Xiegu G90 - O defectiune tipica

 Micul transceiver Xiegu G90 este unul foarte indragit de radioamatorii carora le place sa iasa in natura. 

Forma sa compacta, cei 20 de W in toate benzile HF, antena-tunerul incorporat si plaja de tensiuni de alimentare pentru care a fost proiectat (10-17V) il fac un fel de "Swiss Army Knife" al SOTA/POTA/WFF ori pur si simplu, al /P.

Daca luam in considerare si ca este un transceiver SDR pur-sange cu un receptor cu performante exceptionale si pretul ce il face accesibil, intelegem de ce este atat de raspandit in lumea radioamatorilor.

Am fost ceea ce se cheama un "early adopter", unul dintre primii care s-au aruncat pe noua jucarie; imi aduc aminte ca, la acel moment, pe "lista de cumparaturi" erau KX2 si KX3, Xiegu G90 fiind adaugat cumva intr-o doara, cu usoara neincredere...

Am avut anterior ceva experiente cu un transceiver din seria "G" de la Xiegu (108G), in sensul ca am avut la reparat (HI) si nu pot spune ca am fost impresionat. Era ceva intre "Casa Pionierilor" si "Cooperativa Radio Progres".

In ceea ce priveste G90 insa, cele cateva review-uri "de bine" si pretul scazut m-au facut sa ma decid in favoarea lui.

Dupa primirea pachetului, a fost "dragoste la prima vedere" pana cand, la a doua, la aproape doua saptamani de la "intalnire", microcontrolerul panoului frontal a decis sa ma paraseasca ceea ce a dat nastere unui intreg proces de reverse-engineering. 

In pofida experientei mele destul de neplacute, convins de calitatile acestui transceiver, m-am implicat apoi in comunitatea on-line a utilizatorilor de Xiegu G90; am avut astfel ocazia sa intru in contact cu alti posesori sau utilizatori. In general, in astfel de comunitati discutiile sunt orientate in jurul problemelor intampinate iar radioamatorismul are si importanta componenta de laborator si atelier asa ca am incercat aici, pe blog, sa sintetizez tipologiile de probleme cu care fanii Xiegu G90 se confrunta.

Iar una din problemele cel mai des intalnite, este distrugerea etajului de intrare ca urmare a descarcarilor electrice.

Prin eforturi "specifice", am reusit sa intru in posesia unei scheme a producatorului, astfel ca m-am oprit din ridicarea schemei de pe exemplarul pe care il detin. Totusi, ce am primit este schema unei versiuni pe care nu am vazut-o niciodata in "liberate", cel mai probabil este a unui prototip folosit de producator in evaluarea preliminara. 

Cu toate acestea, diferentele nu sunt intr-atat de mari ca sa justifice ignorarea ei.

De regula, descarcarile electrice afecteaza sectiunea intre conectorul de intrare P1 si C33.

Zilele trecute, un amic radioamator m-a rugat sa ma ocup de trx-ul sau ce a trecut printr-un astfel de eveniment, astfel ca pot ilustra cu imagini cum arata efectele unei descarcari electrice asupra transceiverului lasat conectat la un sistem de antene neprotejat.

La prima vedere, elementele de protectie au decis ca sunt depasite de eveniment...

Aparent, cineva a incercat sa intervina pentru remedierea efectelor, fiind vizibile urme de ceea ce pare a fi flux insa, la incercarea de curatare cu alcool izopropilic, componentele, pur si simplu, s-au dezintegrat! De fapt, acela era "fumul magic" condensat pe PCB.

Schema prevede un circuit filtru de banda format de L12, L13 si condensatoarele aferente, ce nu este insa prezent pe echiparea de PCB din versiunea aflata pe masa.

Dupa curatarea PCB-ului si eliminarea componentelor distruse fizic (GDT, Q13, D4, D40, U31), a fost posibila verificarea celorlaltor componente, rezultand ca si R18 si U32 sunt neconforme.

Pentru remediere au fost comandate componentele necesare; D4, D40 - CDSOD323-T03C (original: GBLC03I) , Q13 - DTC114A, GDT- LittleFuse SG75, U31 si U32 - AS179-92LF (original in circuit).

