Hi Moray,

Those drives still specify the 5us Direction Setup Time the same as older leadshine drives.

But KFLOP now provides more than 5us setup time so it should be compatible.

Cnc4pc sells these:

https://cnc4pc.com/differential-line-driver.html

Otherwise this device AM26C31 should work:

http://www.ti.com/lit/ds/slls871/slls871.pdf

Regards

TK

 

I'm hoping somebody on this list could maybe give me a suggestion for some suitable differential driver chips, that I could use to take the single ended step/dir signals from a KFlop, and make full use of the differential step/dir inputs on some Leadshine EM806 stepper drives?

Am I right in thinking the newer drives should be OK regarding direction/step pulse timing?

The manual can be found at http://www.leadshine.com/UploadFile/Down/EMhm_V1.1.pdf

and the signal timing details can be found on the page numbered 3, or 8 as numbered in Adobe.

Thanks,

Moray

Hi Moray,

JP5 pins have 47 ohm series resistors so with a 150 ohm pull down the High Level voltage would be marginal.  Consider a 1K or larger.

Regards

TK

 

Tom,

Thanks for the part suggestion.

Just a quick question, the AM26C31 differential driver chip datasheet advises pulling unused inputs to gnd.

Would the KFlop's JP5 used as step/dir output be OK driving the chips with a 150ohm resistor (as per certain terminals on JP4/6) pulling the terminals to Gnd?

Thanks,

Moray

On Mon, Oct 10, 2016 at 7:47 PM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

 

Hi Moray,

Those drives still specify the 5us Direction Setup Time the same as older leadshine drives.

But KFLOP now provides more than 5us setup time so it should be compatible.

Cnc4pc sells these:

https://cnc4pc.com/ differential-line-driver.html

Otherwise this device AM26C31 should work:

http://www.ti.com/lit/ds/ slls871/slls871.pdf

Regards

TK

On 10/10/2016 3:12 AM, Moray Cuthill moray.cuthill@... [DynoMotion] wrote:

 

I'm hoping somebody on this list could maybe give me a suggestion for some suitable differential driver chips, that I could use to take the single ended step/dir signals from a KFlop, and make full use of the differential step/dir inputs on some Leadshine EM806 stepper drives?

Am I right in thinking the newer drives should be OK regarding direction/step pulse timing?

The manual can be found at http://www.leadshine.com/ UploadFile/Down/EMhm_V1.1.pdf

and the signal timing details can be found on the page numbered 3, or 8 as numbered in Adobe.

Thanks,

Moray

KFlop_JP5_BOB-13903.jpg

Hi Moray,

That is correct.  You may want to invert the step pulse (by adding 0x80) if that give more direction setup time.

But yes, CW/CCW is the preferred method as it eliminates any setup time issue.

Quadrature mode is even better as it requires half the number of edges, eliminates pulse length issues, and is more noise glitch insensitive.

The 3 modes should all work with the same wiring and interface.

HTH

Regards

TK

 

Hi,

just looking to clarify I've got the correct settings.

Going by the KFlop manual, I need to use Output Channels 12-15, and then add the following line to my init -

FPGA(STEP_PULSE_LENGTH_ADD)=63;

That gives me the required direction setup time, with actively driven high/low (LVTTL) signals from the KFlop?

However, having read the drive manual, would running the drive in CW/CCW be a better option?

It'll use the exact same wiring, and it's just a case of changing a setting in the drive tuning software.

I've also attached a photo of the finished line driver board, but having mounted it, I should of really put the in/out connectors the opposite way around!

Thanks,

Moray

On Sun, Oct 16, 2016 at 7:47 PM, Moray Cuthill <moray.cuthill@...> wrote:

Hi Tom,

thanks for the comments. There is a 0.1uF ceramic SMD cap at each chip, it's just the 3D render picked a resistor file instead of a capacitor one. I keep meaning to add something a bit bigger at the power input, which I'll do once I decide on what type of connector to use, as I'd like something that is polarised so it doesn't end up being plugged in the wrong way around.

For the milling machine this is going into, I had considered doing something that would not involve a Kanalog, but if this machine does what I'd like, I'd probably want to upgrade it to servos at some point, so using the Kanalog makes sense.

But it did get me thinking a bit about a more universal board, and what it would actually need to provide.

Take this machine as an example, I'd need 3 differential step/dirs (ideally 4 to allow for 4th axis), a 0-10V output for the spindle drive, 2 encoder inputs (one for an MPG, one for spindle feedback), and then I think around 9 24V inputs and 6 24V outputs (I still need to go through the wiring diagrams to work out the exact requirements).

Then I'd like to use a pendant, so I'd need an extra 4 inputs for the axis switch, 3 or 4 inputs for the step size switch, and an output for the pendant active light.

