Smart crock pot -looking for fan cooled heat sink.

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by JR. »

I didn't stop tweaking the software.. :roll:

I expanded the temp adjustment range 50% to about 18''F per LED (12 temp LEDs) it was 12'F per LED before. I also added a manual mode of operation so I can just turn on the heat with 20% resolution, so 40W, 80W, 120W, 160W and 200W, but still read the temperature in real time open loop. I am not happy with how hot I need to get the resistors under the pot to get the pot hot, but the heat sink compiund should help a little.So far the solder connections on my wiring still looks shiny so I don't think I'm dangerously close to melting solder joints. I confirmed I can boil water now without going off scale but I don't feel like I have a huge headroom margin.

Interesting to see how the heat required to hold food at temp after it is cooked is so much less than before it is cooked... It seems like more than just the thermal mass of the cold food, probably the moisture in food, turning to steam, so as it runs out of moisture, it takes a lot less power to over-heat cooked food... :oops:

Based on this it seems that using the thermostat temp control makes a lot of sense especially for long-slow cooking. It would be cool if i could shoot an IR thermometer through a glass lid and measure the food temp directly, but I don't need to go down another rabbit hole.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by JR. »

After a full winter season of slow cooking with resistance heat and the summer quickly approaching, I am ready to revisit my Peltier approach. With resistance heat it very rarely uses the full 200W of heat available, highest persistent heat is around 40% of that or around 80W and that is not constant but dithering on/off so less than 80W continuous.

I figure my one Peltier chip can make about 50-60W of heat (considering self heat + heat extracted from the room). If I combine two peltier chips in parallel, and up my PS to around 15V @ 10A, I should be able to push more than 100 W of heat.

Now I need to revisit my cold side design. Instead of a small finned heat-sink with fan, I think I need to make something more substantial. Perhaps a metal base for the cooker. I need to fully isolate the hot side from ambient using my 1/4" aluminum plate but have that live inside my thermal insulated box so the heat stays inside.

I need to cleverly clamp my two Peltier devices between two metal plates that are thermally insulated from each other.

I can easily come up with a PWM control, the Peltier devices don't need clean DC. That said chopping 15V @ 10A might make some RF. :roll:

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by JR. »

I wish I would have thought about this sooner.... :lol:

I just did some math on how much cooling I should expect from my cooker.... I always knew it wouldn't be much. The Peltier devices consume several times the power than the amount of heat they move, so only a couple ten watts of cooling per device.

I just looked at the spec for the power supply (15V @ 15A) and it is 85% efficient. So when the cooker is cranking at full tilt, the power supply could be adding 30W+ of heat to my kitchen. :oops: :oops:

My resistance cooker is looking better and better... but I want to finish this just to see how it works in reality.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by mediatechnology »

Resistors are cheaper too.

I have a friend that is building some custom Peltier wine coolers.
He's harvesting the modules and controllers from commercially-made metal-exterior coolers and then putting them in nice wooden cabinets for his client.
I think each cooler has two large modules.
The heatsinks are huge and the two switching PSUs about the size of a 600W PC supply.
The controller is very cool-looking with blue 7 segment displays.
The manufacturer put a lot of engineering into it.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by JR. »

mediatechnology wrote:Resistors are cheaper too.
Yup, I'm not doing this for commercial reasons but to entertain myself. The rest of my parts arrived today, so this weekend I'll rebuild it to replace the resistors with Peltier devices.
I have a friend that is building some custom Peltier wine coolers.
He's harvesting the modules and controllers from commercially-made metal-exterior coolers and then putting them in nice wooden cabinets for his client.
I think each cooler has two large modules.
I am now using two of the largest P modules (from CUI 60440)... My first attempt only used one.It was marginal... after a winter of cooking with resistors, I have adjusted down my understanding of how much heat i need for "slow" cooking.
The heatsinks are huge and the two switching PSUs about the size of a 600W PC supply.
My new PS arrived today !5V at 15A so 225W or so at full tilt while 2x Peltier loads probably won't draw full current available.

