MES Mod Madness Teaser Pics!!!

[tallbm]

Tall,

I just rebuilt a Smoke hollow electric using a HeaterMeter 4.2. You can do better by, tweaking PID and fan (SSR) controls. The first temp overshoot was 20-30 degees. I've got that down to 15 degrees...then it settles right in at set temp. Driving a SSR from that PWM fan out is a bit trickey. SSR turns on at 3 volts...no reason to drive it with 12volts, fan min 100%, fan max 100%. I have mine set 40% / 40%. You can read PID tuning at that other website. I have some more work on mine. I built the control box and it's a little too small. Have to move it to a biggerbox.

Here's a shot of a pork butt cook....on at 8pm, turned temp down at midnight and went to bed. I did port forward on the home router...works great. Now I can control it from the beer store!.

Good luck!

RG

0408171305.jpg

• radioguy
• May 7, 2017

Hi RG, thanks for the input!

Good point on the 3volts vs 12 volts.

My PID settings were taken from an Auber PID autotune I did on a different MES I bought for $40 and rewired as a gift to my mother.  

My first temp overshoot is 2 degrees so far using the PID settings I copied from the Auber PID autotune.  When empty the temp oscillates no more than 3 degrees.

My biggest issue is that when I add any decent load to the smoker the recovery time when I open the door takes a while when I have the smoker probe at rack 1 level or higher in the smoker.

Also I discovered that the "lid open" setting needs to be set to 0%.  With an electric smoker it wants to cut output for 240 seconds by default when triggered.  This is NOOOO good lol.  I lost 100F degrees due to that setting.  I'll know how it recovers from a 30F or so drop soon but a 100F drop with 2 racks of food was a killer.

This is fun stuff and I look forward to sharing any info I gather with you :)

 

[tallbm]

Tall,

I just rebuilt a Smoke hollow electric using a HeaterMeter 4.2. You can do better by, tweaking PID and fan (SSR) controls. The first temp overshoot was 20-30 degees. I've got that down to 15 degrees...then it settles right in at set temp. Driving a SSR from that PWM fan out is a bit trickey. SSR turns on at 3 volts...no reason to drive it with 12volts, fan min 100%, fan max 100%. I have mine set 40% / 40%. You can read PID tuning at that other website. I have some more work on mine. I built the control box and it's a little too small. Have to move it to a biggerbox.

Here's a shot of a pork butt cook....on at 8pm, turned temp down at midnight and went to bed. I did port forward on the home router...works great. Now I can control it from the beer store!.

Good luck!

RG

0408171305.jpg

• radioguy
• May 7, 2017

RG have you done anything special for security while opening up a port on your router?

I'm thinking I would:

• definitely beef up my router password
• use some non standard port number to avoid any bad traffic written specifically for commonly used ports (80, 22, 443, etc.)

Any thoughts on an RSA SSL setup with certificates?  Like have you dug into it much to even know if it is possible with the PitDroid app?

I don't have any concern about a hacker messing with my smoker.  I'm more concerned about them getting to the HM linux system and being able to back track to my router or remote around my network from there.

Feel free to PM me if you don't want to publicly post about your setup.  Thanks! 

 

[BGKYSmoker]

Thanks Dave for finding that.

The Bradley was bad for hot and cold spots. The fan ran CCW and pushed air into the cabinet. I ran my Bradleys with an Auber PID. Did them so much that Auber rebuilt them 6 times. I switched from Bradley back to MES again and still MES POJ. All the $ i spent i went with a better smoker. TSM 30lb. 0 issues.

 

[tallbm]

 
Thanks Dave for finding that.

The Bradley was bad for hot and cold spots. The fan ran CCW and pushed air into the cabinet. I ran my Bradleys with an Auber PID. Did them so much that Auber rebuilt them 6 times. I switched from Bradley back to MES again and still MES POJ. All the $ i spent i went with a better smoker. TSM 30lb. 0 issues.

Thanks for the info Nepas.  I could see myself going to a different brand someday when all of this wears out.... unless I really fall in love with the HeaterMeter or something lol :)

 

[daveomak]

 
Thanks Dave for finding that.

The Bradley was bad for hot and cold spots. The fan ran CCW and pushed air into the cabinet. I ran my Bradleys with an Auber PID. Did them so much that Auber rebuilt them 6 times. I switched from Bradley back to MES again and still MES POJ. All the $ i spent i went with a better smoker. TSM 30lb. 0 issues.

