[bgolash]

Hi

I might not have been clear. Before I attached the oscilloscope to the Geckos I had a horizontal trace line. I was also
able to touch the probe adj and get a square wave. I then made the probe connections to the Gecko. Tried running
the machine with wizard code for surfacing in the Y direction. Could I have shorted out the oscilloscope when I ran
the machine. So now no probes are attached to the Geckos but now there is no horizontal trace. Probing the probe adj
now shows nothing.This looks to be tricky business for me.... the clueless newbie.

Thanks

[RCaffin]

Do you get a trace when the trigger is on AUTO?
If not, check the Vertical adjust knob: is the trace out of sight at the top or bottom?
Can you still get a square wave from the probe adj point?
It's not that easy to kill a Tek.

Cheers

[bgolash]

Hi Roger

I`m getting the feeling the scope is defective. Do you have any recommendations for a new scope.
Is $300 an unrealistic amount to allocate for a scope . I am also into violin making. Scopes are used to help
tune violin plates and look at the frequencies of sound, Some of the frequencies are in the range
of 440 hertz. Would a entry level scope handle that range.

Thanks, you've been extremely helpful!!!!!

[Al_The_Man]

I`m getting the feeling the scope is defective. Do you have any recommendations for a new scope.

For run of the mill things like tuning, have you considered a USB/'Scope?
CGM-101
If you have a Laptop, you have the advantage of a scope where the common probe leads are not referenced to earth ground.
Also you have the ability to store the trace for future comparison.
They don't come with leads, so hang on to your present ones if you consider it.
Al.

[bgolash]

Hi Al
No laptop available. Are you familiar with tuning for the Geckos. I asked a question in an electronics forum
and they mentioned that there maybe high voltage. Here is a quote: I`m wondering if it applies to tuning the
Geckos

One thing about the original problem. Anything with inductance; coils, motor, electro-magnets etc, can cause
high voltage spikes when turned off. If your motor drivers are of a switched type, they may be the cause.

Second, the probes usually have setting slike 1x, 10x, x100 with 10X is the most common. The X is actually a divider,
ie a 10x probe divides the signal by ten, and if you set the scope input to X10 it will multiply the result by ten to get the
right reading. When working with inductors, coils, set the probe to 10x or better 100x until you have seen what the signal
looks like. That will give some protection to your scope from high peaks.

PS You can see an inductor and causing an inertia for electricity. Trough ann inductor, Current is "hard to start, hard to
stop", has inertia. A traditional car ignition coil is simply a magnetic coil that you run current through. When you cut the
current, it has "nowhere to go" but converts into a high voltage spike. In cars, this converts 12V to short ignition spike of several thousand volts.

[RCaffin]

Hi bgolash

That is the problem with the older oscilloscopes: either a big power supply capacitor dies or sometimes the front end nuvistor valve dies. Replacement caps are possible, but nuvistor front ends ??? (Waht is a nuvistor? A very specialised antique glass valve.)

Yes, I think it is worth investing in a small oscilloscope if you have and maintain a CNC. You have some choices:
* Old Tek or HP: lovely machines, but subject to old age decay, especially capacitors, switches, connectors. Repairs possible, but require experience.
* New small Tek 2000 series: lovely machines, but expensive.
* Equivalent Chinese/HK (stand-alone) units like Hantek (copy of Tek): probably almost as good.
* USB-based units like Pico or Hantek: good, but you would need a separate PC to drive them imho.
* Bare PCB units from China: possibly OK, but you would need to provide a case first, just for the safety of the unti.

My recommendations would be
KEEP your Tek probes ! ! ! ! They are the gold standard in this game. Collecting good probes is also worth while.
Avoid anything with a glass CRT - typically old 2nd-hand units, as the power supply will die. They are in long cases.
$300 would be quite enough today - how prices have fallen!
There are real cheapies around that only go to a few MHz - not enough for debugging a CNC. Aim for 50-100 MHz, dual channel. (Yes, more than enough for audio.)

