It looks like I'm good on the input range, now to figure out the output range. The 358 is "designed to operate from a single power supply and it's common mode input range includes ground". It does not say rail-to-rail. I was thinking this may be a good fit for the circuit where I'm really only using the lower three-quarters of the voltage range.
It looks like I'm good on the input range, now to figure out the output range. The 358 is "designed to operate from a single power supply and it's common mode input range includes ground". It does not say rail-to-rail. I was thinking this may be a good fit for the circuit where I'm really only using the lower three-quarters of the voltage range.
hi,
If you amplify the 820mV to 900mV range to say to 1.6V , 3.75V , you will need two comparator circuits in order to detect the very wide threshold differences.
I think the purpose or function has been somewhat lost. I need to detect ice, or ice being present. When the sensor is in free air, I expect it to be near room temperature. As the ice forms and builds it will eventually contact the LM34 and its reading will drop to near freezing 32-35 degF. At which point the dehumidifier will turn off. When the temperature rises, to around 40 deg F it will turn on again. I expect the hystersis will be, worst case, 1.6V (32*10*5) and 2.25V (45*10*5). The low end may be higher and/or the high end may be lower.
Technically I won't really care about the actual temperature about 45 degF but room temperature is around 72 degF and would would like to keep the circuit from saturation or the rails.
Any thoughts on the selection of the LM358 op amp?
Did you see my prely earlier? Rather than detect ice, you could just ensure ice does not occur. By putting the sensor on the coil, and keep ing the coil just above freezing point you will still get the maximum possible cooling effect from the same hardware but ice will never form.
Actually the overall performance should be higher as ice is an insulator, and a system that never ices should get a better overall heat transfer from coil to the air.
I think the purpose or function has been somewhat lost. I need to detect ice, or ice being present. When the sensor is in free air, I expect it to be near room temperature.
Any thoughts on the selection of the LM358 op amp?
The purpose/function has not been lost, you dont want to detect Ice, you want to know when the LM34 output is at or below 820mV and the output is at or above 900mV.
I know what this means, but why do you think it wouldn't be.?? When the sensor is in free air, I expect it to be near room temperature.
Any thoughts on the selection of the LM358 op amp?
As explained the LM358 output will not go higher than Vsupply-1.5V, so for a 5V supply thats a maximum of +3.5V.
The dual supply idea will not make any difference to this +3.5V level.
Did you see my prely earlier? Rather than detect ice, you could just ensure ice does not occur. By putting the sensor on the coil, and keep ing the coil just above freezing point you will still get the maximum possible cooling effect from the same hardware but ice will never form.
Actually the overall performance should be higher as ice is an insulator, and a system that never ices should get a better overall heat transfer from coil to the air.
I don't want to cycle the dehumidifier that often. It takes only minutes to tens of minutes for the coil to reach freezing. Any temperature below the dew point is effective at dehumdification. But too much ice build up blocks off and restricts the air flow. So some ice build up is OK. Too much is a problem.
The purpose/function has not been lost, you dont want to detect Ice, you want to know when the LM34 output is at or below 820mV and the output is at or above 900mV.
820 and 900mV? 82 an 90 degF? I see the direction you're headed. I think your layout may work as well. As the area is not climate controlled and there will be some variation in room temperature. I'm thinking of merging the two and setting the thresholds in the 50-55 degF range. I don't know how much the LM34 will need to touch the ice before it starts to register a decrease in temperature.
In being conservative, I was thinking it may not beexactly room temperature but slightly below as a) it is in front of a fan and b) near an "ice cube".
Any thoughts on the selection of the LM358 op amp?
As explained the LM358 output will not go higher than Vsupply-1.5V, so for a 5V supply thats a maximum of +3.5V.
The dual supply idea will not make any difference to this +3.5V level.
I had not seen the output was limited to Vs-1.5. I have not been able to determine from the datasheet what the maximum output voltage is with V+=5V. It appears to me the input is limited to 0 to Vs-1.5.
I started to draw up the schematic and in selecting a relay to control the dehumidifier found they don't have inexpensive 5V coil relays that can handle this much load. At 12V, I've found a few that can do the job. All the chips can handle V+=12V, so I'm going to bump up the juice, this should also give me the head room on the op amp output. Yipee, a two-for-one. My plan is to use the relay N/C to run, and open when freezing is detected. This should keep power dissapation (ie., my electric bill) down as the dehumidifier will run more than it doesn't.
hi,
Look at this LM34 clip.
