forked from jharvey/DFH_FreeEMS_1.0_hardware
-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathmarcos's notes
83 lines (57 loc) · 4.23 KB
/
marcos's notes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
/*********** Considerations about the analog voltage regulator **************/
This is the maximum current that all analog circuitry drain:
MPX4100 10mA
MPX4250 10mA (overriden by bosch sensor, with 12.5mA)
temp input 1.93mA (x4)
thermocouple driver 0.25mA
WB externally powered
TOTAL 30.47mA
beware, too much current will saturate the regulator. It the voltage drops below the CPU VCC-0.3, bad, bad things happen.
The same applies for the digital regulator, but it is really oversized.
/**************** Thermal considerations *******************/
* CPU:
max junction temperature = 140°C
ambient temperature = 125°C
so, based in the datasheet, the device should produce no more than 15°C, where
15°C = (total chip power dissipation) x (package thermal resistance)
= (total chip power dissipation) x 43 °C/watt // page 1245 of the datasheet
then, our maximum power is 348mW.
if it has a 5v supply, then it'd drain 76mA. <- this has to be checked!
* To be thought later: how much power/temperature is produced inside the case. 125°C outside the case isn't the same as 125°C inside.
The key is to have a good path for temperature to go out, or it will stay and accumulate inside.
* FT232:
It is the only part that doesn't support 125°C. It can survive to that temperature though, as long as it is not being powered. To make the board fully 125°C capable, the FT232 will be left unpowered. It will get power from the USB cable, avoiding extra load on freeems regulator, and hoping no one will plug the cable in if the case is at more than 85°C. Another more robust solution would use the internal temperature sensor of the thermocouple driver to ensure the operational temperature is met for this device.
* capacitors
use tantalum? (for big capacitance). They have an almost plain temperature response. (check $$)
490-1666-1-ND: .1uF, ceramic, 125° capable, cheap
The electrolytics are still to be discussed, at high temps they can become useless for decoupling.
LM2940 recommends to use one aluminum cap to provide big farads, and one tantalum cap to provide the minimum, reliable uFarads.
* resistors
RMCF0805: 0805 stands 155°C max (with power dissipation = 0). At 125°C they dissipate up to 80mWatts. At 70°c they dissipate 125mW.
/************* outputs ***********/
* ignition
Ignition polarity: done with a XOR gate, which can be set to invert or not. Also, the gate protects the MCU from nasty ignition pikes.
Check if ignition drivers are pin-compatible with FETs, to choose between driving a coil directly and using an external module.
DPAK Low side drivers: NID9N05CLT4G, IPS1041RTRLPBF
* IAC
Solenoids are controlled with a PWM pin.
Steppers use a dual H-bridge IC. Check availability, temperature rating, and $
NJM3517: available at digikey & mouser; -20 to 125°C (check wattage, up to 350mA per coil, and 0.85v or 2v drop); $6.7; control logic.
L9339
L9925 -40 to 125°C, bigger SO28.
/************* case *************/
* Preferently extruded aluminum.
* External HDD cases are quite similar in size
* Easy access to the board.
* Space to stack the main board (2.5cm tall), connector board, and connector.
* what should I do with the USB connector? (type A or B, inside or ouside the case, waterproof or not)
/************ P&H **************/
A PLD like the "ATF16V8C-7JU" (20pin), working together with a 4-channel OpAmp (14pin), plus 4 sense resistors can be used as a low Z driver for 4 injectors.
It would allow a nanoseconds-response (pure logic inside, no code running) and current measuring in the high side (differentially with the opamp). A PLD is under $1.5, and the 4 channel OpAmp is $0.5.
It would drive high its output with a logic '1' from the MCU, and generate a PWM (30% duty cycle) after the treshold current is exceeded. The output goes back to '0' when it gets a '0' from the MCU.
A quite big drawback is the code loading, but for 8 outputs its like $7 (this) against $40 (LM1949).
A better one: quad current sensing IC (MAX4378TAUD+-ND) with 296-6606-1-ND voltage comparator, plus a quad open collector NAND gate (568-4551-5-ND) plus, a 1khz oscillator.
/******* Power supplies ********/
MIC29151-5.0WU has an ENABLE
LP38692SD-5.0/NOPB too
MCP1826-5002E/ET-ND