Why not integrate Jaguars with Diodes?

[KrazyCarl92]

Most people in the FRC who have experience with Jaguars know how frustratingly fragile they are. If you hook up the positive lead from the PDB to the V- and the negative lead from the PDB to the V+ on the Jag accidentally, the thing is busted, blown, and destroyed. So I have to wonder, why wouldn't they integrate a diode into the Jaguar's circuitry so that current would only pass through the one direction? Seems like that would be 10 cents to fix a problem rather than a $90 fix for having to purchase an entirely new Jaguar.

Not asking if this would be FIRST legal (obviously you couldn't modify the internal circuitry of a motor controller), but just asking if this is something that could be integrated to make the product more durable, maybe by a manufacturer in the future. I know I'd be willing to pay an extra $10 for a product that wouldn't break when someone on the team unknowingly swaps the black and red screws...

[lemiant]

If they broke less people would buy less

[Joe Ross]

Originally Posted by KrazyCarl92

I know I'd be willing to pay an extra $10 for a product that wouldn't break when someone on the team unknowingly swaps the black and red screws...

Would you pay $10 more for a speed controller that performed 6% worse?

[Jim Wilks]

Originally Posted by KrazyCarl92

Most people in the FRC who have experience with Jaguars know how frustratingly fragile they are. If you hook up the positive lead from the PDB to the V- and the negative lead from the PDB to the V+ on the Jag accidentally, the thing is busted, blown, and destroyed. So I have to wonder, why wouldn't they integrate a diode into the Jaguar's circuitry so that current would only pass through the one direction? Seems like that would be 10 cents to fix a problem rather than $90 fix for having to purchase an entirely new Jaguar.

Not asking if this would be FIRST legal (obviously you couldn't modify the internal circuitry of a motor controller), but just asking if this is something that could be integrated to make the product more durable, maybe by a manufacturer in the future. I know I'd be willing to pay an extra $10 for a product that wouldn't break when someone on the team unknowingly swaps the black and red screws...

A diode to handle the 60A or so that a Jaguar can handle would cost a lot more than 10 cents. As well. it would create a lot of heat that would somehow have to be dissipated. Finally, you would have a power reduction in the output due to the 0.7 volts lost in the diode. This would translate to something in the order of 6% at full output.

These all go together to make the "fix" you suggest a bit more difficult to accomplish.

[MichaelBick]

You can do a mofset instead, I'm pretty sure p-channel. Much more power effficient, and you can wire them in parallel to handle more amps.

[lemiant]

Your explanations of the possible problems make a diode implausible, but I feel it's nothing a little more complicated circuitry couldn't handle. The highly capable EEs who designed the jaguar should be able to come up with solutions better than a couple of kids who aren't yet our of college. For alternatives you could use Michael's solution or:

The diode wouldn't have to affect all the power, just the power going to the control circuitry. If the jaguar's internal computer didn't turn on, it would never send any signals to the internal relay (or whatever they use) and nothing bad would happen. The computing uses a lot less amperage so this is totally feasible.

[Matt Krass]

Originally Posted by MICHAELABICK

You can do a mofset instead, I'm pretty sure p-channel. Much more power effficient, and you can wire them in parallel to handle more amps.

I was contemplating this solution today, as my team destroyed a brand new Jag today with reversed wires. The simplest way to do that would be a P-channel FET with a grounded gate, power to Drain and output from Source. A suitable (set of) FET(s) would have to be as robust and efficient as the N-channel FETs in the H-bridge, which complicates things, it could be done, but my hunch is that it wouldn't be space, power or cost effective. Perhaps, using an N-channel FET or two in parallel with a diode protected driver IC with a charge pump might work better, but its still going to waste some power to heat, and add complexity and cost.

I think overall, despite the very frustrating problem of reverse polarity burnouts (today still irks me!) it wouldn't be worth it for such a reverse polarity protection scheme, I'd rather try to educate our students better so this doesn't happen again.

Matt

[Ether]

Originally Posted by Matt Krass

my team destroyed a brand new Jag today with reversed wires.

Can you share why that happened, and any measures you are contemplating as a result? There might be a lesson here that others could benefit from.

[DonRotolo]

Or hey - here's a thought - let kids learn that if you hook it up backwards, it goes poof?

As far as training classes go, $85 plus shipping isn't all that expensive.

Or maybe tatoo it onto their foreheads: "The world is not idiot-proof. Pay Attention." Backwards, so they can read it in a mirror.

