GearHeads Corner
May 27, 2020, 03:46:00 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Search Login Register  
Pages: 1 [2]
  Print  
Author Topic: M3 blocks changing PoKeys PWM  (Read 4078 times)
0 Members and 1 Guest are viewing this topic.
ArtF
Administrator
Hero Member
*****
Posts: 5633



View Profile
« Reply #15 on: May 28, 2016, 07:24:07 AM »

J:

    The frequenci es they talk about are thr firing frequency . Its about two different laser setups.. 


My laser for example, has one input, TTL, it has no pot or power setting. When it shoots, it shots at full
power, but at 5khz. The PWM is a 5Khz signal, so each time it fires, it fires 100% current, but for a variable
time from 0 - 200us per shot. The net effect of this is "average" power. if you shoot 100% but only 10% of the
time, you get an effective 10% power. the faster you do this, the smoother the average. Some makers
then allowed the operation to vary in frequency . I could set my 5khz to 2khz for example.

  Much of this is because of the responce of various lasers. Mione has a rise time of about 100uS, so if pulsed
at 5khz, there is just enough time for its power to hit 100% before the pulse is turned off. This allows the
gas to recover some and the next pulse can then achieve the same power. If I were to pulse at 2Khz, Id get
actually higher power as the recovery time is longer and the amplitude would get higher, faster. 

   The timing of the shots has a lot to do with instantan eous power vs average power.

  As to frequency of 20Khz, there really is no such limit. It depends on what input your feeding the PWM.
If you feed PWM into the analogue pot input, any frequency is good, the higher the better. (The pot
after all can be considere d to have a frequency of infinity. .). The PWM here is not really used for anything
other than generatin g an "average" voltage. Rather then a pot feeding 2.5Volts from 5volts,
the PWM feeds 5volts at 20 khz at 1/2 duty cycle, the result being an average of 2.5volts. No differenc e.

In the TTL input, its similar but not the same. Here we trigger the power supply to feed, and to feed at a maximum
ma as set by the analogue pot. There is little informati on on how it works, and Ive never seen a schematic
of the high voltage supplies. BUT, there arent too many ways of controlli ng MA in a plasma circuit. That I do
have some experienc e in. In fact one of only ways I can think of, if one cannot control the electron count with a
filament, is to control the HV at a frequency . To pulse it on and off at a frequency where the average current is
 15ma or whatever you have set with the pot.

    Heres what my conclusio n is as to how it works inside a chinese laser hv supply, I could be wrong, but
all the evidence Ive seen points to this conclusio n..

   You have a pot input with an analogue value on it. This is fed internall y to a voltage to frequency
chip that shoots the HV transform er at a frequency that is proportio nal to  the desired average MA.

  This frequency is generated all the time when a voltage is applied to the pot pin, but controlle d
by a gate.

 You also have a TTL input, which is a simple switch , it turns on that gate that allows the
above frequency to feed the HV transform er.  If its on, and a voltage is causing the
volt-freq circuit to run, you get laser proportio nal to the pot voltage.

      The TTL therefore controls the length of any shot. It will have an inherent min on
time and shutoff time based on its own electroni cs and the ability of the tube to dissipate
its stored HV.

   If one ties the TTL to ground, thus turning on the switch, and simply pulses PWM
to the analogue pin, the same effect is seen, but now the electroni cs of the PWM input
, its volt->freq convertor and such will control the minimum on/off times of the laser.

    In essence then, there is no real differenc e in how one does it, other than the
on/off time variance one may see using one method over the other. BUT, you never really
need both, so one PWM is enough. If using the POT on the board, all will work the same,
just a different form of smoothing . We fool ourselves when we use the concept of
"power level" in this regard I think, what we're really controlli ng is on-time, and the effect
of a gas recovery rate, which is important only in that it affects any linearity of
power responce. (Typically a nonlinear power curve );

   If you dont allow the gas in the tube to recover, it will eventuall y get depleted. There are chemical
interacti ons in the gas which must have time to occur, so the 18ma or 22ma limit is basically that,
a timing that allows recovery.

