Difference between revisions of "Rev200 mods"
From DIDEAS Wiki
m |
m |
||
(42 intermediate revisions by the same user not shown) | |||
Line 1: | Line 1: | ||
− | * | + | {{pf_top_nav}} |
+ | |||
+ | =Rev301= | ||
+ | *isolate current sensors from remainder of 5V supply. | ||
+ | *overload latch circuit | ||
+ | *high side current sense to include clamp | ||
=PCB needed Revisions= | =PCB needed Revisions= | ||
− | * | + | *generate PDFs at 8.5x11 and mirror output when necessary. |
+ | ==IMU revisions== | ||
+ | * [1] (DONE) protect switcherwith resistor / TVS | ||
+ | * [2] (DONE) protect ISP - R53-55 added. | ||
+ | * [3] U5 change to 24LC128 (DFN8 still) or larger (512k available from digikey) | ||
+ | * [4] (done) redo battery sensor circuit to not over voltage ADC | ||
+ | * [5] on SC, route VBAT_PROT through newly added 200mA fuse to IMU connector J3 pins 1,2 | ||
+ | * [6] on FET change F1 (1210) to ~300mA - thus no chance of SC or IMU taking down the other. | ||
+ | * [7] rework #4 - to make more accurate voltage measurement: raw "IMU_VIN" to Q7.S, replace R44 w/80k, replace C58 w/1nF resulting in 80uS tau (2khz BW) | ||
+ | * [8] replace Z gyro filter cap C56 w/1nF - resulting in ~800hz BW | ||
+ | * [9] (done) We'll keep R45 as show, which decreases the peek Z gyro rate (increase gain) | ||
+ | |||
+ | == notes Oct 23== | ||
+ | # use red PCBs | ||
+ | # where are we regarding the 12 x 0.8mm JST connectors? | ||
+ | # LDO voltage is 6.15V, use 0.1%, 97.6k, 24.9k | ||
+ | # for D13 use SDMK0340L | ||
+ | # primary coin cells BT1,BT4 can be removed | ||
+ | # pg 3 : ECO 14 - looks good. | ||
+ | # pg 3 : ECO 20 - can use 6 pin x 1mm ???? | ||
+ | # pg 6 : ECO 15 - looks good. | ||
+ | # pg 7 : ECO 6,7,8,10 - looks good | ||
+ | # pg 7 : ECO 11 - disagreement over cut off freq | ||
+ | # pg 8 : ECO 12 - need to use 1nF for C148 | ||
+ | # pg 8 : ECO 13 - populate R220,R222, R232, R233 - none should be DNP | ||
+ | # pg 11 : check drop cross D14, and coin cell voltage. | ||
+ | # pg 14 : wonder about this circuit and diode to 3.3V MC - should be SC to avoid loading. | ||
+ | # pg 15 : ECO 17, 18 implmented - but... | ||
+ | # pg 15 : wonder about this circuit and diode to 3.3V MC - should be SC to avoid loading. | ||
+ | # pg 17 : verify full bridge | ||
+ | |||
+ | == discussion Mon Oct 5== | ||
+ | |||
+ | * regarding ECO 2016: | ||
+ | **C2 can be decreased to 0.1uF, 50V - and will be a better high frequency filter as a result. | ||
+ | **C19-21 have a maximum working voltage of ~9V. These could be replaced with 16V to 25V devices. | ||
+ | **C60 won't exceed ~5V, 1.0u, 16V would be fine. | ||
+ | **C101, C121, C136 = 0.1uF, 50V (same as C2 above) | ||
+ | **C40, C62 has a 3.3V working voltage. 1.0uF, 16V would be fine | ||
+ | |||
+ | |||
+ | * Regarding : ECO 2018 : | ||
+ | ** this is an input to the micro and requires no change to the associated input circuitry. | ||
+ | * I don't believe that my ECO #15 was captured. This ECO changes the HALL_C input to a 5V CMOS level output. Thus D10A is removed and some sort of buffer or inverter from the 5V supply is inserted. Probably the Schmitt inverter (1200-000348) would work well for this. | ||
+ | |||
+ | ==discussion Sun Oct 4 == | ||
+ | |||
+ | * overall our goal is it keep to a minimum the number of changes. The changes they we have proposed are absolutely necessary from our point of view. | ||
