Help with QN8027 FM chip

[PM3295]

People,

Is there anyone who had some experience using this FM chip?
https://datasheetspdf.com/pdf-file/761694/Quintic/QN8027/1


I got the chip working, but the RF output is very low. The data sheet states 121 dBuV but no reference impedance. If I assume 50 Ohm, that means that you should get +14 dBm at the full-power  setting. The best I can get is about -14 dBm. There is a register value that determines the output level, but leaving it at default or writing the max level to it, the level never goes above -14 dBm. It is stated that leaving it at the default value should give maximum RF output power.

I have tried a few chips to make sure I am not dealing with a faulty chip, but they all produce a maximum of -14 dBm.

I watched a few YouTube videos of people using the chip and from what I could see from the one video, the person measured the output level on a SA and it looked like a level around the -15 dBm mark.

Can it be an error in the data sheet? I could not find any other/later version of the data sheet.

Thanks in advance.

[h0nk]

Hello,

if You want a cheap and not to weak FM-transmitter, you should
look for a Sony-Ericsson MMR70 module.
If You set the Register 2 with the value 0xC2 they produce around
2 V at the antenna pin. This is approx. 80 mW.
You can buy this modules on Ebay.
If You need further information about the modules, let me know.

Best Regards

[PM3295]

Thanks for the info. I may look at it at a later date. I got about 100pcs of the QN8027 from a friend and would like to implement it in a small (post stamp size) module. According to the data sheet, this chip can also transmit RDS data.

[bigtoy]

Definitely not 121 dBuV into 50 ohms. The datasheet states, in "note 3", that output voltage is presented to a matched antenna (which will have a much higher impedance), and refers you to an application note. I believe this is the application note:

https://cdn.hackaday.io/files/4149177510816/QN8027_SANB_Hardware_Application_Note_V_0_25_102209.pdf

The application note doesn't help too much. Without either the datasheet or app note describing the output impedance of the RF power amp, how are you supposed to calculate power, match to it, etc? I suspect that the -14 dBm you're getting is the truth.

[vern]

Definitely not 121 dBuV into 50 ohms

I disagree, if it can deliver 121 dBuV into a 50 Ohm antenna, it will deliver the same power into a 50 Ohm load if that load is suitable for the frequency.

that output voltage is presented to a matched antenna (which will have a much higher impedance)

50 Ohms is 50 Ohms

[bigtoy]

I disagree, if it can deliver 121 dBuV into a 50 Ohm antenna, it will deliver the same power into a 50 Ohm load if that load is suitable for the frequency.
50 Ohms is 50 Ohms

Perhaps I wasn't clear. Yes of course 50 ohms is 50 ohms - that's not the question. The question is: "when the datasheet says the QN8027 can output 121 dBuV, into what load impedance is it outputting into?" The datasheet doesn't say what the load impedance is, and my read on it is "definitely not 50 ohms". It looks to me like it's probably a far higher load impedance; my guess is in the hundreds of ohms.

[PM3295]

As a follow-up, I made up a test board with an SMA connector directly connected to the output via a 10 nF coupling capacitor. I used the recommended 180 nH inductor to GND. I measured the output impedance at 100 MHz on my HP VNA, and it appears to be about 39 ohm in parallel with a small capacitance (18 pF), that varies with frequency. This is to be expected due to the internal varicap mentioned in the data sheet.

Next I took a CATV transistor (MPSH17) that is still readily available. It provides between 18 and 20 dB gain, with the gain rising with frequency when tested between 88 -108 MHz. I tested with matching several other transistors (MPSH10/20, SH9018, 2N5179), but they all had lower power gain than the MPSH17.

So it appears that the chip's output impedance is supposed to be 50 ohms or near this figure. That makes the 121 dBuV entry even more puzzling.

[bigtoy]

I wonder if the 121 dBuV number is a typo in the datasheet.

[PM3295]

That was my thought as well, but I can't see how they got to 121 if -14 dBm gives about 93 dBuV. Unless they meant to print something like 92 dBuV and somehow an extra digit also slipped in the printing error. I don't think they would print 92.1 because the maximum output level is not that accurate requiring an extra digit of resolution.

