GPS LNA overheating on custom PCB

[JulioCalandrin]

Hi! This is basically an update on another question I made a while ago (link down under), but I have a different problem now.

For context, I'm designing a low-cost GPS tracker/radiosonde for my electrical engineering course conclusion project. I'm using the ATGM336H geopositioning module because it is the cheapest one I could find, and as suggested by Kean, who answered my initial post, I went with a passive ceramic patch antenna and an external LNA; I used the AT2659S because that was suggested in the GPS module datasheet (all links are below).

I used the exact same circuit suggested in the ATGM336H datasheet, which agrees with the circuit on the AT2659S datasheet. I also found that this LNA is widely used on those cheap active antennas you can get from AliExpress.

But after manufacturing the PCB and soldering the components, the LNA heats up quite dramatically (nearly too hot to touch).

The GPS fails to get a lock (hanging out of a window with clear sky), so I assume the LNA is not in fact working, and as a desperate test, I removed the LNA and tested directly connecting the passive antenna to the module, which, to the surprise of no one, does not work particularly well.

I checked the pins for shorts and the pinout multiple times and even thought the component was soldered backwards, but I checked and it was not. I also thought the LNA was a faulty unit, but I bought 5 units and tested 3 of them, all presented the same behavior.

I am planning to make a breakout PCB to test the individual pins of this LNA and investigate if it is actually the LNA's fault (perhaps the GPS module is faulty with a low input impedance), if some pin is drawing too much current, and also test if the on/off pin on the LNA does anything. But before I do that, I would like to ask if there is something clearly wrong with my circuit (photos attached below) and if someone has another theory on why this is happening.

Please help, I need this project to get a degree.

Thanks in advance!


Description of the attached pictures:

Circuit from the ATGM336H datasheet: atgm336h_circuit_example_from_datasheet.png

My circuit: my_circuit.png

My layout (zoomed so that track names are legible): LNA_circuit_layout.png


Links:

The initial post: https://www.eevblog.com/forum/rf-microwave/ceramic-patch-antenna-vs-smd-multilayer-chip-antenna-for-gps-tracker/msg5049955/#msg5049955

Link for the antenna: https://www.lcsc.com/product-detail/Antennas_Shenzhen-Kinghelm-Elec-KH-GPS121204-WY_C962213.html

Link for the LNA: https://www.lcsc.com/product-detail/Low-Noise-Amplifiers-LNA-RF_ZHONGKEWEI-AT2659S_C883443.html

Link for the ATGM336H english datasheet: https://www.icofchina.com/d/file/xiazai/2016-12-05/b5c57074f4b1fcc62ba8c7868548d18a.pdf

[Mr Simpleton]

Looking at the datasheet, the chip uses 4.3mA @3.3V and this would not cause even a luke warm chip! So in your case something is wrong, my bet is a broken chip, replace it and watch the current consuption.
And as the pins are nicely placed on the chip, the only thing I would be worried about is if pin 3 (input) is grounded, may cause problems. But remove the old chip, and measure on the free pads on PCB to verify everything is OK before replacing the new chip.

[berke]

How about inserting a cap between RFOUT and RF_IN?  Maybe it can't be DC coupled.
Also why is VCC_RF connected to NRESET?

[JulioCalandrin]

What cap value would you suggest I use between RF_OUT and RF_IN?

And the 0 ohm resistor connecting VCC_RF and NRESET is there because I saw a few reference designs that shorted them, probably so that NRESET is always high. Since I am using a GPIO to switch that pin, I added a 0 ohm resistor that was not soldered on the PCB, but it's there just in case I need it.

[eb4fbz]

Are you using AT2659 or AT2659S? Pinout is different (mirrored).

[berke]

What cap value would you suggest I use between RF_OUT and RF_IN?

I don't know, depends what you have in your drawers.  You want something that is still a capacitor at 1.5 GHz.  Maybe try a 12 pF 0805 C0G?  Yes it's -10j ohms but judging from some Kyocera curves larger Joe random caps won't do the trick as they are likely to be inductive above a couple hundred MHz.

