47 – Understanding Backlash and Scanning Reverse Interval (29:45)

1
00:00:05.32 –> 00:00:12.970
[Music]

2
00:00:16.13 –> 00:00:22.800
welcome to another light leg learning

3
00:00:18.6 –> 00:00:25.439
map today I’ve forsaken my cup of black

4
00:00:22.8 –> 00:00:27.060
coffee in favor of something more

5
00:00:25.439 –> 00:00:30.419
suitable to the lovely sunny weather

6
00:00:27.06 –> 00:00:30.930
outside this is just what it says on the

7
00:00:30.419 –> 00:00:34.620
can

8
00:00:30.93 –> 00:00:38.149
a curious brew it’s a lovely light beer

9
00:00:34.62 –> 00:00:41.640
that’s fermented with a champagne yeast

10
00:00:38.149 –> 00:00:42.390
it’s a mixture of wine and beer curious

11
00:00:41.64 –> 00:00:49.140
yes

12
00:00:42.39 –> 00:00:51.930
enjoyable very much so now today we’re

13
00:00:49.14 –> 00:00:55.590
going to touch on a subject that is not

14
00:00:51.93 –> 00:00:57.540
very well understood I certainly thought

15
00:00:55.59 –> 00:00:59.790
I understood it but realize that didn’t

16
00:00:57.54 –> 00:01:01.410
know too much about it until I had to

17
00:00:59.79 –> 00:01:05.850
look into it quite deeply to help

18
00:01:01.41 –> 00:01:09.600
somebody else just recently the subject

19
00:01:05.85 –> 00:01:13.860
is backlash and associated with that we

20
00:01:09.6 –> 00:01:16.560
have scanning reverse offset or scanning

21
00:01:13.86 –> 00:01:18.899
reverse interval as I think it says in

22
00:01:16.56 –> 00:01:22.319
our t-works now anybody that used

23
00:01:18.899 –> 00:01:27.329
machinery will have probably come across

24
00:01:22.319 –> 00:01:29.880
the term backlash but scanning reverse

25
00:01:27.329 –> 00:01:31.709
interval that’s a different matter so

26
00:01:29.88 –> 00:01:33.869
what we do we try and define both those

27
00:01:31.709 –> 00:01:35.819
terms and separate them out because on

28
00:01:33.869 –> 00:01:38.369
this machine they have completely

29
00:01:35.819 –> 00:01:41.459
different means although they may look

30
00:01:38.369 –> 00:01:43.499
the same now I’ve got a couple of pieces

31
00:01:41.459 –> 00:01:46.289
of scrap perspex here which I pulled out

32
00:01:43.499 –> 00:01:49.459
the pin which very conveniently will

33
00:01:46.289 –> 00:01:54.060
allow me to demonstrate what backlash is

34
00:01:49.459 –> 00:01:56.310
now here I’ve got a driving system and

35
00:01:54.06 –> 00:01:59.880
here I’ve got something that I want to

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

36
00:01:56.31 –> 00:02:03.630
move now this could be the head of this

37
00:01:59.88 –> 00:02:05.249
machine and this could be I don’t want

38
00:02:03.63 –> 00:02:08.880
to call it the belt but let’s call it

39
00:02:05.249 –> 00:02:10.890
the drive system when the drive system

40
00:02:08.88 –> 00:02:12.700
could be right the way back to the

41
00:02:10.89 –> 00:02:16.810
stepper motor

42
00:02:12.7 –> 00:02:21.580
the pulleys the belt anything it’s just

43
00:02:16.81 –> 00:02:24.790
a system now if I move this system X

44
00:02:21.58 –> 00:02:29.310
amount look what happens to the blue

45
00:02:24.79 –> 00:02:31.780
rectangle it moves X amount as well

46
00:02:29.31 –> 00:02:34.930
now that’s brilliant while I’m moving in

47
00:02:31.78 –> 00:02:39.130
that direction but when I reverse and

48
00:02:34.93 –> 00:02:43.440
come back in the other direction and get

49
00:02:39.13 –> 00:02:48.540
back to my start point I haven’t brought

50
00:02:43.44 –> 00:02:48.540
the Blue Square back to its start point

51
00:02:49.65 –> 00:02:58.959
it’s failing to make this dimension here

52
00:02:53.799 –> 00:02:59.790
Y which is the same dimension that we’ve

53
00:02:58.959 –> 00:03:02.790
got here

54
00:02:59.79 –> 00:03:02.790
y

55
00:03:04.069 –> 00:03:10.959
so the clearance in the system

56
00:03:06.849 –> 00:03:14.799
represents something called backlash now

57
00:03:10.959 –> 00:03:18.620
this is all very well provided you have

58
00:03:14.799 –> 00:03:21.950
friction in your system because whenever

59
00:03:18.62 –> 00:03:25.099
I move forward now I shall pick up and

60
00:03:21.95 –> 00:03:27.170
take that head to there and when I move

61
00:03:25.099 –> 00:03:29.150
backwards I shall pick it up and

62
00:03:27.17 –> 00:03:32.120
predictably bring the head back to that

63
00:03:29.15 –> 00:03:34.639
same position if I get a predictable

64
00:03:32.12 –> 00:03:38.060
backlash like that I could probably use

65
00:03:34.639 –> 00:03:41.200
some sort of electronic correction to

66
00:03:38.06 –> 00:03:45.950
deal with it but the problem is this is

67
00:03:41.2 –> 00:03:49.250
mechanical backlash and it’s of a form

68
00:03:45.95 –> 00:03:51.859
which is uncertain because look I’ve got

69
00:03:49.25 –> 00:03:54.560
all this uncertainty in here so if I

70
00:03:51.859 –> 00:03:56.889
move from this position to this position

71
00:03:54.56 –> 00:03:59.780
here quickly

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

72
00:03:56.889 –> 00:04:03.109
look what happens the inertia of this

73
00:03:59.78 –> 00:04:06.439
piece has taken it away from its driving

74
00:04:03.109 –> 00:04:08.449
edge into a zone of uncertainty we do

75
00:04:06.439 –> 00:04:11.239
not know where the head actually has

76
00:04:08.449 –> 00:04:13.699
finished up it’s anywhere in this Y

77
00:04:11.239 –> 00:04:15.650
dimension and we can’t correct for that

78
00:04:13.699 –> 00:04:19.120
electronically because it’s totally

79
00:04:15.65 –> 00:04:21.680
random and uncertain this is mechanical

80
00:04:19.12 –> 00:04:24.220
backlash and something that you will

81
00:04:21.68 –> 00:04:26.750
have to deal with on the machine itself

82
00:04:24.22 –> 00:04:29.270
you’ll have to find out where that error

83
00:04:26.75 –> 00:04:31.099
is if you come across such an error and

84
00:04:29.27 –> 00:04:33.289
you know you’re gonna have to physically

85
00:04:31.099 –> 00:04:35.570
do something to fix it now this

86
00:04:33.289 –> 00:04:37.970
particular machine it’s turned on at the

87
00:04:35.57 –> 00:04:40.340
moment and this is a servo controlled

88
00:04:37.97 –> 00:04:42.320
machine with very heavy-duty stepper

89
00:04:40.34 –> 00:04:46.400
motors that locks into position

90
00:04:42.32 –> 00:04:48.440
extremely well and I can move of putting

91
00:04:46.4 –> 00:04:52.010
quite a lot of force on this head and I

92
00:04:48.44 –> 00:04:54.620
cannot move it but that’s good because

93
00:04:52.01 –> 00:04:57.229
if I could move it if I could feel any

94
00:04:54.62 –> 00:04:59.870
sort of wobble in there at all a little

95
00:04:57.229 –> 00:05:02.900
bit of stitches then it means that I’ve

96
00:04:59.87 –> 00:05:04.789
got this mechanical backlash there’s no

97
00:05:02.9 –> 00:05:07.039
wobble on that head at all so I have

98
00:05:04.789 –> 00:05:08.960
zero mechanical backlash in the system

99
00:05:07.039 –> 00:05:11.300
which is brilliant and that comes about

100
00:05:08.96 –> 00:05:13.520
probably by mainly tensioning up the

101
00:05:11.3 –> 00:05:14.160
belt to get rid of the backlash there is

102
00:05:13.52 –> 00:05:16.020
another sort

103
00:05:14.16 –> 00:05:18.600
backlash that occurs on this machine

104
00:05:16.02 –> 00:05:20.580
which I believe is what the backlash

105
00:05:18.6 –> 00:05:24.120
compensation on this machine is all

106
00:05:20.58 –> 00:05:26.280
about and it basically comes down to a

107
00:05:24.12 –> 00:05:30.990
design weakness that’s built into all of

108
00:05:26.28 –> 00:05:33.600
these belt driven machines now at this

109
00:05:30.99 –> 00:05:37.080
end of the system here we’ve got the

110
00:05:33.6 –> 00:05:39.270
drive for this X now this one this

111
00:05:37.08 –> 00:05:41.070
smaller pulley has probably got about 24

112
00:05:39.27 –> 00:05:43.500
teeth on it we’re gonna have a little

113
00:05:41.07 –> 00:05:45.990
very close look at what’s going on just

114
00:05:43.5 –> 00:05:48.810
here where my finger is now I’m gonna

115
00:05:45.99 –> 00:05:52.020
run a scanning program and I want you to

116
00:05:48.81 –> 00:05:54.210
watch what happens to this belt at the

117
00:05:52.02 –> 00:05:56.430
top here watch it in relation to

118
00:05:54.21 –> 00:05:58.950
something in the background and you’ll

119
00:05:56.43 –> 00:06:00.780
see this belt rise up and go down and

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

120
00:05:58.95 –> 00:06:03.230
come up and go down as we change

121
00:06:00.78 –> 00:06:03.230
direction

122
00:06:03.56 –> 00:06:10.430
so I think you can probably clearly see

123
00:06:06.62 –> 00:06:13.220
the belt rising as it goes onto the

124
00:06:10.43 –> 00:06:15.400
pulley and falling as it comes off the

125
00:06:13.22 –> 00:06:15.400
pulley

126
00:06:16.1 –> 00:06:19.880
you can see it by watching the belt

127
00:06:18.35 –> 00:06:23.140
rising and falling just at the back

128
00:06:19.88 –> 00:06:23.140
there relative to something

129
00:06:23.4 –> 00:06:27.860
no the belt on this machine is a

130
00:06:25.53 –> 00:06:30.120
fixed-length it can’t change its length

131
00:06:27.86 –> 00:06:31.710
now although I’ve shown you the top of

132
00:06:30.12 –> 00:06:33.690
the belt because it’s much easier to see

133
00:06:31.71 –> 00:06:35.520
that exactly the same thing is happening

134
00:06:33.69 –> 00:06:37.889
on the bottom of the belt where the head

135
00:06:35.52 –> 00:06:40.979
is attached to so when this pulley

136
00:06:37.889 –> 00:06:43.199
changes direction and this belt all of a

137
00:06:40.979 –> 00:06:45.479
sudden decides it’s going to do that I

138
00:06:43.199 –> 00:06:48.259
hope you can see what’s happening to the

139
00:06:45.479 –> 00:06:51.210
head as that belt moves up and down

140
00:06:48.259 –> 00:06:54.090
that’s what happens as the belt rises

141
00:06:51.21 –> 00:06:56.370
and falls can you see how it’s changing

142
00:06:54.09 –> 00:06:57.360
the position of the head now it won’t be

143
00:06:56.37 –> 00:06:59.190