The controller is very cool-looking with blue 7 segment displays.
I repurposed a drum tuner PCB so 12 bicolor leds for temp display (red for target and green for actual temp). The 5 clear mode LEDs indicate rate of heat... Other than first turn on, getting up to heat, I rarely see more than 2 LEDs on (80W) and these 2 LEDs don't stay on steady.

I have two modes... "manual" where I can drive constant heat power into the cook pot (40w, 80W, 120W, 150W, or 200W), or automatic "temp" mode where I program a temp and it manages the rest.
The manufacturer put a lot of engineering into it.
Peltier devices are mature technology, while I think there was some recent noise about new different material. As usual with such projects I now know what i should have known before I started the project. This is not a great idea. As I've shared before the Peltier devices are not very efficient at transferring heat (they consume power to move heat), and there are multiple practical gotchas. Too hot on the hot side and the devices delaminate, too much temperature differential and it can't push much heat/cold. In my application I can take advantage of the inefficiency since the extra power I need to use to move heat, creates waste heat I can use for cooking.

My new version has a huge 1 foot square (0.120" aluminum) cold-side heat sink. This will be the base of my cooker (with feet). My device cooler fan, bolts to the bottom of this this to blow room temp air across it. I will have 1/4" cork thermal insulation between the cold side and hot side... the hot side still uses the 6"x6" 0.250" aluminum heat spreader.

I will use power mosfets to switch the Peltier devices on/off with PWM duty cycle, they don't need smooth DC. Since I have two Peltier devices I may drive them separately to give me a lower minimum heat, just one on. I guess I'm being optimistic. :D

I will build another thermal igloo to capture the hot side heat and keep it inside the cooking chamber.

The real lesson I've learned about slow cooking is that insulation and heat loss matters.

I repeat this is mostly a science fair project, don't try this at home. :lol: :lol:

That wine cooler will heat up the room, probably throwing 2x-the waste heat or more, as any cooling it provides. OK for a picnic not so sure for the living room. While a refrigerator will heat up the room too.

JR

[edit] Good news for the wine cooler is the working temp differential is smaller (say 50'F wine vs 75' F room so 25') the lower temp differential helps effectiveness of Peltier... My cooker goes from room temp up to 150'F or so,,, harder, [/edit]
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by JR. »

No battle plan survives contact with the enemy, and my design plans often bump against reality.

My plan to chop DC to feed the Peltier devices has a flaw... according to the apps data feeding a P device with 10% duty cycle is not the same as feeding it with 10% of the voltage... The P devices are happier with lowest voltage necessary.

So next pivot is to make a crude DC to DC convertor but this adds another vector for efficiency loss. I was worried about the PS only being 85% efficient, but adding another switcher after it just compounds the problem.

#1 I suspect the efficiency spec is worst case at full output (15A for my supply) which i will never draw ,I expect maybe 10A max. So next investigation is hacking the PS to make it programmable output.

It is a nice (cheap) thru-hole technology design, and they didn't even bother to conceal the chip numbers, so they use an old fairchild PWM chip. The chip uses an internal 5V reference so fairly straightforward to trick it for 5V-xV, I could even drop in a DPOT and push pot settings from my micro to vary the voltage,, but that seems like a lot of work and is not optimal, so what if i try KISS?

Since the power supply has all the important parts to deliver 15A at 15V or higher? What if i send a PWM signal from the micro to control the high side switch and loosely regulate the voltage, or maybe just close the loop by regulating P-temperature,

I need to poke around inside the PS to see what switching frequency it is already running at, and how much unregulated power supply it has. The PS looks like a conventional 2-stage.. One switcher from the mains that provides transformer isolation and makes roughly 250W, a second PWM that turns that 250w in to different voltages. They sell a series of supplies running from 5V to 48V output.

I am reluctant to ask the manufacturer for schematic and assistance.. well not reluctant to ask, but don't expect any support. 8-)

If the unregulated PS is high enough, I can run the two P-devices in series to keep current and losses lower,,, If not 30V or more, I can run them in parallel...

Next step is to open up the supply and scope around for voltages and switching frequency... they use pretty common fairchild KA7500 pwm chip.

This idea may not be as crazy as it first seemed.... :D

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by JR. »

OK some more data/inspection.