I got lucky...   Now if I could only remember where I put my cocktail down...    

 ...   

 ......

 

[BGKYSmoker]

 
I got lucky...   Now if I could only remember where I put my cocktail down...    

 ...   

 ......

I think i did that yesterday.

tcup4.jpg

• BGKYSmoker
• May 8, 2017

 

[radioguy]

 
RG have you done anything special for security while opening up a port on your router?

I'm thinking I would:

• definitely beef up my router password
• use some non standard port number to avoid any bad traffic written specifically for commonly used ports (80, 22, 443, etc.)

Any thoughts on an RSA SSL setup with certificates?  Like have you dug into it much to even know if it is possible with the PitDroid app?

I don't have any concern about a hacker messing with my smoker.  I'm more concerned about them getting to the HM linux system and being able to back track to my router or remote around my network from there.

Feel free to PM me if you don't want to publicly post about your setup.  Thanks! 

Doing port forwarding is not truly opening up a hole in your network.  You have to trust your firewalls and password complexity.   Don't skimp on firewall software.  Adding RSA makes it much more complex, it would require a lot more work in the software.  I'm not sure about PitDroid supporting SSH.   You need a public IP / URL address for internet DNS.   I signed up for a "free' one at No-IP.  There are a bunch of services out there.  No-IP offers a free one that you have to renew every 30 days and tolerate their emails to upgrade to a paid account.   The IP service will supply you with a URL (advertised on DNS)  You have some some control over the name.  bbq1.abcd.org for example.  In your home router forward port 80 to your URL.  It mat be pretty easy to set it up for an obscure port number I did not investigate this.  I went one step further and made it MAC dependent.  So only my HM (WiFi adapter) can talk to that web URL.  With it i can use my phone browser or PitDroid APP.  

My electric is truly set and forget. Have Fun!

RG   

 

[tallbm]

 
Doing port forwarding is not truly opening up a hole in your network.  You have to trust your firewalls and password complexity.   Don't skimp on firewall software.  Adding RSA makes it much more complex, it would require a lot more work in the software.  I'm not sure about PitDroid supporting SSH.   You need a public IP / URL address for internet DNS.   I signed up for a "free' one at No-IP.  There are a bunch of services out there.  No-IP offers a free one that you have to renew every 30 days and tolerate their emails to upgrade to a paid account.   The IP service will supply you with a URL (advertised on DNS)  You have some some control over the name.  bbq1.abcd.org for example.  In your home router forward port 80 to your URL.  It mat be pretty easy to set it up for an obscure port number I did not investigate this.  I went one step further and made it MAC dependent.  So only my HM (WiFi adapter) can talk to that web URL.  With it i can use my phone browser or PitDroid APP.  

My electric is truly set and forget. Have Fun!

RG   

Yep all that info seems to fall in line with what I have read/researched.

I was eyeballing Duck DNS for a free DNS but saw were many used and reported the same things as you have about No-IP.

I also did not find much in the way of setting up SSH and such and I don't think PitDroid supports it.  I don't want to write my own app just for that because that would be a whole other project to start on but who knows I may just do it.  It would be a good reason for me to play around with mobile app development which I lack experience in.

I guess if I take the steps to expose my HM over the internet I need to go ahead and update my Router firmware and then look into some simple software firewalls for both the HM and my PC's on the network.

As for using an obscure port number, I found information on how to do that.  It all seemed very straight forward is just a few commands to the HM to run things on a different port and a couple of reboots or so and things should be ready to go.

The last thing I was looking to do was to get a smart plug so I could then have TRUE remote on/off capabilities but I think that is probably overkill in the grand scheme of things hahahha.  The idea of getting on my phone while at the grocery store and turning on my smoker seems very appealing but in all practicality I can just turn the smoker on after I walk in the door at home and then go mess with meat and it should be up to temp within the 30 minutes of me unpacking groceries, prepping meat and starting the AMNPS.

Anyhow, thanks for all the info.  I may have some time here in June to do whatever I can think of :)

 

[radioguy]

When the set point is at zero its virtually off, but not really.   Check this out, all it needs is 3-60vdc from the RPi. 

https://www.adafruit.com/product/2935

RG

 

[tallbm]

 
When the set point is at zero its virtually off, but not really.   Check this out, all it needs is 3-60vdc from the RPi. 

https://www.adafruit.com/product/2935

RG

Very interesting!

