11 – RF Engraving: Making Dots, How Hard Can It Be?

0:00welcome i’m glad you could join me for another session with my little tangerine tiger

0:05um i’ve been topping up with coffee because i’ve already got a headache

0:10and we haven’t even started today we’re going to look at the response time it’s one of the most

0:16interesting aspects of the machine for me because i’m really quite interested about

0:22how it can make dots dots are the basis of photo engraving

0:30and it’s very difficult on a glass tube machine to get a fast

0:36enough response out of the hv power supply to put down a single dot cleanly the hv power supply

0:43is the limiting factor on a glass tube machine it won’t be a limiting factor on this

0:49machine because this machine has got very fast switching capabilities now we’re throwing away one of the most

0:57basic parts and one of the most important parts of a glass tube machine

1:02and that’s pre-ionization now the pre-ionization phase of a glass tube

1:07machine is extremely useful because it’s a bit like this machine but

1:13totally random it produces very very high power very very high frequency pulses of energy

1:20so you look at your ammeter and you’ll say yeah but it’s only it’s only outputting one milliamp well that’s only because

1:28the amp meter cannot do anything about the high frequency it doesn’t know what’s going on all it looks at the

1:34average and the average might be one milliamp this machine does not have access to the

1:40pre-ionization zone it has got a pre-ionization zone but

1:46they operate something called a tickle current or tickle frequency even which keeps

1:52the tube operating at just below the switch on point so the moment you

1:58ask for something to happen it does not have to go through the pre-ionization phase

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

2:04it turns on instantly so we’ve got a great advantage here in one respect we’ve got a fast switching machine and

2:11we’ve got a disadvantage in another respect that we haven’t got access to the pre-ionisation zone because it’s being controlled for us and taken away

2:17from us let’s go and look to see if we can find any real positives with this machine potentially where i’m looking today we

2:24should have real positives but trust me it’s a headache

2:30and i can begin to understand why the big companies that make and use these machines

2:36hide everything behind menus i think at the end of this session you’ll probably understand why

2:41this stuff is not for the faint-hearted should we say come on in and let’s have a look at some

2:46of the problems i perceive here’s the test program that we’re going to use today because

2:51one of the most important things that i want to find out is how well the machine can produce

2:58single dots we need single dots for photo engraving yes i know i could produce ordinary scans

3:05backwards and forwards and i could switch on and off very quickly but that’s not going to tell me much about what’s going on with this machine

3:12a single dot is the most powerful thing that tells me how this machine works and how i can

3:18control it we’re really interested in what’s happening at this bottom line if you’ll notice i’ve got my settings so that it

3:24will scan from the bottom up and when i look at oscilloscope pictures which we shall do shortly

3:29we shall be looking at these dots here this is an image which is done at 254 pixels per

3:36inch and that means each one of those pixels is 0.1 of a millimeter square

3:43i know that i can get point one diameter dots and the idea of photo engraving is very

3:50simple one pixel one dot there’s nothing more complicated than that about photo engraving

3:56 it sounds like a very simple task for a pulsing machine like this to be able to get single dots the more i dive into

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

4:03how this machine could work the more complicated it might be this machine as we spoke

4:10about last time is controlled by something called a pwm signal now i’ve got a little section here that i’ve

4:17drawn pre-drawn which allows me to scale a pwm signal if i want but i’m going to do this freehand

4:23at the top here and here’s a pwm signal which is

4:32let’s just call that 10 kilohertz and it’s 50 on and 50

4:39off this is zero watts and this is 30 watts okay now there is

4:47nothing in between we can only switch this machine on to have 30 watts or nothing

4:54that’s the way that this mode of operation works how much damage can we do

5:00with a pulse the answer is very difficult to define for you because we are not

5:07holding this 50 power in one place actually not 50 as i said

5:13it is 30 watts and we’re applying 30 watts over that distance because

5:20while the power is on the head is moving so we’re spreading 30 watts

5:28over a distance that means we do not get 30 watts in one place

5:33we start sharing the 30 watts out over this distance x now i tried to

5:39explain this last time with the candle flame analogy whereas if i put my hand over a candle flame

