Project

General

Profile

Mainboard Revisions » History » Version 4

jaap, 2012-11-09 08:31

1 1 jaap
h1. Mainboard Revisions
2 3 jaap
3 2 jaap
{{>toc}}
4
5
h2. Board design and schematics
6
7 3 jaap
h3. for Rev1
8 2 jaap
9 4 jaap
 PDF of the schematics: https://github.com/LaosLaser/Hardware/tree/master/laos-board-rev1/output/pdf/laos-board.pdf
10
PDF of the board: https://github.com/LaosLaser/Hardware/tree/master/laos-board-rev1/production/PI_E382816.pdf
11 3 jaap
12 2 jaap
h3. for Rev2
13 3 jaap
14 4 jaap
 PDF of the schematics: https://github.com/LaosLaser/Hardware/tree/master/laos-board-rev2/output/pdf/laos-board.pdf
15
PDF of the board: https://github.com/LaosLaser/Hardware/tree/master/laos-board-rev2/production/PI_E404084.pdf
16 1 jaap
17
h3. for Rev3
18
19 4 jaap
 PDF of the schematics: https://github.com/LaosLaser/Hardware/tree/master/laos-board-rev3/output/laos-board-rev3.pdf
20
PDF of the board: https://github.com/LaosLaser/Hardware/tree/master/laos-board-rev3/production/PI_E453774.pdf
21 1 jaap
22 4 jaap
h3. for Rev4
23
24
 SVG of the schematics: https://raw.github.com/LaosLaser/Hardware/master/laos-board-rev4/laos-board.svg
25
SVG of the board: https://raw.github.com/LaosLaser/Hardware/master/laos-board-rev4/laos-board-brd.svg
26
27 3 jaap
h2. Known bugs
28
29
h3. Known bugs in Rev1
30
31
These bugs are only relevant to board version 0.1.
32
33
!Board_rev1_suda.jpg!
34
35
h4. Pololu microstep problem
36
37
The microstep step settings does not work as expected. If you mount all the jumpers, and the motors are still making full-steps, instead of 1/16steps, this is caused by incorrect resistor values.
38
 R6, R11, R14, R20: '''use 1K resistors''' instead of the 10K resistors mentioned in the schematics.
39
40
h4. SD socket problems
41
42
!Sdrepair.jpg!
43
!DSCF0386.JPG!
44
* The pads of the SD socket are placed slightly too close too far from the socket. They should be moved 0,5 mm into the socket.
45
* Three pins are swapped:
46
** PIN 1, chipselect, is now not connected but should connect to p14 on the MBED
47
** PIN 2, MOSI, is now connect to p14 but should connect to p11 on the MBED
48
** PIN3, GND, is now connected to p11 on the MBED but should bridge to PIN6
49
*Note that the pin order is a bit strange: Pin 9 is the first Pin, next to Pin 1, and is not in use!*
50
51
!Sdrepair.png!
52
*Workaround:*
53
* On the top side of the board, cut the track from SD pin 2 to MBED p14 near SD pin 2
54
* On the top side of the board, cut the track from SD pin 3 to the VIA that leads to MBED p11
55
* On the top side of the board, connect SD pin 1 to the track that used to go to SD pin 1 (so SD pin 1 now connects to p14)
56
* On the top of the board, solder a bridge from SD pin 3 to SD pin 6.
57
* On the top of the board, solder a wire from the VIA that leads to MBED p11 to SD pin 2
58
59
*Final solution*
60
* updating part file in KiCAD
61
** Pin 1 SD (Chipselect) connects to p14 MBED (was N/C)
62
** Pin 2 SD (Data In) connects to p11 MBED (was p14)
63
** Pin 3 SD (-Vss/GND) connects to GND (was p11)
64
** Pin 4 SD (VCC) connects to +5V (was ok)
65
** Pin 5 SD (CLK) connects to p13 MBED (was ok)
66
** Pin 6 SD (-Vss2/GND) connects to GND (was ok)
67
** Pin 7 SD (Data Out) connects to p12 MBED (was ok)
68
** Pin 8 SD - N/C
69
** Pin 9 SD - N/C
70
71
h4. USB Connector
72
73
The holes of the USB connector are not pretinned on the bottom side of the board. This is probably a mistake in the USB connector component drawing.
74
75
*Workaround*: Since the pins are pretinned inside the holes, the connector will work. Just doublecheck your soldering!
76
77
H4. Ethernet
78
79
Network connector J4 is available in different types from different vendors. Pinning is different between these types.
80
The pin-out of this board is made for the MAGJACK SI-60002-F connector (Farnell), not for the HALO HFJ11 that is on the part-listing!
81
82
<pre>
83
  status: need to update the BOM. No PCB change required
84
