3
\$\begingroup\$

After a few good years, my CrockPot gave up. It looks to me that there may be something wrong with electronics, and being and Arduino enthusiast, I thought I could re-build it. The symptoms are that the whole device switches on but whenever the actual program is set, it just switches itself off again (goes dark).

From what I can see there is one heating element and a temperature probe attached to the main body (white wire):

enter image description here

.. another view at the connections:

enter image description here

The main circuit board features connection of the temperature probe (NTC1) and a transistor to switch the heating element on/off, as well as all of the buttons, display and power processing components:

enter image description here

The markings on the main driver are can seen on a closeup here: (probably this one: http://www.nxp.com/documents/data_sheet/BTA208X-600E.pdf) Closeup on the transistor

Rear view of the circuit board:

enter image description here

Is my assumption correct that there is probably something wrong with the controller and the heating elements are most likely fine & can be reused?

My main hurdle is to understand exactly what the circuit may look like in this case and how much of it do I need to re-create for basic control. I am happy to design the control/arduino part but for exact understanding what are the requirements of the heating elements, that is where I need help!

Update

I've now also measured resistance between the red and black wire (heating element) - 250Ω. The thermistor seems to be clocking at 13.8kΩ.

\$\endgroup\$
  • \$\begingroup\$ I wouldn't be surprised if the crock-pot was open loop. \$\endgroup\$ – whatsisname Jan 6 '15 at 23:31
  • \$\begingroup\$ Too many assertions which may not be true. If not 100% certain say what you see, not what you conclude. eg 3 elements- how do you know. Looks like 1. My be 3 but why do you think so.| Mo dedicated temperature sensor - PCB says "NTC1" = just that and white-white wire through metal loop is about 103% likely to be said NTC | REMOVE MAINS. Get Ohm Meter (DMM) and measure heating element(s) resistance . OC = dead. 3 large caps on board are ALL different value and probably part of input filter. | Main switch device (3 leads - readable by you - possibe but harder from photo) may well be dead. | .... \$\endgroup\$ – Russell McMahon Jan 7 '15 at 0:24
  • \$\begingroup\$ .... Photos are good - much better than usual. Well lit and in focus. || TRacing out circut and giving us a rough diagram would help. These are quire simple overall. If element(s) dead t's probably history. Anything else is probably fixable. | Appliance repair is off topic here BUT making a new controller isn't. \$\endgroup\$ – Russell McMahon Jan 7 '15 at 0:26
  • \$\begingroup\$ That thing with the red wires in your hand is just a closed end crimped connector (a splice) \$\endgroup\$ – Spehro Pefhany Jan 7 '15 at 0:30
  • \$\begingroup\$ @RussellMcMahon Thank you very much, will provide the further detail! You are correct, my previous conviction that there is no temperature control blinded me from seeing that the white one really most likely is a thermistor. Will measure and provide more info when I get home! \$\endgroup\$ – petr Jan 7 '15 at 9:54
1
\$\begingroup\$

I can think of a number of ways to answer your question. One source of information might be a service manual (if one is available), but perhaps an even easier source of information is the US Patent Office. For example, US patent 6573483 is a patent for a slow-cooker accessory that shows both a thermistor and a triac for controlling the heating element. That particular patent is rather light on details, but you might look at the referenced patents for more information if you need it.

As for details, triac manufacturer application notes are likely a good source of information for redesigning this circuit. In particular, you may find Power Control with Thyristors and Triacs from NXP useful for this purpose since it describes not only the basics of triac operations, but also a specific section titled "Design of a Time Proportional Temperature Controller" describes a control algorithm that is likely to be of interest.

Finer control of the temperature, including temperature ramp and tight regulation, might be an application worthy of throwing a microprocessor at it!

Update

I would be skeptical of the version 1 module mentioned in the comments, but it appears that some of the problems are address in version 2, including making the traces on the PCB wide enough for the expected current. Before proceeding too much farther, it would make sense to measure the AC current used by the heating element and/or seeing if there are markings on it indicating the current. Since the original triac was rated for 8A, I would be cautious about substituting one with a significantly lower current rating (as with the BT136S that comes on that module.)

\$\endgroup\$
  • \$\begingroup\$ Sadly, there is no easy way to obtain service manual. I was thinking along the lines of obtaining this: inmojo.com/store/inmojo-market/item/digital-ac-dimmer-module and driving it via Arduino - does this seem reasonable? \$\endgroup\$ – petr Jan 7 '15 at 15:50
  • \$\begingroup\$ I don't think that module will work. The original triac is rated for 8A and is specifically a hi-commutation triac. A surface mounted triac in a SOT-423 package seems unlikely to be up to the job. \$\endgroup\$ – Edward Jan 7 '15 at 16:00
  • \$\begingroup\$ .. I know that I may not be able replicate precisely the inner workings but would the dimmer not produce similar effect, even though through simpler inner workings? Regarding the load - the whole appliance is rated at 200W, and the dimmer claims to support up to 1000W. (I also plan to mount the outside of its original position to not be affected by the heat as the original controller was) \$\endgroup\$ – petr Jan 7 '15 at 16:05
1
\$\begingroup\$

This looks like fun :-).