+ | |||
+ | * changing 0.8mm headers to 1mm | ||
+ | ** We need to stay with the 12 x 0.8mm header for the pyramid, EC30 MILE, and IMU. Fortunally we need only 1 part number. The 5 position linear spring header can become 1mm | ||
+ | |||
+ | *4 Channel DAC | ||
+ | ** We would like this to remain. | ||
+ | |||
+ | *Primary Li cells | ||
+ | ** We would like these to remain. We may wish to locate SMT cells for both primary and 2ndary, and ones that have lower height profile. | ||
+ | |||
+ | * Current sensors : | ||
+ | ** We're aware of the linearity issue. | ||
+ | ** The present design with the current sensors can actually cause an over-voltage condition on the dsPIC. When this happens there are offset and cross coupling issues on all channels. The slight lose in resolution is minor compared with the upside. | ||
+ | |||
+ | * Hall sensor should remain. The location of the sensor may change slightly. | ||
+ | ** The purpose of the connection to TP252 is so that we may experiment with alternative sensors. At present we don't have any. | ||
+ | |||
+ | * IMU | ||
+ | ** I believe we would like to stay with the right ankle 12 pin x 0.8mm header. KS would like it moved down a few MM. We need to see the full V2 design to finalize this connector. | ||
+ | |||
+ | * Temperature of coin cells: | ||
+ | ** We've monitored the temperature. However, I wasn't aware of the narrow temperature limitation of the cells. Clearly this could be an issue. | ||
+ | |||
==FET== | ==FET== | ||
− | * populate coin charger as default | + | * (1) w/clamp circuit - use thermal vias for R4 - this device is impossible to replace |
− | * | + | ** will try low temp solder and send to MPD |
− | * change D13 to diode that actually has 200mV drop @ 1uA. BAS116-F-7 appears to have 427mV @ 0.3uA | + | |
+ | * (2) height of coin cells is too great. need the cells from this original coin cell specification | ||
+ | * (3) pg 4/8 : populate coin cell charger as default | ||
+ | ** populate BT2,BT3 as default, DNP BT1,BT4 (but keep R18, D21) | ||
+ | ** DNP D13 only if primary coins are installed. | ||
+ | ** Would like the coin cells to have 6.0 to 6.1V to be fully changed. ABS maximum is 6.2V | ||
+ | *** It doesn't appear to be possible to obtain this tight voltage specification with 1% resistors and the [http://www.micrel.com/_PDF/mic5235.pdf (PDF) MIC5235] LDO. The LDO has up to 2% variation in its output over temperate. Thus a total variation of upto +- 3.65% could be expected. | ||
+ | *** 6.2 +- 3.65% = 5.98V to 6.42V | ||
+ | *** expect a 150-200mV drop across D13 | ||
+ | |||
+ | *(4) change D13 (and D21 ?) to diode that actually has 200mV drop @ 1uA. BAS116-F-7 appears to have 427mV @ 0.3uA | ||
+ | ** See : [http://www.diodes.com/datasheets/ds30233.pdf (PDF)] - Diodes Inc BAS116-7-F fig 3, pg 2] | ||
+ | |||
+ | |||
+ | *(5) pg 4/8 : ability to measure coin cell voltage without charger enabled? Connect pin 3 of U10 to something like the SFTYSNS_EN, or MTR_EN signal at TP97 or U20B | ||
==SCMC== | ==SCMC== | ||
− | * populate so | + | *(6) pg 7/18 : change cutoff freq to ~15Khz for temp circuit (U22) R102=1k, C73=0.01u |
− | + | ||
− | * | + | *(7) pg 7/18 : remove R232 - 10k for batt_temp_mon pullup (CEB - 10-13-09 - i think I mean R242) |
− | * need to determine the accuracy of the RTC. Seems to be drifting. | + | *(8) pg 7/18 : remove R245 - 220 ohm for mtr_temp_mon pullup |
+ | |||
+ | *(10) pg 7/18 : R206, R216 -> 6.66k to scale range 0-5V to 0-3.0V (current sensors) | ||
+ | ** presently 5V @ current sensor becomes 3.283V, but ADC reference is 3.0V so of the 75.76A range (0->5V), only 69.23A are viable. Thus range is -37.88A to +31.34A. The scale factor is 65536 / (5/.066 * 3/(5*10/15.23)) = 946.6789 | ||
+ | |||
+ | *(11) pg 7/18 : bridge current sensor use same values phase U,V | ||
+ | |||
+ | |||
+ | *(12) Pg 8/18 : increase size of resisters (by 10x) for supply voltage measurement (R91, R94, R221, R88). Keep cut off freq the same. | ||
+ | |||
+ | *(13) pg 8/18 : populate so measurement of supply voltage is default (R222, R233 installed). its easier to remove components than to add them. | ||
+ | |||
+ | |||
+ | *(14) pg 3/18 : replace 1.25mm header for EC340 enc/hall w/ 12x0.8mm header | ||
+ | ** [[http://mediawiki.dideas.com/index.php?title=Upd_2x0_218_controller#pinout 12x8 pinout]] | ||
+ | ** [http://dhcp-17-197.media.mit.edu/ankle/for_mpd/prelim-EMILE3_070709.pdf interface cable] | ||
+ | |||
+ | *(15) pg 6/18 : replace HALL_C input with SSI_CLK (CMOS 5V) output. | ||
+ | ** Remove D10A | ||
+ | ** Insert Schmitt inverter (1200-000348) between MC and R86 | ||
+ | |||
+ | *(16) <strike>pg 11/18 : (CB) need to determine the accuracy of the RTC. Seems to be drifting. </strike> | ||
+ | |||
+ | *(17) pg 18/18 : Remove LSSGA before TP56 | ||
+ | |||
+ | *(18) pg 18/18 : Connect TP56 to LSSGA connector. needs to protect dsPIC, but no other circuity. | ||
+ | ** adjust C95 for 15khz cutoff freq | ||
+ | ** appears on page 15 | ||
+ | |||
+ | *(19) pg 14/18 : Connect TP252 to LSSGA connector. Needs to protect dsPIC, but no other circuity. | ||
+ | |||
+ | |||
+ | *(20) 3/18 : LSSGA connector signals: | ||
+ | ** AGND | ||
+ | ** A3V ref (analog 3.0V reference) | ||
+ | ** Various power source options such as Vbat(fused), 5.0V, 3.3V. The reference can't (shouldn't) | ||
+ | ** TP56 analog in | ||
+ | ** TP252 analog in | ||
+ | ** Hall_B input | ||
+ | |||
+ | *(21) 11/18 : RTC coin cell isn't fully charged. Recommend connected D14 to the 3.3V supply. | ||
+ | |||
+ | |||
+ | * locate IMU header for easier access. possibly rot180 - or move 2mm away from edge. | ||
+ | ** KS - lower position for V2. | ||
+ | ** CB - in V2 appears to be much more space in the shell, no other changes may be needed. | ||
==IMU== | ==IMU== | ||
− | * probably want to upgrade dual axis IMU | + | * probably want to upgrade dual axis IMU |
+ | * battery measurement circuit needs to be re-done. |
Latest revision as of 13:03, 9 November 2010
PF Users Navigation:
- PCB; Lifefix_users; Pf_users; benchtest_users; CalibFix Users; Hardware; Assembly ; iochan ;
- PCA : PCB ; AKENC SC/MC 218 ; FET 217 ; IMU219 ; Swifi ; Rev200_mods ; PCA Inventory
- Special Commands to the Ankle (PFCMD) : State Controller Commands; Motor Controller Commands; Python Examples; IMU Commands ;PFCMD_PY; Pf_calb_table_py; Virtual spring test; PF EEPROM
- DOC: Pf_users; Powerfoot Keyboard User Interface; Steps for Manual Tuning; "Dashboard" Program For Assisting with Tuning
- NEW (CEB) WIKI
- Torque Feedback Controller Guide
Contents
Rev301
- isolate current sensors from remainder of 5V supply.
- overload latch circuit
- high side current sense to include clamp
PCB needed Revisions
- generate PDFs at 8.5x11 and mirror output when necessary.
IMU revisions
- [1] (DONE) protect switcherwith resistor / TVS
- [2] (DONE) protect ISP - R53-55 added.
- [3] U5 change to 24LC128 (DFN8 still) or larger (512k available from digikey)
- [4] (done) redo battery sensor circuit to not over voltage ADC
- [5] on SC, route VBAT_PROT through newly added 200mA fuse to IMU connector J3 pins 1,2
- [6] on FET change F1 (1210) to ~300mA - thus no chance of SC or IMU taking down the other.
- [7] rework #4 - to make more accurate voltage measurement: raw "IMU_VIN" to Q7.S, replace R44 w/80k, replace C58 w/1nF resulting in 80uS tau (2khz BW)
- [8] replace Z gyro filter cap C56 w/1nF - resulting in ~800hz BW
- [9] (done) We'll keep R45 as show, which decreases the peek Z gyro rate (increase gain)
notes Oct 23
- use red PCBs
- where are we regarding the 12 x 0.8mm JST connectors?
- LDO voltage is 6.15V, use 0.1%, 97.6k, 24.9k
- for D13 use SDMK0340L
- primary coin cells BT1,BT4 can be removed
- pg 3 : ECO 14 - looks good.
- pg 3 : ECO 20 - can use 6 pin x 1mm ????
- pg 6 : ECO 15 - looks good.
- pg 7 : ECO 6,7,8,10 - looks good
- pg 7 : ECO 11 - disagreement over cut off freq
- pg 8 : ECO 12 - need to use 1nF for C148
- pg 8 : ECO 13 - populate R220,R222, R232, R233 - none should be DNP
- pg 11 : check drop cross D14, and coin cell voltage.
- pg 14 : wonder about this circuit and diode to 3.3V MC - should be SC to avoid loading.
- pg 15 : ECO 17, 18 implmented - but...
- pg 15 : wonder about this circuit and diode to 3.3V MC - should be SC to avoid loading.
- pg 17 : verify full bridge
discussion Mon Oct 5
- regarding ECO 2016:
- C2 can be decreased to 0.1uF, 50V - and will be a better high frequency filter as a result.
- C19-21 have a maximum working voltage of ~9V. These could be replaced with 16V to 25V devices.
- C60 won't exceed ~5V, 1.0u, 16V would be fine.
- C101, C121, C136 = 0.1uF, 50V (same as C2 above)
- C40, C62 has a 3.3V working voltage. 1.0uF, 16V would be fine
- Regarding : ECO 2018 :
- this is an input to the micro and requires no change to the associated input circuitry.
- I don't believe that my ECO #15 was captured. This ECO changes the HALL_C input to a 5V CMOS level output. Thus D10A is removed and some sort of buffer or inverter from the 5V supply is inserted. Probably the Schmitt inverter (1200-000348) would work well for this.
discussion Sun Oct 4
- overall our goal is it keep to a minimum the number of changes. The changes they we have proposed are absolutely necessary from our point of view.
- changing 0.8mm headers to 1mm
- We need to stay with the 12 x 0.8mm header for the pyramid, EC30 MILE, and IMU. Fortunally we need only 1 part number. The 5 position linear spring header can become 1mm
- 4 Channel DAC
- We would like this to remain.
- Primary Li cells
- We would like these to remain. We may wish to locate SMT cells for both primary and 2ndary, and ones that have lower height profile.
- Current sensors :
- We're aware of the linearity issue.
- The present design with the current sensors can actually cause an over-voltage condition on the dsPIC. When this happens there are offset and cross coupling issues on all channels. The slight lose in resolution is minor compared with the upside.
- Hall sensor should remain. The location of the sensor may change slightly.
- The purpose of the connection to TP252 is so that we may experiment with alternative sensors. At present we don't have any.