And the 0 ohm resistor connecting VCC_RF and NRESET is there because I saw a few reference designs that shorted them, probably so that NRESET is always high. Since I am using a GPIO to switch that pin, I added a 0 ohm resistor that was not soldered on the PCB, but it's there just in case I need it.

OK

[JulioCalandrin]

Are you using AT2659 or AT2659S? Pinout is different (mirrored).

Oops. I'm using the AT2659S. I corrected the initial post. Thanks!

[geggi1]

Have you checked for oscillations?

[Kean]

Are you using AT2659 or AT2659S? Pinout is different (mirrored).

Oops. I'm using the AT2659S. I corrected the initial post. Thanks!

Maybe the parts you received are incorrect.  Does the part actually say "AT2659S" on it? (like in the LCSC photo)
I was going to suggest you may have got AT2659 instead, but I see that is a tiny 1x1.6mm DFN6 instead of the SOT-26.

[MarkT]

But after manufacturing the PCB and soldering the components, the LNA heats up quite dramatically (nearly too hot to touch).

ESD damage?  Some RF amp chips are exquisitely sensitive to this as protection circuitry would destroy the RF performance.

[JulioCalandrin]

Update here friends.

I reinstalled the LNA, and after uploading a GPS test firmware everything worked flawlessly, GPS lock and cool LNA. Obviously, I was intrigued and thought I had made some stupid mistake.

I continued working on other sensors on the board, and then I noticed the LNA was suddenly hot again. The LNA only got hot when a different code was uploaded, and the code in question had a piece of the GPS test code that initialized the GPS module by enabling the ON_OFF pin (that activates both the GPS module and the LNA) and performed a reset via the NRST pin on the GPS module.

After an afternoon of debugging and burning the tips of my fingers touching the LNA to check if it was toasty, I noticed the problem occurred at a line of code where I disabled the 3V3 DC converter that feeds all the peripherals (done like that for power-saving reasons) and enabled the 3V3 power again AFTER I had initialized the GPS.

So the magic sequence to turn the LNA into a toaster is:
1) Have the 3V3 power off;
2) Enable the ON_OFF pin (HIGH signal from an STM32 - powered by a separate 3.0V LDO Linear Regulator);
3) Turn 3V3 power on.
4) You now have active thermal control

After that, the only thing that cools down the LNA is turning 3V3 off again. Disabling the ON_OFF pin does nothing.
Also, I thought it might be a transient in which the GPS module draws too much current, but in that case the LNA would only heat up momentarily, and in reality it stays hot for as long power is on.

Note that if I FIRST turn the 3V3 on and then turn the ON_OFF pin HIGH everything works fine. So yes, I can easily solve it in firmware and move on, but I really want to know what on earth is going on. It makes total sense to first power up the device and then enable it, but I don't think the heating up part should happen.

Does anyone have an explanation?

And thank you very much for everyone that showed interest in the issue and sent suggestions

[radar_macgyver]

ATGM336N module may be providing 3.3V power at pin 11 (RF_IN) to provide bias power for an external active antenna - there should be a cap between it and the LNA. When the LNA's pin 5 is pulled high it may be getting phantom power from the STM32 (as you say it has its own supply). Just a guess, but could the LNA be going into latchup? Not clear from the DS if it's a CMOS process, which is likely to latch up.

[Kean]

Yes, this sounds like a classic example of latch up caused by an input voltage exceeding a power rail.
https://en.wikipedia.org/wiki/Latch-up

Is there a reason that you are turning off power to the GPS and LNA rather than just disabling them?
There might be some power saving, but it will mean losing any ephemeris data and leading to a cold start and delay in getting a fix.

[JulioCalandrin]

I had never heard of latch-up. It makes total sense now. Thank you very much for the insight and all the help from everyone.

The reason I'm turning it off is to completely eliminate power consumption from sensors during long power-saving periods. My project involves designing a device that can last weeks on a single 18650 cell, so disabling the 3V3 rail was the easiest way to accomplish that. Not elegant, but it works.

During operation, I'll definitely just do short sleep cycles via the CASIC/NMEA command interface (the enable pin does not work, apparently; the ATGM336H module still screams NMEA messages with the pin either high or low, but I need to investigate further).