Printable View
What's the formula for converting a (saturated) temperature reading into a pressure reading for R290? The application is to allow an embedded control computer to read the high and low side pressures with just temperature sensors. If possible, it should be implemented in 16 bit fixed point arithmetic. (I'm using a dsPIC processor.)
Hi, NiHaoMike :)
Unfortunately, (as far as I know;)) that is not possible ... due to suction and discharge superheat ... temperature readings do not correspond to actual saturated temperature i.e. pressure ... without pressure sensors (giving pressure and saturated temperature for that pressure) it is not possible to calculate superheat both on suction and discharge side ...Quote:
Originally Posted by NiHaoMike
Furthermore, capacity regulation via temperature is too slow and inaccurate ....
Best regards, Josip :)
The low side is actually very easy - just attach a temperature sensor just after the expansion valve. The high side is going to need some experimentation, but the middle of the (water cooled) condenser should give pretty good results.
Hi, NiHaoMike :)
Sorry, but I do not follow above explanation :confused: .... for me it is not that easy, but please, try and give us some feedback ... always ready to learn something new ...Quote:
Originally Posted by NiHaoMike
When I was a student I've been told that we always need two variables within equation to calculate the third .... but that was in the last millennium , with nowadays education who knows...
Best regards, Josip :)
If all you had was a thermometer and a PT chart, you can very easily find the low side pressure by measuring the temperature just after the expansion valve. You can also get a good idea of the high side pressure by measuring the condenser temperature at various points and using the reading that seems constant over a large area of the condenser.
The problem is that a PT chart is made for manual use. I need a PT formula that an embedded computer can use to convert temperature reads to pressure.
Hi, NiHaoMike :)
Thanks for explanation ...Quote:
Originally Posted by NiHaoMike
anyhow, please, check this
http://www.refrigeration-engineer.co...essure-Formula
http://www.refrigeration-engineer.co...rature-formula
Hope this above will be of some help to you ...
Best regards, Josip :)
I worked on these formulas at a university research facility about 25 years ago. IIRC, the basic equation was a polynomial curve fit of 15 or 20 terms with built-in triggers to detect each end of the saturation bubble. Back then it was proprietary stuff. I'm guessing it is public domain now. The programs available today are based on that work.Quote:
Originally Posted by NiHaoMike
Unfortunately, I was a lowly keyboard serf at the time and not privy to the inner workings. In any case, the formulae of that era would not be suitable for a PLC.
If you only need data points at 0.5C or 1C intervals you might be better off using a lookup table. They generally adapt well to integer math solutions. Polynomials are a bugger if you are doing stack manipulations.
Is that so. As far as I know about this or i though about this this was possible but by reading yor replies I am surprised right now.
I'm running it on an 80MHz dsPIC, which is a pretty capable processor. It only needs to translate 2 readings every second or so. (And my first PC was only 25 MHz!) I expect most of the processor resources to be used on the "Tiffany Yep" DSP magic that is used to decode the signal from wireless temperature sensors (for sensing hot water tank temperatures), but I expect lots of left over processor time.
Ahh, completely different animal. Curve fit equations in a PLC are a nightmare. It is much easier in a PIC, assuming you aren't programming in assembler.Quote:
Originally Posted by NiHaoMike
I thought about this over the weekend and still think a lookup table is the best solution. The project I worked on was trying to develop data for laboratory standards, overkill for industrial applications. I believe all the trade software available today, as well as smart gauges, use simple algorithms to interpolate between points on a table. It is fast, compact, easy to program, and well within the requirements of HVAC.
You could probably do what you want with less than 20 lines of code. The formula is basically the same thing you would use when doing hand calculations from a table in the back of a thermodynamics textbook.
Here is a link to the equation:
https://www.eng.fsu.edu/~dommelen/co...pol/index.html