 

[tallbm]

UPDATE:

So I did a smoke this weekend and I was able to work on the tuning of the HeaterMeter (HM) PID controller.  I think I have it dialed in but it took some values I wouldn't have been able to guess at and may seem a bit extreme.  That is why am posting the pic of the tuning and smoke below to back up my findings! :)

These electric smokers are slow performing systems compared to other systems that PID controllers seemed to be used with.  This means I had to turn things up for very fast response times.  Every second the heating element is off can result in many degrees in change.

So in all I started out with my "P" (Proportional) doing well and not overshooting by more than about 5 degrees.  The settle temp though was more like 12 degrees under though.

I tuned the I (Integral) to where "I" was able to get the settle time down to about 7 seconds while the overshoot stayed about 5 degrees over with no real change.  I could not do any better with the "I" setting so I moved on to the "D".

Now is where things got really tuned in.  I began tuning the "D" (Derivative) setting which is based on the RATE of temperature change to trigger the power to switch off/on to the heating element.  So before as the temp would rise quickly and overshoot to 5 degrees above, I was able to tune the "D" to where my overshoot was within 2-2.5 degrees!!!

Also the settle was within 2-2.8 degrees!!!

In all I was able to keep my temp between 272/273 as a low and 277.5 as a high!  Keeping within 3 degrees is aswesome!

When I bumped the temp up to 325F I was able to keep within 320-327 so 2 degrees overshoot and 5 degrees settle.  I can live with that as well.

I thought I had disabled the Lid Open mode but apparently I was wrong so I took another shot at "disabling" it and hopefully I have it resolved once and for all so I don't wind up with the element being shut off for such a long period of time to drop 100F+ degrees!

HmTuning.png

• tallbm
• May 15, 2017

I did run into another issue.  I discovered that I am burning up my 2nd On/Off Rocker switch 

warmer-switch-7-8-x-1-1-8-dpst.jpg

• tallbm
• May 15, 2017

After doing some reading it seems that the issue may be a few things but mainly that I am having peaks of drawing too much amperage through the switch towards it's upper limit.  My 15A fuse does not blow so I think I am continuously drawing near the upper limit and the switch really doesn't handle it well and melts down after these 5hr+ smokes.

I'm going to simply bypass the switch and my On/Off will be when I plug or unplug the controller.

The alternative would be to step down the voltage and amperage to the switch and then run another Relay like the SSR where the switch then sends the lower voltage signal to the SSR to allow the real power to flow through when I flip the switch.

This is way more than I really want to do so I'm just going to eliminate the switch and my a plug/unplug will do the job lol.

Anyhow thats what I've got for now.  I'll continue to post updates as I come across more findings.  Once everything has settled down I plan to do a step by step post for a HeaterMeter PID job with components needed like SSR, wires, wifi adapter, etc. as well as the HM PID settings, lid open mode settings, etc. etc.

 

[johnmeyer]

The problem with the switch may be caused by the "inrush" current when a heating element is first turned on. This is a fairly typical problem. It's been a long time since I spec'd switches, but you might look for ones that are designed for inductive loads (like motors). A second thing you can do is to put a 250V non-electrolytic capacitor of at least 0.1 uF directly across the switch. It will help absorb some of the inrush current and reduce the arcing across the switch contacts, which is probably what is causing the switch to self-destruct. Finally, you can simply "derate" the swtich, which means getting one that has much higher specs than called for. This is a typical engineering solution, where you design in a buffer.

I'm sure there are thousands of articles about this, but a quick Google search turned up this one. Just skim the first page.

Switch Ratings, What's it all Mean?

It is for airplane switches (they typically use 24V), but the same principles apply to what you're doing.

[edit] Here's a derating calculator for switches. You'll want to use the "lamp load" column, because the MES heating element is basically the same thing (electrically) as a big incandescent light bulb:

Rerating Current For Switches With 125v AC Original Ratings

As you can see, you'll need a switch that has between 3 and 5 times the current capacity as the steady-state current. For a 1,200 watt heading element, the current is nominally 10 amps, so you would need a 30 to 50 amp switch.

 

[tallbm]

 
The problem with the switch may be caused by the "inrush" current when a heating element is first turned on. This is a fairly typical problem. It's been a long time since I spec'd switches, but you might look for ones that are designed for inductive loads (like motors). A second thing you can do is to put a 250V non-electrolytic capacitor of at least 0.1 uF directly across the switch. It will help absorb some of the inrush current and reduce the arcing across the switch contacts, which is probably what is causing the switch to self-destruct. Finally, you can simply "derate" the swtich, which means getting one that has much higher specs than called for. This is a typical engineering solution, where you design in a buffer.