Cu privire la acest circuit de comutare, el este destinat utilizarii de la 20 MHz la 4 GHz; am incercat inlocuirea cu unul (aparent) destinat frecventelor mai scazute, cu aceeasi topologie si compatibil ca amprenta (SKYA21001) insa  mi s-a parut ca aduce in receptor unele componente de intermodulatie! Ciudat si contraintuitiv dar ramane ca fenomenul sa fie explorat suplimentar cu alta ocazie.

Pentru GDT a fost nevoie de refacerea unui "pat" corespunzator sustinerii de curenti mari intrucat acesta este primul element de protectie, ce descarca curenti mari fiind importanta rezistenta circuitului in special pe traseul de GND.

Un pic de rasina cu intarire la UV a fost de folos aici:

Masurarea celor doua condensatoare, C28 si C174 a aratat ca sunt decalibrate, avand in jur de 2 nF!

Circuitul Q13-C174 este un circuit care pune la masa tensiunea reziduala de RF la emisie, blocand trecerea semnalului de nivel mare catre etajul de intrare.

Datorita vaporizarii traseului, am refacut circuitul "Manhattan style" cu imobilizarea mecanicacu ajutorul unei picaturi de epoxy sub condensatorul C174.

Dupa remedierea si refacerea circuitului, la testare a reiesit ca pinul central din conectorul UFL mama instalat pe sectiunea de coaxial care vine de la releul comutator Rx/Tx realizeaza contact imperfect pe conectorul notat pe PCB cu "Rx". Datorita deselor manipulari, in cele din urma a cedat si conectorul de pe PCB astfel ca am ales sa il conectez direct, fara conector UFL.

Am verificat si functionarea filtrelor trece-banda din etajul de receptie si acestea functioneaza corect.

Reparatia a fost un succes si un QSO m-a convins de asta.

Pe langa aceste probleme am descoperit ca mufa RJ45 de la microfon sufera de binecunoscuta problema a ruperii din PCB astfel ca, daca tot era pe aici statia, am zis sa rezolv si problema asta. Din pacate, microscopul abia a sosit de dimineata si am gresit undeva astfel ca nu a capturat imaginile cu acea parte. Nu a fost simplu, reparatia a fost facuta "in situ", fara a scoate PCB-ul din sasiul panoului frontal intrucat mastile butoanelor sunt lipite cu superglue, din fabrica, de axele encoderelor! Nu e prima statie la reparat care are aceasta caracteristica si solutia este taierea mastilor pentru inlocuirea encoderelor. Nu merita aplicata solutia in acest caz asa ca repararea mufei a fost destul de complicata dar a reusit pana la urma!

Termistorul etajului final era, de asemeni problema cunoscuta,  la distanta de tranzistorul final; si aceasta problema a fost rezolvata, preventiv.

VFO Encoder selection for Xiegu G90

From time to time someone start searching for part number of the VFO encoder.

Here is a selection of part numbers for encoders without detent (no clicks) and with detents (like the original one).

A NOTE FROM  PA2HJK, received by e-mail:

Hello Adrian, 

Thank you for all the information about the Xiegu G90.  I used it to replace my VFO encoder but I had to reverse the outer connections  ( of the 3 pins side) of the encoder. The switch worked fine but the rotation was inversed. I have tried several encoders from Bourns and ALPS but they where all reversed.

Maybe you can add this information to your website to help other Hams.
 

73’s  Harm Jan  PA2HJK

 

An email exchange with Paul VE7NRI showed the same problem. 

 A NOTE FROM ME:

When I searched for the proper replacement I assumed that Xiegu G90 use standard phase in it's rotary encoders and focused on dimensions and mechanichal criterias; never imagined that they are not delivering the ssame way. 

Therefore, it is necessary to cross the terminals (small pieces of wire) for the below encoders in order for it to function properly.

 Taken the above into account, you must cross the pins to match the rotation of the new encoder with the result of the VFO or to search for the cheap Chinese-made replacements.

To match the desire of the user, ICOM use a lever to "add" detents to some of their radios. When using in mobile, detents may be convenient while using it stationary, no detents might be the choice.

There are a lot of manufacturers that make cheap PC11 type encoder but the reliability is not very good; while they can be used in home-made projects, in the radio a more sturdy one is needed. 

Also, they might fail and most hams try to find a faster way to repair the radio than to send it back to China and wait a few months.

Therefore, here are a few options for those "brave hams" that will take the things into their hands and will repair the radio by themselves.

As a personal observation, with more than 18 pulses per revolution it is a pain to use a no detents encoder...

