Recently, I had the opportunity to electrically and thermally examine 4 wall ovens at no monetary expense to myself. Despite the fact that I couldn't do this under laboratory grade conditions, I none the less thought that my readers would enjoy a write up on the subject, so here goes.
There were no markings indicating the model numbers so I have no idea what models they were. One was a Bosch model HBL8453UC, still being sold in 2024. One was a GE, about 10 years old from what I understand. One was a Kitchen Aid Superba, significantly older than the GE. And one was a Frigidaire of unknown age. They were 30", 27", 24", and 27" respectively.
I was, unfortunately, unable to do a teardown of these ovens. I did get to take a peek inside of the GE one. I saw the oven's capacitors were Rubycon and it used relays for switching. The other ovens, based on the sounds they made, all used relays for switching, which I find to be a bit on the antiquated side for the WiFi capable Bosch oven.
Incidentally, if you ever want to test an oven, please bear in mind that it will take a more than 2 hours bear minimum. Of course, the oven must also be willing to participate, otherwise you'll end up spending even more time trying to outsmart a "smart" oven.
It should be noted that all of these ovens output their temperature in imperial units. Though it may have been possible to switch them to metric I had no manuals for them, so I did not attempt it. Therefore you'll all have to enjoy the imperial readings I present to you below though I did convert many of them to metric for your viewing pleasure. Incidentally, dividing an imperial unit by 2 and subtracting 16 will typically get you close to the metric value.
As I said, the test conditions were not laboratory. The particulars that come to mind are that I had only one calibrated DMM to test with, the incoming AC voltage varied between ~231v and ~246v, the temperature outside of the ovens was uncontrolled for, as was wind direction and speed, and the K-type thermocouple's position was not fixed to one location on all of the ovens. That last bit was due to the oven doors putting pressure on the wire for the thermocouple thus causing it to move about a bit. That being said, the thermocouple was always within about 3 inches of the oven doors and about half way between the interior's top and bottom.
I would have just affixed the thermocouple to one spot in the ovens, but I was concerned that anything I would use would cause the reading of the thermocouple to change. Of course, at no time was the thermocouple ever touching anything more than the door liner and possibly some of the door frame as it exited to the outside world.
Tests were conducted with a K-type thermocouple, a 90A shunt resistor, and a GDM-9061 DMM. The DMM is within calibration. All ovens were set to regular bake mode, though some support convection mode. All ovens were set to heat to 350F (~176.7C).
When heating, the ovens all pulsed their elements. The only exception was the Superba. Considering the lack of temperature accuracy you will soon see in the histograms, its unclear why they didn't just hold the temperature steady by pulsing the elements instead of waiting to turn the elements on until the ovens fell below a certain temperature and subsequently raising the temperature well above 350F before switching the elements off.
The temperature which the ovens reported was often off from the actual temperature. Differences varied, but I noted up to 65F more or less), difference between the temperature observed with the K-type thermocouple and the Superba's own thermometer. Alas, I don't have the equipment to readout the DMM and the ovens' own thermometer in realtime.
One would expect some degree of variation between a thermometer in the back of the oven by the heating element and the front middle of the oven, but 65F is a bit much. One would also think that a thermometer sitting just underneath a heating element would read high, but oddly enough the ovens actually all read their temperatures too low.
To be clear, the thermocouple was placed into an ice bath and then boiling water prior to these tests and it did read 0C and 100C as expected.
While I'm commenting on the oddities of these ovens, I should note that the Superba signaled that it was up to temperature at 300F (~148.9C). And that is the temperature according to the thermocouple. I did not witness this behavior with the GE, or Bosch, or Frigidaire ovens.
Now when you test a "smart" something, life becomes more difficult, because you're trying to get an expected behavior out of a machine that now has far more possibilities for it's operation than it would otherwise possess. Such is the case with the Bosch oven, which preheats to 389F (~198.3C), cycles once, then allows the temperature to drop below 300F (148.9C), at which point I opened the oven door to simulate someone putting something in to bake; which triggered the oven to begin cycling normally.