This will be a rolling answer - or hopefully others will join the fray - I'm about to have pieces hacked out of me early next week * and will hopefully be back online some days later. Hopefully ... .

Where are you?
Looks like your mains is 230 VAC.
Power = V^2/R.
For 230 VAC power = 230^2/250 = 212 Watt ~= 200 Watt you mentioned.

You can check TRIAC is OK by finding trigger cct .
The diagram on page 8 here shows the genera form it will take. Controlling the trigger "switch" however it is done should turn TRIAC on.

WARNING: 230 VAC

Units like this are often not isolated between controller and processor.
Odds are the controller is all effectively at mains potential.
Try not to die.

The easiest safest way to control a mains TRIAC is to use an optically coupled zero-crossing TRIAC driver. Thse are often small TRIACS themselves. (Zero crossing means it turns on when mains voltage passes through zero so minimises EMI (electromagnetic interference) amd load on TRIAC for resistive load. For an inductive load where V & I are out of phase TRIC may be less happy, but thems the breaks. Here the element is almost certainly close to pure resistive.

The TRIAC driver I referred to above is a Liteon MOC3020-MOC3023 series driver. t is 400V rated which is more marginal than I'd like for 230 VAC.
230 RMS x 1.414 = 325V peak. Add some noise spikes and ... .
Better to use a 600V driver.

Another 400V rated part - TLP160G

Liteon MOC3063 - 600 V min, zero crossing, 5 kV isolation, $US0.63/1 in stock Digikey - a bargain. This needs 5Ma LED drive max for guaranteed switching - some need 15 mA - something to check in every case.

You'll find a zillion application circuits here.
Some samples that look OKish include

A question but looks useful

In Russian? circuit self explanatory. How hard can it be.Looks like he is having fun.

Russian again but looks good and Google translates well enough (right click - "Translate to English").
Yielding this circuit.

enter image description here

And similar here

enter image description here

http://optoelectronics.liteon.com/upload/download/DS-70-99-0019/S_110_MOC3020%20thru%20MOC3023%20SERIES.pdf


Temperature control.
Thermistor will drop in resistance with temperature.
To get a reasonable idea of performance, connect ohm-meter to thermistor with clips.
Orient unit as usual. Fill crockpot with boiling water and insert thermometer.
Record temperature and resistance as water cools.
Later on when it is heating you can do this with cooking oil (or motor oil :-) ) to record higher temperature results BUT there are more analytical ways that should suffice when you get that far.


  • C4-5 discectomy & cage fusion. What fun.
\$\endgroup\$
  • \$\begingroup\$ Update - gave up on the re-flowing of the PCB - a lot of little surface-mounted resistors under that heatshield, probably not worth the effort. Will focus now on building the "dimmer" for the heating element, as well as figuring out the right resistances \$\endgroup\$ – petr Jan 12 '15 at 17:02
1
\$\begingroup\$

Before you go trying to duplicate the controller, at least take a basic stab at fixing what you've got. You've measured the heating element and the thermistor, and those resistances seem reasonable, so they are probably fine.

One thing that jumps out at me is the single-sided circuit board. That's the one with the orange color and copper traces only on the bottom. Those things are notoriously unreliable. There is no plating in the holes, so all the leads that stick thru the board are held only by the solder meniscus on the other side. On "real" boards, they are held by a thin layer of solder making a good connection inside the cylinder formed by the plated hole, then again by the menisus on the other side. That's much more solid than what your single-sided board can do.

Take a soldering iron and re-flow all (yes, every last one of them) the solder connections on the bottom side of that single-sided board. I have seen such boards where over time the solder meniscus cracked, but it still looked fine unless you examined it with a microscope. Even then it can be hard to tell. Heat each of the joints until the solder visibly flows. Add a little solder if you can do that without making a blob.

After reflowing the solder joints, reassemble the unit and see if it works now. The fault could well be somewhere else, but a failed solder joint on a single-sided board that has been subjected to regular temperature cycles is likely enough to be worth the attempt.

\$\endgroup\$
  • \$\begingroup\$ Hmm, interesting point, for the sake of an exercise, will try it out! \$\endgroup\$ – petr Jan 9 '15 at 9:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.