The Tempr range starts at -50F and increases by 10mV/F, so at +32F thats 82F degree change * 10mV =820mV and at 40F that 90F change *10mV = 900mV.
The LM358 will NOT output higher than Vs-1.5V, look at this clip for 30Vs
That changes things a bit. I (obviously) misread the LM34D d/s. Instead of trying to detect 32 degF at 320mV and 72 degF at 720mV, I'll be looking at 0V and 400mV. So I'm definately going to want that 5x gain.
Can you help me understand how the d/s output spec relates to no outputing higher than Vs-1.5? The clip you provided looks like Vs-4.0 worst case and Vs-2.0 typical. Am I correct in understand that the input cannot exceed Vs-1.5.
That changes things a bit. I (obviously) misread the LM34D d/s. Instead of trying to detect 32 degF at 320mV and 72 degF at 720mV, I'll be looking at 0V and 400mV. So I'm definately going to want that 5x gain.
Can you help me understand how the d/s output spec relates to no outputing higher than Vs-1.5? The clip you provided looks like Vs-4.0 worst case and Vs-2.0 typical. Am I correct in understand that the input cannot exceed Vs-1.5.
I'm looking at the D version, 32 to 212 degF. I guess the comment "Definately going to want the 5x gain" was being over zealous. More appropriately I should have said "I can definately use the 5x gain". The sensor sensitivity hasn't changed, so the need for the additional gain has changed. I was just noticing the increased headroom you provided by correcting my design error.
I'm looking at the D version, 32 to 212 degF. I guess the comment "Definately going to want the 5x gain" was being over zealous. More appropriately I should have said "I can definately use the 5x gain".
EDIT:
Just realised the +3F minimum will NOT be a problem for Fig #1 circuit, I am so used to thinking Centigrade units.
Perhaps too much Brandy on my Xmas pudding!
The LM34D should give320mV at 32F and 400mV at 40F
I see that now. Single supply will not get below 3-5 DegF output. Bummer. I'd really like to avoid dual supplies. I'll have to think on it a while, but the 3-5 degF limitation may not be a problem (but it would be better if it did read to 32F instead of 35F). If I set the thresholds for 50-55 degF not reading to 32 won't be a problem.
I see that now. Single supply will not get below 3-5 DegF output. Bummer. I'd really like to avoid dual supplies. I'll have to think on it a while, but the 3-5 degF limitation may not be a problem (but it would be better if it did read to 32F instead of 35F). If I set the thresholds for 50-55 degF not reading to 32 won't be a problem.
The output will be 40degF=400mV instead of 80mV ([40-32] *10) as we thought earlier.
Based on your d/s clip fo the LM34, the D may not be accurate below 3-5degF? Or it won't read below 3-5degF?
In the LM358 app note, I saw the CM Vi parameter first. I now see the Vo parameter further down. Vo max =Vs-1.5. With Vs=12V that shouldn't be a problem.
The output will be 40degF=400mV instead of 80mV ([40-32] *10) as we thought earlier.
Based on your d/s clip fo the LM34, the D may not be accurate below 3-5degF? Or it won't read below 3-5degF?
In the LM358 app note, I saw the CM Vi parameter first. I now see the Vo parameter further down. Vo max =Vs-1.5. With Vs=12V that shouldn't be a problem.
The 3F minimum on the LM34D means an absolute minimum V of~ +30mV, which in Centigrade is -16Cdeg. I would say your application will never get that cold, I hope.!!
Can you update me on how you plan to use the 'amplified' of say 5Vmax.??
The 3F minimum on the LM34D means an absolute minimum V of~ +30mV, which in Centigrade is -16Cdeg. I would say your application will never get that cold, I hope.!!
Oops. I think I just got my Cs and Fs screwed up as well. I don't think it's going to get to 29 degrees below freezing. Interesting that the part is listed as 32F to 212F but has a minimum accurate output of 3F. I'll take it.
Depending on how close in degrees I want to set the comparator thresholds/hysteresis, I may want more voltage swing. If I set it for 35F/40F that's only 50mV, but 5x would be 250mV. I'm thinking 250mV is easier to set with just a pair of voltmeters. With a 12V supply I might even be able to do 10x gain [75degF*10mV/degF*10=7.5V which is less than 12(Vs)-1.5=11.5V]
I don't want to cycle the dehumidifier that often. It takes only minutes to tens of minutes for the coil to reach freezing. Any temperature below the dew point is effective at dehumdification. But too much ice build up blocks off and restricts the air flow. So some ice build up is OK. Too much is a problem.