Support for Alientek T90 Series Irons

[bvngee]

Hey All,

I acknowledge that this is a pretty low-effort issue that offers no technical support and only a request. Nonetheless, Alientek recently came out with a new USB-C iron which is arguably one of the best out there now. It's rugged, comfortable, supports up to 140W PD, uses C245/210 JBC tips, and has a nice big color LCD display.

Just leaving this here in case anyone is interested in porting the OS to this new Iron!

Of noteworthiness is the new color display, which may be a first for this OS, not sure.

[joric]

Alientek T90 Disassembly video: https://youtu.be/Xl0IKpvlF3Y

• MCU: Nation N32L403 (KBQ7 HHF 15006), 32-bit Cortex-M4 (64 Mhz, FPU, DSP, 24K SRAM, 256K Flash)
• Motion Sensor: QMA 6100P 3-axis Accelerometer (9443 RP2)
• Display: 0.99" IPS LCD 160x60 pixels, GC9D01, SPI, 8-pin (N099-0416THBIG13-C08 W25083003)
• Buzzer: MLT-7525 Passive Buzzer (Externally Driven) 2.5V~4.5V 2.7kHz 80dB SMD, 7.5x7.5mm
• CH224Q USB PD Driver - PD3, EPR, AVS, SPR, BC1.2 (224Q F16)
• RY8411 Sync Step-Down Regulator Standoff 42V 1.2A 800 KHz (HOIAW)
• SPX3819 500 mA Low-Noise LDO Voltage Regulator (L309)
• RS92XM Operational Amplifier (RS92 MS51A114)
• TL431 Precision Reference Shunt Voltage Regulator (431)
• AET4102AP 40V P-Channel EpicMOS (tm) in PDFN3030 package (427F1ND)
• Aishi PH-series 35V 100 uF Solid Capacitor (100 35V PH), 6.3x8mm, 2 pcs

[PanKleszcz]

It looks very similar to Fnirsi HS02 (and uses same MCU - it's actually N32L403). After I'm done with HS02 this looks like a low-hanging fruit.
The only issue here is Alientek is known to implement all kind of obfuscations in their update files/bootloaders.

[andrewthecat]

My friend and I made a schematic and an example of decoding and flashing a custom program with a display driver: https://4pda.to/forum/index.php?showtopic=1061134&view=findpost&p=141786542 (and the next post)

[PanKleszcz]

My friend and I made a schematic and an example of decoding and flashing a custom program with a display driver: https://4pda.to/forum/index.php?showtopic=1061134&view=findpost&p=141786542 (and the next post)

Thanks, I've actually stumbled across your schematic already few days ago, it greatly helped understanding parts of the device (especially the ADC inputs).

Currently I want to finish HS02 before I go completely mad (last few days I'm struggling with the link-time-optimizations messing SPI HW).

I've run few experiments on T90 as well, I must admit it's much more mature HW than Fnirsi.

• It has CH224Q instead of the dumb CH224K,
• Vref is actually connected to a reference voltage, not just 3V3 rail,
• power path is similar to HS02, but large capacitors make significant difference

Interestingly it appears MCU is able to perform QC negotiations, @andrewthecat would it be possible for you to measure this four resistors, I want to confirm these are dividers used to supply 0.6V on the USB data lines:

The biggest pain so far is the GC9D01 display controller, but we'll get there :)

[ma-ale]

Has anyone measured how fast it switches the power mosfet to make use of those capacitors on the output? It seems to me like most chargers should complain with the 11A spikes.

[PanKleszcz]

I've ported msd_mem to T90 (programming N32 feels like first STM32 some 15 years ago, where partially documented silicone bugs are least of pains...)

Now I understand why Fnirsi decided to go for SW SPI implementation.

Anyway, if somebody wants to backup their T90, here's the .atk file: t90_msdmem_0.0.5.zip

[PanKleszcz]

Has anyone measured how fast it switches the power mosfet to make use of those capacitors on the output? It seems to me like most chargers should complain with the 11A spikes.

From what I've seen it should be 2.7KHz, but I had enough soldering iron disassembly after the HS02 so this info should be verified by somebody that has actually connected an oscilloscope to the iron's tip.

[andrewthecat]

Interestingly it appears MCU is able to perform QC negotiations, @andrewthecat would it be possible for you to measure this four resistors, I want to confirm these are dividers used to supply 0.6V on the USB data lines:

@PanKleszcz

I have revision 1.1, so the resistors are located differently. I measured them without desoldering them, so they might not be accurate. I can check later to see which resistors are connected where.

[vshuraeff]

I successfully launched IronOS on the T90B. Everything works, but some details require clarification. PR soon