Seeking: SI derived units with higher base powers.

  • Thread starter Thread starter Knaapje
  • Start date Start date
  • Tags Tags
    Base Si Units
AI Thread Summary
The discussion revolves around the use of SI derived units in a chemistry protocol, specifically seeking units with powers exceeding +7 or below -8. The original poster is limited to a range of -3 to +4 and expresses frustration in finding suitable units. Several participants recommend against using only 4 bits for data descriptors, emphasizing that the potential issues from limited space outweigh any minor bandwidth savings. They suggest that using more bits or human-readable formats, like ASCII for unit names, would simplify communication and debugging. Overall, the consensus is that prioritizing clarity and flexibility in data transmission is more beneficial than strict adherence to bit limitations.
Knaapje
Messages
8
Reaction score
0
Hey guys,

I'm currently working on a chemistry setup and we're building some protocols. We decided that every communication on the protocol between two devices will be in SI units. However, for a data descriptor I plan to use 4 bits per standard SI unit. This means that units for variables can go from -8 to 7.

Now, I'm wondering if there is any measurable SI derived unit which has a higher power than 7 or lower than -8?. I can only get to -3 or +4, but that's it. Anyone suggestions?

For instance, if you have pressure in hPa, this is kg * s^-2 * m^-1, the lowest is -2, highest is 1. Farad is kg^-1 m^-1 A^2 S^4, highest 4.
 
Last edited:
Physics news on Phys.org
Welcome to PF, Knaapje! :smile:

My recommendation: don't use just 4 bits!
It's just not worth all the problems you'll get.
 
I like Serena said:
Welcome to PF, Knaapje! :smile:

My recommendation: don't use just 4 bits!
It's just not worth all the problems you'll get.

In total I'll be using 7 * 4 = 28, or 32 bits because alignment. I'm just curious if there is anything which we can actually measure has more than 7 or less than -8 powers.. Went through Wikipedia as well and just get stuck on 4. Also, it has to be as little as possible to keep everything at reasonable speed. Sending a message that's twice as long just to be sure seems a bit silly.
 
Knaapje said:
In total I'll be using 7 * 4 = 28, or 32 bits because alignment. I'm just curious if there is anything which we can actually measure has more than 7 or less than -8 powers.. Went through Wikipedia as well and just get stuck on 4. Also, it has to be as little as possible to keep everything at reasonable speed. Sending a message that's twice as long just to be sure seems a bit silly.

As I said, it's likely not worth the trouble.
Sending a message that is twice or 4 times as long will (almost) never be a problem.
Handling problems because you have too little space to transfer what you want to transfer really is a problem.
 
Bits are cheap as chips, as they say. Neither processor power nor channel bandwidth are likely to be a serious problem for you unless you are communicating with a submarine of a deep space probe.
 
If you really want to save bandwidth, why not just use the ASCII text for the standard abbreviations of unit names?

"F" for farad is only 8 bits, not 28 :smile:

And when things go wrong, human-readable messages make debugging easierr.
 
Dude! This is the 21st century! Nobody quibbles over bits. That's what they did in like the stone age.
 
Back
Top