5:45there’s 30 watts coming from there and hey it’s gonna hurt i will feel it’s warm if i do that and

5:52if i hold it still it’ll be hot because i’m allowing the energy that’s available from that candle flame

5:58to build up in one spot and do damage so the faster i move my hand over the candle flame the less it hurts

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

6:06or the other way around the faster i move my laser head the less damage i’m going to do to the

6:12material now that is a very very important concept as far as this machine is concerned

6:18because speed is a vital part of power control this system is

6:24controlled by this pwm and we’ve got various frequencies that we could use changing the frequency of the pwm signal

6:3250 on and 50 off my pulse time now the time that the power is on let’s just

6:38call that y now i’ve still got 30 watts being applied to the work that’s one way

6:45of controlling the duration of the power on this one has 50 on and 50 off this might be for

6:53example 25 kilohertz so the frequency is a lot higher here

6:59we’ve got 10 kilohertz and the frequency is lower so now we’ve got a much longer duration

7:05for our pulse we could get back to y

7:10by doing this we could change the power to 10 percent so now we’ve got

7:17a signal there a pulse which is the same length as y even though it’s a

7:2410 kilohertz signal there is no difference between this y and that y having understood the concept

7:32that we can get the same pulse out of different frequencies and speed has an effect so if the power is

7:40at 25 kilohertz 25 000 and i’ve got 50 on and 50 off the 50 percent on represents

7:4820 microseconds and if i’m running at a speed of a

7:54thousand millimeters a second then during that 20 micro seconds

8:00the head is going to move by .02 millimeters that’s what this chart is

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

8:07all about it tells me how far the head is going to move for each one of these combinations can i

8:15get to a pulse of 0.1 well first of all a thousand millimeters a

8:21second is where i’d like to do my engraving with 0.1 whether or not this system will handle it i can’t

8:26tell you but let’s just change that number there for example to 500 what effect does it have it’s gone from

8:330.02 to 0.01 now logical because we’re slowing down therefore our

8:41head is not traveling as far now we’ve still got 25 000 pulses but each pulse

8:4820 microseconds travels 0.01 let’s change the percent power from say

8:5550 to 90 now we’ve gone back up to nearly 0.02

9:04because we’re on for 36 micro seconds now and not 20 and we’re off

9:11for four because we’ve got the frequency fixed at 25 000 we’ve still got the same overall pulse

9:16length it’s just that we’ve changed the ratio of on to off to 36

9:22microseconds on and four microseconds off you think ah that means we’ve got more power because

9:28we’ve got more time on no come back to what i just talked about earlier

9:34we’re on for a longer period of time but we’re still traveling at the same speed

9:39we’re still putting down the power into the surface at the same rate if i keep the power on my dot is going

9:46to get longer that’s what that means why don’t we change the frequency let’s drop the frequency down to 10 so

9:53now we’ve got 50 milliseconds on and 50 milliseconds off have we changed

9:58anything if we run at 500 millimeters a second we’re running slower so we should put down twice as much

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

10:04power per millimeter this hurts your head a little bit to think about all these possible combinations here

10:10now look pulse distance 0.025 and you might say hang on 0.025 so in a

10:18point one pixel i’m going to get four pulses no one two

10:25three four each one of those pulses is point zero two five so i’ve got

10:32.025 high point zero two five low point zero two five high and point

10:39zero two five low i’m only going to get two of those pulses because i’ve got my remember my percent power is

10:46set to 50 so it’s 50 high 50 low so i’m going to get two pulses of 30

10:54watts into that pixel now if your head isn’t already hurting

11:00i’m going to just add another little interesting part to the equation for you ideally i’m looking for a point one

11:08pulse or am i let’s just see what happens when i put down a 0.1 pulse how do we

11:15get to a 0.1 pulse we could set the frequency to less let’s set the frequency to 5000

11:23and that should double it 5000 enter and it has we could reduce the

11:30speed to 250 millimeters a second

11:35hang on this is supposed to be a high speed engraving machine

11:40now i can’t do that i can’t compromise that that dream and go down to 250.