  So whether you use the TTL input or the PWM input to control the supply its basically the same.

  I could be all wet on this, I have no detailed schematic, but observing my 3 lasers and what
happens with the K40 as they get hooked up has led me to the above mental model of how
it all works. Seems to work for me so far. Smiley

Art






 

 




Logged
Joakim
Jr. Member
**
Posts: 56


View Profile
« Reply #16 on: May 28, 2016, 08:38:01 PM »

Art:

Thank you for explainin g the inner workings of the power supply and CO2 gas lasers. I sounds right to me!
High voltage in transform ers can only be generated with an AC input or pulses, as we know it from switch mode power supplies.
The voltage and current on one side of the transform er is proportio nal with voltage and current on the other side. That make sense if the current control is handles with PWM or pulse frequency modulatio n (pulse length constant) internall y in the power supply.

If we send a PWM signal to the current control pin, it takes over the job of setting the current for the power supply, bypassing it to the transform er - could be confirmed by connectin g an oscillosc ope on the primary winding of the transform er.

Having the TTL as an on/off gate seems as the most logical way to have the supply to work both in "DC" and PWM mode.

I see the trouble in my current configura tion, pulsing TTL and having a DC value on the potentiom eter input. That input voltage is internall y converted to a "unknown" frequency / pulse modulatio n to feed to the transform er. At the same time I am chopping that signal through the gate (TTL) with a high PWM frequency to turn the laser beam on/off.
These two frequenci es are mixed and may give unexpecte d end results.. .
It is like driving both inputs with PWM signals, and one of them is unknown.

For now, I will keep the analog input from the potentiom eter and see if it gives trouble. Just received the ordered 6N136 fast optocoupl ers from eBay, I was planing to use at the potentiom eter input. They could now come in use Smiley

As I remember, both GlennD and YaNvrNo are switching the potentiom enter input between 0 and 5V, not between the two other inputs from the potentiom eter plug.
This is important as the PWM signal must be limited to avoid burning the tube with too much current.
Earlier I have measured that the DC on that input must not exceed 3.5V (15 mA).

I found this technical article on CO2 lasers http://www.repairfaq.org/sam/laserco2.htm by  Samuel M. Goldwasse r. Be aware it is technical, but it answers a lot of questions on how CO2 lasers work.

It helps a lot, having an good inner model of the workings. ..

Joakim
Logged
ArtF
Administrator
Hero Member
*****
Posts: 5633



View Profile
« Reply #17 on: May 29, 2016, 09:12:30 AM »

Hi J:

>>These two frequenci es are mixed and may give unexpecte d end results.. .
>> It is like driving both inputs with PWM signals, and one of them is unknown.

   Yes, this was my problem with it, using 2 PWMs would seem to have a sync problem, though it still works..

>>For now, I will keep the analog input from the potentiom eter and see if it gives trouble. Just received the ordered 6N136 fast optocoupl >>ers from eBay, I was planing to use at the potentiom eter input.

  Sounds good.

>>As I remember, both GlennD and YaNvrNo are switching the potentiom enter input between 0 and 5V, not between the two other inputs from the potentiom eter plug. This is important as the PWM signal must be limited to avoid burning the tube with too much current.
Earlier I have measured that the DC on that input must not exceed 3.5V (15 mA).

   True, the average DC shouldne exceed a set alue, though 15ms is low, Id suspect 20ma is more reasonabl e an upper limit for a K40.
Feeding the analog input with 5 volt PWM is fine, youd just need to not exceed the duty cycle where the voltage averages that 3.5-4 volts..

