STM32WL matching circuit included in the chip ?

[emil28]

Hello,

I am very interested in using the STM32WL for a wireless application. Is the matching circuit already included in the chip (so I just have to add an antenna) ? I don't find any information about this.

Thanks.

[dbgarasiya]

which part number you are working on it?

[roman239955]

I have the same question and I've searched for "impedance" in all the documents that I've downloaded relative to the STM32WL and SX1261/2. I'm interested in PA impedance on both the HP and LP RF outputs on all 3 STM32WL chips. Semtech doesn't mention it in the SX126x data sheet.

There are no impedance matching inductors in the example designs, the PA supply inductor is the same in relative example designs, but the lead filter cap changes.

I want to use this chip with an external PA and would like to lower the max power output further than what is given.

Does ST or Semtech have documentation similar to SiLabs "AN627 Si4X6X AND EZR32 LOW-POWER PA MATCHING"?

[roman239955]

I found what I was looking for in Semtech app note AN1200.40

[Andrew Larkin]

The app note describes the process of measuring the PA output impedance using a "load pull analysis" but I am unclear how to actually do this. Likewise a "source pull analysis" for the receiver.

What I am not entirely clear on is why these values can't be documented somewhere. For a production chip, they have to be pretty stable otherwise it would be necessary to adjust the matching network for each individual chip which would be untenable for a production board.

I have to assume that the actual values discussed in AN1200.40 will not exactly match the STM32WLE chips.

You mention example designs - I haven't found them yet. Any links or hints?

[roman239955]

Hello Alark,

I think what you're looking for and what I was looking for 2 days ago are not here yet.

I've found that I must either pay to get the load matching done or I'll need to do it myself. So far it looks like it's about $3k US to have someone do the load matching here.

ST will have to fill you in on why the values aren't immediatey available right now.

I'm not going to comment on why you we don't have an example design to copy for a product that's barely released but I can tell you that Tech support sent me to this 'community' to get support on this product and there were only two posts when I first looked at it.

Reference designs at:

https://www.semtech.com/products/wireless-rf/lora-transceivers/sx1261#download-resources -look under Reference Designs.

AND

ST AN5408

I agree that the AN1200.40 will probably not exactly match the STM32WLE chips and I don't like the filter layout in the Semtech examples either but I do have chips in hand and my first rev proto boards will be done in a couple weeks.

Cheers ,

Roman

[Andrew Larkin]

Thanks, Roman.

I have been digging through the Semtech designs this afternoon. I thought when you said "There are no impedance matching inductors in the example designs," that you had found something specifically on the STM32WLE. I would be surprised if ST didn't do something like a discovery board equivalent to what they did with the Murata module.

This will be my first BGA board design. BGA is a swamp I have avoided to date but this chip makes it almost worth it. Having said that, I am hoping that other footprints might appear based on this comment: "Four ECOPACK2® compliant packages, featuring three different memory sizes, are available to address different levels of complexity to ensure a cost-efficient fit for application requirements". We have only seen one package with three memory sizes as of now.

AN5408 looks interesting and I hadn't picked up on that. Looks like this is the missing piece!

Cheers back!

Andrew

[lz1im]

Hello, Andrew, Roman, and Emil.

I think that ST buy die of SX126x, ST /with or without Semtech/ make some modifications concerning Power Amplifier, get some formulae of what some parameters means /say, Fh and Fl of calibration/, but good part of the ST datasheet, concerning Radio is that from Semtech. Now, I asked Semtech about LNA and PA impedance, they told me to contact sales representative!! ha!, anyway I wrote to him three weeks ago and still have no answer. So, my hope is in the ST, based of our previous conversations at the forum, about SPIRIT1. Semtech logic is simple: LORA is allowed at 868 MHz, and rare at 433 MHz, so here is our reference design. About rest of whole range 150-960 MHz, they say nothing, except: Buy VNA, if You have not, and make measurements, et voila!, You will have impedances, we omitted at Datasheet. That is shame!

Cheers!

Pavel

[Andrew Larkin]

The material I have seen so far from both Semtech and ST points to using VNA to match the antenna to 50 ohms. What hasn't been discussed is the source impedance of the chip itself which is clearly not 50 ohms.

There is two matching networks on the Tx path. One between the chip and the RF switch. The second between the RF switch and the antenna. It is the first network that is thin on documentation.

I suspect the impending Nucleo reference board will resolve initial component values for the first network at various frequencies.

But I would love to have the source impedance value so I can test my own PCB with network values basically from the Tx pin out to the antenna.

Same thing is needed on the Rx path.

Perhaps ST is waiting for the first full production version of the chip to characterise these values and document them?

In the mean time, if you have any clues as to how to use a VNA to perform a "load pull analysis" (as per Semtech docs) on the chip, I would love to see it. The only materials I have found so far on this assumes you have access to both the input and the output of the amplifier and we just don't here.

[SHATE.1]

THE MATCHING CIRCUIT IS NOT INCLUDED ON THE CHIP SO YOU HAVE TO ADD IT EXTERNALLY.

YOU HAVE THE BOARD LAYOUT RECOMMANDATION ATTACHED

THANKS

[lz1im]

Hello, SHATE.1. Well, that AN is Excellent aboutPCB design, it is more intuitive, with practical formulae. I think, rules and advices in that AN are enough for RF PCB design, at least below 1-1.5 GHz. The main problem, concerning this thread, is, that there is no information about impedance of receiver differential input at frequencies 150-960 MHz, which is operating range if STM32WL. Same situation is at transmitter output. Without that impedances, there is no possibility to apply that excellent AppNote,.

[Hamilton DC]

Hi @lz1im​ ,

Thanks for your appreciation of the AN5407. I'm finishing the next AN for the matching network of this series. It will be called AN5457 "RF Matching Network Design Guide for STM32WL Series". In this AN you will find the information that you're looking for. This AN will be available in few weeks.

Best regards,

Hamilton

[Hamilton DC]

Hello,

You can find more information in this document: AN5457 - RF matching network design guide for STM32WL Series

Kind regards,

Hamilton

[TWard.1]

Hello @Hamilton DC​ - sorry to jump on this thread but I'm looking for help with RF tuning of the PA.

The approach we are taking is to remove our STM32WL from the board and to measure the impedance with a VNA looking from the HP output, then search for a complex conguagate match to the values provided in Appendix A of your app note. We're able to find this match with some significant tuning (12.52+j3.67 at 915MHz) but this does not give us the power we expect. I wanted confirmation that my approach is valid beofre ripping up the design...