a huge amount

144
00:06:57.36 –> 00:07:02.039
I mean I’m grossly exaggerating it with

145
00:06:59.19 –> 00:07:03.990
what I’m doing but the error is there

146
00:07:02.039 –> 00:07:07.250
and that’s what I believe

147
00:07:03.99 –> 00:07:10.110
really backlash is all about it’s not

148
00:07:07.25 –> 00:07:13.440
really backlash there are two problems

149
00:07:10.11 –> 00:07:17.820
associated with these pulleys number one

150
00:07:13.44 –> 00:07:21.270
is that the actual teeth on the belt are

151
00:07:17.82 –> 00:07:24.509
not a perfect match for the tooth form

152
00:07:21.27 –> 00:07:27.300
that’s in there there is a very small

153
00:07:24.509 –> 00:07:29.070
amount of clearance between these teeth

154
00:07:27.3 –> 00:07:32.400
and the belt

155
00:07:29.07 –> 00:07:35.760
sideways because the belt has to locate

156
00:07:32.4 –> 00:07:38.520
right in the bottom of its groove and on

157
00:07:35.76 –> 00:07:41.610
the top of the teeth it can’t locate on

158
00:07:38.52 –> 00:07:44.280
the tapered sides of the teeth both

159
00:07:41.61 –> 00:07:47.580
sides at once it has to locate on one

160
00:07:44.28 –> 00:07:51.060
side or the other now okay that backlash

161
00:07:47.58 –> 00:07:53.340
that clearance is very small but that’s

162
00:07:51.06 –> 00:07:55.199
one sort of clearance that exists in

163
00:07:53.34 –> 00:07:57.659
this system and there’s nothing we can

164
00:07:55.199 –> 00:07:59.639
do about it the other thing that exists

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

165
00:07:57.659 –> 00:08:02.370
in this self system is what you’ve just

166
00:07:59.639 –> 00:08:05.849
seen me demonstrate here belt climb and

167
00:08:02.37 –> 00:08:08.820
that’s because the first tooth being

168
00:08:05.849 –> 00:08:11.550
rubber or any material it will tend to

169
00:08:08.82 –> 00:08:13.590
get stuck on the side of the on the

170
00:08:11.55 –> 00:08:16.409
flank of the pulley tooth and it will

171
00:08:13.59 –> 00:08:18.240
lift up and it’s not until it gets 2 or

172
00:08:16.409 –> 00:08:20.639
3 teeth into engagement that it fully

173
00:08:18.24 –> 00:08:23.099
sits down where it should do around the

174
00:08:20.639 –> 00:08:24.870
pulley and that’s what’s causing this

175
00:08:23.099 –> 00:08:27.840
effect now

176
00:08:24.87 –> 00:08:30.389
it’s a fairly small error and it’s not a

177
00:08:27.84 –> 00:08:32.430
big issue for the simplicity of this

178
00:08:30.389 –> 00:08:35.640
drive but this drive system was never

179
00:08:32.43 –> 00:08:37.979
designed for very precision control of

180
00:08:35.64 –> 00:08:41.849
systems like this it was designed for

181
00:08:37.979 –> 00:08:44.760
power transmission in one direction only

182
00:08:41.849 –> 00:08:47.880
and it’s excellent at doing that job and

183
00:08:44.76 –> 00:08:50.010
it’s 98 99 percent good for doing this

184
00:08:47.88 –> 00:08:52.290
job but there are some little weaknesses

185
00:08:50.01 –> 00:08:55.380
that I’ve just demonstrated to you now

186
00:08:52.29 –> 00:09:00.200
this offset that occurs every time we

187
00:08:55.38 –> 00:09:02.940
change direction is predictable and

188
00:09:00.2 –> 00:09:06.510
therefore we can do something about it

189
00:09:02.94 –> 00:09:07.770
now in case you don’t believe me what

190
00:09:06.51 –> 00:09:11.450
I’m going to do is just a very quick

191
00:09:07.77 –> 00:09:11.450
demonstration at the moment for you

192
00:09:13.59 –> 00:09:25.620
we’re going to set the maximum power to

193
00:09:16.34 –> 00:09:30.660
about 10% I’ve very carefully driven the

194
00:09:25.62 –> 00:09:35.460
head so that the X is exactly 300 and

195
00:09:30.66 –> 00:09:41.460
I’m going to draw a line down there like

196
00:09:35.46 –> 00:09:43.710
that okay so that’s a line at 300 in X

197
00:09:41.46 –> 00:09:49.140
I’m now going to press the reset button

198
00:09:43.71 –> 00:09:51.930
and I want you to note which way the

199
00:09:49.14 –> 00:09:55.640
head is moving as it stops on that line

200
00:09:51.93 –> 00:09:56.790
because that’s the origin that I set

201
00:09:55.64 –> 00:09:59.339
okay

202
00:09:56.79 –> 00:10:00.270
it came in cleanly and stopped it didn’t

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

203
00:09:59.339 –> 00:10:03.540
go like that

204
00:10:00.27 –> 00:10:06.029
it stopped on the line I’m now going to

205
00:10:03.54 –> 00:10:11.670
move down the tree down the line a

206
00:10:06.029 –> 00:10:13.200
little bit and I’m gonna do a pulse I’m

207
00:10:11.67 –> 00:10:17.370
now going to move down the line a little

208
00:10:13.2 –> 00:10:19.500
bit further and I’m going to again press

209
00:10:17.37 –> 00:10:24.240
the origin button bearing in mind the

210
00:10:19.5 –> 00:10:26.160
keyboard still says exactly 300 and I’m

211
00:10:24.24 –> 00:10:28.830
now going to run a very small program

212
00:10:26.16 –> 00:10:33.860
and I want you to watch which direction

213
00:10:28.83 –> 00:10:33.860
