Data transfer error-- likely cause?

[tlewick1]

The project involves reading data from a Mitutoyo Digital Indicator to an Arduino Uno. I've found a few people posting about projects similar to mine. This example is comparable. A brief of what's going on:
The Arduino pulls a data req. pin high; the indicator (a height gage with linear encoders) responds with a numerical value. However, occasionally the value received is way off.
I think I know why, but it made me curious about the code which I used from aforementioned examples:

digitalWrite(req, HIGH); // generate set request
for( i = 0; i < 13; i++ ) {
  k = 0;
  for (j = 0; j < 4; j++) {
    while( digitalRead(clk) == LOW) {
      } // hold until clock is high
    while( digitalRead(clk) == HIGH) {
      } // hold until clock is low
    bitWrite(k, j, (digitalRead(dat) & 0x1));
  }
  mydata[i] = k;
}
sign = mydata[4];
value_str = String(mydata[5]) + String(mydata[6]) + String(mydata[7]) + String(mydata[8] + String(mydata[9] + String(mydata[10]))) ;
decimal = mydata[11];
units = mydata[12];

value_int = value_str.toInt();
if (decimal == 0) dpp = 1.0;
if (decimal == 1) dpp = 10.0;
if (decimal == 2) dpp = 100.0;
if (decimal == 3) dpp = 1000.0;
if (decimal == 4) dpp = 10000.0;
if (decimal == 5) dpp = 100000.0;

value = value_int / dpp;
digitalWrite(req,LOW);

This picture might be helpful:

Shouldn't there be an interrupt set up on the clock pin? It looks to me like-- the clock pin triggers when the data starts, but what is keeping the two devices in sync? Is it just coincidence that the two devices are at a similar enough frequency that the 9 bytes of data are polled at the right times? Could this be why I occasionally get bad data? Or is the compiler smart enough to know what I want?

I can later share my thoughts why I think I get bad data, which may be unrelated to the questions posed above, but does anyone have any comments?

[wildbill]

It looks to me as though the clock is generated by the tool and that tells you when you can read data. No need for an interrupt, although you could do it that way if you wish.

That code is odd though, from the pulse diagram, it looks like you just need to read data when there is a falling edge on the clock but waiting for low, then high, then low, it appears that the first bit will be missed.

[tlewick1]

Thank you for you’re response. It made me realize I was misunderstanding the code. I was thinking that “for j” loop was waiting for the d1-d4 to finish— now I realize that the “for j” loop reads the first four bits of each d1-d13 bytes, because it only needs 4 bits of information.