From BOURNS:

No detens:

PEC11R-4020K-S0018 with knurled shaft, 20mm lenght, no detents, momentary switch and 18 pulses per revolution.

PEC11R-4025F-S0018 with flatted shaft. From ALPS, EC11E, 20mm long flatted shaft (only) momentary switch.

With  detents:

PEC11R-4120K-S0018 - 20mm long shaft, knurled (4120F for flatted shaft), momentary switch.

PEC11R-4125K-S0018 - 25mm long shaft, knurled (4125F for flatted shaft), momentary switch.

 

From ALPS: 

No detents:

EC11E153440D with 15 pulses per revolution. Flatted shaft. 20mm long, momentary switch.

EC11E18344OC with 18 pulses per revolution. Flatted shaft. 20mm long, momentary switch.

With detents:

EC11E18244AU with 32 detents, 18 pulses per revolution. Flatted shaft. 20mm long, momentary switch.

EC11E15244B2  with 30 detents, 15 pulses per revolution. Flatted shaft. 20mm long, momentary switch.

 

Serrated (knurled). Please check the dimensions of the shaft for proper selection of the knob.

Without detent:

EC11G1574402 - 15 pulses per revolution, momentary switch.

With detent:

EC11G1564411 - 30 detents, 15 pulses per revolution, momentary switch.

For both producers I reccomend study their product selection datasheets which can be downloaded as pdf clicking on their names.

The pinout of the above encoders is compatible with the one in the radio.

I usually like to have longer shafts and cut them to my needs so, if you want a shorter one, just check the datasheets for the part number. 

Both manufacturers are producing compatible encoders for VFO, VOL and FUNCTION. 

 

XIEGU G90 - FW update problem solution (bricked radio)

Very often in the latest months the same problem pops out in social media. 

For other problems I have a dedicated page on my blog but for this one,  I made a distinct one.

Xiegu G90 owners complain about bricking the radio while trying to downgrade it to 1.77 FW revision (or earlier revision). The "COMM LOST" after uploading the firmware is the main indication for this kind of problem. 

While the problem and the solution is simple, the problem is not enough explained by neither XIEGU or their representatives in the market.

So, this is how the things are:

First, the Chinese manufacturer switched from the  original STM32F103 to GD32F103 chip in the Front panel unit (they also made the same thing into the Main Unit MCU).

GD32xxx are perceived as cheap Chinese clones of ST original ones but the reality is a little bit more complicated and the GD MCU "appears to be a faster, more capable version of the STM32F103".

There are significant improvements in the inner architecture that makes GD MCU more attractive for this application (and the price factor also is a plus).

 

 But the choice of the GD chip have some consequences.

One of them is the way that the FW is uploaded...

The upload process is managed by a small program called BOOTLOADER which resides in an area of memory to which the user has no access. It determines the sequence of operations that the chip performs when it is powered. 

In the case of STM, a sequence is automatically executed at power-up which, for a short time, waits for an upload command on the UART port accessible to the user. Once the correct command is received on this port, the BOOTLOADER executes some predefined commands, such as clearing the existing FW from memory and loading the new FW.

In the case of the GD chip, in order to access the FW loading procedure, the user must execute a preset combination of buttons when turning on the radio.
Most likely, this method was chosen to eliminate errors caused by accidentally starting the radio with the data cable connected.

The differences in the internal architecture of the two chips and the modification of the BOOTLOADER required some changes in the FW; since it is proprietary and I don't have access to the source code, it is hard to say how the SW engineers solved these problems concretely.
But it is obvious that the SW versions released after the implementation of the HW modification take into account both versions because ST chip platforms can use all SW versions without problems. 

As for models equipped with GD chip, they are identified by the letter "V" in the serial number to differentiate them from those with ST chip.
In the FW update archives, the manufacturer was considerate enough to include a note about the improvements to the released version but, as we are radio amateurs, we have the bad habit of not reading them!

If we would just read them, we would learn interesting things, such as those about HW version compatibility. 

 

 What that means?

Well, 1.77b is the FW revision where the compatibility issue was solved and this FW is working on both version of HW, GD and STM.

Above this version, all HW are compatible but below this, the FW can run only on ST equipped radios.

The BOOTLOADER is a little bit dumb; it's job is just to open the UART "door" to the user and let the new FW to get into the FLASH memory of the MCU (eventually decrypt it) but cannot check if the FW is compatible with the inner architecture of the chipset.