I'd like to add physical buttons for cycle start, and  feed hold, so another 2 inputs, and I'd like them back lit appropriately so another couple outputs. Then either two more encoder inputs, or two analogue inputs for SSO, and FRO dials.

From that, I'd be looking at a total of about 18/19 inputs, 9 outputs, 4 step/dir outputs, 2 encoder inputs, and either 2 more encoder inputs or 2 analog inputs.

However, I'm sure most people you ask will come up with totally different requirements.

As for the RJ45 suggestion, I can see the appeal, but it's not an option I'd personally prefer.

Moray

On Sun, Oct 16, 2016 at 12:51 AM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

 

Hi Moray,

Nice.  I don't see any power supply bypass Capacitors.  You might want to add a 0.1uf Ceramic near each IC and a 10uf Tantalum near the power connector.

Actually we have been thinking to do something along those lines as well.  We were thinking to use RJ45 connectors instead of the Screw Terminals and include power and GND and shield.  Also to allow for JP4 and JP5 connections to have an option to pick off 4 Step/Dir Signals from those instead of JP5.  As well as to provide differential encoder input options.

Regards
TK

On 10/15/2016 4:19 PM, Moray Cuthill moray.cuthill@... [DynoMotion] wrote:

 

Hi Tom,

thanks for the info.

I dug through the parts collection today, and I've got some 4k7 0603 resistors already, so I'll use them rather than add to the parts collection.

I've attached a 3D render of what I've designed. I'll hopefully get the circuit boards ordered at the start of the week.

Thanks,

Moray

On Thu, Oct 13, 2016 at 8:38 PM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

 

Hi Moray,

JP5 pins have 47 ohm series resistors so with a 150 ohm pull down the High Level voltage would be marginal.  Consider a 1K or larger.

Regards

TK

On 10/12/2016 4:59 PM, Moray Cuthill moray.cuthill@... [DynoMotion] wrote:

 

Tom,

Thanks for the part suggestion.

Just a quick question, the AM26C31 differential driver chip datasheet advises pulling unused inputs to gnd.

Would the KFlop's JP5 used as step/dir output be OK driving the chips with a 150ohm resistor (as per certain terminals on JP4/6) pulling the terminals to Gnd?

Thanks,

Moray

On Mon, Oct 10, 2016 at 7:47 PM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

 

Hi Moray,

Those drives still specify the 5us Direction Setup Time the same as older leadshine drives.

But KFLOP now provides more than 5us setup time so it should be compatible.

Cnc4pc sells these:

https://cnc4pc.com/differentia l-line-driver.html

Otherwise this device AM26C31 should work:

http://www.ti.com/lit/ds/slls8 71/slls871.pdf

Regards

TK

On 10/10/2016 3:12 AM, Moray Cuthill moray.cuthill@... [DynoMotion] wrote:

 

I'm hoping somebody on this list could maybe give me a suggestion for some suitable differential driver chips, that I could use to take the single ended step/dir signals from a KFlop, and make full use of the differential step/dir inputs on some Leadshine EM806 stepper drives?

Am I right in thinking the newer drives should be OK regarding direction/step pulse timing?

The manual can be found at http://www.leadshine.com/Uploa dFile/Down/EMhm_V1.1.pdf

and the signal timing details can be found on the page numbered 3, or 8 as numbered in Adobe.

Thanks,

Moray

Hi Moray,

I don't think rising or falling makes much difference.  Although rising and falling times can be slightly different and make small timing differences.  If falling times are much faster than rise times then a narrow positive pulse may not work where the same negative going pulse may work fine.

But the main thing is that the drive uses the Trailing edge of the pulse to sample the Direction and Step.  That provides a lot more time from when the direction changes until the drive samples it.  So for example if the drive steps on the rising edge of the step pulse then the pulse should be a normally high negative going pulse.

HTH

Regards

TK

 

Thanks Tom,

I do have the option to flip the active edge on the drive, but I'm not sure if Rising or Falling edge would be better?

I just re-read the manual, and the EM806 only works in Step/Dir mode, and the CW/CCW option isn't available on this model.

Moray

On Sun, Nov 20, 2016 at 10:05 PM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

 

Hi Moray,

That is correct.  You may want to invert the step pulse (by adding 0x80) if that give more direction setup time.

But yes, CW/CCW is the preferred method as it eliminates any setup time issue.

Quadrature mode is even better as it requires half the number of edges, eliminates pulse length issues, and is more noise glitch insensitive.

The 3 modes should all work with the same wiring and interface.

HTH

Regards

TK

On 11/20/2016 1:31 PM, Moray Cuthill moray.cuthill@... [DynoMotion] wrote:

 

Hi,

just looking to clarify I've got the correct settings.

Going by the KFlop manual, I need to use Output Channels 12-15, and then add the following line to my init -

FPGA(STEP_PULSE_LENGTH_ADD)= 63;

That gives me the required direction setup time, with actively driven high/low (LVTTL) signals from the KFlop?