I'm sure there are thousands of articles about this, but a quick Google search turned up this one. Just skim the first page.

Switch Ratings, What's it all Mean?

It is for airplane switches (they typically use 24V), but the same principles apply to what you're doing.

[edit] Here's a derating calculator for switches. You'll want to use the "lamp load" column, because the MES heating element is basically the same thing (electrically) as a big incandescent light bulb:

Rerating Current For Switches With 125v AC Original Ratings

As you can see, you'll need a switch that has between 3 and 5 times the current capacity as the steady-state current. For a 1,200 watt heading element, the current is nominally 10 amps, so you would need a 30 to 50 amp switch.

Thanks johnmeyer, that all makes sense!

I found some marine single pole breaker toggle switches by Blue Sea Systems that should do the trick.  I can find switches in all kinds of AC Amperage ratings and go up to the 100A for 125V AC switch if want.  The 100A trips at 67.5A.    It should also fit in my controller box as well which would be a huge win.

61N-5Dgx6BL._SX522_.jpg

• tallbm
• May 16, 2017

Also do you feel it would be wise to upgrade the wire that connects to this switch to be 10AWG? 

Thanks for the info.

 

[radioguy]

Johnmeyer,  I have to respectfully disagree with you comment on "In-rush" current.  Heating elements are resistive, their load is constant.  The resistive value may vary a slight bit due to temperature change, has to do with what they are made with.  In rush current only occurs on inductive or capacitive loads IE motors, coils.

What is killing the switch is the continuous duty, duty cycle.  Those cheap Chino switches don't cut it in a constant high amperage application.  The materials used are inferior, contacts heat up, casing more resistance, more heat and so on.  What you need is a bad a$$ switch made in the USA.   You are only switching 1200watts (mine is 1500).  That is a one time switch per use.  The SSR takes over from there.  Your current through the switch will vary from 0 to about 10amps.  It will be 10amps for nearly 100% of the time at the start but the duty cycle will decrease as you reach set temp.  Once at set temp, duty cycle should be at it lowest (per set temp), I hope that makes sense.  I'm using a similar switch in mine 1500 watts, removed from some old equipment dated 1962, about 20 hours of usage no problems.  

Glad to see you got it tuned in better.  Are you using a thermo couple for pit probe?  I decided not to try and save and bought some good quality TC and RTD from ThermoWorks.  

Happy Smokin'

RG

0517170912.jpg

• radioguy
• May 17, 2017

 

[tallbm]

 
Johnmeyer,  I have to respectfully disagree with you comment on "In-rush" current.  Heating elements are resistive, their load is constant.  The resistive value may vary a slight bit due to temperature change, has to do with what they are made with.  In rush current only occurs on inductive or capacitive loads IE motors, coils.

What is killing the switch is the continuous duty, duty cycle.  Those cheap Chino switches don't cut it in a constant high amperage application.  The materials used are inferior, contacts heat up, casing more resistance, more heat and so on.  What you need is a bad a$$ switch made in the USA.   You are only switching 1200watts (mine is 1500).  That is a one time switch per use.  The SSR takes over from there.  Your current through the switch will vary from 0 to about 10amps.  It will be 10amps for nearly 100% of the time at the start but the duty cycle will decrease as you reach set temp.  Once at set temp, duty cycle should be at it lowest (per set temp), I hope that makes sense.  I'm using a similar switch in mine 1500 watts, removed from some old equipment dated 1962, about 20 hours of usage no problems.  

Glad to see you got it tuned in better.  Are you using a thermo couple for pit probe?  I decided not to try and save and bought some good quality TC and RTD from ThermoWorks.  

Happy Smokin'

RG

0517170912.jpg

• radioguy
• May 17, 2017

Hi RG thanks for that input as well.

I ordered that 100A marine switch breaker so I should be covered for continuous duty, inrush, or anything short of a lightning strike :P

My pit probe is a K Type Thermocouple (TC).  The HeaterMeter (HM) I bought is created for 1 TC and 3 Maverick 732-3 hybrid Thermisters. 

I tested all probes twice in boiling water.  I had to wrap the TC in a plastic sandwich bag to waterproof it. 