Therefore, the user will see a "UPLOAD SUCCESSFULL" onto the PC screen but a bricked radio on his desk!

The solution is very simple because, being dumb, the BOOTLOADER will act in the same way when the BOOT FW sequence is applied at the start but it will be a good choice to completely shut down the power supply.

To sum up:

-CHECK THE HARDWARE VERSION of your radio from it's serial number. 

-If the serial number is above X04G211600001 or have a "V", do not use FW version below 1.77.

-If you "bricked" the radio or you have a "COMM LOST" after uploading the FW into the MAIN UNIT, upgrade the FW at least to 1.77b. Make sure you have a compatible Front Panel - Main Unit FW combination.

Don't be afraid to use the official page of XIEGU; it is in Chinese but these days automatic translation makes wonders!

Cheers,

Adrian

Some considerations about voltage readout on radios

 From time to time, in discussion groups some fellow hams start worring about the voltage drop on the radio readout.

This is from a Xiegu G90 group:

> However, the 0.9 voltage difference was still there. I am fairly sure now the

> difference is due either to inaccuracy of G90 volt meter. (I do know it reads

> 0.2 volts low in receiver mode) or there is something internal to the G90

> causing the drop

I think some theory must be exposed to help users to understand what it is about the voltage readings in these radios (and others as well).

The voltage is measured with an ADC (Analog to Digital Converter) which "translates" variable voltages into digital variables.

One important thing to understand is "resolution" which is the lenght of the number that store the analog voltage value.

In our case, the ADC input of STM32F4xx can operate in 6-bit, 8-bit, 10-bit, and 12-bit configurable resolution.

Another important value is the maximum voltage that can be applied to the ADC input, which, in our case is 3V3.

Based on the datasheet of the uC, we can safely assume that the voltage is measured in 8-bit resolution (best resolution without some tricks that involve supplemental processor cycles which are precious because they are time-consumer in a uC which also have to do DSP things), 

Resolution = ( Operating voltage of ADC ) / 2^(number of bits ADC).

Therefore, in our case:

Resolution = 3.3V/2^8 = 3.3/255 =12 mV.

This are the "steps" in which the voltage is measured in 8 bit resolution. 

BUT! There is a big caveat here...

We cannot measure with this resolution the input voltage as is much over the 3.3V after which the input of the ADC will be destroyed!

So we put in line a voltage divider!

The divider will have to accept at least 20V (because the radio accepts input as much as 17V in normal operation.

Let's find out, what is the voltage divider ratio in our radio...

If we look on the schematic,  Xiegu G90, the voltage divider is made with R63B and R67B 3.3KOhm and 470 Ohm respectively, which gave a ratio of 1:8 which means the resolution of the internal voltmeter is 0.096 (roughly 100mV) and the maximum voltage is 26.4 V!

So, any variation in the input voltage of more than 101 mV will be shown as a ... surprise, 200mV or 0.2V!

Simply said, the radio cannot show variations less than 0.2V!

As for the big variations when transmitting, again, from the schematic we can observe that the whole PARF components are tied to +13V.

The voltage tap used to measure the voltage is well beyond some components that will present a certain resistivity:

-power cord;

-fuse receptacle;

-RFI choke with both ground and positive leads;

-two MOSF-FETs used for reverse polarity protection and PowerON.

So, a 0.2-0.5 Ohm is a decent value for all of these and all of the above could explain the voltage swing measured by the internal DMM.

I think this will give a reason to enjoy the radio without worring about that voltage readout!

Cheers, 

Adrian

Xiegu G90 Full schematic

 Finally, the schematic of Xiegu G90 is available!

Here they are, in pdf format.

1 PA RF Board

2 Small Signal Board

3 Display Unit

4 Microphone - Remote DTMF Microphone

The board revision is 2.0.6. I have 2.0.7 revision and there are some minor modifications, especially in the power supply line (reverse protection).

Some other resources.

Xiegu G90 CAT Tuner ideea

 The Xiegu G90 fans desperately need a CAT command to enable the TUNE on the built in ATU.

I think we will not see such an improvement too soon, maybe never.

So, just an ideea for a workaround:

The Rx DATA (input from PC) can be buffered to a uC to listen for CI-V standard command:

FE FE 00 00 1C 01 01 02 FD

When this is received, the uC will issue a command on the TUNE switch input of the Main Unit uC, thus enabling the TUNE.

