 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Microcontroller and PC projects : AHT10 Humidity and Temperature Sensor
| Author | Message | ||||
phil99 Guru  Joined: 11/02/2018 Location: AustraliaPosts: 2135 |
The AHT10 Humidity and Temperature Sensor is marketed as a more accurate and cheaper substitute for the DHT sensors. It uses I2C for communication so so does not require a firmware driver or CSub to use on MMBasic. Here is some test code for the PicoMite. It can be daisy-chained from the top of an RTC module, though the pads don't align so the leads must cross over. 'Read AHT10 using System I2C Dim Integer dat(6), addr, bytes5, n Dim Float humid, temp 'Find address I2C write &h38, 0, 1, 0 ' write zero to that adress If MM.I2C=0 Then ' check for errors addr = &H38 Else addr = &H39 EndIf Print ;" AHT10 address ";Hex$(addr,2);" Hex" Print "dat()" Print "0 1 2 3 4 5 40 bit word humid temp rel. humidity temp" Do I2C write addr, 0, 3, &b10101100,&b00110011,0 'trigger measurement Pause 75 'wait for data I2C read addr, 1, 6, dat() 'read 6 data bytes Print Hex$(dat(0),2);" "; ' "State" byte = &H1C bytes5 = 0 For n=1 To 5 bytes5 = bytes5 + (dat(6-n)<<8*(n-1)) 'Assemble 40 bit word Print Hex$(dat(n),2);" "; 'print data bytes Next Print ,Hex$(bytes5,10),Hex$(bytes5>>20,5), Hex$(bytes5 And (&Hfffff),5), ' humid = (bytes5>>20) / 2^20 * 100 'extract humidity humid = bytes5 / &H10000000000 * 100 'equivalent result to the line above ' temp = (bytes5 And (2^20 - 1)) / (2^20) * 200 - 50 'extract temperature temp = (bytes5 And &Hfffff)/&H100000*200-50 'equivalent result to the line above temp = Int(temp*10)*0.1 'truncate temp Print humid; "%", temp;" C" Pause 2000 ' read every 2S Loop End And the output Saved 1277 bytes AHT10 address 38 Hex dat() 0 1 2 3 4 5 40 bit word humid temp rel. humidity temp 1C B6 C0 45 02 1F B6C045021F B6C04 5021F 71.38713007% 12.6 C 1C B6 47 55 02 05 B647550205 B6475 50205 71.20259409% 12.5 C 1C B6 4B 05 02 1A B64B05021A B64B0 5021A 71.20822077% 12.6 C 1C B6 58 05 02 00 B658050200 B6580 50200 71.22805719% 12.5 C 1C E6 BB E5 0B F9 E6BBE50BF9 E6BBE 50BF9 90.13045458% 13 C 1C FF FF F5 25 C2 FFFFF525C2 FFFFF 525C2 99.99993531% 14.3 C 1C FF FF F5 3E 12 FFFFF53E12 FFFFF 53E12 99.99993588% 15.5 C The humidity readings differ significantly from a DHT11, breathing on it rapidly takes it to 99.9%. Temp. response is faster too. |
||||
TassyJim Guru Joined: 07/08/2011 Location: AustraliaPosts: 6098 |
It will be interesting to see how well it recovers from prolonged high humidity. I now have a couple on the slow boat from China. Jim VK7JH MMedit MMBasic Help |
||||
| phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2135 |
Have had success reviving a DHT11 in a jar with some silica gel (salvaged from many pill bottles) and sitting it on the wood heater for a few hours. The AHT10 datasheet says 10 hours at 80 - 85 C. Two addresses are available so 2 can use the same I2C. A solder bridge at the top selects the alternate - &H39. Edited 2023-10-04 17:04 by phil99 |
||||
| Amnesie Guru Joined: 30/06/2020 Location: GermanyPosts: 396 |
Thank you @ phil99 for another option of temp & humidity sensors!  Sounds promising!Greetings Daniel |
||||
| lizby Guru Joined: 17/05/2016 Location: United StatesPosts: 3150 |
Thanks for the post. If you know, what are the differences between the AHT10, AHT20, AHT21, and AHT30? PicoMite, Armmite F4, SensorKits, MMBasic Hardware, Games, etc. on fruitoftheshed |
||||
| phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2135 |
