Re: [Emc-users] (Off topic) A Question Re: BOBs and Opto-Isolators

[Gene Heskett]

On Monday 14 October 2019 13:36:24 John Dammeyer wrote:

> > On Sunday 13 October 2019 19:46:55 Kenneth Lerman wrote:
> > > Hello All -- especially Gene,
> > >
> > > I've seen many posts on the lists regarding breakout boards and
> > > dealing with issues regarding slow opto-isolators. I have a
> > > question:
> > >
> > > Is anyone using: ISO776x
> > >  devices?
> >
> > I've seen that pdf before, but in the cases where I've simply
> > bypassed them, I had good solid, noise free sources to start with. 
> > If I use a BoB, its a Saintsmart, simply because there is not any
> > opto's in their output path, and most outputs goto the input opto's
> > of a stepper driver, which I've found is a bit faster than the
> > average bear. A 2M542 for instance can, if the drive is good rail to
> > rail stuff, be driven to about 375k steps a second before it
> > miss-behaves!  That will turn a nema-23  motor rated for 270 oz/in,
> > to over 3000 rpms if the supply voltage is pushing the 2M542 at
> > around 42 volts.  Nearly 2000 revs at 28 volts.
>
> Hi Gene,
> I think you have your units mixed up here.  There is no way you will
> get a stepper motor to turn at 3000 RPM with a 48V supply.  They just
> can't do that physically.
>
Why not? I've done it on the table, driving a 2M542 with a function 
generator. The point of increased voltage is to overcome the inductance 
faster to get a usable current thru the coils. That is what the motor 
steps on, the rotating magnetic field. Give me a driver that hasn't any 
opto's in its inputs to slow it down, and can tolerate a 150 volt supply 
voltage, and I'll turn a nema 23 at 10k rpms laying loose on the table. 
The motor can generally stand several hundred volts, so 150 shouldn't be 
a problem.

The limits of speed on a stepper are determined by:

1. applied voltage, must be high enough to get a usable current into the 
coils for working torque in the allocated step durations time between 
steps.  I might add that the more costly 3 phase motors have an 
advantage in that dept because of the longer time allocated for currant 
reversals in their coils in the 3 phase design.  They are a nice idea, 
but the cost needs to come down before they'll get any takeup in new 
designs. but they are very close to a servo with a builtin encoder, 
except they don't have one.

2. The steadyness of that clocking so the motor isn't being asked to play 
catch during the current step, then told to slow down on the next one. 
I've seen some stuff written where a 5% wobble in the step rate can cost 
you 50% of your torque.

This also applies to vfd's. The top speed limit you can run a motor at 
with a vfd is the motors coil inductance. I can drive this 1hp rated 
motor at up to about 120hz, effectively getting 2hp out of it.  But at 
180 hz, 3x its rated speed, the motors coil inductance has limited the 
coil current to just over an amp, and with bronze bushings the drag is 
so bad thats as fast as it can turn the spindle. But at the same time I 
can run it at a 1/4 speed or 15 hz, drawing nearly its nameplate FLA, at 
no cutting load and do that for long enough to get the job done without 
its getting so hot I can't touch it.  Properly programmed, a vfd is 
magic, and it can do it with no damage to the motor.

Any vfd that takes 30 seconds to get to speed has never been programmed 
since it came out of the box new. 

I can turn this 8" chuck at 400 revs, punch in an m4, and just a bit over 
a second later its turning 400 revs backwards.  The drive belts are 
screaming like Michelin tires, but its doing it. Punch in an m3, and its 
running fwd with the same sound effects as it made before. Yet its all 
sequenced in the hal file and the motor is never subjected to more than 
its nameplate FLA, DC for dynamic braking or AC for running. Never gets 
anywhere near its LRA, which can destructively heat it pretty fast.  And 
the vfd never gets hot as its staying inside of the motors ratings.  
Whats not to love?




