0
\$\begingroup\$

I'm prototyping a device whose schematics is open source and available on-line. I have modified the layout to fit it into an Arduino Uno form factor and now I got the PCBs along with components. Once soldered altogether, the device is not getting detected.

It uses several chips I suspect has burnt off but when I check the ground and Vdd connections, it doesn't show any shorted paths on the PCB. Device has a MCP2200 chip as the UART bridge and the datasheet states that its ESD protection is > 4kV HBM. It also uses a PIC24EP256GP204 micro-controller and when I try to connect it to a PicKit3 using MCLR, VCC, GND, PGD, PGC pins, it throws the device ID 0x0 is not valid issue.

I tried removing the PIC IC and still pickit3 shows the same error. It kind of tells me that the PIC is also burnt, isn't it?

The MCP2200 chip should be identified by the PC when connected. So I desoldered the IC using hot air and mounted it in an extra PCB and made the connections similar to a breakout board;

MCP2200-Breakout

But that is also not detected by the PC. I have a couple of questions in this regards.

  1. Does high ESP protection mean higher tolerance to short circuits? (I'm not sure if the PCB has shorts at all, because when I check every pin to PIC IC I could only read the correct voltages at each pin)

  2. Is there a chance to get the MCP2200 IC and the PIC IC to be burnt while I am testing path connectivities (Using diode check in multimeter)?every pin to PIC IC I could only read the correct voltages at each pin)

\$\endgroup\$
1
\$\begingroup\$

Consider the following points:

  • what is the component source you used? If you bought them on Alibaba or from another non-trusted source, then they must just be defective on arrival (DOA);
  • you say not much about your soldering techniques. You may have just overheated the chips during hot air soldering. Depending on the package type, I would solder chips manually with soldering iron with controlled temperature grounded to the soldering station. Remember: quality of tools 80% define success of your works. If you use bad, faulty, incorrect tools, or use tools improperly, you will get what you get.
  • you did not solder chips with hot air properly. Take magnifier and examine soldering. Take a pin and test soldering quality putting pin between chip pins (if not BGA). Or look for shortages on the pins. Using multimeter you already have proven that there's no shortages between power rails, but there could be between signals, and power rails and signals. There could be correct voltage while signals are still shorted. Visual inspection and checking with multimeter in voltrop (diode sign) mode required.

PIC IC to be burnt while I am testing path connectivities (Using diode check in multimeter)?every pin to PIC IC I could only read the correct voltages at each pin)

Most probably not; and how do you know that read is correct for specific test points? You do not say that board is not detected - and it is logical issue rather than electrical (so far), and this logical issue may be caused by some electrical issue - either problem in circuit design, quality of soldering or faulty components.

And finally... did you supply correct power to the chips? Is your PickIt3+software operational?

Update:

I bought the components from Mouser

Sounds good!

using the diode check and when I measure resistance between Vdd and GND pins, it says 600 Ohms

It is not Ohms, it is Volts, a voltage drop. 600 is good value for semiconductor, if you would measure in another direction it usually should measure with higher value (1200 or infinity).

I used temperature at 200 C

You must know melting temperature of solder you use, and the nominal soldering temperature to use with hot air for this solder type. You may find that you just did not solder the pins well. As I said - perform visual check and if you can resolder with soldering iron.

would it say that the MCP chip is faulty? Or could it be the 12 MHz oscillator?

You can definitely say that chip is faulty only in two conditions

  1. you replace the chip with new one, and board starts working;
  2. you use advanced diagnostic device(s) (multimeter and scope) to prove that under defined conditions chip misbehaves.

Until that is done I would suspect quality of your work rather than chip fault.

And again, are you sure that circuit you designed is going to work at all? That you did not make mistake in circuit and did not make mistake on the board? This is also a good question, which you must make clear before anything else.

\$\endgroup\$
2
  • \$\begingroup\$ I bought the components from Mouser. The soldering technique I used is solder paste with hot air (SMD rework station). I used temperature at 200 C and I cleaned the board with IPA after each component is soldered. I checked each pin for shortages using the diode check and when I measure resistance between Vdd and GND pins, it says 600 Ohms. Is that normal or shouldn't it be infinity? \$\endgroup\$
    – Padmal
    Jan 25 '18 at 8:30
  • \$\begingroup\$ And if I just isolate the MCP2200 chip from the circuit design I had and connect pins according to a breakout board as I've shown in the question, would it say that the MCP chip is faulty? Or could it be the 12 MHz oscillator? \$\endgroup\$
    – Padmal
    Jan 25 '18 at 8:33
1
\$\begingroup\$

Does high ESP protection mean higher tolerance to short circuits?

ESD protection is not intended to protect against short circuits. Since short circuits are current, and ESD protection acts on voltage.
If the ESD protection also provides a series filter, then maybe this filter acts as fuse. But it's not intended to.

Is there a chance to get the MCP2200 IC and the PIC IC to be burnt while I am testing path connectivities (Using diode check in multimeter)?

Unlikely, unless your meter creates a higher than average voltage in diode mode. Fluke typically uses 2 to 3 Volts.
However, you might be able to damage ultra low voltage parts or very sensitive analog circuits.

\$\endgroup\$
1
\$\begingroup\$
  1. No, ESD protection does nothing to prevent short circuit damage. If anything, it creates new paths for currents to flow. Protection against a short is suppored to cut or limit the current going out of control.

  2. No, you cannot damage MCU pins by probing them with a (reasonable) multimeter, which should not apply more than a couple of volts to the probes and limit the current to a few milliamps.

If you suspect you have accidentally shorted some pins during testing, or connected something with a wrong polarity, or applied voltage to MCU pins while ground and power were not connected (without limiting the current, like the multimeter does), it is possible that this has damaged those pins, or the whole chip.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.