I think the others have slightly higher accuracy than the AHT10, but haven't looked into the datasheets very closely. To avoid a delay between initiating a read and outputting the data many sensors do it back to front. When you query them they output data from the previous read then get new data and save it for the next read. The AHT10 has separate initiate and read commands so can work either way. The Sub in this program avoids the 75mS pause by using the reverse method. > LIST 'Read AHT10 using System I2C Dim Integer dat(6), addr = &H38, bytes5, n Dim Float humid, temp Print "rel. humidity temp" I2C write addr, 0, 3, 172, 51, 0 'trigger first measurement Pause 75 'wait for data Sub AHT10 I2C read addr, 1, 6, dat() 'read 6 data bytes bytes5 = 0 For n=1 To 5 bytes5 = bytes5 + (dat(6-n)<<8*(n-1)) 'Assemble 40 bit word Next humid = bytes5 / &H10000000000 * 100 'extract humidity temp = (bytes5 And &Hfffff) / &H100000 * 200 - 50 'extract temperature humid = CInt(humid) 'round humid temp = CInt(temp * 10) * 0.1 'round temp I2C write addr, 0, 3, 172, 51, 0 'trigger measurement for next read End Sub Do AHT10 Print humid;" %",, temp;" C" Pause 2000 ' read every 2S Loop End > Edit. The AHT10 provides 20 bits each for humidity and temperature, which greatly exceeds the accuracy of it. This streamlined version uses 8 bits for humidity and 12 bits for temperature. 'Read AHT10 using System I2C Dim Integer dat(4), addr = &H38 Dim Float humid, temp Print "rel. humidity temp" I2C write addr, 0, 3, 172, 51, 0 'trigger first measurement Pause 75 'wait for data Sub AHT10 I2C read addr, 1, 5, dat() 'read first 5 data bytes - last one not needed humid = Cint(dat(1) * 0.390625) temp = Cint((((dat(3) And 15) << 8) + dat(4)) * 0.48828125 - 500) * 0.1 I2C write addr, 0, 3, 172, 51, 0 'trigger measurement for next read End Sub Do AHT10 Print humid;" %",, temp;" C" Pause 2000 ' read every 2S Loop End > RUN rel. humidity temp 63 % 14 C 63 % 14.1 C 63 % 14 C 63 % 14 C > Edit 2 Was curious to know what the first byte - the "State" byte (= &H1C) was for. The data sheet doesn't say so ran a loop continually reading the device and printing dat(0) and the timer. Between initiating a read and getting new data dat(0) = &H9C then changes to &H1C. This takes about 42mS on my unit so the recommended pause of 75mS is longer than needed. If you want the shortest possible read time keep reading until dat(0) changes then you have new data. Sub AHT10 I2C write AHaddr, 0, 3, 172, 51, 0 'trigger measurement for next read Timer =0:Do :I2C read AHaddr,1,5,dat():Loop Until dat(0)=28:Print Timer,, humid = Cint(dat(1) * 0.390625) temp = Cint((((dat(3) And 15) << 8) + dat(4)) * 0.48828125 - 500) * 0.1 'read first 5 data bytes - last one not needed. End Sub Timer rel. humidity temp 41.548 71 % 13 C 41.655 71 % 13 C 41.694 71 % 13.1 C 41.682 76 % 13.1 C 41.724 77 % 13.1 C 41.562 77 % 13.1 C 41.679 77 % 13.1 C Edited 2023-10-16 14:28 by phil99 |
||||
| phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2135 |
Have been comparing the AHT10 with a DHT11 and a DS3231 RTC chip sensor for a few weeks. To ensure they are all at the same temperature they are all sealed in a jar, close together. According to the datasheets their temperature accuracies are:- AHT10 +/- 0.3C DHT11 +/- 2C DS3231 +/- 3C However in practice the DHT11 and DS3231 are usually within 0.3 C of the AHT10 If more people get similar results it may be worth asking Peter if he can add tenths of degrees to the DEVICE HUMID command output. The data sheet Peter was using indicated only whole degrees were available and with +/- 2C accuracy that seemed reasonable. A different translation of the datasheet showed tenths were there but for the above reason there seemed no reason to change it at the time. The hack below uses the DHT12 command and an extra pin to get the tenths from byte 4 of the output. ' AHT10 DHT11 DS3231 and VL53LOX .bas ' PicoMite V5.08.08b2 'Read AHT10, DS3231, and VL53LOX using System I2C and DHT11 using GPO and GP1 '======================================================================= 'AHT10 Dim Float humid, temp Const AHaddr = &H38 ' AHT10 I2C address I2C write AHaddr, 0, 3, 172, 51, 0 'trigger first AHT10 measurement Pause 45 'wait for data Sub AHT10 I2C write AHaddr, 0, 3, 172, 51, 0 'trigger measurement for next read Pause 45 'wait for data I2C read AHaddr,1,5,dat() 'read first 5 data bytes - last one not needed. humid = Cint(dat(1) * 0.390625) temp = Cint((((dat(3) And 15) << 8) + dat(4)) * 0.48828125 - 500) / 10 End Sub '======================================================================= ' DHT11 SetPin gp1, dout 'provides extended "wake" pulse for DHT11 Pin(GP1) = 