11:47so let’s change the power what do we need to do to the power to get up to 0.1 well the answer i think

11:55is to make it a hundred percent there we go so we’re running a frequency of five thousand hundred

12:01percent power 500 millimetres a second and we get a 0.1 pulse would that work we should be

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

12:08getting a signal that looks like this 200 milliseconds on and 200 milliseconds off

12:14and we should have 30 watts applied over that pixel for that period of time

12:23that’s exactly what we want isn’t it remember we’re starting off with the dot which is diameter 0.1 of a millimeter

12:31i don’t think this system has enough intelligence to say ah that’s a pixel and i want you to

12:36apply the power at the center of that pixel i think that what will happen is this the controller comes along and it

12:44sees the edge of that pixel and says right we need some power and then it will turn the power off

12:50when it gets to the other end of the pixel so at the moment we shall have exactly

12:55this situation here where we switch the power on have 30 watts over the duration of that

13:00pixel how much of that 30 watts is spread over that pixel depends on the speed that

13:05we’re running at are we going to finish up with the dot look we switch on here

13:11and we switch off here but here is our dot so we haven’t produced a dot

13:21we’ve produced a sausage so i’m losing the will to live almost

13:28the more we dig down into this the bigger problems we are finding we’re getting the same amount of power

13:36all the way across that scan let’s call it a scan i know it’s only a an incredibly small

13:42amount of distance but it still represents a scan we’re looking at this at such a micro level that people don’t

13:49normally look at they just look at a picture and say oh i think this picture at 600 dpi looks

13:54better than this picture at 300 dpi that’s not understanding what’s going on

14:00we’re trying to establish now can we produce a dot that somehow

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

14:06matches the frequency of these pixels how can we achieve that goal hopefully

14:13the controller is going to come along and it’s going to say right we’ve reached this edge here switch on 30 watts and so the pwm

14:21will switch on 30 watts for whatever frequency we’ve chosen i’ve chosen a frequency here which purposely matches

14:27the width of my pixel so we’ve got one two and a half power spikes

14:34per pixel then we’ve got one two and a half power spikes with

14:41nothing in it and then we’ve got two and a half power spikes again pixel two and a half with nothing in it

14:47the first question i don’t know or understand because i’m not an electronics person

14:52is whether or not the signal that i’m seeing here is the

15:00result of a background signal which looks like this being switched on

15:06and off in other words if this signal has switched on i shall see that much of the signal so

15:13it’ll switch on and then it will switch off and then it

15:19will switch on and it will switch off

15:26and on and off but in the background we’ve got this call it a time base the

15:32pwm signal is running continuously in the background and what i’m doing i’m switching the signal on and off

15:38now if that is the case then fine i’ve got matching frequency that

15:45matches perfectly my pixels because every time i reach a

15:50pixel i’ve reached a switch on point a switch on point now if i choose a frequency that doesn’t

15:57match the pitch of my pixels like this example here

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

16:02[Applause] i’ve stretched the frequency and now the first one

16:08starts off the same and so on if i’ve got a switching system where i switch on and then i switch off

16:17i’ve got my two and a half pixels i’ve got my two and a half pulses but ish but now because the

16:26signal is running in the background when it comes to switching on to the next pixel i’m not going to

16:31switch on at a rising point i’m going to switch on part way through and this one switches

16:38on at part way off and this one switches on nearly the whole distance so if i’m doing a

16:46switching on and a switching off against a background time base i will get this sort of signal

16:54like this where i will not get a repetitive signal for every pulse

17:02or one question i’ve got to ask i don’t know how this system works or

17:08does it work this way where every time i encounter a black edge

17:16here for my black pixel it switches the pwm on

17:23for nothing white it switches the pwm off and for black edge it starts the pwm

17:31offer game from scratch so i get complete signals every one of

17:36these being two and a half pulses as i’ve shown here i don’t know

17:41how this machine works if i set the frequency to 25 kilohertz 25 000

17:47speed to a thousand and the percent power to 50 percent um i should get a pulse distance

17:55in other words i should get a distance traveled by each pulse of point zero two of a millimeter

18:02where we’ve got point two high point two low point two high point two point two so we