the head arrives back at this line

214
00:10:34.03 –> 00:10:40.730
they were dope very quick but it arrived

215
00:10:37.13 –> 00:10:44.210
back in that direction the opposite

216
00:10:40.73 –> 00:10:47.990
direction to which it reset so let’s

217
00:10:44.21 –> 00:10:49.370
press the pulse button again and let’s

218
00:10:47.99 –> 00:10:53.060
have a look at the results that we’ve

219
00:10:49.37 –> 00:10:56.210
produced okay so here’s our scan line

220
00:10:53.06 –> 00:10:58.010
and just to put some sort of proportion

221
00:10:56.21 –> 00:11:01.970
onto it we will just check what

222
00:10:58.01 –> 00:11:05.210
dimension the line width is and there’s

223
00:11:01.97 –> 00:11:08.390
a point to reference line there point to

224
00:11:05.21 –> 00:11:10.670
wide and I think we can yeah it’s pretty

225
00:11:08.39 –> 00:11:12.860
reasonable about a point to wide line

226
00:11:10.67 –> 00:11:17.240
now when I did the drawing of that line

227
00:11:12.86 –> 00:11:20.180
I came in from a reset position and when

228
00:11:17.24 –> 00:11:23.330
I did that dot I came in from a reset

229
00:11:20.18 –> 00:11:25.390
position so technically the dot should

230
00:11:23.33 –> 00:11:29.360
sit on top of the line and it does

231
00:11:25.39 –> 00:11:32.030
within a very very small amount but when

232
00:11:29.36 –> 00:11:36.200
we did the second dot we came in from

233
00:11:32.03 –> 00:11:38.630
the opposite direction and I think you

234
00:11:36.2 –> 00:11:41.810
can clearly see know that we’re half a

235
00:11:38.63 –> 00:11:44.480
dot out of position and because the line

236
00:11:41.81 –> 00:11:47.480
thickness is point two I think we can

237
00:11:44.48 –> 00:11:51.530
probably safely say that that dot is 0.1

238
00:11:47.48 –> 00:11:53.990
of a millimeter out of position now

239
00:11:51.53 –> 00:11:56.990
that’s a point one millimetre

240
00:11:53.99 –> 00:11:59.570
backlash because coming in from one

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

241
00:11:56.99 –> 00:12:02.570
direction it does not settle in the same

242
00:11:59.57 –> 00:12:07.190
position as when it comes in from that

243
00:12:02.57 –> 00:12:12.380
direction the keyboard says 300 for both

244
00:12:07.19 –> 00:12:17.420
of these positions they can’t both be

245
00:12:12.38 –> 00:12:21.730
right so this is where we have to now go

246
00:12:17.42 –> 00:12:25.610
in and look at the backlash correction

247
00:12:21.73 –> 00:12:31.570
now before I started this I connected my

248
00:12:25.61 –> 00:12:34.490
PC up and I have set my backlash to zero

249
00:12:31.57 –> 00:12:36.160
so there is an X backslash hiding here

250
00:12:34.49 –> 00:12:38.770
in the user settings

251
00:12:36.16 –> 00:12:40.950
so what I’m now going to do is to change

252
00:12:38.77 –> 00:12:40.950
that

253
00:12:41.95 –> 00:12:44.460
[Music]

254
00:12:43.36 –> 00:12:47.120
two point one

255
00:12:44.46 –> 00:12:50.559
[Music]

256
00:12:47.12 –> 00:12:50.559
nope yep point one

257
00:12:51.08 –> 00:12:55.990
and I’m going to write that back to the

258
00:12:52.76 –> 00:12:55.990
machine point one

259
00:12:56.14 –> 00:13:02.110
and now we run the same test again and

260
00:12:58.51 –> 00:13:03.779
we’ll see whether we’ve corrected so my

261
00:13:02.11 –> 00:13:07.060
ex is still at 300

262
00:13:03.779 –> 00:13:09.519
we’ve not origin set I will do a reset

263
00:13:07.06 –> 00:13:12.209
to make sure I bring the head in from

264
00:13:09.519 –> 00:13:12.209
this direction

265
00:13:17.06 –> 00:13:23.509
[Music]

266
00:13:18.229 –> 00:13:23.509
and I’ll draw my line at 300 in X

267
00:13:25.07 –> 00:13:35.210
[Music]

268
00:13:27.62 –> 00:13:37.370
and do I need to go back and set I can

269
00:13:35.21 –> 00:13:39.760
put a dot on there because I’ve already

270
00:13:37.37 –> 00:13:42.260
come in from the right hand direction

271
00:13:39.76 –> 00:13:45.470
we’ll go down and we’ll set a new origin

272
00:13:42.26 –> 00:13:47.990
point just they’re still at 300 on the

273
00:13:45.47 –> 00:13:50.800
keyboard and then we’ll do my little

274
00:13:47.99 –> 00:13:50.800
test

275
00:13:53.24 –> 00:14:00.559
and frame and now I’ll do another pose

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

276
00:13:58.389 –> 00:14:02.089
well here’s our first attempt at

277
00:14:00.559 –> 00:14:05.509
correcting the backlash

278
00:14:02.089 –> 00:14:08.839
there’s the dot coming in from the right

279
00:14:05.509 –> 00:14:12.350
and then here is the corrected result

280
00:14:08.839 –> 00:14:15.139
and it’s not bad but it looks as though

281
00:14:12.35 –> 00:14:17.420
we’ve overcorrected it with a point one

282
00:14:15.139 –> 00:14:22.160
backlash setting I’m now going to change

283
00:14:17.42 –> 00:14:25.540
the backlash to point zero five and now

284
00:14:22.16 –> 00:14:25.540
I’m going to run my little program

285
00:14:26.61 –> 00:14:30.170
press the post button again

286
00:14:28.8 –> 00:14:31.579
[Music]