In no particular order:
The GE drew:
2306.7W on bake.
3499.3W on bake and high broil.
The broil element was periodically switched on in bake mode. Its unclear why.
The oven initially preheats to 427F (~219.4C).
I'm of the opinion that they expect you to open the door and let some of the heat out, so some ovens, such as this one, preheat well above the selected temperature.
It cycles at 341F - 371F (~171.7C - ~188.3C).
Variation 30F (16.6C).
The Frigidaire drew:
2127.2W on bake.
3448.5W on high broil.
There is no preheat max.
It cycles at 347F - 380F (175C - ~193.3C).
Variation 33F (18.3C).
The Kitchen Aid Superba's power draw wasn't recorded due to the circumstances.
The preheat behavior was already discussed above.
It cycles at 339F - 391F (~170.6C - ~199.4C).
Variation 52F (28.8C).
The Bosch HBL8453UC oven drew:
4555.47W on bake.
5346.7W on fast preheat bake. If the voltage would have stayed at 246v, it would have drawn ~6Kw!
4097.5W on high broil.
The preheat behavior was already discussed above.
It cycles at 324F - 396F (~162.2C - ~202.2C).
Variation 72F (40C).
As you may have noticed, of all these ovens, the modern "smart" Bosch had the greatest temperature variation at almost ~2x the average of the other 3. One wonders if they *un*-improved anything else. Like, it's expected lifespan...
And now for what you've all been waiting for, the histograms of the ovens cycling! They were taken as the ovens heated after cooling down. Unfortunately, I was of the option that 340F - 360F (~171.1C - ~182.2C) should be more than a big enough window to capture their cycling behavior. This proving to be false, I had to adjust each histogram for each oven as each proved too be insane in it's temperature control. Ordinarily, I'd have redone everything so that all of the histograms lined up with each other, but such was not to be.
GE:
Frigidaire:
Superba (I got too many bins on this one.):
Bosch:
Well, I hope you enjoyed!
Disclaimer: the image of the Bosch oven is used under the fair-use doctrine. It doesn't belong to me.
There were no markings indicating the model numbers so I have no idea what models they were. One was a Bosch model HBL8453UC, still being sold in 2024. One was a GE, about 10 years old from what I understand. One was a Kitchen Aid Superba, significantly older than the GE. And one was a Frigidaire of unknown age. They were 30", 27", 24", and 27" respectively.
I was, unfortunately, unable to do a teardown of these ovens. I did get to take a peek inside of the GE one. I saw the oven's capacitors were Rubycon and it used relays for switching. The other ovens, based on the sounds they made, all used relays for switching, which I find to be a bit on the antiquated side for the WiFi capable Bosch oven.
Incidentally, if you ever want to test an oven, please bear in mind that it will take a more than 2 hours bear minimum. Of course, the oven must also be willing to participate, otherwise you'll end up spending even more time trying to outsmart a "smart" oven.
It should be noted that all of these ovens output their temperature in imperial units. Though it may have been possible to switch them to metric I had no manuals for them, so I did not attempt it. Therefore you'll all have to enjoy the imperial readings I present to you below though I did convert many of them to metric for your viewing pleasure. Incidentally, dividing an imperial unit by 2 and subtracting 16 will typically get you close to the metric value.
As I said, the test conditions were not laboratory. The particulars that come to mind are that I had only one calibrated DMM to test with, the incoming AC voltage varied between ~231v and ~246v, the temperature outside of the ovens was uncontrolled for, as was wind direction and speed, and the K-type thermocouple's position was not fixed to one location on all of the ovens. That last bit was due to the oven doors putting pressure on the wire for the thermocouple thus causing it to move about a bit. That being said, the thermocouple was always within about 3 inches of the oven doors and about half way between the interior's top and bottom.
I would have just affixed the thermocouple to one spot in the ovens, but I was concerned that anything I would use would cause the reading of the thermocouple to change. Of course, at no time was the thermocouple ever touching anything more than the door liner and possibly some of the door frame as it exited to the outside world.
Tests were conducted with a K-type thermocouple, a 90A shunt resistor, and a GDM-9061 DMM. The DMM is within calibration. All ovens were set to regular bake mode, though some support convection mode. All ovens were set to heat to 350F (~176.7C).
When heating, the ovens all pulsed their elements. The only exception was the Superba. Considering the lack of temperature accuracy you will soon see in the histograms, its unclear why they didn't just hold the temperature steady by pulsing the elements instead of waiting to turn the elements on until the ovens fell below a certain temperature and subsequently raising the temperature well above 350F before switching the elements off.
The temperature which the ovens reported was often off from the actual temperature. Differences varied, but I noted up to 65F more or less), difference between the temperature observed with the K-type thermocouple and the Superba's own thermometer. Alas, I don't have the equipment to readout the DMM and the ovens' own thermometer in realtime.
One would expect some degree of variation between a thermometer in the back of the oven by the heating element and the front middle of the oven, but 65F is a bit much. One would also think that a thermometer sitting just underneath a heating element would read high, but oddly enough the ovens actually all read their temperatures too low.
To be clear, the thermocouple was placed into an ice bath and then boiling water prior to these tests and it did read 0C and 100C as expected.
While I'm commenting on the oddities of these ovens, I should note that the Superba signaled that it was up to temperature at 300F (~148.9C). And that is the temperature according to the thermocouple. I did not witness this behavior with the GE, or Bosch, or Frigidaire ovens.
Now when you test a "smart" something, life becomes more difficult, because you're trying to get an expected behavior out of a machine that now has far more possibilities for it's operation than it would otherwise possess. Such is the case with the Bosch oven, which preheats to 389F (~198.3C), cycles once, then allows the temperature to drop below 300F (148.9C), at which point I opened the oven door to simulate someone putting something in to bake; which triggered the oven to begin cycling normally.
In no particular order:
The GE drew:
2306.7W on bake.
3499.3W on bake and high broil.
The broil element was periodically switched on in bake mode. Its unclear why.
The oven initially preheats to 427F (~219.4C).
I'm of the opinion that they expect you to open the door and let some of the heat out, so some ovens, such as this one, preheat well above the selected temperature.
It cycles at 341F - 371F (~171.7C - ~188.3C).
Variation 30F (16.6C).
The Frigidaire drew:
2127.2W on bake.
3448.5W on high broil.
There is no preheat max.
It cycles at 347F - 380F (175C - ~193.3C).
Variation 33F (18.3C).
The Kitchen Aid Superba's power draw wasn't recorded due to the circumstances.
The preheat behavior was already discussed above.
It cycles at 339F - 391F (~170.6C - ~199.4C).
Variation 52F (28.8C).
The Bosch HBL8453UC oven drew:
4555.47W on bake.
5346.7W on fast preheat bake. If the voltage would have stayed at 246v, it would have drawn ~6Kw!
4097.5W on high broil.
The preheat behavior was already discussed above.
It cycles at 324F - 396F (~162.2C - ~202.2C).
Variation 72F (40C).
As you may have noticed, of all these ovens, the modern "smart" Bosch had the greatest temperature variation at almost ~2x the average of the other 3. One wonders if they *un*-improved anything else. Like, it's expected lifespan...
And now for what you've all been waiting for, the histograms of the ovens cycling! They were taken as the ovens heated after cooling down. Unfortunately, I was of the option that 340F - 360F (~171.1C - ~182.2C) should be more than a big enough window to capture their cycling behavior. This proving to be false, I had to adjust each histogram for each oven as each proved too be insane in it's temperature control. Ordinarily, I'd have redone everything so that all of the histograms lined up with each other, but such was not to be.
GE:
Frigidaire:
Superba (I got too many bins on this one.):
Bosch:
Well, I hope you enjoyed!
Disclaimer: the image of the Bosch oven is used under the fair-use doctrine. It doesn't belong to me.