Let's disassembly and reassembly the Front panel Unit of a Xiegu G90

Karl   DJ5IL, asked me about the access into the Front Panel unit of Xiegu G90.

Instead of writing a long story I decided to play a little with the Camera and here is the result:

Can ICOM IC-7000 microphone be used with Xiegu G90?

 Yes, it can be use!

As a matter of fact, any RJ-45 ICOM microphone can be used with Xiegu G90 transceiver.

Of course, the ICOM remote microphone is using a different serial protocol to communicate with the radio so the function buttons WILL NOT work.

Basic PTT and modulation WILL work. 

I also tried with the microphone from other radio, from ic-703 and IT DOES WORK.

This is the proof, in a small video I made this evening after a fellow ham said on FB his combination is not working.

Here are some comments about the Xiegu G90 RJ45 microphone pinout:

ATTENTION!!!!

The RJ45 is the female connector as seen when you look to the front panel!

MDATA - This is the serial Data from Microphone to Front Panel Unit. Pretty similar to CAT protocol but with some differences. It is not ICOM Compatible but the voltage level is the same.

GND - Ground. Not Shield.

MIC - Audio signal from microphone to radio (HOT signal).

MICE - Microphone ground, the shield of the microphone signal wire.

PTT - PTT Switch, goes to Ground to switch the radio into TX.

MSVSW - This is a pin used on ICOM radios to identify the microphone type. Not used in Xiegu G90.

NC - Not Connected

8V - 8V from the 9V LDO. Actually, around 8.7V on the G90.

Xiegu G90 Faults

 I will try to post here faults from various sources and how the problems were mitigated.

The post will be periodically updated. 

Last update: October 31, 2023

 

 7. Rx Front End and ATT section on 2.0.1 rev. Small Signal Board

While trying to help a fellow ham I annotate the PCB layout according to the known schematic released by the factory for a previous version.

I was convinced that they kept the High pass filter formed with L12/L13, C32/C37/C209/C210/C211 but those are not on the PCB!

 

Here is the PCB component layout, according to the known official schematic:

 

 

6. Bricked radio after trying to downgrade to 1.73 FW revision.

    I made a separate post about this kind of problem because I think there is a real need for some details.

You can read here what cause this upgrade/downgrade problem and how to solve it.

 

 5.  Loss of sensitivity albeit ATT works normal

From Michele, IZ2EAS:

Premise: my G90 has the latest version of the SSIGBD (ver. 2.1.1), and the front-end is totally
different from the previous one (ver. 2.0.7); unfortunately, the schematics that are available on-
line refer to the previous version only, so you’re on your own.

Symptoms: sudden loss of about 15 dB of sensitivity, on all bands and modes; the MDS rises from
about -130dBm to -115 dBm. Apart from this, everything else is fully functional, both in TX and RX.
 

Diagnosis: the RF switch AS179 in the first front-end stage (circled in red in the picture) is faulty;
this switch is used to ground the RX line when the radio is transmitting, to protect the front-end (I
suppose). In my radio the switch was stuck grounding the RX line all the time, hence the loss in
sensitivity. Replacing the component fixed the problem. The reasons for the fault are unknown.

 

 

Photos of the repair phases; the switch is small (2mm x 2mm), you will need a good magnifier and
a steady hand.

Below a schematic of the front-end, as I reconstructed it from visual observation of the board:

73, Michele IZ2EAS

Some sidenotes:

 It looks like the ATT section was moved after the front end Rx filters. Michele pointed me to this but i was a little bit reluctant to this untill I looked on the front end pictures he sent to me:

This might improve the SNR and the sensitivity on the DCC SDR receiver because the RF switches which are active components can add noise from out of band strong signals.

Another thing is that with the cost of a third AS179 the protection of the RF Rx section is improved a little bit.

Overall, it looks Xiegu is watching the customer's feedback and try to improve their products.

Here is a sample from the AS179 RF GaAs RF switch:

Thank you Michele!

4. Failure at Power On /  Intermittent Power On

From Groups.IO

" When I press the G90's power button it turns on, and then turns off right
after. All this started to happen after I turned off the equipment by the
power supply, instead of having turned off the G90 first through its power
button."

Possible solution:

Here is the internal power supply and Power On/Power Off switching transistor.