  MagJack  HALO    PCB modification if you want to use HALO
85
  p1 RD+   TX+     RD+ (OK)
86 1 jaap
  p2 VC    TX-     RD- (from MBED pin RD-)
87 3 jaap
  p3 RD-   RX+     TD+ (from MBED pin TD+)
88
  p4 TD+   VC      VC  (from original connection to p2)
89
  p5 VC    VC      VC  (OK)
90
  p6 TD-   RX-     TD- (OK) 
91
</pre>
92
93 4 jaap
To make your board function correctly: Use the Stewart MagJack part, or modify the board, using these 4 modifications: https://github.com/LaosLaser/Hardware/tree/master/laos-board-2/notes/notes/ (eth-mod1 to eth-mod4).
94 3 jaap
95
h4. LEDS
96
97
The holes of ALL LEDS are not pretinned on the bottom side of the board. This is probably a mistake in the component drawing.
98
99
*Workaround*: Since the pins are pretinned inside the holes, the connector will work. Just doublecheck your soldering!
100
101
h4. Phoenix connector holes
102
103
The holes for the Phoenix connectors are too small (all of them!). They are now 1mm (?) but the pins are 1mm squares, so the size should be 1.5mm.
104
105
*Workaround*: Use a file to make the pins round, then they will fit
106
107
h4. Resistors
108
109
Some mistakes in resistor values:
110
* R23, R24,R26,R27 should be 1K
111
* R28, R29, R31, R32 should be 330R
112
* R2, R3 should be 1K
113
* R7,R8 should be 2K2
114
Resistors should be renumbered so it's easier to find them on the board
115
116
h4. Capacitors
117
118
C7 has a non-standard value (33nF), so change it to 22nF.
119
120
h4. POLOLU J17
121
122
Pin 1 from POLOLU J17 (ENABLE) connects to MBED P20, but it should not. It should be connected to all the other POLOLU ENABLE Pins.
123
124
*Workaround*: On the top side of the baord, cut the track going from POLOLU J17 pin 1 near pin 1. On the top side, connect Pin 1 to the track going to all other Enable pins (it's 4mm to the right of pin 1).
125
126
h4. Enable Solder Pads
127
128
The Enable solder pads (under the Pololu's) Connect to Pololu pin 2 (MS1), but they should connect to Pin 1 (ENABLE). Also, in the Schematic, the names of the solder pads (EN / DIR) are swapped.
129
130
*Workaround*: If you are using Pololu's this is not a problem at all. If you don't use Pololu's, bridge Pin1 and Pin2 of the Pololu headers with a wire.
131
132
h4. Plus / Ground signs
133
134
Where possible, ground and +V should be marked on the print. Some obvious examples:
135
* J38 (Power in)
136
* J5 (VMOT in)
137
138
h4. J10 (I2C)
139
140
The middle two pins should be swapped to make a flat cable to the LCD/Keyboard
141
142
*Workaround*: Swap the pins in the wire from J10 to the LCD/Keyboard
143
144
h4. Ground circuit of Power 5V IN
145
146
The ground of the Power 5V IN circuit is not connected to the main ground. This should be the same NET
147
148
*Workaround*: Scratch the GND around the power connector GND (pin 2) and make a solder connection between the two GND's
149
150
h4. MBED Pin Vin
151
152
MBED pin Vin is connected to the Net +5VD, but this Net should not exist. This pin should be in the +5V Net.
153
154
*Workaround*: Use a jumper wire on the back to connect MBED Vin to the solder pad J3 (Vusb). Connect it to the +5V without joining the solder pad!
155
156
h4. Missing jumper near J5
157
158
There should be a jumper pad to connect Pin 1 of J5 (VMOT, Stepper driver power) to VAA.
159
160
*Workaround: To get +5V on VMOT, connect a jumper wire from connector J2, pin 1, to connector J5, pin 1.
161
162
h4. Resistor R12
163
164
The value of Resistor R12 is very dependent on the function of J14 and the value of VMOT, because a wrong value could blow up p19 on the MBED! We use it with switch to detect Lid Open/Close with VMOT=5V. In that case it should be 22K, so p19 is about 3.4V when J14 pin 3 is not connected.
165
166
h4. Network connector
167
168
J4 p11 was connected to p18 on MBED for no reason. 
169
 status: removed line in schema and pcb rev 0.2
170
171
h3. Bugs we noted but are not show-stoppers
172
173
h4. Phoenix connector spacing
