The only problem I can see is that you are assuming the other bits in the CCP1CON are always set correctly? For the sacrifice of a few extra instructions I would prefer to have ALL register bits set as in: CCP1CON=$0C|((pwm_pars&$3)<<4)
|
Re: TEC controller
The only problem I can see is that you are assuming the other bits in the CCP1CON are always set correctly? For the sacrifice of a few extra instructions I would prefer to have ALL register bits set as in: CCP1CON=$0C|((pwm_pars&$3)<<4)
Dave Purola,
N8NTA
EN82fn
|
Re: TEC controller
Hi Dave,
You are making a good point. It comes with lock of experience on my side.
As I said before I have two good qualities: learning extremely slow and, in contrast, forgetting extremely fast.
At this point I’m not looking to save program space so I guess playing safe will be a good idea.
Any input on the hardware side?
Nick
|
Re: TEC controller
I got all the hardware working and wrote a simple temperature control program. Nothing fancy.
The code drives the TEC with different power output levels depending on how far the real temperature is from the set temperature.
I don’t have yet a mechanism of setting the target temperature but this is just to prove that the system works.
Since the major hardware is in place and the cooler is basically functional I will settle for now for a fixed, preprogrammed target temperature.
The ICSP connector for the 16F819 is accessible and I will play with the code to get more features.
I’m forced to take a break from the project but I will get back to it when time permits.
For now 62⁰F should be fine on most of the wines I drink so I’ll make it default.
Anyway it has been a fun weekend project and what makes it special is the fact that I did not have to order any parts. Recycling old junk makes me feel really good.
For all interested below is the simple control code, not fully tested since I just finished it last night. There are few things in the code that are not fully implemented (serial output will be used for a LCD display, hysteresis, id (to diferenciate between the two chambers). These variables will be used for future development when time comes.
Regards,Code:@__config_device_pic16f819 @__config_wdt_off @__config_hs_osc @__config_pwrt_on @__config_mclr_off @__config_lvp_off @__config_protect_on DEFINE OSC 8 DEFINE DEBUG_REG PORTB DEFINE DEBUGIN_REG PORTB DEFINE DEBUG_BIT 6 DEFINE DEBUGIN_BIT 7 DEFINE DEBUG_MODE 1 DEFINE DEBUG_BAUD 38400 DEFINE ADC_BITS 8 DEFINE ADC_CLOCK 3 DEFINE ADC_SAMPLEUS 50 ADCON1=%01001110 pwm_out var PORTB.2 adval var word val_1 var byte val_2 var byte val_3 var byte new var word act_temp var byte set_temp var byte new_temp var byte temp_index var byte i var word pwm_pars var word id var byte start_ch con 126 end_sym con 251 samples con 39 divider con 20 hysteresis con 2 min_pwm con 30 max_pwm con 50 CCP1CON = %00001100 PR2 = 15 CCPR1L = 0 CCP1CON.5 = 0 CCP1CON.4 =0 High pwm_out Pause 500 Low pwm_out T2CON.2 = 1 set_temp=62 goto main send_out: debug start_ch,id,val_1,val_2,val_3,end_sym return set_pwm_pars: CCPR1L = pwm_pars >> 2 CCP1CON=$0C|((pwm_pars&$3)<<4) return read_act_temp: new = 0 For temp_index = 1 TO samples ADCIN 0,adval new = new + adval Next act_temp=new/divider If act_temp <= set_temp then pwm_pars = 0 else pwm_pars = min_pwm+(act_temp - set_temp) if pwm_pars>max_pwm then pwm_pars=max_pwm endif gosub set_pwm_pars val_1=act_temp val_2=pwm_pars val_3=set_temp Return Main: gosub read_act_temp gosub send_out pause 1500 goto Main end
Nick
|
Re: TEC controller
Nick, The statement:
if pwm_pars>max_pwm then pwm_pars=max_pwm
endif
can be made as such: pwm_pars = pwm_pars min max_pwm
That way no if then is needed.
Dave Purola,
N8NTA
EN82fn
|
Re: TEC controller
Hi David,
Thank you for the hint. It will be first on my “To do” list.
I noticed it in your first post and mark it as a good candidate for the time when I will do code optimization. I just slapped the code together to see how the system works in a dynamic scenario and prove the concept. The way I thought it is that when the actual temperature is way higher than set temperature the system will push maximum power. The closer the two temperatures are the system reduces its output. Only when the actual temperature is lower than set temperature the TEC will be turned OFF. This reduces the number of ON / OFF cycles which I understand the TECs are not happy with.
I am open to any other suggestions that will improve the project.
Nick
|
Re: TEC controller
I’m in between two trips and had a little time to add a new feature to the code: hysteresis.
Adding few more variables and a call to te_control subroutine (code below) in the Main loop I can smoothly control the temperature with a ±2⁰F (good enough for me now).
One added bonus is that when restarting the TE after it was OFF I’m starting it with a safe voltage level output (start_pwm = 40) that will make sure the fans will not hesitate starting. After one cycle the regular rule, drive the TE as hard as the difference (act_temp – set_temp) is, kicks back in place. Short bench testing shows that it works like it should. When returning from next trip I will do more testing and also play with the values of min_pwm, start_pwm and max_pwm in order to get the desired temperature range with maximum efficiency and minimum ON / OFF TEC cycles.
My questions are:Code:te_control: If act_temp < (set_temp-hysteresis) then pwm_pars =0 pwm_flag=1 endif If act_temp > (set_temp + hysteresis) and pwm_flag =1 then pwm_pars = start_pwm pwm_flag=0 goto jump endif If act_temp > set_temp and pwm_flag =0 then pwm_pars = min_pwm + (act_temp - set_temp) endif jump: gosub set_pwm_pars Return
Is there a more elegant way of doing it?
Am I covering all logical possibilities or some more code is needed?
Any input will be appreciated.
Nick
|
Re: TEC controller
according to this a pid style of control may extend the life of your peltier device, thermal cycling is detrimental
https://thermal.ferrotec.com/technol.../thermalref10/
Warning I'm not a teacher
Members who have read this thread : 0You do not have permission to view the list of names.
Posting Permissions
Bookmarks