 
The microphone is fed with 9V from the first LDO from the 13.4V.
The Front Panel is fed with 9V and 3V3 from an inside LDO.

It might be a good starting point to check te second LDO 7809 on the lower PCB inside the Main Unit.
You must feed the first LDO with 13.4V through a 500 Ohm from a lab power supply limitd to 200 mA and check the presence of 9V at the output of the LDO.

Before trying to find a solution for the problems, it's a good practice to try to understand where, how and why the fault is ocurring.

A short version of what is happening when you push Power:

a - A short signal is applied to the POWER SWITCH MOSFET;
b - The Main CPU is "talking" to the Front Panel CPU
c - If Main CPU receive valid comms from Front Panel CPU and check is everything is OK, then go step d.
    c' - If there is no valid comms from Front Panel CPU, "COMM LOST" is shown, go to e'
d - The Main CPU send a signal to the same POWER SWITCH MOSFET to keep in energised.
e - Radio boot, radio ON stays ON when Power button is released

   e' - Radio shut down

The first noise you hear is from 9V being applied to the Audio amplifier.
After that, because the Front Panel has no 9V it cannot communicate with the Main Unit CPU which cannot keep POWER SWITCH ON.

Solution A: Probably will be a good option to check if the silicone rubber is in the right place inside the Front Panel, I remember someone on Facebook had the same problem, solved with this. The single differente to that case is the microphone backlight in your case.

Solution B: Another possible solution is to reflash the FW (a correct pair of FW on both main unit and front panel unit).

If none of the above restore the normal functionality of the radio, further investigation must be done around the power switch transistor.

 

3. Various sources. Lightning discharge near antenna.

This problem is fairly common and affects the receiving front end of the radio.

The fault is located on the Small Signal Board (the PCB that is exposed when the upper panel of the Main Unit is removed) and the fault can be repaired without removing it from the radio as the PCB is one side populated with components.

First, the RF is coming from the antena through a Rx/Tx relay (PA RF Board) to that small connector.

Let's explain what is near the RX connector. Here is the relevant part of the schematic: 

On the left there are the two ESD protection components, D4 and D40. These are not regular diodes but specific protection semiconductors. 

The original components are GBLC08I-LF-T7 but might be very well replaced with PESD2V8R1BSF. In order the protection provided to be effective, the replacement components must be connected in anti-paralell mode across the D4-D40!

On the right side there is Q13 which role is to cut the RF that might came from the PA RF board.

At least D4, D40 and the Q13 must be removed to isolate the problem of No Rx!

If Rx is still not working, we must check the proper working of the Attenuator circuit, which is made with U31 and U32.

Those two circuits are similar to SPDT relays but they are semiconductors. In vast majority of cases I saw the issue, the problem is here also, at leat U31 is burned but U32 is still working.

U31 and U32 are located under the metallic shield.

The shield is not soldered and can be lifted with a sharp tweezer. Pay attention when you are removing it, be gently! 

When you put it back, pay attention to the alignement of retaining clips on the PCB!

Pin 5 on U31 is on the middle of the LEFT side.

Pin 5 on U32 is on the middle of the RIGHT side.

You can test the RF then by putting a 22-100 nF capacitor from the pin 5 of U31 to the pin 5 of the U32 (or before and after the separation capacitors).

Here it is how:

On an another burned radio (assisted repair via e-mail, HI):

On EVERY radio I see affected by strong discharge on the Rx, this solved the problem and re-established the capacity of receiving. Of course, with this, the ATT will not work so proper replacement part must be provided to restore the radio to fully working state.

U31 and U32 Original part: 

PE4259: https://www.psemi.com/products/rf-switches/low-il-rf-switches/pe4259

Compatible:

AS213-92LF

https://uk.farnell.com/skyworks-solutions/as213-92lf/ic-switch-rf-spdt-0-1-3-0ghz/dp/1753766 

Here is a picture sent by one of fellow hams with this kind of problem...

2. From Facebook Xiegu G90 Users, Ronan Cantwell

I finally got a bit of time this evening to sit down and repair my Xiegu G90 radio that developed a very strange fault. I posted about this a couple of weeks ago but here's a link to a video of the misbehaving radio anyway .

Before I start, I'll just mention that there are NO schematics or service data available for this radio as Xiegu seem to think that the world will collapse if they release any useful info like that. There is a very good block diagram (re-posted below for reference) produced by Adrian Florescu when he was looking at his own G90. His radio is a different revision than mine but the main blocks are the same.