287
00:14:30.17 –> 00:14:34.639
then let’s have a look what we’ve

288
00:14:31.579 –> 00:14:40.040
produced now so here’s our reference

289
00:14:34.639 –> 00:14:43.760
mark coming in from the right and here’s

290
00:14:40.04 –> 00:14:45.699
our mark coming in from the left I think

291
00:14:43.76 –> 00:14:51.350
you’ll see they’re both pretty well

292
00:14:45.699 –> 00:14:52.010
matched so 0.05 is my backlash on this

293
00:14:51.35 –> 00:14:54.320
machine

294
00:14:52.01 –> 00:14:57.430
now point oh five is about the thickness

295
00:14:54.32 –> 00:14:59.660
of a human hair that’s not a great deal

296
00:14:57.43 –> 00:15:03.079
it’s something you should be aware of

297
00:14:59.66 –> 00:15:04.850
this machine is particularly good not

298
00:15:03.079 –> 00:15:08.360
all machines are this good it really

299
00:15:04.85 –> 00:15:10.880
depends on the belts and the police and

300
00:15:08.36 –> 00:15:14.240
the number of teeth on your pulley that

301
00:15:10.88 –> 00:15:15.709
will determine this sort of offset now

302
00:15:14.24 –> 00:15:17.990
although I have to copy this program

303
00:15:15.709 –> 00:15:19.940
onto my PC next to the machine and run

304
00:15:17.99 –> 00:15:22.040
it from the machine while I’m in the

305
00:15:19.94 –> 00:15:24.860
office I will actually show you how to

306
00:15:22.04 –> 00:15:26.360
draw the next program which is a very

307
00:15:24.86 –> 00:15:29.720
important test program that you should

308
00:15:26.36 –> 00:15:31.610
probably keep on your machine so we’ve

309
00:15:29.72 –> 00:15:33.560
got to do a square and we’ll hold down

310
00:15:31.61 –> 00:15:35.120
the control key to make it a square and

311
00:15:33.56 –> 00:15:38.209
then I’m going to put a marquee around

312
00:15:35.12 –> 00:15:40.490
that that’s a control C for control copy

313
00:15:38.209 –> 00:15:42.230
and then control V we’ll put it at the

314
00:15:40.49 –> 00:15:45.290
cursor there we go

315
00:15:42.23 –> 00:15:50.050
so now I’ve produced a copy now the

316
00:15:45.29 –> 00:15:53.089
first one I’m going to do is a reference

317
00:15:50.05 –> 00:15:57.140
square and we’re going to put that onto

318
00:15:53.089 –> 00:16:00.980
a blue layer and we’re going to make

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

319
00:15:57.14 –> 00:16:04.160
that reference square let’s close the

320
00:16:00.98 –> 00:16:09.070
padlock click we’ll set the X or Y

321
00:16:04.16 –> 00:16:09.070
dimension up here to 20 millimeters

322
00:16:09.91 –> 00:16:15.890
we’ve just seen that the line that we’re

323
00:16:12.89 –> 00:16:20.930
likely to produce when we burn a line is

324
00:16:15.89 –> 00:16:25.420
about 2.2 diameter or point too wide now

325
00:16:20.93 –> 00:16:28.790
if I draw a line there will be a point 1

326
00:16:25.42 –> 00:16:30.980
round end on the end of the line that

327
00:16:28.79 –> 00:16:33.680
goes point 1 further than the line

328
00:16:30.98 –> 00:16:36.440
itself because the line will stop in the

329
00:16:33.68 –> 00:16:38.269
center of a dot and the other half dot

330
00:16:36.44 –> 00:16:41.959
will hang out beyond the end of the line

331
00:16:38.269 –> 00:16:43.370
that I’ve drawn so what I want to do is

332
00:16:41.959 –> 00:16:46.520
produce a gap

333
00:16:43.37 –> 00:16:50.270
I want to draw lines inside that square

334
00:16:46.52 –> 00:16:55.430
and I need the lines to be about point

335
00:16:50.27 –> 00:16:57.620
one clear on left and right so I’ve got

336
00:16:55.43 –> 00:17:00.529
to allow knowing that this is 20

337
00:16:57.62 –> 00:17:02.720
millimeters square I’ve got to take

338
00:17:00.529 –> 00:17:04.939
point 1 off each side for line thickness

339
00:17:02.72 –> 00:17:06.770
which brings it back down to 19 point

340
00:17:04.939 –> 00:17:08.900
eight then I’m going to take another

341
00:17:06.77 –> 00:17:13.030
point one off each side for a gap which

342
00:17:08.9 –> 00:17:15.199
brings it down to nineteen point six and

343
00:17:13.03 –> 00:17:18.800
then I’m going to have to take another

344
00:17:15.199 –> 00:17:21.740
point one off each end of the line which

345
00:17:18.8 –> 00:17:23.990
brings it down to nineteen point four so

346
00:17:21.74 –> 00:17:26.900
I want lines that are nineteen point

347
00:17:23.99 –> 00:17:29.990
four millimeters long so what I’m going

348
00:17:26.9 –> 00:17:34.190
to do now is to set this rep this square

349
00:17:29.99 –> 00:17:37.550
to nineteen point four okay now that one

350
00:17:34.19 –> 00:17:44.240
is black and we’re going to set that

351
00:17:37.55 –> 00:17:46.190
layer to scan not cut so we’re gonna set

352
00:17:44.24 –> 00:17:48.110
that layer to scan and we’re going to

353
00:17:46.19 –> 00:17:50.000
run that at four hundred millimeters a

354
00:17:48.11 –> 00:17:53.090
second or the fastest your speed will

355
00:17:50.0 –> 00:17:56.420
your axis will go maybe it’ll go 500 but

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

356
00:17:53.09 –> 00:18:05.270
400 is good enough 400 millimeters a

357
00:17:56.42 –> 00:18:08.870
second and probably 25% maybe 30% power

358
00:18:05.27 –> 00:18:13.730
to get a reasonable depth of burn at

359
00:18:08.87 –> 00:18:16.420
that speed 30% it may be too dark we

360
00:18:13.73 –> 00:18:22.100
might have to come back in it at that

361
00:18:16.42 –> 00:18:24.559
3030 scan yes yes yes and the interval I

362
00:18:22.1 –> 00:18:26.780
want the interval this time to be quite

363
00:18:24.559 –> 00:18:32.300
large point five because I want to see

364
00:18:26.78 –> 00:18:34.520
gaps between the lines and X swing so

365
00:18:32.3 –> 00:18:36.980
that we get scans in both directions we

366
00:18:34.52 –> 00:18:39.920
don’t want independent output so nothing

