Jaguar 0.4-alpha PCB Status (was Jaguar A4 PCB Status)
Posted: Thu Jul 05, 2012 8:35 pm
I have been working on the changes to the Jaguar A3 PCB based on feedback from my testers around the world (literally), so I have created a "dev" branch to my github Jaguar repository located at:
https://github.com/DeuceEFI/Jaguar/tree/dev
And here is a list of the changes since the A3 PCBs were produced:
Fix R12 and C1 pads from being connected to the ground plane on the Through Hole component side - DONE 03/2012
Delete JP6 (not necessary) - DONE 04/15/2012
Change XOR chip from DIP to SMD - Changed Footprint, still need to regenerate PCB file - DONE 05/13/2012
Make Silk Screen lines thinner for outlines on SMD side
Move R15 label from inside XOR (U5) outline to outside near R15's position -DONE 07/04/2012
Add D37, D38, D40 and D42 to BOM - DONE 04/14/2012
- Improve the silk screen for the LEDs to show anode/cathode orientation - DONE 07/04/2012
- Too much copper around pads, looks like it is set to 8 per hole, 4 would be better/sufficient/good - DONE 07/04/2012
- Along with #9 above, revisit ground plane thermal relief rules to keep from sucking all the heat out of the pad while soldering - DONE, increased Thermal Relief Antipad Size from 0.0200" to 0.0300", Copper Width from 0.0200" to 0.0150" 05/13/2012
- C26, C42 and C43 need to be different part #'s with different lead spacing or the footprint changed back to 3mm disc - DONE 07/05/2012
- Move C17 label outside of U3 - DONE 05/13/2012
- Delete D43 and D44 as the LM2937 is already reverse protected - DONE 04/15/2012
- Fix J2 footprint so that the pin out connections are correct - DONE 05/13/2012
- Re-route power feed to MOV first, then to the input capacitors, then to the regulators - DONE 07/04/2012
- Move FAN and FUEL labels and outlines to observe the 1/16" space from the edge of the PCB for the case slot - DONE 07/04/2012
- Re-route some of the ground plane traces onto the Through Hole side of the PCB to clean up the SMD side (signal side) - DONE 07/04/2012
- Add more space around the BDM header - DONE 07/04/2012
- Remove the 5.6v Zener diode (D3) as it is not necessary since there is no chance of voltage feeding back from the MCU - DONE 05/26/2012
- Add 1k resistor to each output of the MAX9926 to limit current to MCU Port T0 and T1 - DONE 05/26/2012
- Change 5vdc-analog to 5vdc-cpu at CEL (D1) so that the CPU supply is used to power up all the necessary MCU functions. (Analog supply should only be for inputs from the engine) - DONE 05/26/2012
- FTDI USB IC decouple caps are on the wrong side of the feed from R79, and also pretty far from the IC - DONE 07/04/2012
- Check MAX9926 decouple caps to make sure power feed is on the correct side and they are close to the IC - DONE 07/04/2102
- Check the MCU decouple caps are close to the MCU and power is feeding from the correct side - DONE 07/04/2012
- Label GND/RESET/+5v/BKGD near BDM (possibly), might have to shift BDM over a little - DONE 07/04/2012
- C10 and C11 (crystal load capacitors) need to be changed to 20pF to match the manufacturer specifications. - DONE 05/26/2012
- Update BOM to use S912XDP512MAL MCU not the VAL variant as listed. (I used the MAL on all my prototypes). DONE 07/02/2012
- Add git hash to PCB design file for tracking - DONE 07/04/2012
- Update the Jaguar Assembly Manual with photos of the Serial #1 build. - DONE 07/05/2012
- Add instructions to use High Temperature Silicone or something similar to brace the upright components - DONE 07/05/2012
And I have updated the Jaguar A3 Assembly Manual with pictures from my Jaguar A3.S1 build and it can be found in my github repository at:
https://github.com/DeuceEFI/Jaguar-Assembly-Manual
Have a look at the files (in KiCAD) and let me know if you see anything that needs to be addressed that I may have missed.
