So, I've got this heater, here's what is on the label: Ceratech. Model KHT-A15ST. AC120V 60HZ Up To 1500W.
Kuo Horng Electric and Co LTD, Made in Taiwan.

Problem is that the fan never comes on. I can't find much online about it, there certainly are not any schematics. Some interesting ways to use OpAmps like the LM324 that this heater uses.

I've checked the heating elements, they work when the switch is on Heat. The fan never comes on in any mode. I've connected the fan directly to 110v and it works great. Realized one side of the fan connects to a triac (BCR1AM), I jumped the triac pins 1 and 2 and the fan came on.

I've tested all of the components in circuit with my multimeter, all diodes seem to test OK, the 47uf Capacitor tests OK, the resistors seem to be OK. No burnt marks on the board or anything physically obvious.

I get 110V at the circuit when the fan switch is on and the unit is plugged in.

I've measured the voltage between pins 4 and 11 of the LM324 and I only get about 1/2 volt. I have to think that chip needs more voltage to run, so perhaps the Zener has something to do with it. However, it tests OK on my multimeter on diode test (.7 Volts, reverse is infinity). I'm not sure what value it is though, I think I can see 2 and 24 on it.

Seems like there should be some better way to feed the electronics on this thing, no rectifiers or transformers on it. Just a few random diodes and that one Zener. I'm not really sure how the circuit operates.

Anyone have ideas on what to test next or a different way to approach this?

Thanks for your input.

DG

 

Can you check the voltage across that zener while things are powered up? (Be careful!) I think you're on to something. It may have failed. If you have any qualms (you should), desolder one end of it from the board and look up how to test a zener.

 

Checked voltage across zener while circuit was powered up: 1/2 VDC across it.

I also connected a known good zener (7 volts) to my 16volt variable power supply with current meter and voltage connected as a test, got about 7 volts as expected. Connected this zener in the same manner and couldn't get any voltage reading at all. It seems suspect, but I'm not sure what value it should have. Since it has a "2" and "24" printed on the Zener, I wonder if it is 24 volts? Note that the electrolytic capacitor is 47uf and 35 volts. I saw in the specs that the LM324 can work up to 32 volts, but the minimum seems to be 16v. I could be reading that wrong as it has a +/- by the 16 for minimum voltage.

Anyone know if a 2 and 24 on a zener indicates 24 volts? I'd love to order one and try it.

Thanks Wayneh for the guidance.

 

Can you get it under a dissecting microscope or a good magnifying glass? Sometimes you can see a lot more under good lighting and magnification. Otherwise you might have to wing it.

It's worth pointing out the sad truth: The zener might have failed because something else failed. It'll be very cheap and easy to try just replacing the zener, but just be prepared for the worst while hoping for the best.

 

Here's two pictures. I finally got to use that "usb microscope" that I picked up for cheap a couple of years ago.

Just for fun, I put my 7v Zener in it's place. I measured across that Zener and got the same 1/2 volt across Zener.

If this is in fact a 24volt, I could just put three of my 7V Zeners in series and that might be close enough.

There are so few components in this thing, I keep thinking I'm overlooking something obvious. I've seen where you put something under a load and it quits working, even if the voltage is there. But a simple circuit like this doesn't seem like much of a load .

 

I agree that this was likely a 24V zener.

It's disturbing that your 7V was also "shorted" out when put in place of the 24V one. Maybe I'm not understanding your statement - did you mean that you put a 7V into the PCB to replace the 24V one, and saw only 0.5V across it?

How did you test that blue electrolytic, in place? I'd replace it for good measure.

This thing uses a transformer-less power supply, where AC is coupled - probably through that green capacitor, is regulated to 24V by the zener and another resistor, and is stored in that blue capacitor, for supply to the LM324. That's my guess, anyway. I'd say the first goal is to try to get that power supply working again. This will be difficult unless you can sketch out the circuit. It would be tough from the photo but it looks like you should be able to do it.

 

I did mean to say that I removed the 24V zener and soldered in my 7v zener. Did the same test. Put meter across zener and fired the circuit up. Observed the voltage start out small and within a few seconds rise to settle at about 1/2 volt. I imagine that short observable rise is due to the capacitor charging.

As for the electrolytic, I did replace it yesterday with a 47uf 50Volt. No change. My multimeter does have a capacitance measurement, it did test at around 47uf, but I replaced it anyway.