Test 1 was to see what the probes were reading.  The TC was 4F high and the other probes 5 degrees low.  I then set the offsets for the probes in my HM and retested.  With Test 2, everything was spot on to 212F in the boiling water :)

I just looked at the Thermoworks TC's and I have my eye on one of the two alligator clip ones should this TC fail.  I know the one I bought from Auber may be more cheaply made.  Whomever assembled it decided it was a good idea to twist the slack out of the two wires that get screwed down to the prongs in the yellow case.  This meant it always read 80F.   I had to undo the yellow case, untwist the wires, spread them, tape them off with electrical tape for "insualtion" and then screw them back down.  The probe worked properly after that. 

As for now I'm try and work all the kinks out before I do any additional tweaking.

Oh, by the way.  After I had dialed in the P-I-D values I decided to switch the fan to 40% output like you suggested.  After there the change the temp quickly creeped up to over 280F and continued upward on a 275F set point.  One or more of the P-I-D tunings seems to take into account the output percentage because 4V vs 12V output should make no difference for the SSR to switch off and on and cause the heating element to receive any more or less power.. 

Since the HM is designed for a blower fan my guess is that the P-I-D logic/calculations take the fan output into consideration and resulted in my temp creep. 

I think the Derivative (D) was affected most by the change in fan output because I turned the output back to 100% and it didn't recover.  Only after I turned the D back to 0 did the temp stop climbing.  Once the temp was back in the valley of the P-I cycle from before did I set the D back to the working settings and sure enough I was back in my tight loop.

This was an interesting learning experience with the fan output lol.  For now I'm going to leave the output at 100% and simply waste the extra 8V of the 12V being sent to switch the SSR lol.

 

[johnmeyer]

 
Johnmeyer,  I have to respectfully disagree with you comment on "In-rush" current.  Heating elements are resistive, their load is constant.  The resistive value may vary a slight bit due to temperature change, has to do with what they are made with.  In rush current only occurs on inductive or capacitive loads IE motors, coils.

You are correct that the higher startup current for an inductive motor is causes by not having "back EMF" from the motor when it first starts and is not turning. Once turning, the current drops because the magnetic fields in the motor reduce the current. However, for a lamp or heating element there is also a much higher initial current because the resistance (as opposed to impedance) changes substantially as the element heats. It is not a small change. You can prove this by measuring the resistance of a common incandescent light bulb with a common ohmmeter. If you plug that resistance into an ohm's law calculator, it will predict a current, and a wattage, that is way higher than what the bulb actually consumes.

But you are probably correct that the word "inrush" may not be the correct one to describe this initially higher current.

Regardless of what you call it, the additional initial current does need to be taken into account when specifying the switch.

 

[tiros]

I think the "no back removal" MES rewire tips will be super helpful and I will include the measurements for adding a back panel for the Rollout Limit Safety switch, which I did not show in the pics above.  I swapped mine with a 350F manually resetable switch and used high temp steel connectors.  I don't like the idea of the one time use 302F switch being there without me being able to access it.  Plus the higher 350F threshold will give me some margin for error as I manually tune my HeaterMeter PID to get things heating properly. 

I hope you enjoyed and thanks for stopping by!

Are you saying here the over temp limit is one use?

I thought they reset when they cool down...Can you verify as I am currently experimenting with designing a new controller.

 

[tallbm]

 
Are you saying here the over temp limit is one use?

I thought they reset when they cool down...Can you verify as I am currently experimenting with designing a new controller.

Hi there and welcome!

I cannot 100% verify but recently a couple of other reputable members claimed they were auto reset.  I believe on of them may have even found the exact replacement on Amazon.

I would say I am about 95% confident in the information from these other members.

I will go back and fix the info in that post to make sure it highlights that the rollout limit switch is more than likely an auto reset type.

Let me know if you have any other questions and best of luck on the controller design :)

 

[dr k]

Are you saying here the over temp limit is one use?
I thought they reset when they cool down...Can you verify as I am currently experimenting with designing a new controller.

The safety snap disk sensor resets itself per DaveOmak. Unless it's damaged/defective.
-Kurt

 

[tallbm]

I still think it is a good idea to make a panel to access the Rollout limit switch as it's connectors are one of the known week spots in the MES wiring circuit.  This way you can easily access and repair/swap the Rollout limit switch when the connectors corrode away... and it seems they almost always do.

I replaced my connectors with Hi Temp stainless steel ones at both the Rollout limit switch and the heating element sites.  I think I will be good to go for a while without corroding connectors :)