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

18:08should have three pulses high and then nothing three pulses high

18:13nothing so that’s a pattern that i hope i’m going to be able to see fifty

18:18percent power fifty percent a thousand millimeters a second

18:25[Music] the interval doesn’t matter but the frequency does

18:3025 kilohertz there we go so we’ve got everything set up correctly

18:36now we’ll go on to the oscilloscope and i’ve got this already set up and

18:41running waiting for it to be triggered right well let’s see what we get

18:51nearly but not quite good news is the pwm is switching on for every pixel

18:59there it is the first pulse seems to be short

19:04now the time base across the bottom here is 0.1 of a millisecond 1000 millimeters a

19:11second basically a second divided by a thousand is one

19:16millisecond equals one millimeter but we’re not talking about a millimeter

19:23distance we’re talking about 0.1 of a millimeter for a pixel therefore one pixel should represent

19:310.1 milliseconds and so that is the time base that we’ve got set up on the

19:37scope so that is point one of a millisecond and should represent one pixel it looks

19:43as though we’ve lost a little bit on the first pulse the time base is switched on perfectly

19:49but it stopped early the very first pulse has got some sort of electronic delay on it

19:55this is only the electronics this is not what’s happening down in the real world and so we’ve got to

20:01relate this to that later on we’ve still got this fundamental problem of how are we going to get a pulse

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…) down

20:09that’s the right size instead of 50 power i’m putting down 100 power in other

20:14words i should have a straight line will the pwm still switch on and off

20:20so 100 on 100 off thousand millimeters a second 25 kilohertz so hopefully we’re gonna

20:27see point one on point one off point one on point one off

20:34we’ve got point one on absolutely with no time delay we’ve got an exact point one

20:40and off point one but we’ve still got our 25 kilohertz switching in the background

20:47look so we found a nice clean way of getting 0.1 pixels but it’s also a nice clean way of

20:54producing sausages so yeah we’re making headway we’re understanding how the system works but

21:00we still haven’t found a way to get a 0.1 pixel what happens if i run that at a thousand

21:06one thousand enter oh i get 0.5

21:14so if i change that to 5000

21:22i’ll get my 0.1 pulse i want a short time on and a long time off

21:28so let’s change this power here to 10 percent so why don’t i go for

21:35something like about 5 000 now i’ve got a 0.02

21:40pulse distance 0.02 so now my sausage will only be

21:49that long let’s go and have a look what signal is when we program that in

21:57that’s about twice what i expected

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

22:02i was expecting one two one two something a bit strange is going

22:10on because at five thousand render settings look what we’ve got here

22:15pre-ignition frequency 5k i’m wondering whether that’s actually a lower limit

22:20for sensible control doesn’t mean to say we might not be able to control it below that but

22:26who knows whether it goes out of control see what we can do with a slightly different value see whether i can just get away with

22:32something like about five thousand and fifty five thousand and fifty okay so now we’ll see whether or not

22:38that’s changed things ah-ha yes it was something to do with

22:44the pre-ignition frequency now it’s rather interesting that when we look at this

22:51we can see that it gets short longer longer longer and then it’s about stable

22:58so it’s taken four pulses to get longer the first pulse didn’t really make it to 30 watts okay now i’ve added another

23:07factor into my uh formula here because i want to make sure that my

23:14dots are pitched at 0.2 because i’ve got a dot then a gap then a dot then a gap so

23:24this is probably more important to me than this dimension here i mean this dimension here is now very

23:30very small and it’s virtually a circle i mean we’ve only got point zero

23:35two sausage if you like we’ll change the

23:43parameter for frequency to five

23:50point zero zero one that’s one hertz above the critical

23:56value now it works one hertz

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

24:04the difference between working and not working 0.1.2 milliseconds 200

24:09microseconds that looks like about 20 microseconds and here it is look

24:1420 microseconds so it’s doing exactly what we’ve asked it to do now but hey this is a major fiddle to get

24:21the correct parameters for doing 0.1 dots now the quick next question is

24:27what sort of dots am i going to get and i’ve got 30 watts for 20 microseconds we’re gonna have to just

24:35give it a try right now last time i was using some white card we’ve got two other things

24:40that we could try one of them i thought i might try is slate because it’s a nice hard material and

24:48it might take quite a nice fine dot

24:55well i saw some sparks which is good news well after all that faffing around