Thanks,
[Edit 10-25-2012: Changed subject from Jaguar A4 PCB Status to Jaguar 0.4-alpha PCB Status (was Jaguar A4 PCB Status)]
https://github.com/DeuceEFI/Jaguar/tree/dev
And here is a list of the changes since the A3 PCBs were produced:
Fix R12 and C1 pads from being connected to the ground plane on the Through Hole component side - DONE 03/2012
Delete JP6 (not necessary) - DONE 04/15/2012
Change XOR chip from DIP to SMD - Changed Footprint, still need to regenerate PCB file - DONE 05/13/2012
Make Silk Screen lines thinner for outlines on SMD side
Move R15 label from inside XOR (U5) outline to outside near R15's position -DONE 07/04/2012
Add D37, D38, D40 and D42 to BOM - DONE 04/14/2012
- Improve the silk screen for the LEDs to show anode/cathode orientation - DONE 07/04/2012
- Too much copper around pads, looks like it is set to 8 per hole, 4 would be better/sufficient/good - DONE 07/04/2012
- Along with #9 above, revisit ground plane thermal relief rules to keep from sucking all the heat out of the pad while soldering - DONE, increased Thermal Relief Antipad Size from 0.0200" to 0.0300", Copper Width from 0.0200" to 0.0150" 05/13/2012
- C26, C42 and C43 need to be different part #'s with different lead spacing or the footprint changed back to 3mm disc - DONE 07/05/2012
- Move C17 label outside of U3 - DONE 05/13/2012
- Delete D43 and D44 as the LM2937 is already reverse protected - DONE 04/15/2012
- Fix J2 footprint so that the pin out connections are correct - DONE 05/13/2012
- Re-route power feed to MOV first, then to the input capacitors, then to the regulators - DONE 07/04/2012
- Move FAN and FUEL labels and outlines to observe the 1/16" space from the edge of the PCB for the case slot - DONE 07/04/2012
- Re-route some of the ground plane traces onto the Through Hole side of the PCB to clean up the SMD side (signal side) - DONE 07/04/2012
- Add more space around the BDM header - DONE 07/04/2012
- Remove the 5.6v Zener diode (D3) as it is not necessary since there is no chance of voltage feeding back from the MCU - DONE 05/26/2012
- Add 1k resistor to each output of the MAX9926 to limit current to MCU Port T0 and T1 - DONE 05/26/2012
- Change 5vdc-analog to 5vdc-cpu at CEL (D1) so that the CPU supply is used to power up all the necessary MCU functions. (Analog supply should only be for inputs from the engine) - DONE 05/26/2012
- FTDI USB IC decouple caps are on the wrong side of the feed from R79, and also pretty far from the IC - DONE 07/04/2012
- Check MAX9926 decouple caps to make sure power feed is on the correct side and they are close to the IC - DONE 07/04/2102
- Check the MCU decouple caps are close to the MCU and power is feeding from the correct side - DONE 07/04/2012
- Label GND/RESET/+5v/BKGD near BDM (possibly), might have to shift BDM over a little - DONE 07/04/2012
- C10 and C11 (crystal load capacitors) need to be changed to 20pF to match the manufacturer specifications. - DONE 05/26/2012
- Update BOM to use S912XDP512MAL MCU not the VAL variant as listed. (I used the MAL on all my prototypes). DONE 07/02/2012
- Add git hash to PCB design file for tracking - DONE 07/04/2012
- Update the Jaguar Assembly Manual with photos of the Serial #1 build. - DONE 07/05/2012
- Add instructions to use High Temperature Silicone or something similar to brace the upright components - DONE 07/05/2012
And I have updated the Jaguar A3 Assembly Manual with pictures from my Jaguar A3.S1 build and it can be found in my github repository at:
https://github.com/DeuceEFI/Jaguar-Assembly-Manual
Have a look at the files (in KiCAD) and let me know if you see anything that needs to be addressed that I may have missed.
Thanks,
[Edit 10-25-2012: Changed subject from Jaguar A4 PCB Status to Jaguar 0.4-alpha PCB Status (was Jaguar A4 PCB Status)]