YX8018 IC Solar Lawn Light not working - bad chip or bad installer?

Thread Starter

davidw963

Joined Jun 24, 2019
3
Good Morning - first post here!

I am in the process of repairing a hard-to-replace solar lawn light that packed it in a few weeks ago. I disassembled it and tested away, finding the chip was unserviceable (not to mention almost impossible to find unless I wanted to purchase a batch of 10,000). So I decided to gut it and rebuild the entire circuit with simpler and easier to find components. I opted to use the YX8018 IC as it was plentiful, inexpensive, simple and could use either 1.2v or 2.4v as input voltage as the old light used 2AA cells. I downloaded the datasheet (attached), built the circuit on a breadboard in accordance with the datasheet (1.2v config) using a red LED and 100 microhenry inductor. I hooked up the power and......nothing. Reversed the LED in case of installer error - still nothing. Checked the wiring - good. Checked the circuit construction - should be good (disclaimer: I have a technical background, however breadboarding and this specific application is pretty rudimentary). Checked all the components (see the pictures) - looks good. Replaced the YX8018 chip with another - still nothing.

I scoped out the output of the inductor on the drain side and got a pattern, however it was not amplified to the 3.5v that was supposed to be present (see pictures). This was consistent for both chips. I am unsure if I got a few bad chips or if this is my bad / ignorance. I would sure appreciate the wisdom of you guys here on this one.

Cheers
 

Attachments

Marcus2012

Joined Feb 22, 2015
383
Hi

Welcome to AAC!!

I have had an excessively high fail rate with these before (about 6/10!!) and they are very sensitive to thermal shock. That being said though if yours has an output signal something is working and you have not soldered it yet. Can I ask what you are using as solar panel input for your bread boarding? Afaik these chips work by using the solar panel output voltage as an ambient light detector. It might not be working if it thinks it is daylight.
 

Thread Starter

davidw963

Joined Jun 24, 2019
3
Thanks, Marcus2012. I left the CE (chip enable) pin floating, as I believe that if it senses voltage (from the panel) it shuts the oscillator down and lets the battery charge. I could try to ground it out and see what happens. I guess I will have to be more disciplined with my handling of these - didn't think they were that delicate. I appreciate your help.
 

Marcus2012

Joined Feb 22, 2015
383
Well that sounds like it should be an adequate substitute and all seems correct. Are the LEDs ok? Vf for Red ones are quite low.

EDIT

I did notice this though, the LED is the wrong way around, but guessing this was corrected prior?

Breadboard 1.jpg
 

Thread Starter

davidw963

Joined Jun 24, 2019
3
Marcus - I did reverse the LED, and still had nothing. More on this to follow.

Audioguru - The picture was more to represent the setup on the board in terms of components. I didn't use the power rails, I just connected the power and ground to the applicable row - as demonstrated on the test leads picture.

So here is the final follow up on this: After pulling out more hair than I care to admit, I purchased another batch of YX8018 chips from a different vendor, plugged one in and it was like magic - the LED came on! I even scoped it to see what the wave should look like (see below). Since then I have soldered the circuit on protoboard and ready to install in the old housing.

I have replaced plenty of parts in my experience, with only a few failing as a brand new replacement/install. However, to have ALL 5 chips in the same lot fail was either my bad luck or I got taken for a sucker and given dead components. Either way, the lessons are learned and all is right in the world. I sure appreciate all of the time and help you guys have given me on this. Cheers.
 

Attachments

Marcus2012

Joined Feb 22, 2015
383
Great news, glad it's all working now!! :)

You're not alone, I was in exactly the same situation with some QX5252s and I was certain it was something I did wrong for ages. Not entirely sure why these have such a high failure rate but I never did find out if it was the vendor or the manufacturer. I couldn't seem to find any alternate, more trustworthy source either (RS, digikey etc) so it's a roll of the dice whenever you use one it seems.
 
Top