- IMU
- I believe we would like to stay with the right ankle 12 pin x 0.8mm header. KS would like it moved down a few MM. We need to see the full V2 design to finalize this connector.
- Temperature of coin cells:
- We've monitored the temperature. However, I wasn't aware of the narrow temperature limitation of the cells. Clearly this could be an issue.
FET
- (1) w/clamp circuit - use thermal vias for R4 - this device is impossible to replace
- will try low temp solder and send to MPD
- (2) height of coin cells is too great. need the cells from this original coin cell specification
- (3) pg 4/8 : populate coin cell charger as default
- populate BT2,BT3 as default, DNP BT1,BT4 (but keep R18, D21)
- DNP D13 only if primary coins are installed.
- Would like the coin cells to have 6.0 to 6.1V to be fully changed. ABS maximum is 6.2V
- It doesn't appear to be possible to obtain this tight voltage specification with 1% resistors and the (PDF) MIC5235 LDO. The LDO has up to 2% variation in its output over temperate. Thus a total variation of upto +- 3.65% could be expected.
- 6.2 +- 3.65% = 5.98V to 6.42V
- expect a 150-200mV drop across D13
- (4) change D13 (and D21 ?) to diode that actually has 200mV drop @ 1uA. BAS116-F-7 appears to have 427mV @ 0.3uA
- See : (PDF) - Diodes Inc BAS116-7-F fig 3, pg 2]
- (5) pg 4/8 : ability to measure coin cell voltage without charger enabled? Connect pin 3 of U10 to something like the SFTYSNS_EN, or MTR_EN signal at TP97 or U20B
SCMC
- (6) pg 7/18 : change cutoff freq to ~15Khz for temp circuit (U22) R102=1k, C73=0.01u
- (7) pg 7/18 : remove R232 - 10k for batt_temp_mon pullup (CEB - 10-13-09 - i think I mean R242)
- (8) pg 7/18 : remove R245 - 220 ohm for mtr_temp_mon pullup
- (10) pg 7/18 : R206, R216 -> 6.66k to scale range 0-5V to 0-3.0V (current sensors)
- presently 5V @ current sensor becomes 3.283V, but ADC reference is 3.0V so of the 75.76A range (0->5V), only 69.23A are viable. Thus range is -37.88A to +31.34A. The scale factor is 65536 / (5/.066 * 3/(5*10/15.23)) = 946.6789
- (11) pg 7/18 : bridge current sensor use same values phase U,V
- (12) Pg 8/18 : increase size of resisters (by 10x) for supply voltage measurement (R91, R94, R221, R88). Keep cut off freq the same.
- (13) pg 8/18 : populate so measurement of supply voltage is default (R222, R233 installed). its easier to remove components than to add them.
- (14) pg 3/18 : replace 1.25mm header for EC340 enc/hall w/ 12x0.8mm header
- (15) pg 6/18 : replace HALL_C input with SSI_CLK (CMOS 5V) output.
- Remove D10A
- Insert Schmitt inverter (1200-000348) between MC and R86
- (16)
pg 11/18 : (CB) need to determine the accuracy of the RTC. Seems to be drifting.
- (17) pg 18/18 : Remove LSSGA before TP56
- (18) pg 18/18 : Connect TP56 to LSSGA connector. needs to protect dsPIC, but no other circuity.
- adjust C95 for 15khz cutoff freq
- appears on page 15
- (19) pg 14/18 : Connect TP252 to LSSGA connector. Needs to protect dsPIC, but no other circuity.
- (20) 3/18 : LSSGA connector signals:
- AGND
- A3V ref (analog 3.0V reference)
- Various power source options such as Vbat(fused), 5.0V, 3.3V. The reference can't (shouldn't)
- TP56 analog in
- TP252 analog in
- Hall_B input
- (21) 11/18 : RTC coin cell isn't fully charged. Recommend connected D14 to the 3.3V supply.
- locate IMU header for easier access. possibly rot180 - or move 2mm away from edge.
- KS - lower position for V2.
- CB - in V2 appears to be much more space in the shell, no other changes may be needed.
IMU
- probably want to upgrade dual axis IMU
- battery measurement circuit needs to be re-done.