367
00:18:36.98 –> 00:18:44.510
in there so that’s a nice clean set of

368
00:18:39.92 –> 00:18:49.490
parameters 430 and X wink with 0.5

369
00:18:44.51 –> 00:18:50.780
interval now the next one is a cut now

370
00:18:49.49 –> 00:18:52.700
we’ve got to put the cut on the bottom

371
00:18:50.78 –> 00:18:57.200
because well that will happen naturally

372
00:18:52.7 –> 00:19:00.409
kappa rameters 200 that’s reasonable

373
00:18:57.2 –> 00:19:04.549
power probably not much more than about

374
00:19:00.409 –> 00:19:05.929
15 maybe 20 right so I’ve got one square

375
00:19:04.549 –> 00:19:08.750
which is going to be engraved with a

376
00:19:05.929 –> 00:19:10.789
series of lines and those lines I want

377
00:19:08.75 –> 00:19:12.740
to appear inside that reference square

378
00:19:10.789 –> 00:19:14.690
so what I’ve now got to do is make sure

379
00:19:12.74 –> 00:19:16.399
these two things line up with each other

380
00:19:14.69 –> 00:19:20.750
so we’ll make them line up horizontally

381
00:19:16.399 –> 00:19:22.909
and vertically and there we go one

382
00:19:20.75 –> 00:19:26.750
inside the other okay now before I’ve

383
00:19:22.909 –> 00:19:29.029
run I’ve been to config system settings

384
00:19:26.75 –> 00:19:32.179
and I have removed the tick from the

385
00:19:29.029 –> 00:19:34.399
scanning offset so there’s no scan

386
00:19:32.179 –> 00:19:35.840
offset involved at the moment and we’re

387
00:19:34.399 –> 00:19:38.659
just going to see what the natural

388
00:19:35.84 –> 00:19:41.110
machine does at 400 millimeters per

389
00:19:38.659 –> 00:19:41.110
second

390
00:19:46.62 –> 00:19:51.140
[Music]

391
00:19:48.73 –> 00:19:53.030
we’ll do a series of tests

392
00:19:51.14 –> 00:19:54.200
I will check them and then we’ll go and

393
00:19:53.03 –> 00:19:56.300
we’ll go and have a look under the

394
00:19:54.2 –> 00:19:59.330
microscope now we need to look carefully

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

395
00:19:56.3 –> 00:20:04.250
to work out which way the first line is

396
00:19:59.33 –> 00:20:10.460
drawn that’s very important I didn’t see

397
00:20:04.25 –> 00:20:12.230
that because it happened so quickly but

398
00:20:10.46 –> 00:20:15.170
I should be able to sort it out when I

399
00:20:12.23 –> 00:20:17.210
come to look at the video I know from a

400
00:20:15.17 –> 00:20:21.050
little bit of experience that when I

401
00:20:17.21 –> 00:20:26.630
look at this zero offset picture the

402
00:20:21.05 –> 00:20:28.250
error that I see is not very large now

403
00:20:26.63 –> 00:20:29.540
I’ll draw this in thick marker pen so

404
00:20:28.25 –> 00:20:33.710
you can see what I’m talking about

405
00:20:29.54 –> 00:20:36.950
but my first my square runs down like

406
00:20:33.71 –> 00:20:41.740
that and my first line starts there and

407
00:20:36.95 –> 00:20:41.740
it finishes up with a gap

408
00:20:41.799 –> 00:20:48.489
on the other side and then the second

409
00:20:45.34 –> 00:20:51.519
line starts there and finishes up with a

410
00:20:48.489 –> 00:20:55.509
gap on that side and the third line

411
00:20:51.519 –> 00:20:59.289
starts there and the fourth line starts

412
00:20:55.509 –> 00:21:01.960
there now that’s the offset problem now

413
00:20:59.289 –> 00:21:06.789
if my first line is going in that

414
00:21:01.96 –> 00:21:11.769
direction and it’s touching there I need

415
00:21:06.789 –> 00:21:14.619
to put a positive plus offset to move it

416
00:21:11.769 –> 00:21:16.690
across and at the same time what that

417
00:21:14.619 –> 00:21:19.450
will do that will also shift that line

418
00:21:16.69 –> 00:21:22.989
when I come in the other direction away

419
00:21:19.45 –> 00:21:26.350
from the edge as well so hopefully if I

420
00:21:22.989 –> 00:21:30.029
put a small positive offset in it will

421
00:21:26.35 –> 00:21:33.340
send my patterns the way that I expect

422
00:21:30.029 –> 00:21:34.960
now my first attempt because I know that

423
00:21:33.34 –> 00:21:38.409
the increments are quite a bit going to

424
00:21:34.96 –> 00:21:42.730
be very small I’m going to put something

425
00:21:38.409 –> 00:21:44.499
like about 0.02 in to start with and

426
00:21:42.73 –> 00:21:47.230
then we’ll put some larger increments in

427
00:21:44.499 –> 00:21:49.330
to see what happens

428
00:21:47.23 –> 00:21:52.090
so here we are in the config system

429
00:21:49.33 –> 00:21:54.519
setting window and I’ve already got 400

430
00:21:52.09 –> 00:21:57.610
millimeters per second in my table and

431
00:21:54.519 –> 00:21:59.320
I’ve set that to point zero two

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

432
00:21:57.61 –> 00:22:02.409
let me just show you how I’ve done that

433
00:21:59.32 –> 00:22:04.450
if I double click on it it brings this

434
00:22:02.409 –> 00:22:06.429
little window up where I get the

435
00:22:04.45 –> 00:22:09.669
opportunity to change the settings now

436
00:22:06.429 –> 00:22:12.850
I’ve put point o2 in there and say okay

437
00:22:09.669 –> 00:22:16.809
and it comes back into my table let’s

438
00:22:12.85 –> 00:22:18.600
point o2 and that’s all it is provided

439
00:22:16.809 –> 00:22:21.610
I’ve got my tick at the top corner here

440
00:22:18.6 –> 00:22:24.909
when I specified 400 millimeters a

441
00:22:21.61 –> 00:22:30.639
second in a scanning mode it will put

442
00:22:24.909 –> 00:22:33.669
that offset in if I put 399 in as my

443
00:22:30.639 –> 00:22:35.769
scanning speed it will be ignored it

444
00:22:33.669 –> 00:22:40.000
will be as though that tick is not there

445
00:22:35.769 –> 00:22:44.159
so you have to put in a number for every

446
00:22:40.0 –> 00:22:44.159
scanning speed that you plan to use

447
00:22:51.01 –> 00:22:55.120
[Music]