25:01we’ve got pretty rubbish results really yes we’ve got some dots nothing spectacular

25:07at all in fact particularly they’re particularly weak i’ll tell you what i’m going to do

25:14i’m going to grab some numbers out of the air we’re going to have 50 power

25:21700 millimeters a second and because 7 is a lucky number we can

25:26have seven kilohertz

25:32now out of nowhere

25:37that looked pretty pumped

25:43let’s just chuck another random number in shall we

25:49we’ll put the power on to continuous 100 percent and we’ll slow it down

25:57see if we can make it a bit darker and we’ve got a fairly good they’re not not brilliant i mean the

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

26:03black spots in the middle you can see look but we’ve also got a halo around the outside now that halo could be because i’ve not got

26:11it in focus properly so let me try and focus it up a little bit we’ve got sausage shaped dashes

26:18but to be honest they are pretty thin those lines probably closer to 0.07

26:25the sausages are a bit more than 0.1 and the gaps are a little bit less than 0.1

26:30so it’s not perfect but it’s not bad that would give quite a nice

26:36image

26:42they’re so pathetically faint that they don’t work really so my random numbers work better than

26:49the calculated numbers 500 millimeters a second

26:5810 000 hertz 100 power

27:04is working go figure now that’s not the cleanest set of dots

27:10i’ve ever seen but the black fit in the middle as opposed to the halo around the

27:16outside that’s the bit that you’re going to see so i suspect

27:22if i was to do a picture with that now i would get a pretty good picture at 254

27:28i’ve just done my scanning offset there and fixed that and that looks pretty perfect if you look there’s virtually

27:35zero over travel on it

27:41so even though it’s only going at 500 millimeters a second

27:50so that’s a picture that was prepared for a um for a glass tube so

27:59it’s doing a lot better than i thought it might be i suppose overall that was a bit of a failure because hey we’ve got a half

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

28:06decent picture it’s not a good one but it’s half decent picture much better than i ever thought it’s very dark which is good news

28:12because i was expecting it to be very light and pathetic so

28:18i think we better do a bit of investigation further and find out how this machine works because obviously

28:23i have got no idea at the moment well i think when we look at those in

28:29this light um i mean when i look up

28:35when i look at the very close this one is probably the best in terms of detail

28:41and definition it’s still not a particularly good picture in terms of the

28:46the shading and the the definition but it’s the best picture of the lot and

28:52that came out basically 100 power 10 kilohertz 800 millimeters a

29:00second these are all done at 800 millimeters a second because i can’t see the point of doing

29:05any engraving at less than that it’s you know there’s no benefit but the

29:10great thing is that surprisingly enough there is quite a lot of burn depth

29:16in these pictures and i can only come to one conclusion

29:21on the basis of what we’ve seen and that is the fact that the

29:26below 20 kilohertz is not bad 10 kilohertz seems to be a balance between the number of dots

29:34per pixel because obviously the more kilohertz we have

29:40i just have to look at that and see if we can analyze what’s going on 10 kilohertz is there

29:46anything stark and obvious about that set of numbers

29:5210 000. ninety-five eight hundred

30:00there’s nothing stand out about those numbers that said they match up with anything that i’m attempting to do

Transcript for RF Engraving: Making Dots, How Hard Can It Be? (Cont…)

30:05there’s no point one or 0.2 in there let’s turn that off well it wasn’t quite

30:13the disappointment that i was expecting this hope yet i mean on

30:19paper with a very crisp lens we’ve been able to get quite a lot of

30:25color now whether or not i should be able to change lenses and get different results

30:32something we should be testing in the future i’ve got a long way to go yet but it’s showing more promise than i

30:38anticipated plus the fact that most of the test work that i’ve been doing has been 800 millimeters a second so yeah that’s

30:45substantially faster than i could get on the glass tube machine um i don’t want to push it too far we

30:51still might get up to a thousand or more for engraving it looks possible um but

30:57i’ve got to do a lot more detailed investigation yet so there’s some good stuff to come in the next few sessions i think so

31:04thanks for your time and patience today and i’m going to go in and have a a warm top

31:10up and catch up with you in the next session bye for now