Cheers
Roger

[Al_The_Man]

. Anything with inductance; coils, motor, electro-magnets etc, can cause
high voltage spikes when turned off. If your motor drivers are of a switched type, they may be the cause.

Second, the probes usually have setting slike 1x, 10x, x100 with 10X is the most common. The X is actually a divider,
ie a 10x probe divides the signal by ten, and if you set the scope input to X10 it will multiply the result by ten to get the
right reading. When working with inductors, coils, set the probe to 10x or better 100x until you have seen what the signal
looks like. That will give some protection to your scope from high peaks.
PS You can see an inductor and causing an inertia for electricity. Trough ann inductor, Current is "hard to start, hard to
stop", has inertia. A traditional car ignition coil is simply a magnetic coil that you run current through. When you cut the
current, it has "nowhere to go" but converts into a high voltage spike. In cars, this converts 12V to short ignition spike of several thousand volts.

Most drives whether stepper or servo are PWM based, these produce a square wave of varying width (energy) the resultant current through the motor does not follow the voltage waveform but has a mean level saw tooth in shape.
The voltage in a ignition coil is due to the rapid collapse of the stored induced energy, this also happens when you turn a relay off, as you may have noticed, there is a BEMF diode often placed across the coil to provide a very low impedance path to this collapsing field.
PWM drives also have these types of devices across the switching Mosfets or IGBT's to suppress reverse spikes.
The x10 x100 probe switch is so that very high voltages can be measured that otherwise could not be displayed on the scope directly.
Drives of all types generally have internal power supply capacitors, as well as those provided in the power supply itself.
Al.

[bgolash]

Hi all

Since it seems the old scope is having issues. I popped for a new scope. Now I have a new
project. Learning how to use the device.

[RCaffin]

Bears a strange resemblance to a Tek TDS 220. Bit of a generic standard these days. They are pretty easy to use. I would be interested in hearing how it goes.

Cheers
Roger

[bgolash]

Hi
The more research I do the more confused I get regarding the test points for the Gecko 320. Here are two pictures
of the G320, if you notice the orientation of the test points is different. I`m concerned I will be potentially
using a incorrect reference. Its been also mentioned to use the - lead of the capacitor instead of the official ground.
I`m wondering if the test point placement has changed as the drives were improved. I believe mine were purchased
in 2007

Here is the thread where I retrieved the pictures goes into detail regarding test points
http://www.cnczone.com/forums/gecko-...g-trouble.html

Attachment 269112

Attachment 269114

[RCaffin]

Yeah, looks to me as though Gecko changed the layout slightly between versions. Mine is a 320X, which is slightly different again. But minor changes like that in the layout don't matter.
No problem with using the lead on the capacitor as a test point ground, that's fine. A solid anchor.
If you can get both earth clips onto the cap lead, that is good, but space may be a little restricted. You should get one of them on anyhow.

Cheers
Roger

[bgolash]

Hi
Here it is the new scope. I made what I believe are the trigger settings. Not certain about the slope.
Both channel 1 and 2 are active. Please double check my settings.

Thanks !!

Attachment 269342

[RCaffin]

Pretty good so far.
The sweep is set to auto. That works with a repetitive signal like a sine wave, but for tuning the Geckos I think you will need to switch to 'normal' so it only triggers when DIR changes.
You may also need to use AC coupling - that is not shown on the screen under Trigger of course.
Trigger level may need adjusting slightly with the DIR signal.

To get any sort of useful display you will need to cycle the axis fairly fast, back and forth, over 1 - 2 mm. Only this way will give you the frequent DIR edges needed. You will need a small program to do this. I used both a program and sometimes a test generator. The program approach is fine.

Btw - that front panel is so close to the Tek TDS 220 I have it is not funny. The layout of the knobs is almost identical. Oh well - if it worked for Tektronix, it will work for Rigol!

