"Read fail" from Gotek

[dr4g0nsr]

Hey guys,
I have a problem with CPC 6128 and Gotek drive which suddenly stop working.
Since Gotek is connected to FDD port as "B" drive, when using |b I see that head is moving (on gotek LCD) but it ends with "read fail".
Now, I was looking schematics and I see that there's RDAT signal from floppy which goes to 74HC240 and then SED9420C and back to UPD765AC which is FDC.
When I measure it I see some strange spikes and not much other than that and I notice that pin 2 on 74HC240 is shorted to 5V (pin 20 on 74HC240  and pin 1 on resistor array just below it), even when computer is off. When it's on it's still 5V on both pins.
While there's short I tried to measure output of 74HC240 which is connected to SED9420C and resistance is like 5V.
I also tried with 3.5" floppy made to work with cpc, I hear head movement, identical to what gotek is showing but same "read fail".
Should I replace  74HC240 or there's something else that could be a problem here?
Can I test something more?

[abalore]

A silly question: did you try another cable?

[dr4g0nsr]

Don't have another cable or any other component except 3.5" drive which I am not sure it has floppy correctly formatted.
Cable is new and it was working before so pinout should be correct.

[bart71]

Is original 3" drive working OK?

[dr4g0nsr]

Original 3" drive is broken for other reasons so I can't test it.
I will probably remove 74HC240 and test it.

[dr4g0nsr]

Problem is solved by replacing chip with 74HC240B1, original was 74HC240N.
For future references that came here looking for solution:
Few things to note here - without chip computer works but gives "Isp" instead of brand.
Do not worry about this, it's mechanism to detect certain brands based on detected hardware instead of four contact (LK1, LK2, LK3, LK4). Dump from ROM is this: Arnold Amstrad Orion Schneider Awa Solavox Saisho Triumph Isp - so this is the list of all supported brands. Have nothing to do with ROM, don't change it, it's most likely fine.
Next, this octal buffer/driver is very simple device, but logic from FDC will behave differently based on the state of this chip. If you remove whole chip you get kinda delayed initialization, not hang or anything. If whole FDC is removed you will see no change until you run command, then you will see red screen.
My exact problem was really a short in this chip of 1A1 and 5V VCC. Good thing that levels are the same everywhere so other electronic didn't die because different voltage but instead input signal from RDAT (read data) is overriden by short instead going down and up representing digital levels. Using osciloscope you will read "a try" to low the signal but it will happen for a very short period of time, not enough for digital signal there.
I came to conclusion like this tracing RDAT line from floppy. At first I tought it's FDC pin where this goes, but it's not, it's this buffer first and then SED9420CAC (CMOS data separator) and then to FDC. Good thing that I followed schematics and not trying to trace hard way.
Looking at logic of this chip most likely it will die with short with 5V, other possible case is floating line. So my advice is to test short from 5V to other pins. No other pins is shorted to 5V (when you see datasheet for buffer you will see it cannot work that way).
When soldering chip, first solder 20 pin socket in case it dies again and chance of dying is quite high because it's first in the line from floppy drive and a lot of things can happen during that transfer, from bad/dead floppy to interference (my centronics floppy cable was 1 meter). Take care of resistor array near it, do not push it hard.
When you put replacement in socket first measure if short is gone, then turn it on without floppy and measure signals again, 5v and input/output, and after than you can connect floppy and try to read something from there.
Hope this helps someone.