> The Bergerda AC Servos run 3000 RPM and with a 2500 line quadrature
> encoder (10,000 edges per second) means a 500, hz step rate is
> needed to achieve that.
>
> But, the Bergerda AC Servo drive has a dual purpose stepping input
> that can be wired as RS422 or as opto-isolated up to 24V.  According
> to the Bergerda motor manual the TTL opto isolated inputs are not
> suitable to that high a step rate while the differential RS422 inputs
> that are good to that speed.
>
Makes perfect sense.

> So regardless of what sort of PC based step/dir interface you are
> using, if you want isolated 500kHz step rates then your BoB needs to
> have something like the ISO7760 in front of the RS422 differential
> drivers.  I don't know if the Bergerda differential inputs are
> isolated from the high voltage inside the driver.
>
> The iso7760 is  nice device.   The ground on the input side can be
> connected to the PC ground and the output ground to the low 

Re: [Emc-users] (Off topic) A Question Re: BOBs and Opto-Isolators

[John Dammeyer]

Oh and the reason for all those opto  RJ45 interface modules is because the 
STMBL drive:
http://www.autoartisans.com/mill/STMBL-4.1.jpg
uses the top connector for step/dir/enable and fault as RS422 signals.
The next RJ45 is for the encoder index, A and B as RS422 signals along with 5V 
to power the encoder.
The bottom RJ45 is for the AC Servo motor U, V and W signals also RS422 and 
again 5V to power the encoder.

This photo
http://www.autoartisans.com/mill/HarmonicDrive/TestSetup1.jpg
shows the Harmonic drive (rated at 170VDC) running on 48V with the STMBL.  The 
TTL input step/dir and enable are translated to RS422 on the proto-board.  The 
harmonic drive has a tiny and expensive connector for the encoder signals.  
Again a hacked together interface to provide the RS422 into the STMBL.  That 
I'll be changing to a DB-25 that is identical to the pins from the Bergerda 
DB-25.  In that previous posting with my interface boards I have that small 
DB-25 to RJ45 adaptor that will provide a much cleaner interface.

One of the nice things about the STMBL is that you can test a high voltage 
servo motor at low voltages while getting things connector correctly.  That 
option isn't there with Servo Drives that take 220VAC input for motors that are 
rated at a lower voltage.

John




___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] (Off topic) A Question Re: BOBs and Opto-Isolators

[John Dammeyer]

> On Sunday 13 October 2019 19:46:55 Kenneth Lerman wrote:
> 
> > Hello All -- especially Gene,
> >
> > I've seen many posts on the lists regarding breakout boards and
> > dealing with issues regarding slow opto-isolators. I have a question:
> >
> > Is anyone using: ISO776x
> >  devices?
> 
> I've seen that pdf before, but in the cases where I've simply bypassed
> them, I had good solid, noise free sources to start with.  If I use a
> BoB, its a Saintsmart, simply because there is not any opto's in their
> output path, and most outputs goto the input opto's of a stepper driver,
> which I've found is a bit faster than the average bear. A 2M542 for
> instance can, if the drive is good rail to rail stuff, be driven to
> about 375k steps a second before it miss-behaves!  That will turn a
> nema-23  motor rated for 270 oz/in, to over 3000 rpms if the supply
> voltage is pushing the 2M542 at around 42 volts.  Nearly 2000 revs at 28
> volts.
> 

Hi Gene,
I think you have your units mixed up here.  There is no way you will get a 
stepper motor to turn at 3000 RPM with a 48V supply.  They just can't do that 
physically.

The Bergerda AC Servos run 3000 RPM and with a 2500 line quadrature encoder 
(10,000 edges per second) means a 500, hz step rate is needed to achieve 
that. 

But, the Bergerda AC Servo drive has a dual purpose stepping input that can be 
wired as RS422 or as opto-isolated up to 24V.  According to the Bergerda motor 
manual the TTL opto isolated inputs are not suitable to that high a step rate 
while the differential RS422 inputs that are good to that speed.  

So regardless of what sort of PC based step/dir interface you are using, if you 
want isolated 500kHz step rates then your BoB needs to have something like the 
ISO7760 in front of the RS422 differential drivers.  I don't know if the 
Bergerda differential inputs are isolated from the high voltage inside the 
driver.

The iso7760 is  nice device.   The ground on the input side can be connected to 
the PC ground and the output ground to the low voltage CNC machine control 
ground.  Note the data sheet also lists that half the units inside the package 
(7763) can be reversed so it makes an ideal single isolator package for 
multiple RS485 or RS232 circuits or a BoB where you want to isolate the PC 
parallel port.Had I known about this part I'd probably have used it instead 
of the multiple devices below.

What I've used are  HCPL2631 and initial testing shows they can handle the step 
rates for the Bergerda AC Servo.  This module can be configured with jumpers to 
interface to either RS422 step/dir/enable and fault signals for the STMBL Servo 
Drive  or to RS422 step/dir and opto-isolated enable/fault for the Bergerda AC 
Servo Drive.  Additionally this little board also creates the 5V for the MESA 
7i92H and two Open Collector signals to reset the HP_UHU DC servo drives in 
case of a Fault from one of them.  Those are simple, and very slow, 4N25 
devices.

http://www.autoartisans.com/mill/StepDir2STMBL_SCH.pdf

I built 3 of them.
http://www.autoartisans.com/mill/StepDir2STMBL.jpg

Should be installed this week once I make up some cables.  Too many projects.  
Never enough time.

John




___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] (Off topic) A Question Re: BOBs and Opto-Isolators

[Gene Heskett]

On Monday 14 October 2019 08:14:38 Kenneth Lerman wrote:

> They are actually rated from zero Hz (DC) to 100MHz.
>
Then they have stuff in there that not in the diagrams in the 
pdf.  'twon't be the first time howto's don't show proprietary stuff :)

> Ken
>
> Kenneth Lerman
> 55 Main Street
> Newtown, CT 06470
>
> On Mon, Oct 14, 2019 at 5:27 AM Gene Heskett  
wrote:
> > On Sunday 13 October 2019 19:46:55 Kenneth Lerman wrote:
> > > Hello All -- especially Gene,
> > >
> > > I've seen many posts on the lists regarding breakout boards and
> > > dealing with issues regarding slow opto-isolators. I have a
> > > question:
> > >
> > > Is anyone using: ISO776x
> > >  devices?
> >
> > I've seen that pdf before, but in the cases where I've simply
> > bypassed them, I had good solid, noise free sources to start with. 
> > If I use a BoB, its a Saintsmart, simply because there is not any
> > opto's in their output path, and most outputs goto the input opto's
> > of a stepper driver, which I've found is a bit faster than the
> > average bear. A 2M542 for instance can, if the drive is good rail to
> > rail stuff, be driven to about 375k steps a second before it
> > miss-behaves!  That will turn a nema-23  motor rated for 270 oz/in,
> > to over 3000 rpms if the supply voltage is pushing the 2M542 at
> > around 42 volts.  Nearly 2000 revs at 28 volts.
> >
> > I suspect the reason they aren't used may have something to do with
> > the cost, or possibly the fact that they are capacitively coupled
> > and very low re-rate signals might drift back to their default
> > state, without any change in the input voltage.
> >
> > But thats, to me, just a SWAG.
> >
> > The most glaring example of too slow in my stuff is the 1000 line
> > encoder sitting on top of the G0704 spindle motor, an ABX encoder,
> > and the resultant "scale" is over 14,000 in the .ini file, and thats
> > way too fast for an opto.  So I have 2, $2, rs485 convertors to make
> > single ended rail to rail logic out of the encoders differential
> > outputs, Those 2 single ended signals are now being fed to a
> > sainsmart bob whose optos I have bypassed, and on to the 5i25. Works
> > perfectly.
> >
> > I have switches on the gearshift knob that tell my .hal file which
> > gear its in, or no gear in case its between gears, in which case I
> > feed the pid just enough offset to make the motor spin maybe 25 rpm
> > when its out of gear, making gear shifts slick and painless while
> > its running! When its in gear by the last degree and one of the
> > switches is closed, the circuit then scales the "scale" by the gear
> > and restores the speed input from lcnc, and 200ms later the spindle
> > is back up to the set S speed even if its 3000 revs. Thats 800 revs
> > faster than OEM.
> >
> > The 5i25 encoders x input is still the x input from the original
> > optical encoder wheel I made.  And with that high a resolution,
> > quantization noise is essentially gone, and I can run the spindle
> > pid's PGain at 40 or so, but 20 is unconditionally stable.
> >
> > With 126 volts in the motor supply feeding one of Jon's pwm-servo
> > amps to that 90 volt, 9.7 amp rated motor, I'm getting close to 2hp
> > at the amplifiers 17 amp limit, rigid tapping in steel with a 3/8's"
> > tap can be done by pecking in steel or cast, going about 1/2 turn
> > per peck. No change in the speed until that 17 amps limit makes its
> > iron chirp.
> >
> > And I'm amazed that the plastic gears in that 2 speed head are still
> > working, by all rights I should have stripped the teeth  off them by
> > now.  Tough stuff.
> >
> > > They are 100 MHz data rate devices, have high isolation, have six
> > > in a package, and look like they would work just fine. The only
> > > downside might be that they are limited to 5.5 volts on the
> > > output, so you might need to add a transistor driver after them.
> > > And all of the ones in a single package share a ground.
> > >
> > > I hadn't seen them before, so I thought I'd mention them.
> >
> > They look like they'd be perfectly usable in an active circuit, but
> > I'd have reservations about putting one in a home or limit switch
> > circuit until we find out what their dc characteristics are.
> >
> > > Regards,
> > >
> > > Ken
> > >
> > > Kenneth Lerman
> > > 55 Main Street
> > > Newtown, CT 06470
> > >
> > > ___
> > > Emc-users mailing list
> > > Emc-users@lists.sourceforge.net
> > > https://lists.sourceforge.net/lists/listinfo/emc-users
> >
> > Cheers, Gene Heskett
> > --
> > "There are four boxes to be used in defense of liberty:
> >  soap, ballot, jury, and ammo. Please use in that order."
> > -Ed Howdershelt (Author)
> > If we desire respect for the law, we must first make the law
> > respectable. - Louis D. Brandeis
> > Genes Web page 
> >
> >
> > ___
> > Emc-users mailing list
> > 

Re: [Emc-users] (Off topic) A Question Re: BOBs and Opto-Isolators

[Kenneth Lerman]

They are actually rated from zero Hz (DC) to 100MHz.

Ken

Kenneth Lerman
55 Main Street
Newtown, CT 06470



On Mon, Oct 14, 2019 at 5:27 AM Gene Heskett  wrote:

> On Sunday 13 October 2019 19:46:55 Kenneth Lerman wrote:
>
> > Hello All -- especially Gene,
> >
> > I've seen many posts on the lists regarding breakout boards and
> > dealing with issues regarding slow opto-isolators. I have a question:
> >
> > Is anyone using: ISO776x
> >  devices?
>
> I've seen that pdf before, but in the cases where I've simply bypassed
> them, I had good solid, noise free sources to start with.  If I use a
> BoB, its a Saintsmart, simply because there is not any opto's in their
> output path, and most outputs goto the input opto's of a stepper driver,
> which I've found is a bit faster than the average bear. A 2M542 for
> instance can, if the drive is good rail to rail stuff, be driven to
> about 375k steps a second before it miss-behaves!  That will turn a
> nema-23  motor rated for 270 oz/in, to over 3000 rpms if the supply
> voltage is pushing the 2M542 at around 42 volts.  Nearly 2000 revs at 28
> volts.
>
> I suspect the reason they aren't used may have something to do with the
> cost, or possibly the fact that they are capacitively coupled and very
> low re-rate signals might drift back to their default state, without any
> change in the input voltage.
>
> But thats, to me, just a SWAG.
>
> The most glaring example of too slow in my stuff is the 1000 line encoder
> sitting on top of the G0704 spindle motor, an ABX encoder, and the
> resultant "scale" is over 14,000 in the .ini file, and thats way too
> fast for an opto.  So I have 2, $2, rs485 convertors to make single
> ended rail to rail logic out of the encoders differential outputs, Those
> 2 single ended signals are now being fed to a sainsmart bob whose optos
> I have bypassed, and on to the 5i25. Works perfectly.
>
> I have switches on the gearshift knob that tell my .hal file which gear
> its in, or no gear in case its between gears, in which case I feed the
> pid just enough offset to make the motor spin maybe 25 rpm when its out
> of gear, making gear shifts slick and painless while its running! When
> its in gear by the last degree and one of the switches is closed, the
> circuit then scales the "scale" by the gear and restores the speed input
> from lcnc, and 200ms later the spindle is back up to the set S speed
> even if its 3000 revs. Thats 800 revs faster than OEM.
>
> The 5i25 encoders x input is still the x input from the original optical
> encoder wheel I made.  And with that high a resolution, quantization
> noise is essentially gone, and I can run the spindle pid's PGain at 40
> or so, but 20 is unconditionally stable.
>
> With 126 volts in the motor supply feeding one of Jon's pwm-servo amps to
> that 90 volt, 9.7 amp rated motor, I'm getting close to 2hp at the
> amplifiers 17 amp limit, rigid tapping in steel with a 3/8's" tap can be
> done by pecking in steel or cast, going about 1/2 turn per peck. No
> change in the speed until that 17 amps limit makes its iron chirp.
>
> And I'm amazed that the plastic gears in that 2 speed head are still
> working, by all rights I should have stripped the teeth  off them by
> now.  Tough stuff.
>
> > They are 100 MHz data rate devices, have high isolation, have six in a
> > package, and look like they would work just fine. The only downside
> > might be that they are limited to 5.5 volts on the output, so you
> > might need to add a transistor driver after them. And all of the ones
> > in a single package share a ground.
> >
> > I hadn't seen them before, so I thought I'd mention them.
> >
> They look like they'd be perfectly usable in an active circuit, but I'd
> have reservations about putting one in a home or limit switch circuit
> until we find out what their dc characteristics are.
>
> > Regards,
> >
> > Ken
> >
> > Kenneth Lerman
> > 55 Main Street
> > Newtown, CT 06470
> >
> > ___
> > Emc-users mailing list
> > Emc-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/emc-users
>
>
> Cheers, Gene Heskett
> --
> "There are four boxes to be used in defense of liberty:
>  soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author)
> If we desire respect for the law, we must first make the law respectable.
>  - Louis D. Brandeis
> Genes Web page 
>
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>

___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] (Off topic) A Question Re: BOBs and Opto-Isolators

[Gene Heskett]

On Sunday 13 October 2019 19:46:55 Kenneth Lerman wrote:

> Hello All -- especially Gene,
>
> I've seen many posts on the lists regarding breakout boards and
> dealing with issues regarding slow opto-isolators. I have a question:
>
> Is anyone using: ISO776x
>  devices?

I've seen that pdf before, but in the cases where I've simply bypassed 
them, I had good solid, noise free sources to start with.  If I use a 
BoB, its a Saintsmart, simply because there is not any opto's in their 
output path, and most outputs goto the input opto's of a stepper driver, 
which I've found is a bit faster than the average bear. A 2M542 for 
instance can, if the drive is good rail to rail stuff, be driven to 
about 375k steps a second before it miss-behaves!  That will turn a 
nema-23  motor rated for 270 oz/in, to over 3000 rpms if the supply 
voltage is pushing the 2M542 at around 42 volts.  Nearly 2000 revs at 28 
volts.

I suspect the reason they aren't used may have something to do with the 
cost, or possibly the fact that they are capacitively coupled and very 
low re-rate signals might drift back to their default state, without any 
change in the input voltage.

But thats, to me, just a SWAG.