448
00:22:53.59 –> 00:22:56.890
well I think when we look at that under

449
00:22:55.12 –> 00:22:58.540
the microscope you’ll see that we can’t

450
00:22:56.89 –> 00:23:02.890
get much more perfect than I’ve got

451
00:22:58.54 –> 00:23:05.470
there so I’ve now set my backlash and my

452
00:23:02.89 –> 00:23:07.360
scanning offset the backlash won’t

453
00:23:05.47 –> 00:23:11.020
change but the scanning offset will

454
00:23:07.36 –> 00:23:14.320
change for different speeds so let me

455
00:23:11.02 –> 00:23:18.430
just prove by putting in a silly number

456
00:23:14.32 –> 00:23:20.350
either side of my perfect results this

457
00:23:18.43 –> 00:23:25.240
one has now been corrected with a naught

458
00:23:20.35 –> 00:23:27.160
point oh two millimeters now point out

459
00:23:25.24 –> 00:23:30.160
two millimeters there’s only half hair

460
00:23:27.16 –> 00:23:33.060
but it makes a huge difference to the

461
00:23:30.16 –> 00:23:33.060
look of this pattern

462
00:23:35.42 –> 00:23:40.710
so now what I’m going to do I’m going to

463
00:23:37.89 –> 00:23:42.720
go plus one millimeter and – one

464
00:23:40.71 –> 00:23:44.460
millimeter and we’ll see what effect it

465
00:23:42.72 –> 00:23:47.010
has on the pattern this is the first

466
00:23:44.46 –> 00:23:49.950
test when we had the scanning offset set

467
00:23:47.01 –> 00:23:53.100
to nothing it was not ticked at all so

468
00:23:49.95 –> 00:23:57.450
it was not operative and our first line

469
00:23:53.1 –> 00:23:59.580
ran from right to left now I’m going to

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

470
00:23:57.45 –> 00:24:02.700
keep stressing this you must understand

471
00:23:59.58 –> 00:24:05.550
which way that top line goes because it

472
00:24:02.7 –> 00:24:06.870
helps you to sort out what Corrections

473
00:24:05.55 –> 00:24:09.510
you’re making and whether you’re over

474
00:24:06.87 –> 00:24:12.150
correcting because you can easily send

475
00:24:09.51 –> 00:24:16.170
that line out the other way and get

476
00:24:12.15 –> 00:24:18.210
confused the second line should not be

477
00:24:16.17 –> 00:24:20.460
touching the pattern and the first line

478
00:24:18.21 –> 00:24:23.490
appears to not be touching the pattern

479
00:24:20.46 –> 00:24:27.950
fine but in reality it probably is

480
00:24:23.49 –> 00:24:30.270
because if we look at lines 3 5 7 etc

481
00:24:27.95 –> 00:24:31.890
you’ll see that they are all touching

482
00:24:30.27 –> 00:24:33.810
the pattern the only reason why the

483
00:24:31.89 –> 00:24:35.850
first line wasn’t touching the pattern

484
00:24:33.81 –> 00:24:38.550
is because that is where it actually

485
00:24:35.85 –> 00:24:41.010
started from the head came across and

486
00:24:38.55 –> 00:24:42.780
started to accelerate in there once the

487
00:24:41.01 –> 00:24:45.570
head got to the right hand end of this

488
00:24:42.78 –> 00:24:47.880
line it was running at high speed it

489
00:24:45.57 –> 00:24:50.880
zipped off to the end here stopped

490
00:24:47.88 –> 00:24:52.590
turned round and accelerated back up and

491
00:24:50.88 –> 00:24:56.130
so it was already running at full speed

492
00:24:52.59 –> 00:25:00.330
before the beam switched on just here ok

493
00:24:56.13 –> 00:25:01.620
so when we get to lines 3 5 7 etc

494
00:25:00.33 –> 00:25:04.080
they’re the ones that you need to look

495
00:25:01.62 –> 00:25:06.540
at to see what’s happening on like what

496
00:25:04.08 –> 00:25:09.870
should be happening on line 1 now this

497
00:25:06.54 –> 00:25:13.890
second test I ticked the box to make the

498
00:25:09.87 –> 00:25:15.090
scanning offset operative but I set 400

499
00:25:13.89 –> 00:25:17.160
millimeters a second

500
00:25:15.09 –> 00:25:20.370
I set the scanning fact to the offset

501
00:25:17.16 –> 00:25:22.440
factor to zero so technically this

502
00:25:20.37 –> 00:25:25.890
should look exactly like the first

503
00:25:22.44 –> 00:25:28.470
pattern and it does so then we went on

504
00:25:25.89 –> 00:25:33.060
and I tried to choose I guesstimated

505
00:25:28.47 –> 00:25:36.750
what would be the perfect offset for my

506
00:25:33.06 –> 00:25:40.590
machine and here it is now with the

507
00:25:36.75 –> 00:25:43.620
eyeglass it looked a pretty good pattern

508
00:25:40.59 –> 00:25:45.990
nice and neat gaps down the left and the

509
00:25:43.62 –> 00:25:46.860
right hand side but now I say under the

510
00:25:45.99 –> 00:25:48.750
microscope I

511
00:25:46.86 –> 00:25:52.860
think I probably could have pushed

512
00:25:48.75 –> 00:25:56.130
another 0.01 in there and made it point

513
00:25:52.86 –> 00:25:57.840
Oh three offset as opposed to 0.02 now

514
00:25:56.13 –> 00:26:00.330
you might think that we’re being ultra

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

515
00:25:57.84 –> 00:26:01.890
fussy but this can have quite an

516
00:26:00.33 –> 00:26:05.370
interesting effect when you’re doing

