3

I've noticed this a lot, including this recent example. \$\text V_{OH}\$ is specified at 3V and 4.5V, for the maximum and a small output current. I assume this is to allow for 10% load regulation on the 3.3/5V, since those are common rail voltages. This doesn't really make sense. Why would the manufacturer assume some amount of load regulation rather than just specifying them at the voltages they're commonly used at? Making sure the load regulation won't effect logic levels is my job. Is there a historic reason for this, or is my assumption entirely wrong?

Matt Young
  • 13,844
  • 5
  • 34
  • 61
  • I thought it was to allow for the case where the output is loaded to the maximum allowable level. – user57037 Dec 09 '15 at 20:07
  • @mkeith That is what is spec'd. It really has nothing to do with why the spec is at two common supply voltages - 10%. – Matt Young Dec 10 '15 at 00:20
  • 1
    I see what you are saying. I think when people look at these numbers, they are just looking for compatibility with the input for the next stage. Low VCC is generally the worst case. So the manufacturers spec it there. Makes sense to me. – user57037 Dec 10 '15 at 02:00

2 Answers2

4

I think your assumption of why 3V and 4.5V were chosen (common supply voltages less 10%) is probably correct.

But your statement that "making sure the load regulation won't effect logic levels is my job" might not resonate with everyone depending on how conservative they are. If the datasheet did specify min Voh only at 3.3V, then how would you know exactly what the min Voh is at Vdd = 3.0 or 3.1V? It would probably be OK to assume the same drop, but how sure are you? You could check on some parts that the Vdd-Voh is the same at 3V, but you can't be sure how well it will hold up over process and temperature variations. You can just use a value of 2.9V, and you are guaranteed that it will be OK.

A more useful spec that I've seen on some parts is Voh relative to Vdd, stratified by Vdd ranges, but it's a bit more work for the manufacturer to characterize and test.

Justin
  • 5,984
  • 22
  • 34
  • Although many characteristics behave monotonically with respect to supply voltage, it seems somewhat sloppy not to either say so or else define operation for voltage ranges that combine include the entire supply-voltage range for the part. On the other hand, many other things are unclear as well. For example, if a part has an absolute maximum of 100mA on an I/O pin, and is specified to source 5mA into a 2.7-volt load with a supply voltage of 3 volts, is there any guarantee that feeding it 3.01 volts and having it drive a hard 2.70-volt clamp won't cause it to source... – supercat Dec 10 '15 at 00:23
  • ...enough current to fry it? Such behavior would be unlikely, of course, but does anything specify that it won't happen? – supercat Dec 10 '15 at 00:23
1

The idea is to make sure you have enough voltage on the output to exceed the threshold of the circuit you are driving under worst-case conditions. Generally, it doesn't matter what the exact voltage is, as long as it's above that threshold.

user28910
  • 3,944
  • 15
  • 19
  • That really doesn't have anything to do with the question. – Matt Young Dec 10 '15 at 00:18
  • This doesn't help explain why the output voltage would be specified under the low supply voltage operating condition? I guess I don't understand the question, then. – user28910 Dec 10 '15 at 16:33