The family of 240W power supplies specify a range of efficiencies as low as 79% for a 40A supply and 87% for their 5A supply so efficiency is worst case and resistive losses (logical).

I opened up the supply and probed around, a pretty mundane 25kHz switching frequency and what looks like a 40V+ unregulated supply. A 25kHz PWM period should give me plenty of resolution using a PIC PWM output.

With 40V of PS to work with it looks like my best choice is to stack two P devices in series, they will draw about 5A with 15V across each so 5A for 30V across two in series. This should give me good efficiency and control.

I just need to carefully figure out how to drive the high side switch.

I will ask them to share a schematic, but don't expect much help.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by mediatechnology »

Couldn't some of the lost heat from switcher inefficiency be funneled into the hot side?
When you're trying to make heat why does inefficiency matter?
I realize that it may be hard to recover the waste heat from the switcher's inductors, diodes, capacitor etc. but couldn't the switching transistor's losses be heating the pot roast as well?
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by JR. »

mediatechnology wrote:Couldn't some of the lost heat from switcher inefficiency be funneled into the hot side?
When you're trying to make heat why does inefficiency matter?
I realize that it may be hard to recover the waste heat from the switcher's inductors, diodes, capacitor etc. but couldn't the switching transistor's losses be heating the pot roast as well?
Yes for a scratch design I could attach all the heat generating devices to my heat spreader plate (6" square 1/4" aluminum, not unlike the power resistors that were attached before... but looking at the PS there are at least three different heat-sinks not to mention the other lossy parts in the path.

I am looking at this as a closed hot side system...heat spreader plate, insulated from the cold side heat-sink by 1/4" cork layer, inside an insulated igloo, so heat goes in and mostly stays in...

The cold side system is a 1' square aluminum plate with a forced air fan (with chip cooler heat-sink) the cold side is sitting out in ambient kitchen air.

At full tile power the two P-devices are moving some 20+ watts of heat each + dissipating 50-60W of heat to do the work of moving the heat... so total for both P-devices 140-150 W of heat going to the hot side, and maybe 40W of cool sucked from the cold-sink.

If I can run the switching PS to drive the two P-devices in series the efficiency of the PS is close to 90% so 15W of heat to ambient. If I locate the PS near the cold plate the 40W of cold suck should easily swallow the 15W of PS heat... OTOH if I have to drive the 2x P devices in parallel the efficiency drops to more like 85%,so 22W of heat.... [edit] I just checked and the output caps are 25V so parallel it is [/edit]

Net net, I should be making more cool than heat into my kitchen air and this is at worst case full tilt... when running at less than full tilt the PS and P devices will be more efficient/effective, so cooler.

From my experience cooking with resistors, the thermal igloo to prevent heat loss is very important... the less heat loss from the cooking operation, the less heat I need to add back in to the system to hold temperature up. The food consumes small amounts of actual heat to "cook" the food, but generally while cooking on a stove top, massive amounts of heat are lost to the room.

I am pretty optimistic about this direction... If I add another chopper stage and filter between the 15V PS and the P devices that could be another X percent of power loss.. better fro me to just hack the PS to accept PWM from my micro-controller. .

JR

PS: Note that these are all very brown numbers ;) (pulled from my butt..). The effectiveness of the cold side fan/sink is a variable I do not have a handle on... If the cold side gets too cold, the heat from the PS may be a good thing.
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Smart crock pot -looking for fan cooled heat sink.

Post by JR. »

In no great surprise it has been several days and not a peep from the PS company about my request for a schematic so I can modify it. Such is life.

I am probably over thinking this. Most simple thermostats are on/off. With two P-devices I can configure both in series , or both in parallel so two power levels... I might consider a third as just one full voltage, as that will be slightly more power/heat but less cold than the two in series, but I suspect two power levels will be more than enough range.

I am waiting for some more metal and longer screws. The cork insulation I bought for between the hot and cold sinks is thinker than they specified. So I need a thicker spacer, and longer screws.

I should probably get something together by this weekend... May have to cook on my stove this week... :lol: :lol:

JR
Cancel the "cancel culture", do not support mob hatred.
Post Reply