Cheers
Roger

[bgolash]

Hi all

I have the scope connected but am having a tricky time getting the settings correct. I thought I`d add
this link and post from the Gecko Yahoo Forums. This info is from the master himself: Mariss


https://groups.yahoo.com/neo/groups/...s/topics/21407

1) Use scope channel 1 for the test point and channel 2 connected to DIR for the trigger.

2) Scope setup:
Set CH1 input coupling to 'GND'
Set CH1 vertical gain to '0.5V / div'
Set trigger mode to 'AUTO'
Set trigger source to 'CH2'
If you have the option, set display mode to 'CHOP'
Set trigger slope to '+' or "RISING EDGE'
If you have the option, set trigger input filter to 'NOISE REJECT'
Set timebase to '20 milliseconds / div'
Move CH1 vert pos to have the trace on the bottom graticule line
Set CH2 input coupling to 'GND'
Set CH2 vertical gain to '2V / div
Move CH2 vert pos to have the trace on the center graticule line

Have the motor turning back and forth at about 1 rev / sec, once a second.

Now set CH1 input coupling to 'DC'
Now set CH2 input coupling to 'DC'
Now set the trigger mode to 'NORMAL' The trace may disappear.
Move the trigger level until a trace appears on the scope.
Turn off the CH2 trace

You should see a small repetitive + or - going glitch within the first horizontal division.

Increase the time base speed to 5 millisec / div to comfortably see the servo loop response 'glitch'.
Increasing the motor speed will increase the size of the response glitch. The drive will fault if the glitch reaches the 0V or 3.6V limits.

[RCaffin]

Seems about right to me.
I skip some of his setting up steps, but that's OK.

> Have the motor turning back and forth at about 1 rev / sec, once a second.
That's the important bit: cycling.

Cheers
Roger

[bgolash]

Hi
Here we have a video of the machine and scope. The waveform looks to be like what I`ve seen on the web
regarding servos. Is this correct.

[RCaffin]

Hi bgolash

Very hard to tell. You have the axis cycling going fine, but the triggering on the CRO does not seem to be working properly.
You need Normal trigger, not Auto, if that is any help. And maybe slow the time base down one notch - maybe?

IF I interpret the screen shots correctly, there is a bit of ringing there which you may be able to tune down some more. Either increase the D term or decrease slightly the P term. I THINK. NOT sure.

Anyhow, the CRO seems quite nice. Useful beast.

Cheers
Roger

[bgolash]

Hi Roger

I was looking at the cncookbook site. It has good info regarding the tuning. After looking at the waveforms that
are typical. Mine do not match those waveforms. That`s good info about the ringing. That looks like the waveform I have.
I`ll work on refining the scope settings.

[RCaffin]

Thing is, you can now see the waveforms and know what to do. That's real progress.
let me know if you find setting the pots really difficult. You may need to change the gain switches away from the default. i did.

Cheers
Roger

[bgolash]

Hi All
I was somewhat frustrated with my progress tuning this drive. I sent a email to Gecko.
Mentioning that no matter what I did with the gain and damping I could not change the
the waveform.
The reply: " It looks like you are underdamped, but it would be best to detach your motor
from the actual mechanics of your machine. Right now you could be experiencing poor
damping combined with the inertial load of the machine, which can spoil the waveform while tuning.
For now, go ahead and disconnect the motor from the axis you are tuning and try it while on the
bench."
I sent another email to them and mentioned facts about the axis that are different then the x and Z.
When I converted from steppers to servos I never added timing belts and pulleys to the Y axis. Gecko informed
me that could be a part of the problem since servos like to have significant amount of gearing.
So I`ll try Geckos recommendation. I just ordered capacitors for terminal 1 and 2. The old caps could
not be reused since the new power supply has a 72 VDC supply. The new caps are rated for 100 volts
I would add that even though the axis has no gearing I haven't had fault issues with the drive.All axes
at the moment can run 3d parts at 90 ipm. Before the control box upgrade I was lucky to get a reliable
50 ipm.