Problems with HuanYang VFD in 2.7 and limiting switch in 2.8./2.8.1

I am running into a puzzling situation and hope someone can help.

I had version 2.7 running (Debian Wheezy 7) and the functions I needed worked (including limiting switches and user jog buttons under pyvcd panel). I tried to upgrade spindle control to RS-485 (manual control before), but Linuxcnc could not communicated with the hy_vfd (I tried linuxCNC 2.7, 2.8 and 2.8.1 under Wheezy 7, none of them worked). Under halrun mode, I got "ERROR Communication time out (-12), error reading setup from VFD, retrying" after setp vfd.enable 1 command. I tried all the speeds, no parity and even parity, made no difference.

Without doing anything to the hardware, there was no problem controlling the spindle via the vfd under LinuxCNC 2.8.0 Debian 10 Buster PREEMPT-RT ISO, but LinuxCNC would not read the limiting switches and user pyvcd buttons do not work either. The BOB is powered by a PC power source 12v supply (slightly less than 12V). I read slightly less than 9V when the limiting switch is open. Any insight anyone can provide would be greatly appreciated. I have messed with this for a while. Thank you.

New Information:

I forgot to state two computers were involved. The RS-485 issue was tried on an old computer (Pentium Dual with built-in parrallel port). The failure to connect user button was also likely using this computer (I don't remember, tried so many combinations).

Due to pulse frequency limitation with the old computer, LinunCNC 2.8 was installed on a newer computer, I7-6700 with a PCIe parrallel card. I made some voltage measurements on the BOB last night. It seems the input pin voltage at the optical isolation chip, on the computer side, was 5V with the old computer and only 2.23V with the new computer attached. The 2.23v was barely enough to produce a logical signal in open circuit, but the leaky current from my limiting switch (10k resistance in open state) reduces the voltage at the computer pin so it does not register (voltage to low).

Can anyone tell me if I have a defective parrallel port board? or is this a new standard on the new computer? If I install a 50k resistor to the limiting switch it does register, but I don't like it since the threshold is near the limit, rather than in the middle of the range.

Is it common for 3.3V to drop to 2.3V at the optical chip? The logical high threshold is like 2V and so not a lot of room to work with.

I also thought the voltage is driven by the BOB, according to a sample schematic in this forum. Apparently not with this cheap product. The BOB is connected to both a USB and a 20V external power source.

So I need to solder on a wire to every input pin? There is not a lot space on the BOB and I don't have the confidence I can do it right. Do 5V PCIe parallel boards exist (I searched but they don't specify)? Alternative is to find a different BOB? Thank you.

I measured the voltage right on the optical isolator chip, on the C-E terminals when the limiting switch is in an open state, no power to the led and C-E should not be conductive. Same measurement connected to the old computer || port yielded 5 volts and so I guess the power came from the || card rather than the BOB.

For now I am going back to the old computer, reducing micro stepping from 16 to 8, to keep the speed. I can still return the || card, but would do it if I can buy a 5V version.

If anyone else has the same problem, I have connected 5V from the USB source directly to the optical isolation chips. I have used 2k ohm resistors. Switch closed voltage hovers at around 100 mV and open at 2.4V. This solves the low voltage on the input pins problem.