[dennis78]

Can someone (I don't have such board so I can't test) test this IC with resistor instead output inductor (10, 50, 100Ohm or...) and look with scope (it's not too high freq). Maybe experimets like this can give move approximate info about some parameters output amp: voltage and current capability, some type of "dynamic" resistance,...

I don't have experience with this IC so I can't imagine what we can expect from these tests.

[PM3295]

We can assume that the output stage is using a type of PNP device with a tuned tank circuit in the collector, made up from the internal varicap and the external inductor. Using a resistor instead of the inductor will obviously not work.

[dennis78]

Maybe not, but why not to try? Of course, internal system can't trim Cd to get maximum Vpp in that case, but I expect sine (or similar, maybe square but less posible beacause of harmonics...) with some amplitude. I expect output is like current source with several mA Ipp + dc bias. On (1 - 500Ohms or...) resistor i can produce about xx mVpp-1Vpp + Vbias?

Varicaps have function to adopt total impedance of output cirucits(include all external components with antenna) to get maximum Vpp, but not for create oscillations. If with total range register for trimming(00- 0x1f, maybe I wrong about values) if you got maximum -> circuit is trimmed best as can, if not find maximum it mean that is RF circuit out of fm frequency range.

[PM3295]

Looking at the data sheet again, there is a note that the TX mode current at FCC part 15 levels is typically 7.7 mA. Now, FCC part 15 level is really low, most probably around -20 dBm from what I can remember measuring on some certified FM devices. They state that the maximum TX mode current to be 9.9 mA. From this it is reasonable to assume that you could go from -20 dBm @ 7.7 mA to -14 dBm @ 9.9 mA. I can't see how you could reach much higher levels with this current limit. The only other explanation is that there is a printing error in the data sheet.

[dennis78]

Very possible. This ic is suitable for USB-SDcard/radio converter for car radio, range about several meters or similar app.

[PM3295]

There is a company selling what they call "Whole-House House Transmitter." If you read some of the comments by angry customers, it is suggested that they should have called it the "Whole Room Transmitter" instead.

[Sideshow Bob]

Very possible. This ic is suitable for USB-SDcard/radio converter for car radio, range about several meters or similar app.

You are quite correct about this. This is also stated in the datasheet i think. Range is very limited https://en.wikipedia.org/wiki/FM_transmitter_(personal_device)

[PM3295]

I wonder if the error occurred during some electronic document translation from maybe a Chinese factory test data into PDF, because the 121 appears in the key features and somewhere else in the data sheet? The one line shows a minimum level of 82 dBuV (-25 dBm) and a maximum of 121, which is obviously bogus. It is surprising that nobody in the quality control department or document review picked it up.

[dennis78]

Datasheet is like other from Asia manufacturers, not very detail.

[Sideshow Bob]

I guess people did not bother. As it worked for its purpose

[bobcat1]

Dennis78 is right ,

I try to avoid working with Asian chips ,as data sheet is never complete or even close, I wish they learn from European and American semiconductor manufactures and produce a more reliable data sheets.
So we can use there chips more.

All the best

Bobi   

[Jerry]

I use these tx/rx boards from china,and they are realy pretty good,& respond to AT comands,they use the same chip you have i think? and if you cut the tracks going to the LCD controler you then have i2c control. and they cost very little.See link.
https://www.ebay.co.uk/itm/PLL-Digital-FM-Radio-Transmitter-Receiver-Module-Wireless-Microphone-87-108MHZ/293141328444?hash=item444095b23c:g:~joAAOSwktFdHc5E

[PM3295]

I have done some more work on his chip and designed an RF amp for it. I used a cheap common on-semi CATV transistor (MPSH17) which is perfect for the FM band, and still readily available.

Below is the design and performance plots from the actual prototype. It has good input/output matching (S11/S22) over the band, and will provide a minimum gain (S21) of 19dB at the high end of the band. Stability is unconditionally-stable with a K-factor of 1.25.  This will boost the chip  output from -15dBm to about +4dBm. If you want more power, you can easily follow this by a mini-circuits  MMIC amplifier block with another 20 dB gain, to give you 100 mW output.

Second harmonic is down at least 45dB in my measurements. I used a 10pF trimmer capacitor to peak the output gain. It measured 3.9pF after adjustment.