ESP8266 ESP07 module DoA Fix

It seems the Far-East factories can’t churn out ESP8266 based modules fast enough to feed all the world’s hackers. Well, Pick-n-Place machines are human too, so it’s not too long before you end up with a messed up batch from a factory. [Tracker Johnny] found a bunch of ESP07 modules which had their resonator mounted the wrong way around, effectively making them DoA. The resonator mounting isn’t consistently wrong too – most have reported them 90 deg offset, while others had them 180 deg. off.

Unfortunately, you need some tools and skills to fix the error. The ESP07 modules have a metal shield which needs to be removed to access the resonator. This is best done using a hot air gun. With the cover removed, you need to de-solder the resonator, and put it back in the right orientation as shown in the pictures on [Tracker Johnny]’s blog. You can find other people reporting the same fault at this forum thread. Coming in the wake of the problem with magic smoke from ESP8266 based ESP01 modules we reported earlier, it seems obvious that quality comes at a cost.

27 thoughts on “ESP8266 ESP07 module DoA Fix

      1. We therefore have a “soft” design rule not to place the interconnect of such pins between the pads. Too often you get the “oops we have a bridge there”, when in fact it’s intentional.

        1. Responsibility to realize that: 1.) the marks are likely unauthorized, 2.) the responsibility of compliance would rest upon the user, and 3.) just buying something doesn’t mean it’s going to do what it claims to do in complete compliance with every law/statute/regulation. Whatever happened to “buyer beware”? There’s nothing magical here. Common sense remains uncommon.

    1. Not surprising at all. People buying them, while these modules are cheap. There is a constant demand. This demand is not related to the certification, or quality check. Why then somebody should spend additional money for testing? =)

  1. These look hand assembled, or at least partially hand assembled and therefore would probably be hand tested too.

    Due to the high demand, probably everybody within 100 miles of the chip source, and who thinks they can operate a soldering station is trying to bang these out as fast as possible.

    Bear in mind that assembly and testing both cost money.

    Assembly = 2 mins = 10 cents.. bash out 30 per hour, 6 dollars.. x 10 hours = cost 60 dollars per day…
    Testing = 2 mins = 10 cents…. cost = another 60 dollars per day

    Most Sales = cheaper than the rest… since you cant supply them unassembled, but you can supply them untested, it stands to reason that you dont bother, that way you ship them all and if your DOA rate is low, you still make more monet than they guy who tests them… possibly as much as twice what the other guy makes.

    Capitalism, red in tooth and claw, and in this case red in Dragon too.

  2. 1. if the lack of testing is about money then why not do a walmart? (“oh we need you to do this as soon as you get punched out”) and work them off the clock chinese dont have the american labor rights so it should be ok.

    2. depending on the type of oscillator plain crystal like used in cb and rc cars of the 80’s it may not be a problem but today’s oscillators have electronics built into them so they wont work if reversed.

    3. is the wifi frequencies and power levels controlled like as with other radios? could it be a way to sell you a transmitter that is too high power for the license and simply dont allow it to work out of the box so then by you fixing it it now becomes modified thereby putting it into experimental or some other unlicensed status?

    1. 1) You can “force” someone to work 10 or 20% extra “off the clock”, but not 100%. If testing would be 10-20% of the job, you can guess how “skipping the tests” started.
      2) Above about 20MHz, crystals are overtone crystals. A 25MHz crystal would be an 8.33MHz crystal that can be made to resonate at 25MHz too. Give this crystal to the random hacker and he’ll complain that it runs at 8.33Mhz and not the stamped 25MHz. The solution is to sell the crystal with the oscillator included in the package.
      3) The module is already a “component” and therefore the responsibility to stick to regulations belongs with the entity that integrates it into a product.

    1. Resonators do not have the low tolerance or frequency for this application. It would have to be a crystal or a hybrid crystal resonator. Some resonators need the capacitors too while the 3 pins one have integrated capacitors.

      Resonator =/= crystal oscillator.

  3. Hi, I’m an embedded engineer, I bought a unit of these modules (esp-7), removed the metal case, and found the resonator in the correct position, as it shows in the schematic of the module. The picture that you say it shows the right position of the resonator is in fact incorrect, you’re probably running the module on its internal resonator. Pin one of the resonator should be connected to XTAL_IN of the chip, and 3 to XTAL_OUT, as it exactly was when I removed the metal and looked at it. The device works great with its original firmware, flashing the device with the open source firmware, I found out, slows it down a bit.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.