But the connector FX23-100S-0.5 [1] has only 104 usable pins… (100 Signal + 4 power pins)
so where are the ground pins? Maybe the holes for the mounting pins?
Why not
101 GND
102 VCC3.3
104 GND
105 V5
?
Also, the pinout of the ReTerminal does not seem to match the pinout of the TecSpec from Hirose.
This leaves questions open… Could someone check this and if necessary add a picture that clearly shows how the pins have to be counted?
This schematic looked careless in places, and I had similar questions.
But fortunately I lived in the country where the connector manufacturer is located, which was helpful.
The GND of the FX23 connector is only connected to a few small contacts.
Schematic pins 105-108 seem to be intended for FX23 pins without contacts.
Also, pins number are inverted between schematic and FX23…
That is, 101- (schematics pin No.) = (FX23 pin no.)
I just created a PCB to break out PCIe signals and finished testing so there may be no mistake.
Yes, it’s a shame because the case and concept is really great…
The layout of the connector also looks quite reasonable on paper. But something clearly went wrong with the layout of the PCB. Because that the PINs are not correct, the designer should have noticed. GPIOs between the differential lines of the PCIe? That’s not good quality that you can sell with a clear conscience.
The power button also doesn’t work as it should.
Seeed would have to revise the PCB and send the customer a new one.
That’s not good quality that you can sell with a clear conscience.