The radio is an SDR design with incoming RF converted down to two I/Q signals that have a 90 degree phase difference. These are then digitised and processed by various DSP witchcraft and sorcery. On this radio, and many other SDR designs, the I/Q signals are available on a rear socket so they can be fed into a computer soundcard and processed with software like SDR Sharp or HDSDR etc. When tested, the I/Q signals were present and correct at the rear socket (see photo) which proved the front end of the radio was OK, as was the Quadrature Sampling Detector (QSD). This led me to suspect the CODEC chip that digitises these signals and pipes them to the radios main processor.

I checked for the I/Q signals at the input of the CODEC chip and they appeared to be OK there too.... Excellent. Just order up a replacement CODEC chip and that should be it.... Unfortunately, these turned out to be rare as unicorn poop, with none of the online suppliers having ANY in stock at all. There was one eBay seller that had two of them so I grabbed one for £20. It arrived a couple of days later and I fitted it to the radio and fired it back up..... The fault was still there! What?

So... This evening I sat down with a Guinness West Indies Porter or two and worked through things a bit more methodically than before.

I injected some RF into the radio and traced the I/Q signals from the QSD through to the CODEC chip. The signals were fine at the output of the QSD and also at the following buffer stage which is where they are tapped off for the rear socket. After this each of the two signals is fed to another op-amp stage. These op-amps take the I/Q signals and produce a differential output signal from each one. This is the bit I hadn't noticed before. These differential signals are fed to the CODEC chip. Looking at the outputs of the "I" channel's dual op-amp showed the signal to be present and correct on both outputs. Checking the "Q" channel showed only one output to be present. Thankfully, the two channels are identical so I was able to take a couple of measurements from the good channel and compare them to the faulty one.

Across the + and - inputs of each op-amp half were a resistor and capacitor. On the good channel, I could measure a resistance of 2.2k across the pair. On the dead channel - ZERO OHMS! So, it's either a dud chip or an issue with one of the two components. I removed the op-amp chip as gracefully as I could and measured across the resistor/capacitor pair again. Still zero ohms. So, the fault was either the almost invisible capacitor or the almost invisible resistor. On a hunch, I re-flowed the connections to the capacitor. Not an easy task due to its size. I don't mind working with surface mount stuff but these parts were almost invisible. Anyway, after a bit of swearing I re-checked with the meter and the short had gone. Fantastic!

I re-fitted the surface mount op-amp chip, partially re-assembled the radio and gingerly powered it up...... IT ONLY BLUMMIN WORKED !!!!! No more mirroring on the display and the tuning behaved as it should.

Is this a finished repair? No. The fault may have been a tiny solder bridge under the capacitor that was cleared by re-flowing. It may have been an issue with the part that was resolved by heating. Who knows.... To be certain, I'll replace the capacitor but that will mean trying to figure out its value. That will probably mean removing the same microscopic part from the other channel and measuring its value. I'll maybe get round to doing that at around quarter past never o'clock...... Meanwhile... RADIO.

1. From Facebook Xiegu G90 Users, Kyle Arrowood

I have a Xiegu G90 and the receive just started messing up the other day. I have done everything I can think of to remedy the problem with no luck. I have used an online sdr to listen to my signal and it was clear and clean, but when receiving the other stations sound like robots. I used the same antenna on my Yaesu FT-818 and it works perfectly. This happens on every Band on USB and LSB. Does anybody have any suggestions?

I bought it from Bridgecom and would rather not have to deal with them again.

Xiegu G90

Version

Base: 1.77

APP: 1.76

Antenna

G5rv

I have tried:

Factory Reset

Re flashing the firmware

Tried older firmware

A different antenna

Using the extension cable for the head

Changing the rf gain

Changing the filter settings

Checked Vox is off

Checked RIT was set to 0

Plugging in headphones to make sure it wasn't the speaker

Moving the vfo to make sure I wasn't off frequency

Toggled agc, NB, attenuation and pre amp,mode,band

Vfo A and B, compression

Made sure it was on Mic input

Tuned and checked SWR

Left it unplugged from everything for a day

Swapped coax

Even tried Battery power.

The problem was solved after checking the Reference Clock which was bumped by accident to -2 value!