However, having read the drive manual, would running the drive in CW/CCW be a better option?

It'll use the exact same wiring, and it's just a case of changing a setting in the drive tuning software.

I've also attached a photo of the finished line driver board, but having mounted it, I should of really put the in/out connectors the opposite way around!

Thanks,

Moray

On Sun, Oct 16, 2016 at 7:47 PM, Moray Cuthill <moray.cuthill@...> wrote:

Hi Tom,

thanks for the comments. There is a 0.1uF ceramic SMD cap at each chip, it's just the 3D render picked a resistor file instead of a capacitor one. I keep meaning to add something a bit bigger at the power input, which I'll do once I decide on what type of connector to use, as I'd like something that is polarised so it doesn't end up being plugged in the wrong way around.

For the milling machine this is going into, I had considered doing something that would not involve a Kanalog, but if this machine does what I'd like, I'd probably want to upgrade it to servos at some point, so using the Kanalog makes sense.

But it did get me thinking a bit about a more universal board, and what it would actually need to provide.

Take this machine as an example, I'd need 3 differential step/dirs (ideally 4 to allow for 4th axis), a 0-10V output for the spindle drive, 2 encoder inputs (one for an MPG, one for spindle feedback), and then I think around 9 24V inputs and 6 24V outputs (I still need to go through the wiring diagrams to work out the exact requirements).

Then I'd like to use a pendant, so I'd need an extra 4 inputs for the axis switch, 3 or 4 inputs for the step size switch, and an output for the pendant active light.

I'd like to add physical buttons for cycle start, and  feed hold, so another 2 inputs, and I'd like them back lit appropriately so another couple outputs. Then either two more encoder inputs, or two analogue inputs for SSO, and FRO dials.

From that, I'd be looking at a total of about 18/19 inputs, 9 outputs, 4 step/dir outputs, 2 encoder inputs, and either 2 more encoder inputs or 2 analog inputs.

However, I'm sure most people you ask will come up with totally different requirements.

As for the RJ45 suggestion, I can see the appeal, but it's not an option I'd personally prefer.

Moray

On Sun, Oct 16, 2016 at 12:51 AM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

 

Hi Moray,

Nice.  I don't see any power supply bypass Capacitors.  You might want to add a 0.1uf Ceramic near each IC and a 10uf Tantalum near the power connector.

Actually we have been thinking to do something along those lines as well.  We were thinking to use RJ45 connectors instead of the Screw Terminals and include power and GND and shield.  Also to allow for JP4 and JP5 connections to have an option to pick off 4 Step/Dir Signals from those instead of JP5.  As well as to provide differential encoder input options.

Regards
TK

On 10/15/2016 4:19 PM, Moray Cuthill moray.cuthill@... [DynoMotion] wrote:

 

Hi Tom,

thanks for the info.

I dug through the parts collection today, and I've got some 4k7 0603 resistors already, so I'll use them rather than add to the parts collection.

I've attached a 3D render of what I've designed. I'll hopefully get the circuit boards ordered at the start of the week.

Thanks,

Moray

On Thu, Oct 13, 2016 at 8:38 PM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

 

Hi Moray,

JP5 pins have 47 ohm series resistors so with a 150 ohm pull down the High Level voltage would be marginal.  Consider a 1K or larger.

Regards

TK

On 10/12/2016 4:59 PM, Moray Cuthill moray.cuthill@... [DynoMotion] wrote:

 

Tom,

Thanks for the part suggestion.

Just a quick question, the AM26C31 differential driver chip datasheet advises pulling unused inputs to gnd.

Would the KFlop's JP5 used as step/dir output be OK driving the chips with a 150ohm resistor (as per certain terminals on JP4/6) pulling the terminals to Gnd?

Thanks,

Moray

On Mon, Oct 10, 2016 at 7:47 PM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

 

Hi Moray,

Those drives still specify the 5us Direction Setup Time the same as older leadshine drives.

But KFLOP now provides more than 5us setup time so it should be compatible.

Cnc4pc sells these:

https://cnc4pc.com/differentia l-line-driver.html

Otherwise this device AM26C31 should work:

http://www.ti.com/lit/ds/slls8 71/slls871.pdf

Regards

TK

On 10/10/2016 3:12 AM, Moray Cuthill moray.cuthill@... [DynoMotion] wrote:

 

I'm hoping somebody on this list could maybe give me a suggestion for some suitable differential driver chips, that I could use to take the single ended step/dir signals from a KFlop, and make full use of the differential step/dir inputs on some Leadshine EM806 stepper drives?

Am I right in thinking the newer drives should be OK regarding direction/step pulse timing?

The manual can be found at http://www.leadshine.com/Uploa dFile/Down/EMhm_V1.1.pdf

and the signal timing details can be found on the page numbered 3, or 8 as numbered in Adobe.

Thanks,

Moray