the swap on the statement in post 1 and it does make sense to me to do this. Also with this done the LED's for Enable works correctly. Yes it does work without the change, but the out looks better on my scope with the change done.

2) Driver IC ENABLE input improperly driven: The opto-coupler for the ENABLE input is wired incorrectly having the npn emitter connected to supply. Although the circuit does function, the improper wiring of the npn will cause very low transistor gain opening the possibility of system noise corrupting the ENABLE signal. Solution: Rewire the npn of the opto-coupler with collector to supply and emitter to output.

[interzis]

page 2, posts #18 and #19

but keep the mod if your board have inverted optos...

[stragenmitsuko]

I wonder why the manufacturer states that a
nema 23 needs 24V and a
nema 34 needs 35V

As far as I know , the size of the stepper should be irrelevant .
The current setting is important , but I could drive a nema34 at 2A
and a nema 23 at 3A if that is the rated motor current .
Ususally a higher voltage results in a higher max rpm .

What I do know is that you just can't hook up a stepper to a driver and expect everything to work perfect .
A stepper driver must be tuned to the motor characteristics .
Chopper frequency , blanking time , sense resistors and ref voltages , fast and slow decay settings , It all matters and it all has a great deal of influence on the motor performance .
A scope is an invaluable tool to do this .

Anyway , I've ordered a 3-axis 2.5Amp version to drive 3 nema23's at 1.2Amp . I'm building a small pcb router , don't need much power .

I'll make sure to post the results here .

Pat

[jmchris]

Since I'm not an electronics engineer and HAVE to have better accuracy machining 316SS than these boards allow... I have two for sale as well as several 40 amp external relays.

$50 - 3 axis TB6560 (connectors, and LPT cable included)
$60 - 4 axis TB6560 (connectors and LPT cable included)
$10 - each for 3 - 40 amp electronic relays
$25 - each for2 - 24V power supplies

The boards work fine for wood work or hobby CNC. Just not accurate or reliable enough for tight tollerance stainless steel machining.

PM / Email

Jeff

[tanglewoods]

Has anybody used the TB6560-T4-V4. It has Schmitt triggers all ready on the board. If you have used it could someone please post the pinouts for the DB25 parallel port and the DB15 Manual ports.