   A much safer way is to use the PWM of the pokeys to fire a transisto r as Glen does, where you could feed the pot voltage to the transisto r so that 100% PWM is only 3.5Volts average DC, thus keeping current to within limits. Some have claimed the pot voltage
imput is too slow, but tests on YaNvrNo's show that to be untrue, he runs at 20Khz on the pot input line, 0-5voltPWM, and
has done engraving s that show timing isnt an issue. It IS possible to tell Auggie with MaxPWM to limit the current anyway, if set to 80% for example, the DC voltage on the pot input will never exceed 4 volts , so the current is properly limited.

  Most of your tests in the K40 have been done on that setupo, using PWMMin and PWMMax to find the best range for his K40,
which we find fires at about 50% PWM, with no radiation below that value with TTL jumpered to ground.

 This isnt totally unusual, my Chinese laser fires at 17%, with nothing below that.. point is though, that the analogue input line
seems plenty fast enough to control on/off as well as power simultane ously, though TTL is a fine way to do it as well, it all
goes to show the flexabili ty of a K40, it can be run in many ways, and they all seem to give pretty good performan ce..

  I appreciat e your feedback, Ill bet its ready many times over the next few years by K40 users looking for data on hooking
their units to software of any kind.. Its hard informati on to find.. I know, Ive looked. Smiley

Art

Art



 
Logged
Ya-Nvr-No
Moderator
Full Member
*****
Posts: 220



View Profile WWW
« Reply #18 on: June 06, 2016, 10:43:57 AM »

I was test cutting thru some 1/4" teak and found I could do it with two passes at two different power levels per pass. So to test I wrote just a quick multiline MDI script that turns on the tickle, sets the feed, turns on my air assist, and burns twice.

m3 f8
m103 (air assist on)
g1 x1.5
g0 x0 s70 (rapid back set pwm to 70%)
g1 x1.5
g0 x0 s100 (reset back to 100%)
m105 (air assist off)
m30


I edited in the mcode script for setting speeds
 
//Sets the Spindle speed, triggered by S3000 fo example
global SetSpindl eSpeed = function( speed )
{
    //SpindleSpeed( speed );
  //Control power level using SpindleSp eed Settings
  if (speed>100){
    speed = 100;
    print("Speed set to: "+speed);
  };
    GlobalSet("LaserMaxP ower",speed);  // passing the value to the laser power variable
 
};
Logged

BobL
Gearotic Motion
Global Moderator
Hero Member
*****
Posts: 858



View Profile WWW
« Reply #19 on: June 06, 2016, 11:15:38 AM »

Thanks Ya-Nvr-No..

 I'm just about finished my home built version of a laser cutter using Auggie, so much to try? 

Cheers
Bob
Logged

Gearotic Motion
Bob
BobbyW
Jr. Member
**
Posts: 58


View Profile
« Reply #20 on: February 17, 2020, 08:09:08 AM »

Art:

I don't know that much about laser power supplies, but I was of the impressio n that the potentiom eter just gave a DC input between 0 and 5V to set a current limit for the tube. That input react very slowly and I read that it is recommend ed to use a 20kHz for that DAC.
If I measure the input voltage, it must be below 3.5 to limit current to max 17mA and give the tube long life.

Need to read some more on laser power supplies and how they work...
Joakim
Hello .
I has a lot of problem with my power source . My input impedance was very low almost 1kR and the creator decide to let some residual input voltage as 1.2V , and that made my work very difficult .
I solved finaly , i eliminate almost entire chinese cheap control and now is how should be .In that time i made a lot of researche s about , and finaly i open and modify myself .
What i found close with what i saw inside of my source was this schematic found on some russian forum .
If help someone ...
http://mir-cnc.ru/topic/5872-назначение-подстрочников-на-блоке-питания-dy10-к-т/
Thanks
Bobby


* laser source.jpg (80.96 KB, 1200x849 - viewed 38 times.)
Logged
ArtF
Administrator
Hero Member
*****
Posts: 5633



View Profile
« Reply #21 on: February 17, 2020, 01:26:06 PM »

Bobby:

 Thx, the more informati on we have the better.

Art
Logged
Pages: 1 [2]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!