Spidey Sense v2 board schematic

[lvmaoa]

Would also like to know if connecting all the gnds on this board will be bad. Any advice regarding board design will also help.

---

This change is 

[lvmaoa]

In regards to the first comment, how should I approach point 2, should I just hooking up multiple screw terminals to the LT1025 and have multiple amplifier circuits? Secondly, I'm not too sure on how to add a full scale trim resistor to reflect the gain calculated.

More importantly, I added another thermocouple design (Spidey_Sense_v2.1) which is significantly cheaper (though it only offers the chip in SM). It used the AD8495 which provided more easy to read documentation and was in general more understandable (as it is a thermocouple amplifier with a cold junction compensator built in it). On the other hand, it does not specify what the error is outside of its measurement temperature range (-25°C to 400°C) and only says error will be nonlinear. Any comments about it would be helpful!

[BluCodeGH]

V2: 2, Yep multiple input screw terminals and amplifier circuits exactly.

3. The gain is set by the resistor divider made of up R2 and R3. I think in this case the way to think about the circuit is that the op amp sets its output voltage such that the thermocouple voltage is equal to the output divided by 255.
4. If you need a -5V source you should put the regulator to provide that on the board rather than using a screw terminal. Also looks like the input screw terminal is still labeled -15V but the rest of the circuit uses -5V.
5. Looks like you've got the screw terminal direction flipped, the (at least rocketry) convention is for the wires to go to the right for inputs.

V2.1:

9. The REF pin only sinks a few uA of current, I don't think you need an op amp there.
10. I was originally looking at this chip but the problem is that it's out of stock everywhere. LMK if you've found a supplier but otherwise I think we might be stuck with V2.

[lvmaoa]

Updates to names: changed V2 to V2.0.

V2.0: Not sure whether there exists a device that takes in a positive supply and outputs a negative supply. From my understanding, I used a negative voltage regulator as suggested (the -12V is assumed to come from a potential difference in the GND rails). Did not see a message from Jack, working on V2.0 now.

V2.1: The op amp is for a cleaner supply (recommended in the datasheet, not sure what a regulator was not suggested).

V2.2: New circuit that uses the LTC2997 and LTC6078 that should be able to measure from -200°C to 1300°C (with ~ -0.5°C error for most of the spectrum). Not sure if we should continue with the 2.1 or 2.2 circuit, would like a recommendation on what to do next.

I'm also not sure about notation, to notate 1.8V power symbol in kicad should I use the +1V8 power symbol.

[lvmaoa]

Update on V2.0 Switched the negative voltage regulator to a DC/DC switching regulator as per suggested.

[lvmaoa]

2.0 Addressing the comments,

12. Fixed, not sure what happened here
13. Fixed, each Vout screw terminal is also in the amplifier sheet (I'm not sure if there is some sort of convention on where to put it)
14. Added
15. Fixed, gain calculations in the Amplifier sheet
16. Changed the output voltage to ~0.2V to 1V
17. Fixed

2.2 I will stop working on this one.

[lvmaoa]

13. Fixed
14. Added a test point
15. I'm not too sure the calculations made for V_L are correct. The gain is given in the datasheet is G = (SF)(T_H - T_L)/(V_H - V_L) Since for a thermocouple, V = (41uV/C)(G)(T) and substituting into G, G = (SF)(T_H - T_L)/[(41uV/C)(G)(T_H - T_L)] Which arrises an interesting situation where, G² = SF/(41uV/C)
16. Changed
17. Fixed
18. Removed the global labels

[lvmaoa]

PCB has been routed

• Will add a comment near the LTC1983 about the specifications of the capacitor
• Will adjust the footprints to the built in 8-soic footprints

[lvmaoa]

1. Changed so that J1 and J2 are together while J3,J4,J5,J6 are on the other side of the board. Also added silkscreen for each terminal
2. Changed to a rectangular hole pattern
3. Added GND fill zones
4. Each 'amplifier' schematic is the same, though some of the trace paths do differ
5. Moved everything to front of board

Additional Comments:

• Changed mounting holes to the ones found in the I2C pancake
• Added silkscreen on C4 about being low ESR

[lvmaoa]

1. Fixed!
2. Made K- and K+ slightly bigger, the other silkscreens seem too cramped when increased in size
3. Changed!
4. Fixed!
5. Did some rerouting to remove all the daisy-chaining that I was able to find

[lvmaoa]

Removed both gitignore.

bom_html_with_advanced_grouping does not exist in KiCAD 6. Used bom_csv_grouped_by_value_with_fp and added the parts to the master electrical orders spreadsheet. LT1001 seems to have ran out of stock on digikey so I added the part from mouser.

spidey_sensev2 and other variants seem to be gone on my side and the branch doesn't seem to be showing them on GH. If someone could double check this on their side.

[lvmaoa]

Added back the root gitignore. The gitignore only ignores the files in the spidey_sense backup, should I change this to all zips?

[BluCodeGH]

You should leave the root gitignore untouched in this PR except for adding the line *-backups/. The rest of the lines there are important and should stay.

[lvmaoa]

My bad, I didn't realize I had wiped the root gitignore. I have now added the line *-backups/.

[ZTeper]

Please add vias here to eliminate ground loops:

[ZTeper]

Please also add a pdf of the schematic

[lvmaoa]

Changed.

• Added vias to avoid GND loop
• Added pdfs

[LanaTomlin]

heyaaaaa this bad boi is bound to be prettttttyy noisy, might wanna add some more decoupling caps, & mayyybe a filter on the output that cap on the non-inverting input is gonna slow down your response & potentially create oscilations (well it definitely would if it were on the inverting input, but bc its on the non inverting im not sure) also are you aware youve got a filter in here? its prob good to have, but have you done any bandwidth calcs? you have decent bandwidth, but the filter is gonna cut that off and you might end up losing gain oh yeah,,, youreeee definitely gonna want more filtering on your power supplies, this is gonna be sensitive to noise. I'd also shove a diode on te end of the filter so the capacitance of the filter is seperated from the power on the op amp and it doesnt fuck with your speed too much where did this design come from?

oh boy i reaaaaallly shouldve reveiwed this earlier

the pcb looks great tho!

[BluCodeGH]

Just talked with Lana at the review beer, we realized that most of the things she pointed out aren't actually issues (we're dealing with low frequencies and a charge pump). That being said, you should reduce the value of C5 to 0.1uF to reduce its noise and link the LT1025 datasheet above the amplifier circuit so its clear what's going on.

[lvmaoa]

Changed the value of c5 (will change the master spreadsheet too) and added a link to the 1025 datasheet.