[efoote868]

Originally Posted by Matt Krass

I was contemplating this solution today, as my team destroyed a brand new Jag today with reversed wires. The simplest way to do that would be a P-channel FET with a grounded gate, power to Drain and output from Source. A suitable (set of) FET(s) would have to be as robust and efficient as the N-channel FETs in the H-bridge, which complicates things, it could be done, but my hunch is that it wouldn't be space, power or cost effective. Perhaps, using an N-channel FET or two in parallel with a diode protected driver IC with a charge pump might work better, but its still going to waste some power to heat, and add complexity and cost.

I think overall, despite the very frustrating problem of reverse polarity burnouts (today still irks me!) it wouldn't be worth it for such a reverse polarity protection scheme, I'd rather try to educate our students better so this doesn't happen again.

Matt

Making a product like this fool-proof takes away an important $90 lesson, which can easily be split between 3-15 people. The other thing is that the lesson learned doesn't only apply to the $90 part, but potentially 1000 dollar or more parts.

When I was still in high school, one of the freshmen on the team fried the camera. If I recall correctly, that camera cost about $200. While they felt bad about it for a week, I'm certain they or anyone there never made that mistake again.

[Matt Krass]

Originally Posted by Ether

Can you share why that happened, and any measures you are contemplating as a result? There might be a lesson here that others could benefit from.

We were rewiring the chassis (second pass for neatness, after testing drive systems) and under the gun of a ticking clock, a student connected the Jaguars power input quick connects backward to the main power connection routed for it and powered on the robot without having the wiring cross checked and the Jaguar bit it. In the aftermath we will be focusing on more stringent examination after any rewiring, and better education of team members so that they are more aware of the dangers.

Overall a lot of people (myself included) feel sheepish about this mistake, I think the lesson will propogate well on its own.

Matt

[EricVanWyk]

Yesterday I saw this post and smiled. There is nothing that makes me happier than knowing that a buck saved some student a couple hundred dollars of grief.

We evaluated every single power input connection for protecting against reversed inputs, and added it only where it made sense. In general, high power devices are quadratically more expensive to protect than low power devices. By expensive, I don't just mean raw dollars: I also include power loss and safety.

For example, the PD only reverse protects its power supplies, and does not protect the normal breaker outputs. The power supplies are ~50 Watts reasonable peak with mild pulse, and it cost a dollar. The entire robot is 2ish kiloWatts reasonable peak (with painfully higher pulse), and would have cost 40 dollars of FETs plus a heat sink plus a fan plus a significant drop in total robot performance.


There are two types of reverse input protection: Pass and Crowbar.

Pass solutions are things like series diodes or series FETs that have to handle (pass) the normal operating current. The control system components in blue plastic all use pass protection.

Crowbars are things like shunt diodes that short out the power input in an attempt to blow an upstream fuse before exploding. These make me nervous: It is a protection against incorrect wiring that assumes that the rest of the robot is correctly wired. In particularly badly wired robots, they can make the situation worse. I typically use crowbar protection in more controlled environments.

[Andrew Schreiber]

Originally Posted by efoote868

Making a product like this fool-proof takes away an important $90 lesson, which can easily be split between 3-15 people. The other thing is that the lesson learned doesn't only apply to the $90 part, but potentially 1000 dollar or more parts.

When I was still in high school, one of the freshmen on the team fried the camera. If I recall correctly, that camera cost about $200. While they felt bad about it for a week, I'm certain they or anyone there never made that mistake again.

So, what you're saying is "Measure twice, cut once" applies to things other than saws? Who'da thunk that being careful and thinking through what you were doing would have practical side effects?

[Matt Krass]

Originally Posted by EricVanWyk

Crowbars are things like shunt diodes that short out the power input in an attempt to blow an upstream fuse before exploding. These make me nervous: It is a protection against incorrect wiring that assumes that the rest of the robot is correctly wired. In particularly badly wired robots, they can make the situation worse. I typically use crowbar protection in more controlled environments.

I was actually just toying with this solution last night in our lab, given the rather unnerving ratcheting click noise coming from the PD board, I thought it served as a fair indicator to turn the robot off immediately while saving the Jaguar, I also used a fairly beefy power rectifier to do it, so heat wasn't an immediate concern. I'm more concerned with the effects of repeatedly shorting the thing out rapidly.

Just curious, because I'm sure there's an angle I haven't thought of, what kind of incorrect wiring are you anticipating causing a crowbar to be an ineffective solution?

(I am mostly toying with idea out of build season curiosity, since I don't believe the current electrical rules would permit such a protection circuit in competition use).

Matt

[Sammy 3278]

We replaced the colored screws with all silver so that people won't assume black means negative and red means positive and will look at the molding on the speed control instead. Has been working so far. We had also toyed with the idea of using red and black nail polish on the speed control.