1 'via diode to GP0 (anode to GP0, cathode to GP1) Dim n%, a(9) : a(9) = 255 ' = "array not full" flag, BASE 0 Dim Float T, H Sub DHT11 'get DHT11 reading with tenths of degrees Pulse GP1, 25 'provide an extended low "wake" pulse for DHT11 Pause 23.9 'overlap the 1mS DHT22 "wake" pulse Device humid GP0, T, H 'start DHT22 command If T <> 1000 Then ' 1000 = checksum error or no response H = H*10 >> 8 ' convert data format from DHT22 to DHT11 T = (T*10 >> 8) + (T*10 And 255) / 10 ' and add tenths of degrees ' byte 1 = humidity, 2 = not used, 3 = temp., 4 = tenths of degrees, 5 = checksum a(n%) = T ' load array to get 10 sample mean Inc n% : n% = n% Mod 10 ' set array index, after 9 jump back to 0 If a(9) < 100 Then T = Int(Math(sum a())) / 10 'if array is full get mean to 1 DP EndIf Pin(GP1) = 1 ' ensure correct start for next run (optional) End Sub '======================================================================= ' DS3231 RTC temp Function RTCtemp() As FLOAT ' DS3231 RTC temp. registers get updated at 64 S intervals. Local D%, Q% RTC GetReg 17, D% ' Degrees are in register 17 RTC GetReg 18, Q% 'and 0.25C steps in register 18, using top 2 bits. RTCtemp = D% + Q% / 256 End Function '======================================================================= 'VL53LOX laser distance Dim Integer dat(4), dst(1), dist Const VLaddr=&H29 'VL53LOX I2C address I2C write VLaddr,0,2,0,1 'trigger first VL53LOX measurement Pause 75 Sub VL53LOX I2C write VLaddr,1,1,30 I2C read VLaddr,0,2,dst() dist = (dst(0)<<8) + dst(1) - 20 I2C write VLaddr,0,2,0,1 'trigger measurement for next read End Sub '======================================================================= 'Main Loop Print "AHT10 humid AHT10 temp RTC temp DHT11 temp DHT11 humid "' distance" Do AHT10 Pause 500 DHT11 Pause 500 ' VL53LOX Print humid;" %",, temp;" C",, RTCtemp()" C ", T" C ", H" %"',, dist" mm" Pause 1000 ' read every 2S Loop Sub mm.prompt Print Cwd$"> "; End Sub End The output to console. For the DHT11 a rolling mean starts after 10 samples to reduce jitter. Saved 3005 bytes A:/> RUN AHT10 humid AHT10 temp RTC temp DHT11 temp DHT11 humid 39 % 19.9 C 19.5 C 20.3 C 48 % 39 % 19.9 C 19.5 C 19.8 C 42 % 39 % 19.9 C 19.5 C 19.7 C 48 % 39 % 19.9 C 19.5 C 19.9 C 48 % 39 % 19.9 C 19.5 C 20 C 48 % 39 % 19.9 C 19.5 C 19.7 C 48 % 39 % 19.9 C 19.5 C 20.1 C 48 % 39 % 19.9 C 19.5 C 19.9 C 48 % 39 % 19.9 C 19.5 C 20 C 48 % 39 % 19.9 C 19.5 C 19.9 C 48 % 39 % 19.9 C 19.5 C 19.8 C 48 % 39 % 19.9 C 19.5 C 19.8 C 48 % 39 % 19.9 C 19.5 C 19.8 C 48 % 39 % 19.9 C 19.5 C 19.8 C 48 % 39 % 19.9 C 19.5 C 19.8 C 48 % 39 % 19.9 C 19.5 C 19.8 C 48 % 39 % 19.9 C 19.5 C 19.8 C 48 % 39 % 19.9 C 19.5 C 19.8 C 48 % 39 % 19.9 C 19.5 C 19.8 C 48 % A:/> Edit Ignore the VL53LOX bit, it was already attached to the RTC so its code was left in. Edit 2 The discrepancy between the AHT10 and DHT11 humidity readings has gradually decreased in the time they have been sealed in the jar. I put a small silica gel capsule in with them so I could compare them at lower humidity. The difference was 18% when they went in a few weeks ago and this afternoon is down to 6%. . Edited 2023-12-05 16:55 by phil99 |
||||
| Volhout Guru Joined: 05/03/2018 Location: NetherlandsPosts: 4234 |
For continued high humidity use the "wet sock" method on a DS18B20 sensor. That works quite good. Volhout PicomiteVGA PETSCII ROBOTS |
||||
| phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2135 |
There are a few new accurate, affordable T & H modules now available. One that got my attention is the TI HDC1080 I2C. At this stage I don't have a use for another so won't get one. After lengthy comparison with the AHT10 it seems my effort to revive the DHT11 have only been partially successful, the humidity is off by 10% to 20%, varying with temperature. |
||||
| Volhout Guru Joined: 05/03/2018 Location: NetherlandsPosts: 4234 |
I have used the TI HDC2080, and it is quite accurate, but also not handling high humidity long time very well. Hence the wet sock. You can “calibrate” the 2080 Volhout PicomiteVGA PETSCII ROBOTS |
||||
 Print this page |
