Possible to Troubleshoot an Imp that stopped working?


I have two Imps. They both have been working for the last few weeks. Last night one of them stopped working.

I know the power supplies and boards are working because the first Imp works in both, but the second (bad) Imp won’t work in either.

Is there anything I can do to troubleshoot an Imp itself?

Can you say what the bad imp is doing? Any LED activity at all?

No, no LED activity at all in either of my two development boards. Same on both whether using USB or battery power. Though, my other Imp is working fine in all scenarios (both boards, USB or Battery).

I figured this is just a random failure, and I ordered another 2 Imps and 2boards. I assumed there is not much I can do to troubleshoot, but figured I would ask.

Thanks for the reply.

How are the boards powered? You should ensure you have enough capacitance local to the board if there are long wires running from the power source. The imp can take some huge current transients and the resulting inductive spikes could be problematic without enough local cap.

Thanks, Hugo.

Well, when powered via USB I use the wall charger that came with my phone - a samsung galaxy s3. It goes USB to micro usb.

When powered by battery I use a 9v with a 1000uF capacitor rated to 25v.

Wire length from the power source isn’t much. Just a few feet at most.

I know you are busy, and this isn’t a huge deal. But I appreciate your responses.

The April board has a Mini-USB connection, so a phone charger with that style plug should work, but make sure it will output at least 400mA, as the Imp card can occasionally draw that much.

Remember that you have to move the Jumper to either USB or BATT when moving back and forth from powering the boards using the P+,P- pads and the USB connection. P+/P- pads = BATT, USB port = USB

My imp stopped working as well. It is USB powered. I got a new April board thinking that was the issue… It wasn’t the issue, still no led activity…,thoughts?

“A few feet” of wire is quite a lot in inductance terms (especially if it’s thin wire). The cap you have sounds great, you should put that between GND and VIN on the board - it’ll then work either for battery or USB supply.

Imps are pretty robust, but in the face of overvoltage they will die quickly. 5v is not good for them, spikes also bad. The CPU is in a 90nm silicon process, so small and fragile especially when compared to (eg) an ATMEGA.

Thanks, Hugo. I should clarify that “a few feet” from the power source is the length of the USB cable to the wall charger. It’s actually about 1.5’ long. I said “a few feet” because sometimes I was using a short light duty extension cord as well.

I know you wrote earlier and asked for the full system…what devices were connected? I attached a sketch.

One thing I did (which may be bad) is that a few times I think I would leave the battery connected to Vin/Gnd and the USB cable plugged in at the same time while switching the jumper. I tried to be methodical about switching power sources, meaning I would disconnect either the battery or USB, then switch jumper, then connect the other power source. But based on what you said about power spikes, I could have been sloppy about this at some point.

I’ve found the Imp to be very, very robust, and I’ve put a few through the wringer accidentally. Here is the difference between connecting a battery to VIN and to the P+P- pads. The P+,P- pads have a FET (That little guy with the resistor right next to the jumper) that protects your imp from reverse voltage, meaning hooking up the battery incorrectly. Connecting the battery at VIN means that your power goes straight in to the buck regulator, it actually doesn’t even matter where your jumper is, or if it is in place. If you look very carefully at the traces you can see this. Because if this, it is safer to use the P+,P- pads with a battery. Did you ever have the battery and the USB power connected at the same time, with the USB jumper pins covered? If so, you were supplying 5V and 9V at the same time. I’m not sure what the result of that would be.

If you are following the Instructable, note that Tom put the cap across VIN and GND, but soldered the battery to the P+,P- pads.

If you don’t have a multimeter, I recommend getting even an inexpensive one. Then you can easily see if you are getting the expected voltages at the expected places.

(If any of that is inaccurate, someone please correct me.)

Thanks, jwehr. Yes, I’m also now wondering if I had USB and the 9v battery connected at the same time. I was trying not to do that, but for a while it was a real flurry of activity.

Thanks for clarifying that the pads should ideally be used with battery, rather than VIN/GND.

So if I understand you correctly, wiring the 9v to the VIN/GND renders the jumpers inoperable (overrides them). But when wiring 9v to the pads the jumpers do indeed function, limiting power input to either the USB or 9v, but not both at same time.

Or am I wrong about that, meaning even if I had connected the 9v to the pads and had the USB connected at the same time, I might still be asking this question about if I may have applied 14+ volts to the Imp?

For now anyway, I am using only the USB…I had abandoned the idea of using battery for now since I’m mainly working on a stationary device for mechanical rooms, where power is always close by. But I may use batteries again for another version soon, and getting best practice information really helps.

By the way, I’m really jealous that you have spent a good deal of time flying over the Finger Lakes. That’s awesome. And, Southern Tier IPA is indeed one of the best beers you can get around here. For sure.

VIN doesn’t render them inoperable, it is just a direct connection to the voltage regulator on the April. When you cover the BATT pins, you are connecting the P+ pad to the same copper trace as VIN. When you cover the USB pins, you are connecting the DC+ pin on the USB port to the same copper trace as VIN. P- and all of the Ground traces are always connected. I don’t know what connecting two different DC supplies would do, but I don’t think you get +14V. Probably depends on the kind of power supply, but someone else would have to answer that.

Using USB is always a good safe solution. Only use VIN if you are really, really sure of your supply and polarity.

Awesome, thanks jwehr. That helps a lot. I really appreciate your time spent helping clarify this.

Nice diagram! Yes, you may have put both 9v and USB on VIN concurrently, however this shouldn’t have caused an issue for the breakout (and hence no issue for the imp either) because most likely the battery would end up powering the imp - most power supplies can’t sink current, and as 9 > 5, the current to power the buck would have come from the battery.

The rest of the wiring looks fine, so I guess this is a bad imp. Do contact info@electricimp.com to return it and get a replacement so we can take a look.