517
00:26:01.89 –> 00:26:07.429
scanning it makes your lines the edge of

518
00:26:05.37 –> 00:26:10.159
your scanning rhaggy

519
00:26:07.429 –> 00:26:14.039
now I’d said that we would go and try

520
00:26:10.159 –> 00:26:16.350
extreme offsets and see what happens

521
00:26:14.039 –> 00:26:18.960
here’s what happens when we put plus one

522
00:26:16.35 –> 00:26:22.919
millimeter in and you can see what’s

523
00:26:18.96 –> 00:26:25.679
happened to the top line it’s pushed the

524
00:26:22.919 –> 00:26:28.950
top line to the left by a huge amount

525
00:26:25.679 –> 00:26:31.169
and the top line is hanging out over the

526
00:26:28.95 –> 00:26:33.600
edge of the square on the left there I

527
00:26:31.169 –> 00:26:35.850
need you to understand again this top

528
00:26:33.6 –> 00:26:40.130
line thing because when I look at the

529
00:26:35.85 –> 00:26:40.130
next pattern where we put minus one in

530
00:26:41.36 –> 00:26:48.240
look the top line is now hanging out the

531
00:26:45.299 –> 00:26:50.190
right-hand side of the pattern but if

532
00:26:48.24 –> 00:26:54.590
you just look at this pattern somewhere

533
00:26:50.19 –> 00:27:05.820
down here in the middle like this is

534
00:26:54.59 –> 00:27:09.179
that a minus one or is it a plus one you

535
00:27:05.82 –> 00:27:12.120
can’t tell which way you’re going unless

536
00:27:09.179 –> 00:27:14.220
you look at that top line I’ve now put

537
00:27:12.12 –> 00:27:18.000
my machine back to that naught point O

538
00:27:14.22 –> 00:27:20.220
two to try and demonstrate finally the

539
00:27:18.0 –> 00:27:21.550
difference between backlash and scanning

540
00:27:20.22 –> 00:27:25.960
offset

541
00:27:21.55 –> 00:27:30.930
scanning offset as it says only applies

542
00:27:25.96 –> 00:27:33.450
to scanning it doesn’t apply to cutting

543
00:27:30.93 –> 00:27:36.600
backlash can have an effect on scanning

544
00:27:33.45 –> 00:27:39.300
and it can confuse the scanning offset

545
00:27:36.6 –> 00:27:42.390
results that’s why you should check out

546
00:27:39.3 –> 00:27:45.270
the backlash before you attempt to do

547
00:27:42.39 –> 00:27:47.490
the scanning offset results but when it

548
00:27:45.27 –> 00:27:52.910
comes to cutting I’m going to show you

549
00:27:47.49 –> 00:27:52.910
something now if you millimeter circle

550
00:27:53.6 –> 00:28:01.179
doesn’t look very pretty does it we’ve

Transcript for Understanding Backlash and Scanning Reverse Interval (Cont…)

551
00:27:56.82 –> 00:28:03.399
got a weird something happening there

552
00:28:01.179 –> 00:28:07.269
and I did that at a hundred millimeters

553
00:28:03.399 –> 00:28:10.059
a second let me just move across and do

554
00:28:07.269 –> 00:28:13.169
the same thing again but at 400

555
00:28:10.059 –> 00:28:13.169
millimeters a second

556
00:28:14.99 –> 00:28:21.250
they both got exactly the same sort of

557
00:28:17.87 –> 00:28:25.630
shape to them here this really strange

558
00:28:21.25 –> 00:28:25.630
offset from one to the other

559
00:28:25.66 –> 00:28:31.610
now that circle there that 50 millimeter

560
00:28:28.79 –> 00:28:36.380
diameter circle there was done with a

561
00:28:31.61 –> 00:28:38.780
two millimeter backlash applied now I

562
00:28:36.38 –> 00:28:41.270
know that’s a stupid amount but I did it

563
00:28:38.78 –> 00:28:44.390
just to prove the point that that’s what

564
00:28:41.27 –> 00:28:47.030
you will get if you add backlash and

565
00:28:44.39 –> 00:28:50.450
you’re over enthusiastic but the other

566
00:28:47.03 –> 00:28:55.790
thing is the one beside it looks exactly

567
00:28:50.45 –> 00:28:58.370
the same but that’s done at 400

568
00:28:55.79 –> 00:29:00.800
millimeters a second and the point I

569
00:28:58.37 –> 00:29:03.550
wish to make there is that speed has no

570
00:29:00.8 –> 00:29:05.440
effect on this offset

571
00:29:03.55 –> 00:29:08.740
[Music]

572
00:29:05.44 –> 00:29:10.809
and finally we’ll go back

573
00:29:08.74 –> 00:29:13.600
and we’ll check out what the correct

574
00:29:10.809 –> 00:29:17.140
offset is point zero five for this

575
00:29:13.6 –> 00:29:18.730
machine so the start-finish point was

576
00:29:17.14 –> 00:29:22.750
there as you can see from the black mark

577
00:29:18.73 –> 00:29:24.850
and it lines up perfectly well hope that

578
00:29:22.75 –> 00:29:27.700
will keep you out of trouble if you ever

579
00:29:24.85 –> 00:29:31.090
do find that you’ve got mismatching cuts

580
00:29:27.7 –> 00:29:33.340
or scans that look vaguely round the

581
00:29:31.09 –> 00:29:35.890
edge so hopefully you’ve got a little

582
00:29:33.34 –> 00:29:38.620
bit of a clearer understanding now about

583
00:29:35.89 –> 00:29:43.299
the difference between backlash and the

584
00:29:38.62 –> 00:29:47.100
scanning reverse interval until the next

585
00:29:43.299 –> 00:29:47.100
time bye

Transcript for Understanding Backlash and Scanning Reverse Interval