-->
Page 1 of 1

Is having no resistor on ch_pd enough to fry the chip?

PostPosted: Sat Jun 11, 2016 4:31 am
by coldpenguin
I was trying to set up an esp-12f with minimal components to be a final prototype for a pcb

Power was from a 9v rms supply, through a full bridge rectifier, then a 47uf cap across the output through to an mp1584 IC/pcb set to output 3.28V DC as measured by an oscilloscope. The output without load had ripple but the AC component was minimal.

To try and confirm the chip was working, I made the following connections:
Vcc +3v, Gnd to Gnd. A 102 capacitor across this at the pins.
A 10k resistor from GPIO15 to Gnd.
I connected ch_pd direct to vcc
Touched GPIO00 to gnd, then reset by disconnect/reconnect vcc

The usb to serial adaptor I used is a 3.3v adaptor I have used a lot before, attached gnd, tx and rx only (not 3.3 to vcc)

I then used the esp flasher to upload the AT firmware from April at about 50% of the way through the chip started 'chirping'
At 89%, the flash failed, and the chip is no longer responding. It is drawing .144A

Re: Is having no resistor on ch_pd enough to fry the chip?

PostPosted: Sat Jun 11, 2016 7:35 am
by xtal
I would have to suspect that maybe you connected the wrong pin to V+, that why I always use a resistor, when prototyping, it dosen,t matter how care full you are, there's always room to mess up.
By chirping I would assume the chip is self destructing inside the metal cover, there is also a possibility that chip enb was partially shorted to another pin.....

Re: Is having no resistor on ch_pd enough to fry the chip?

PostPosted: Sat Jun 11, 2016 9:41 am
by coldpenguin
I should have said, I tack soldered the ch_pd using a breadboard cable through the hole, then to the capacitor leg (which itself went through to VCC). Once in place, there could have been no change in the placement of the pins.
If I had gotten the wrong pin (they are marked on the silkscreen as they go through the board) then I wouldn't have expected the chip to have flashed at all? (The flashtool performs a 'sync' downloading the MAC address of the chip prior to starting the flash, so TX and RX must have been fully functional to the device for this to occur).

I guess from what you are saying though, is that a resistor shouldn't have been necessary, but would have been ideal? I haven't used any on the ESP-210s I have been using, but I guess there may have been one on the boards themselves.