Homeroasters - Home Roasting Coffee Community » ALL ABOUT ROASTERS » Popcorn Popper roasting
Who is here? 1 guest(s)
 Print Thread
Controllable AC dimmer / TRIAC
marcov
Top  #1 Print Post
Posted on 01/21/2019 11:45 AM
Hey,
I'm so glad to have found this great forum, this is my 1st post.

I am thinking about hacking a popcorn popper with the goal of controlling the roasting temperature with a PID.

I see here that most of the projects do this by modulating the power on the heating element using a common (zero-cross) SSR. This is done with some PWM-style control on the SSR.

This kind of control is not well suited for AC loads, so I'd prefer to use a TRIAC + zero cross circuit that compared to PWM + SSR is able to control power over each single AC wave period.

I found this ready to use board with all what I need, but it can drive 2A in continuous and that's not enough for a 1500W load: https://robotdyn....r-5A-1L-EN

Do you know any other board that can driver higher loads? Using separate TRIAC and zero cross board would be ok too, as long as they could directly interface with a 3.3v micro.

Otherwise I guess the other alternative is to use a random turn on SSR + zero cross.
 
renatoa
Top  #2 Print Post
Posted on 01/21/2019 12:47 PM
Heaters inductive component is very weak, even you see them as coils, so you can consider heaters as resistive loads without worry.
ON-OFF is fine for 1 second cycle, you have 1% resolution/precision, which is ok for most heating control application.

RobotDyn dimmer is indeed the cheapest solution anyone can buy for a SSR with zero cross in a single package.
Their 2A rating is a non-sense, the triac is 16A, and the heatsink dimensions allows for more than 5A without exceeding 50C for 15 minutes, which is barely lukewarm for such triac.
I am using the same components in a DIY build for more than one year without issues.

Regarding control styles, please be aware that the RobotDyn library use by default the so called pulse skip modulation, better known as ICC (integral cycle control) in the TC4 dominated realm of coffee roasting.
PSM/ICC could be better suited for applications where full/half sine wave passing is desired, instead sharp edge commutation from zero to full power in the middle of the sine wave, the later generating a lot of harmonics and EMI.
 
btreichel
Top  #3 Print Post
Posted on 01/21/2019 10:21 PM
I've built them all. the easiest solution is the random turn on SSR + zero cross. In the end its what makes you the happiest since it really doesn't effect the roast. You are the significant control variable.
 
marcov
Top  #4 Print Post
Posted on 01/22/2019 11:14 AM

Quote

renatoa wrote:

Heaters inductive component is very weak, even you see them as coils, so you can consider heaters as resistive loads without worry.
ON-OFF is fine for 1 second cycle, you have 1% resolution/precision, which is ok for most heating control application.



Assuming 50Hz, a 1s period would give me a 2% resolution and accuracy, but without a zero cross to do phase syncing the error would be closer to 5%.

Quote


RobotDyn dimmer is indeed the cheapest solution anyone can buy for a SSR with zero cross in a single package.
Their 2A rating is a non-sense, the triac is 16A, and the heatsink dimensions allows for more than 5A without exceeding 50C for 15 minutes, which is barely lukewarm for such triac.
I am using the same components in a DIY build for more than one year without issues.

Valuable info, thanks.

Quote


Regarding control styles, please be aware that the RobotDyn library use by default the so called pulse skip modulation, better known as ICC (integral cycle control) in the TC4 dominated realm of coffee roasting.
PSM/ICC could be better suited for applications where full/half sine wave passing is desired, instead sharp edge commutation from zero to full power in the middle of the sine wave, the later generating a lot of harmonics and EMI.


It's ok, I'm planning to write the code. I was also looking for a microcontroller able to do the duty cycle control purely in HW, using a capture compare peripheral, but given the low frequency involved and the fact that you can re-sync on every zero cross a SW solution would be good, too.

I did not consider the EMI emissions. That could indeed be a problem for the microcontroller or other clock sensitive IC.

Thanks for the great answer!
 
marcov
Top  #5 Print Post
Posted on 01/22/2019 11:22 AM

Quote

btreichel wrote:

I've built them all. the easiest solution is the random turn on SSR + zero cross. In the end its what makes you the happiest since it really doesn't effect the roast. You are the significant control variable.


That's true! Probably though the cheapest solution is just to use the cheap chinese FOTEK SSR (i.e. a zero cross SSR) and do PWM with on the SSR input, with a period bigger than 1s.
 
renatoa
Top  #6 Print Post
Posted on 01/22/2019 1:21 PM
50Hz means 100 half-sines, this is the granularity of control when using a random on-off, because a triac switch off each zero cross.

The Fotek SSR are ok, but you should be aware their specs are fake, at least the 25/40A units. I broke down the 40A model to replace the burnt triac and found inside 16A triac. This guy video, same finding as me... 12A triac in a SSR rated for 25A...
https://www.youtu...xEhxjvifyY
 
Jump to Forum: