YOOCNC 3020Z-D52 losing steps

[Pedro34]

I bought it for a company called Omc stepper online
3 Axis TB6600 CNC Stepper Motor Driver Controller Board For CNC Router Milling|3-TB6600|Stepper Motor Drivers


Sent from my iPad using Tapatalk

[kirkguille]

Thanks!

[Rockfella]

Double

[Rockfella]

Hi everyone. Having some issues with my YOOCNC-NT65-3x card. The Y axis steppermotor all of the sudden started vibrating. I can´t make any sense of it as it has been running smooth for about 3 months. As I turn down the velocity and steps/min there is a quite weird characteristic, if I give it a steady jog it does 17 steps in one direction then changes back 17 steps and this just keeps on going. I checked the capacitors but no faulty there. Any idea on where to look? Thanks!

[japioblanco]

Hallo, I do all the mods with the resuld the system is generating and not working wel and the steppers wil very hot.

[RogerClark]

Guys,

Can you tell me whether the missing steps issue can also be extra steps.

I've been consistently getting around 0.5% too many steps.

I've tried adjusting the calibration in Mach3 from 400 steps per to 398 steps per, and I think this probably helps a bit, but ultimately if its a noise issue on supply or signal lines then this sort of hack is not going to work

So.

I suspect its best to solder on the added caps as detailed in the various postings in this thread, and see if that help ??

[chris9870]

Hi Pedro34,

The DCDC is to separate voltages.

Chris

[cordovabernal]

But I have put 63V capacitor... as shown in the attached picture and it works fine.
Attachment 226412

I bought other model of tb6560 controller, in theory is better that other models because it has hex converter in pin 3: https://www.amazon.com/SainSmart-TB6...ller+sainsmart
I put the 470nF caps between 15 and 20 pin of every chip, is this the same of your model board for mine? I noticed a reduction in noise (just like low decay config) instead of being set on fast decay, but the acurrancy is not good enought, almost the same before. Could you give me a clue of more settings please. PD: sorry for bad english.

Paul

[cordovabernal]

I bought other tb6560 board of $100 for replacing a cheaper one , I will tell you if the problem is in all of them.
This is loosing 1 milimeter in every level I draw, even if the 470nFcaps are soldered:
https://www.amazon.com/SainSmart-TB6...ller+sainsmart

Now I´m trying to prove this:
https://www.amazon.com/RioRand-Alumi...cnc+controller

Let´s see the next week if this is better or maybe I´ll need a Gecko

[helisat]

Hey all, it's been a while since I last posted in this topic. My modified YOOCNC-NT65-3X is still working very well.

I have noticed that different manuals for this controller seem to contradict info regarding the polarity for the STEP inputs. My YOOCNC manual suggest they are active-high, but others that I've seen show active-low.

I think that the TB6560 datasheet appears to suggest that it will step on a rising-edge CLK (going by the rising edge symbol). Working back though the circuitry I get the following:
1. The TB6560 steps on a rising edge of the CLK.
2. The 6N137 inverts the signal, but since its input is fed into the cathode, the net result is non-inverted.
3. 74HC14 inverts the signal.
Result: I think the DB25 STEP signal would need to be active-low.

From my observations, active-low on the step signals seems to generate more heat on the steppers.

Any thoughts?

[john-100]

It could be the high logic level on the step input switches the driver from its idle state to full current mode

another chinese design error ?

I expect they should off used two of the 74HC14 inverters so a positive going input drives the TB6560 step input positive

John

PS

if your board is like this -


it could be easier adding the extra inverter to the base of the emitter follower in the idle current control circuit

[helisat]

Hi John,

You might be right.



Orange = CLK
Green = TQ1 & TQ2

TQ1/TQ2 HIGH = 20% torque
TQ1/TQ2 LOW = 100% torque

Stepping occurs on the rising-edge of CLK, and each time that is 20% torque ...

Quite possibly a fault. Would like to know more before I go modding the board again though.

Does anyone have an oscilloscope to map the waveforms for CLK/TQ1/TQ2 ?

Cheers.

[helisat]

I used the DMM to perform some basic logic checks on the circuit board. The circuit logic appears to be correct, so either there's a bug in the schematic diagram, or it's slightly different to my board.

This is how I perceive the board to be working:
The torque is reduced while sitting idle (CLK=HIGH; TQ=HIGH. When a step command is issued, the CLK goes LOW, which brings TQ LOW (increasing torque), then step occurs on the rising edge of CLK. There is a RC with 8µS time constant that holds the torque pin LOW (100% torque) after the rising-edge clock. I'd say this ensures full torque throughout the step action. I'd need to borrow a logic analyser to watch whats going on during operation.

Orange = CLK
Green = TQ1 & TQ2



To answer the original question .. the step pins are active-high (at the DB25). You'll know when this is correct, as the steppers are a lot easier to freewheel (at idle).

Cheers.

[john-100]

Hi Helisat

the step & tq pulse diagram in post 133 is what I'd expected
although the RC time constant is a bit short -
I would of made it long enough to be still in high torque when the next step pulse arrives


John

[helisat]

I agree with you on the time constant. I'll check the circuit tomorrow, might just be a bug in the schematic (820nF instead of pF).

[ciuzzo]

Salve a tutti ho lo stesso problema di perdita passi asse x
Ho letto tutti i vostri consigli ma sulla mia scheda dove dovrei saldare i condensatori ceramici da 475nf?
Grazie a tutti e aiutatemi
La mia scheda è una jp382c simile alla vostra