The most glaring example of too slow in my stuff is the 1000 line encoder 
sitting on top of the G0704 spindle motor, an ABX encoder, and the 
resultant "scale" is over 14,000 in the .ini file, and thats way too 
fast for an opto.  So I have 2, $2, rs485 convertors to make single 
ended rail to rail logic out of the encoders differential outputs, Those 
2 single ended signals are now being fed to a sainsmart bob whose optos 
I have bypassed, and on to the 5i25. Works perfectly.

I have switches on the gearshift knob that tell my .hal file which gear 
its in, or no gear in case its between gears, in which case I feed the 
pid just enough offset to make the motor spin maybe 25 rpm when its out 
of gear, making gear shifts slick and painless while its running! When 
its in gear by the last degree and one of the switches is closed, the 
circuit then scales the "scale" by the gear and restores the speed input 
from lcnc, and 200ms later the spindle is back up to the set S speed 
even if its 3000 revs. Thats 800 revs faster than OEM.

The 5i25 encoders x input is still the x input from the original optical 
encoder wheel I made.  And with that high a resolution, quantization 
noise is essentially gone, and I can run the spindle pid's PGain at 40 
or so, but 20 is unconditionally stable.

With 126 volts in the motor supply feeding one of Jon's pwm-servo amps to 
that 90 volt, 9.7 amp rated motor, I'm getting close to 2hp at the 
amplifiers 17 amp limit, rigid tapping in steel with a 3/8's" tap can be 
done by pecking in steel or cast, going about 1/2 turn per peck. No 
change in the speed until that 17 amps limit makes its iron chirp.

And I'm amazed that the plastic gears in that 2 speed head are still 
working, by all rights I should have stripped the teeth  off them by 
now.  Tough stuff.

> They are 100 MHz data rate devices, have high isolation, have six in a
> package, and look like they would work just fine. The only downside
> might be that they are limited to 5.5 volts on the output, so you
> might need to add a transistor driver after them. And all of the ones
> in a single package share a ground.
>
> I hadn't seen them before, so I thought I'd mention them.
>
They look like they'd be perfectly usable in an active circuit, but I'd 
have reservations about putting one in a home or limit switch circuit 
until we find out what their dc characteristics are.

> Regards,
>
> Ken
>
> Kenneth Lerman
> 55 Main Street
> Newtown, CT 06470
>
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] (Off topic) A Question Re: BOBs and Opto-Isolators

[N]

> Hello All -- especially Gene,
> 
> I've seen many posts on the lists regarding breakout boards and dealing
> with issues regarding slow opto-isolators. I have a question:
> 
> Is anyone using: ISO776x 
>  devices?

Well, not exactly his one but there are others with different numbers of 
isolation barriers/directions and these I have used. I think they are really 
good.

> They are 100 MHz data rate devices, have high isolation, have six in a
> package, and look like they would work just fine. The only downside might
> be that they are limited to 5.5 volts on the output, so you might need to
> add a transistor driver after them.

It is common Micro controllers have at least some 5 volt tolerant inputs. I 
have seen one of the new Micro controllers have 5 volt externally which 
probably is a rather good idea even if internal logic run on lower voltage.

> ... And all of the ones in a single package
> share a ground.

It's a drawback but only if you have separate supply voltage for each channel 
otherwise they are connected anyway.


___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] (Off topic) A Question Re: BOBs and Opto-Isolators

[Kenneth Lerman]

Hello All -- especially Gene,

I've seen many posts on the lists regarding breakout boards and dealing
with issues regarding slow opto-isolators. I have a question:

Is anyone using: ISO776x 
 devices?

They are 100 MHz data rate devices, have high isolation, have six in a
package, and look like they would work just fine. The only downside might
be that they are limited to 5.5 volts on the output, so you might need to
add a transistor driver after them. And all of the ones in a single package
share a ground.

I hadn't seen them before, so I thought I'd mention them.

Regards,

Ken

Kenneth Lerman
55 Main Street
Newtown, CT 06470

___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users