Xiegu G90 MCU and DB9 reference

## Board revision: V2.0.7 GAI03 

## LCD Display: KD018QQTBN009-RT 

 Link to manufacturer: http://en.tft-tft.com/product/detail?id=19 

### uC pinouts and signals

 

FRONT PANEL UNIT
STM32F103RCT6 - LQFP64
Pin no	Pin name	Description
51	PC10	USART_3_Rx 3.5mm PC_RING
52	PC11	USART_3_Tx 3.5mm PC_TIP
42	PA9	USART_1_Tx to MAIN DB9_2
43	PA10	USART_1_Rx from MAIN DB9_3
MAIN UNIT
STM32F429ZGT6 - LQFP144
Pin no	Pin name	Description
96	PC6	USART_6_Tx to Front Panel DB9_3
97	PC7	USART_6_Rx from Front Panel DB9_2

DB9 connector between MAIN UNIT and FRONT PANEL

Pin	Description
1	9V after POWER ON from PWR Board LM2940-9.0 LDO
2	USART From PANEL to MAIN UNIT
3	USART From MAIN UNIT to PANEL
4	Vcc (13.8V) on stand by /POWER COMMAND/ 3V3 on POWER ON
5	GND

Xiegu G90 power supply section

 Here is the schematic for the power supply section on the PA Board of the Xiegu G90 transceiver.

The board is marked "G90-BASEBD-V2.0.7-GAI01" and contain:

-Power supply section with EMI filter, reverse voltage protection, electronic power switch, linear regulators for 9V (LM2940-9.0) and 5V (LM7805).

-Tx drivers

-Tx PARF

-Dir/Ref RF power bridge

-Band filters

-Antenna Matching Unit (ATU).

Here is the schematic for the power supply section (for high resolution click on image then right click and "save as"):

NOTE: There is an error in the diagram; will be corrected soon. The voltage divider R3/R4 is connected to the main V+ and not to the 9V rail.

Most probably, there is an error in the resistor values, the resistive divider most probably is connected to the ADC input of the main MCU for Battery voltage indicator.  Thanks Jake G1YFF for this pertinent observation!

You can download the schematic as pdf HERE.

Xiegu G90 replacing 5V LDO with SMPS

Owners of Xiegu G90 knows the radio can get pretty hot just on receiving.

Looking inside the radio, on the lower board are two linear regulators.

One is a LM2940T which lower the power supply voltage down to 10 9V followed by a LM7805 which power all the logic circuits in the radio. So, there is a significant current through it.

The LM7805 is one of the worst linear regulators ever...

The thermal protection and short circuit are bad jokes. I never saw one to work!

Also they work hot and often need some sort of current limiter before.
Overall, bad choice from Xiegu.

I had some sort of DC/DC converters in my boxes and one of them in particular work pretty well and, BEING A BRAVE HAM, I thought, why not replace that piece of shit with one I already have?

The D-Sun DC/DC Buck converter is based on MP1584, a compact monolithic high frequency SMPS. In my particular case, the switching frequency is 1.5 MHz which can be conveniently filtered. Also, a major advantage is the overshoot protection which is essential when using it to power sensitive circuits.

The board have common negative ground so everything is OK from electrical perspective.

I test the dimensions and the little board almost fit into the slit of the PCB so I removed the LM7805. The removal was more like an extirpation as I removed it without lifting the lower PCB...

I used a little piece of sandpaper to fit the board, and i scratched a little bit the Ground around the SMPS board to add some capacitors.

After soldering the ground and before connecting the positive wires, I set my lab power supply to 10V and set the output of the SMPS precisely to 5V (actually, 5.002V, HI HI).

One concern was the conducted RFI from the converter so I added some capacitors to filter whatever ripple and noise the regulator is making; 150 nF, 1.5 nF and 150 pF on each side of the board.

Under the board I placed a piece of Kapton tape, just to be sure...

The photos are pretty self explanatory about how the things were done.

After this mod, the radio is much, much cooler; I will replace the LM2940 too, but I didn't find another SMPS. The other one will be place under this one.

Also, a nice result was that the current in Rx mode dropped from abt 650 mA to 520 mA. Probably will go further down with the LM2940 replaced.

I did a quick check in bands and no noise from the SMPS was found.

LATER EDIT.

I got another SMPS and I got into the radio one more time to change the 2940 too.

So, there are now two SMPS, first one is under the PCB and the second one is above the first. Here are the pictures:

The Rx current present now a significat drop.

With ATU:

Without ATU: