Automotive Human switching and power management

[dephwyggl]

TomGeorge:
S2 FUNCTION is ORed with S1 FUNCTION .

Exactly. ;D

TomGeorge:
R2a may need to be much lower in value to turn Q4 ON. Say 470R.
R4 make it 10K.
Place a 0.1uF in parallel with C1.

I was hoping for feedback on the over-voltage stuff, and I appreciate the other details you provided (I just don't have enough experience yet to spot things like this).

Additional question: I was thinking of adding another cap next to C1, but wasn't sure of value or type. From what I've seen for capturing voltage spikes, an larger electrolytic is typically paired with a smaller ceramic, so I should add a 0.1uF ceramic cap?

I'd still like to know what exactly the TVS is going to do here; i.e. specifically, what's the max voltage we'll see after it? (As I mentioned earlier, I can't make heads nor tails of the spec in this regard.) At the moment, there are far too many PMOSFETs which dislike more than 20V at their gates, so I'd like to see a maximum of 18V available after this over-voltage stuff. Will any of what's happening here meet this objective? (Doing so would eliminate the need for zeners at all of these PMOSFETs.)

TomGeorge:
Use S2 through an Opto Coupler to bias Q4 ON.
In fact would it be easier and lower component count to get S1 through another Opto to bias Q4 ON as well?

See? This is exactly why I'm here - to have the obvious (to those with more experience than me) pointed out to me. It only took two reminders for me to think of optocouplers.

Tom, thank you for all your comments.

An update to the schematics should be up soon...

Cheers!
Dirk

[dephwyggl]

Updated schematic:

(Nearly) everything has been re-labeled/numbered.

I'm looking forward to suggestions for a simple way to implement under-voltage protection - something that would disable everything when there is less than 10V at JP1. (Under-voltage stuff (place-holder) has been left out for now.)

I'm starting to re-think using a MOSFET for reverse-polarity protection. I may go back to using a diode. Yes, there's the voltage drop, but I think they'll be more reliable in highly variable voltage situations than MOSFETs.

I look forward to all comments and suggestions - particularly component layout and specific values.

Cheers!
Dirk

[TomGeorge]

Hi,
Try this, you have to make the gate negative with respect to the source to turn the Pch MOSFET ON.

Even get rid of D2 and D3.
How does Q1 work?
Tom....

[dephwyggl]

TomGeorge:
you have to make the gate negative with respect to the source to turn the Pch MOSFET ON.

Mine did that, but. ...BUT:

TomGeorge:
Try this,
[...]
Even get rid of D2 and D3.

You're kidding, right? Reduce cross-overs from 5 to 2, reduce parts count by 4 and pin count by 9? AND do the same thing?

Dude; you rock!

TomGeorge:
How does Q1 work?

Took me a bit to understand as well, but Q1 acts as a low voltage drop diode; current nominally conducts via its body diode, and if reverse polarity occurs, the gate slams the MOSFET shut.

EDIT: I was slightly off with this description. The PMOSFET initially conducts via its body diode - until a ground path is established for the PMOSFET's gate - which then conducts via DS. Infineon has a great paper on this.

Still not happy about what could (maybe) happen if reverse polarity presents more than 20V at its gate. I need to decide what the probability is of this "more than 20V" happening along with the possibility of the TVS not being able to mitigate that; if it's still high enough, I may go back to a diode here...

Still need under-voltage stuff...

Cheers!
Dirk

[dephwyggl]

I just realized those diodes will have to go back into this circuit. The Arduino must know when either or both of S2 and S3 are closed, so I'll have to have some way to differentiate between their closures from the input or output of the optos.

Tom, you likely didn't anticipate this, so this in no way diminishes what you did!

Question: For each opto: would tying an Arduino pin up, and pulling it down with the emitter side work?

Cheers!
Dirk

[dephwyggl]

New question: For each opto: would tying an Arduino pin down, and pulling it up with the collector of the opto work? Like so:

(D2/D3, R2b/R3b and connections to Arduino pins added.)

Cheers!
Dirk

[allanhurst]

Alll the helpful posters here are pointing out the various approaches to your endeavour. I have refrained.

You're obviously interested in the detail... May I suggest you enter a formal degree-level electronics course?

We can give you snippets, but to understand things better you need to get a fuller view..

regards

Allan

[dephwyggl]

allanhurst:
Alll the helpful posters here are pointing out the various approaches to your endeavour. I have refrained.

You're obviously interested in the detail... May I suggest you enter a formal degree-level electronics course?

We can give you snippets, but to understand things better you need to get a fuller view..

Thank you for your suggestion, Allan. I've mentioned this elsewhere: I'm doing this strictly as a hobby; I'm not sure I'd continue with this project with my head stuck in books. I'm having far too much fun learning here.

...and please do comment; I look forward to your feedback.

Cheers!
Dirk

[dephwyggl]

dephwyggl:
New question: For each opto: would tying an Arduino pin down, and pulling it up with the collector of the opto work? Like so:
[...]
(D2/D3, R2b/R3b and connections to Arduino pins added.)

...and I'm already thinking that maybe going back to the original "set pin high and pull it down", but with the cathodes of the optos? This would eliminate the need for the diodes.

Cheers!
Dirk

[dephwyggl]

OK, I'm still trying to wrap my head around the fact that no-one appears to have met the challenge of filtering automotive power to reliably provide something like 10-18V. I've been wracking my noggin, and came up with the following:

I'm quite sure this can't work, but have no idea why. Anyone care to edify?

Note: The above is missing a few things; in particular, it doesn't represent the automotive environment (which is more like -28V to +120V), and I've left out the power filtering stuff (TVS, caps, reverse polarity protection).

Cheers!
Dirk

[dephwyggl]

Mark, I'm learning. I know you made these comments a while back, and yes, it's taken me this long to grok them fully:

MarkT:
You can't connect diodes to a gate like that (D2, D3b), current has to flow both ways, into and out of the gate to switch it. Are you using diode-resistor logic without the resistor perhaps?

Protection for the gate is a zener between gate and source, 13V would be reasonable in an automotive environment (so long as there is a resistor somewhere to limit the zener current)

I finally got these. Thank you.

MarkT:
The TVS1 between gate and drain is going to simply blow the gate oxide on a transient. TVS between drain and source, or between gate and source is fine (although a zener does that fine).

I think I may have figured this one out too. However, wouldn't the TVS blow the gate on either side of the PMOSFET? And, I'm not sure the TVS would be fast enough on the source side to protect the PMOSFET. I'd be more than happy to see further suggestions on this; worst case, I'll go back to using a diode with its attendant voltage-drop hit instead of the PMOSFET.

Thank you again for all of your contributions, help, and suggestions.

Cheers!
Dirk

[dephwyggl]

Here's a more complete schematic for managing over-voltage and reverse polarity circumstances:

• Voltage source: Typical of automotive power systems: nominally, we'll see between 12V (car not running), and 14.5V (car is running). Though extremes of minus 28V up to plus 120V occur infrequently and represent fairly extreme circumstances, they still need to be taken care of.
• D1: The SBR30150CT is a 150V, dual 15A, common cathode SBR which provides reverse polarity protection. (I couldn't figure out how to reliably implement a reversed PMOSFET for this, so I'll have to take the higher voltage-drop using diodes.)
• C1, C2: Capture voltage spikes.
• TVS1: Littelfuse's 3KP16A doesn't do anything under 16V, and shunts everything over 26V.
• D4: Prevents voltage shunted by TVS1 from getting to Q1's gate.
• R1, R2, D2: Provide a circuit to be limited to 18V.
• D3: Provide "regulated" 10V to Q1's gate (shutting it off) when voltage is above 18V.
• Q1: Power PMOSFET which is on by default; it's gate is grounded when voltage is nominal.
  I'd appreciate comments on the challenges this circuit presents.