I guess I can try to sketch out the circuit, maybe that will help in understanding and troubleshooting

Although, if I replace the zener, cap, and resistor, there isn't anything left to replace!

 

It's easy to check the resistance of every resistor on the board. Worth a look with the power off. In my experience they will most often show their nominal resistance despite being in-circuit. Any that do not match their nominal value can be double checked by pulling one end out of the board.

Did you check that green cap?

I'd also be tempted to disconnect all those jumpers and see if the board works on its own.

 

I'll check all the resistors shortly. Did not check that green cap, I'll do that too. As for jumpers, do you mean the wiring harnesses?

 

Ugh. Everything seems to be in order. Some resistors were a little off, but when I removed on leg and tested, all was well. Green cap tested good. I'm beginning to wonder if that LM324 is bad and pulling the voltage down somehow. Not easy to desolder and remove either.

I'll work on sketching out the circuit. Wish there was some software program that was easy to use for that.

Where can I learn about transformer-less power supplies that feed that LM324?

 

If you really suspect the LM324, you could clip one of the power pins and see if voltage comes back to the board. That's not ideal, since it's not cool to have power on the inputs without the power pins connected, but it could be a last resort experiment. If you can't desolder it, you could clip all the pins and poke them out one at a time.

I'm not sure where to point you to the power supply description. We don't talk about them much around here because it's a forbidden topic for DIY projects. (I believe repairing a commercial product is OK. I guess we'll see.) I'll see if I can find a link to what I suspect.

 

I'll desoldered and removed the LM324 but now I get virtually no voltage across the diode. I wonder if I had an oscilloscope and probed around if I could find something interesting. I should just give up and trash it, but the challenge has me hooked. I replaced that rectifier looking diode, but there was no change.

 

Is there line power at the board? I had assumed so but it's worth checking. Are all those jumpers still attached? Have you had any luck tracing out the board?

Sorry - all questions and no answers.

 

Yes, line power at board. I checked from the LED lead to ground and observed 65Vdc. Disconnected and also reconnected all wiring harnesses, same game. I started tracing out the board, but ran out of time. I thought I would hold it up to a light and take a picture, it's cool that you can see the components on one side and the traces on the other side.

Know of any software that could help with making a schematic?

Thanks for the help, it's fun to bounce thoughts off someone with more experience than I have.

 

I started sketching out the circuit. See attached pic. It's not complete but had to start somewhere. Also took a pic of the board showing traces on back. And my chicken scratches on the back of the board.

 

Know of any software that could help with making a schematic?

I usually do it by hand first. I draw a map of the components and then start making the connections by doing what you described - looking through the board - and flipping it back and forth. It's a pain but this board is not so bad.

If you have a computer drawing program, you might make yourself a nicer map. You could import a photo of the board into a lower layer and draw your picture on a layer above it. You could also import a picture of the back, then flip it to overlay over the components. I've never tried that but now that I say it, I wish I had.

 

Ah, cool, I responded before seeing your drawing. I believe you have the polarity of the LED wrong but otherwise I think you're getting it. And the "+" sign on the AC is on the wrong side. The upper one should be shared with board ground, I think.

From this drawing, you might try grounding the side of R11 that faces away from the triac, or applying +24V to it. That's what the op-amp would do to turn the fan on or off.

 

Nothing happens if I ground the triac. If I connect the triac's anodes together, the fan comes on as expected.

Maybe just the triac is bad and all the readings I'm getting with the zener and what-not are misleading. I expect to see DC, but there is so much AC on this board, it's throwing me or my meter off?. Still, seems like I should be able to measure some reasonable DC voltage between V+ and V- on the chip and that just is not the case.

 

Interesting, I just looked up the specs on the triac. It's a BCM1AM 8. The spec seem to indicate the trigger voltage is 6 volts. I suppose if I connect 6 volts to it from an separate 6 volt supply, it should turn on the fan? Makes me wonder if the 24 volt assumption about that zener is all wrong.

 

Yes, I think that's worth a try, as I said in #17, but using 6V based on the new info (9V battery would likely be fine). See if the triac is working.

But even if the triac is shorting the input coming from the op-amp, it's doing it through R11. The op-amp should be able to drive that small current without its supply voltage and power being pulled down. So a bad triac would only be a partial answer.