174
175
The outline of the Phoenix connectors is not in the component drawing. Because of that, they are places too close on the board, both too close together and too close to other components (the SD card). Because of this, it is not possible to place all Phoenix connectors on the board at the same time.  This affects the location of the connectors: J9, J18, J24, J33, J37, J39, J10, J27, J14.
176
177
*Workaround*: This is only a real problem when you want to use all possible pins. Most lasers will not need that, so if you use 4-pin Phoenix connectors instead of 5, it will fit almost everywhere. If you really need all the pins, use other connectors, like the 2.54 spaced Phoenix mini connectors or pin headers.
178
179
h4. 7805 cooler
180
181
There should be some space around the 7805 so it can have a cooler attached.
182
183
*Workaround*: 
184
* Put the 7805 slightly higher or bend the cooler in between the components
185
* Use alternative to 7805 that doesn't get so hot: 
186
 http://letsmakerobots.com/node/3905
187
 http://dimensionengineering.com/DE-SW050.htm?gclid=CMD26bLgq6sCFcGCDgodN3io6w
188
189
h4. Not enough current to drive Enable on steppers
190
191
Apparently the MBED does not deliver enough current to drive the Enable on all steppers from one pin. 
192
193
*Workaround*: An amplifier (BC-547C) should be placed just after MBED p7.
194
195
h2. Differences between Rev1 and Rev2
196
197
* SDcard rewired to fix bugs
198
* Pololu microstepper resistor values updated
199
* Pre-tinning of several holes in the print fixed
200
* Ethernet connector is now always Steward Magjack, see BOM
201
* Hole size for Phoenix connectors fixed
202
* Resistor values updated
203
* Fixed pololu enable line for J17
204
* Fixed pololu enable solder pads
205
* Fixed disconnected ground lines
206
* Fixed 5V net for mbed
207
208 1 jaap
h2. Differences between Rev2 and Rev3
209
210
h3. Missing Vsensor on endstops
211
212
In Rev2, the lines marked Vsensor on J29 and J34 and the + side of the opto-couplers are connected with each other, but somehow the connection to the rest of the Vsensor net has gone missing. They should have been connected to Vsensor on J27 and the middle pin of the VCPU/Vsensor/+5V jumper. In Rev3 this problem has been fixed. Vsensor on J27 now has it's own jumper to select the power level.
213
214
h3. CAN bus not working
215
216
In Rev2, there are a few problems with the CAN bus connection.
217
* U1 pin 8 (Rs) should be connected to GND
218
* U1 pin 4 (Vref) should be left open
219
* U1 pin 3 (Vdd) should be connected to 5V (instead of 3V3)
220
This has been fixed in Rev3
221
222
h3. Swapped label for Direction and Enable
223
224
In Rev2, the (E)nable and (D)irection pin labels are swapped in the design and on the PCB markings. It does not cause functional problems, as long as you make sure your wiring of external drives is correctly connected. In Rev3, this has been fixed.
225
226
h3. Series resistor for analog input
227
228
In Rev3, an extra resistor (R33) was added between AIN (PIN 3 of J14 and the junction of R10/R12. This allows us to make a voltage divider. Typical value for R33 should be 1K. 
229
230
h3. Series resistor for analog output
231
232
In Rev3, an extra resistor (R34) was added between the MBED and AOUT (PIN 2 of J24), in order to limit currents in case of a ground connection of the analog output. 
233
234
h3. Solder joints for Pololu microstep factor
235
236
Instead of jumpers, Rev3 has  solder joints so it is easier to solder the pins to +V to obtain microstepping. Solder joints: J7, J16, J22 and J31
237
238
h3. Moved Ethernet connector
239 3 jaap
240 1 jaap
In Rev3, the Ethernet connector has been swapped with the USB to allow easier connection of the MiniUSB connector of the MBED.
241
242
h3. Added FFC / FPC connector
243 3 jaap
244 1 jaap
Rev3 has place for a FFC / FPC connector, as used in some HPC and Chinese Lasers. This connector combines X-Axis stepper and endstops.