Cheers!
Dirk

[MorganS]

Reading that Infineon paper, it seems like the idea is to protect against the defined ISO fault pulses. Your circuit doesn't have to withstand +120V indefinitely. It has to withstand that for a specified time and it doesn't have to block it 100%.

If it restricts the fault current at the 120V pulse condition to the point where the downstream components don't go over their ratings (eg. 25V capacitors) then it has done the job correctly. It doesn't totally ensure that the voltage is less than 14.5V under all conditions.

On the Pololu voltage regulators page, they suggest using larger capacitors for long cables and high voltage. With these conditions, the inductance of the cables is significant so the inrush current at switch-on tends to raise the voltage at the regulator's terminals above the supply voltage. The capacitor doesn't so much as absorb fault currents but it makes the fault current high enough that there's significant voltage drop on the cables and the regulator doesn't see high voltages.

[dephwyggl]

MorganS:
Reading that Infineon paper, it seems like the idea is to protect against the defined ISO fault pulses. Your circuit doesn't have to withstand +120V indefinitely. It has to withstand that for a specified time and it doesn't have to block it 100%.

If it restricts the fault current at the 120V pulse condition to the point where the downstream components don't go over their ratings (eg. 25V capacitors) then it has done the job correctly. It doesn't totally ensure that the voltage is less than 14.5V under all conditions.

Thank you for your comments, Morgan. I'm not sure I understand all of it; I certainly am not attempting to provide protection against 120V "indefinitely", and I'm trying to design everything to work at 10-18V - and significantly higher where possible. From what I've figured out, automotive environments can see transients (voltage and current) for up to 400ms. This is significant enough that things like PMOSFET gates and lower-voltage caps (like you say) might become prone to releasing magic smoke. Hence the TVS; from what I understand, TVS' will mitigate significant voltage transients long and reliably enough. What remains is trying to protect the PMOSFETs' (there are more than the one identified here) gates from the up to 26V a TVS may allow through.

I have also mentioned under-voltage protection; again, to protect the gates of the PMOSFETs used in this project, I'll need to figure out how to switch off that initial PMOSFET when voltage drops below 10.

Bottom line? My focus in this topic has been on protecting the gates of PMOSFETs; I'm trying to figure out how to get 10-18V at 20A in an automotive environment. Not having a lot of luck so far.

MorganS:
On the Pololu voltage regulators page, they suggest using larger capacitors for long cables and high voltage. With these conditions, the inductance of the cables is significant so the inrush current at switch-on tends to raise the voltage at the regulator's terminals above the supply voltage. The capacitor doesn't so much as absorb fault currents but it makes the fault current high enough that there's significant voltage drop on the cables and the regulator doesn't see high voltages.

I've read that Pololu page. Through-out my attempts at managing power in this topic, I'll always employ decoupling caps before much of anything else; the ones in this section will be 200V items. There will also be decouplers (maybe 35-50V) further down the circuit where warranted (like at the 5V regulator's Vin and the Arduino's Vin or +5V).

I look forward to any suggestions and pointers on getting to a reliable 10-18V.

Again, thank you for your contributions, Morgan.

Cheers!
Dirk

[TomGeorge]

Hi,
If D2 and D3 are zener diodes,can you designate them ZD2 and ZD3.
If D1 and D4 are diodes, can you use the symbol for a diode, you are using a zener diode symbol.

Thanks .. Tom...

[dephwyggl]

Thank you for the feedback, Tom:

TomGeorge:
If D2 and D3 are zener diodes,can you designate them ZD2 and ZD3.

Good idea. I'm still doing everything in Fritzing; it has specific symbols for diode, Schottky, and zener. The symbols for these two in my schematics are how Fritzing represents zeners; I'll try to remember to label them ZD#.

TomGeorge:
If D1 and D4 are diodes, can you use the symbol for a diode, you are using a zener diode symbol.

Technically, D1 is a Super Barrier Rectifier - basically, an enhanced Schottky; I use Fritzing's Schottky symbol for it. (Fritzing does not have a "dual anode, common cathode, TO-220 diode" component to properly represent these SBRs.) D4 is a Schottky. (There are no "plain" diodes in this schematic.) I'll try to remember to indicate Schottky diodes as SD#.

Cheers!
Dirk

[MorganS]

dephwyggl:
(Fritzing does not have a "dual anode, common cathode, TO-220 diode" component to properly represent these SBRs.) D4 is a Schottky. (There are no "plain" diodes in this schematic.) I'll try to remember to indicate Schottky diodes as SD#.

Well then please don't use F**ing. Using the wrong symbol on a schematic is worse than drawing a box that says "stuff goes here".

I like the schematics that Eagle produces. I use it for plumbing, high-level overviews and actual circuits. I'm sure other PCB software makes adequate schematics.

[dephwyggl]

MorganS:
Well then please don't use F**ing. Using the wrong symbol on a schematic is worse than drawing a box that says "stuff goes here".

The only components (in this topic so far) Fritzing can't represent "as-is" is the dual-element SBR (which I've represented as two Schottkys in parallel - which is effectively what it is), and the TVS, which is effectively a zener.

FWIW: I'm aware of - and have encountered - a number of Fritzing's limitations, but it uses symbols for diodes that easily distinguish the various types:

While I'm aware that Wikipedia is not definitive, these Fritzing symbols for diode types are exactly the same as those in its entries for diode/zener and Schottky types. Also, I've been contributing Fritzing schematics in these forums since I joined; yours and Tom's recent comments are the first I've heard that it may not be appropriate for this project.

MorganS:
I like the schematics that Eagle produces. I use it for plumbing, high-level overviews and actual circuits. I'm sure other PCB software makes adequate schematics.

I'm happy that you've found tools that work for you, Morgan. I have not tried anything (yet) that comes even remotely close to Fritzing's "user-friendliness" and speed for creating and modifying schematics - despite its significant short-comings. I'm still looking, though (on Linux, the most promising so far is KiCad), and will switch as soon as I find that Fritzing cannot reasonably do the job I need from a schematics editor, and I sincerely hope that you - and others in these forums - can see your way past any concerns you may have about the use of Fritzing.

NB: I'm much more concerned about my lack of understanding than I am about using a crude-ish but effective tool to illustrate my - at times very fundamental - questions to chip away at that lack.

Cheers!
Dirk

[dephwyggl]

I'm still stuck on preventing over-voltage (of primary concern are the gates of the PMOSFETs used in this project; they don't tolerate more than 20V). My goal is to get maximum voltage down to 18V or less. Here is what I've come up with so far:

I don't understand all of it entirely. Here are a few questions:

• From what I understand of the 3KP16A specs, TVS1 starts shunting at just above 16V, but doesn't clamp completely until 26V. Does anyone with experience with TVS' know what the voltage at B is expected to be if the voltage at A is, say, 40?

• The intent with R1, R2, ZD2 and ZD3 is that, at less than 18V, C will be conducting to ground. When voltage exceeds 18, ZD2 avalanches and C should see 10V. With this circuit, will that happen?

• I can't seem to get the calculations right for the components to switch Q1's gate. Am I even close? If not, where and/or what needs adjusting?

• I am concerned that the over-20V shunted by TVS1 may get to Q1's gate, hence the presence of SD4. Is this (SD4) even needed?

• Will the IRF4905 survive 5V at its gate while passing 20A for periods in the multiple of seconds?
  Assumptions that can be made:

• This circuit will typically see +11 to +14.5V, but droops to 5V - though more likely 6V - can occur for multi-second periods

• This circuit is unlikely to ever see anything in the 0 to +5V range

• Current required will generally never exceed 20A

• Power transients (below 0 and above 20V) can occur for up to 400ms

• Reverse polarity protection is handled by SD1 (30A, 150V Schottky)
  I appreciate any and all answers anyone can provide.

Cheers!
Dirk

[TomGeorge]

Hi,

1. The voltage at A will always equal the voltage at B.

2. If you want to turn Q1 OFF, the voltage at C has to equal the voltage at B, that is the gate has to be at zero potential with respect to the source.

Tom...