[ger21]

I have an instant electric water heater, and fried some resistors on the control board. I had a spare board, but fried them both. I think I need to replace the triacs??, as there are really no other parts in it. I'd like to see if changing these resistors will get me back to full power, as right now, it's only working at half power and it's just hot enough to shower. The coming cold weather will drop the water temp, and it won't be hot enough.

I know that the two (R3, R5) are 1K ohm and it looks like R1 may be bad as well. I think it's 10K ohm. When I try to measure it, I get a correct reading for a half second then it starts jumping around.

Thanks.

[ViperTX]

Gerry,

If you have a soldering iron and some solder wick you can change them yourself....wick some of the solder...and using a pair of tweezers lift one end of the resistor up....if you have a friend and another soldering iron you can accomplish this task much quicker.

Unless you have some sort of active control circuitry....you probably just have a diac which is used to trigger the triac....if the triacs are operating at all (indication by the half power mode) they are not the problem.

Paul

[Al_The_Man]

They appear to be MOC3063? if so these are triac opto's and the resistor is connected to pin 4 which is the output side to a external triac, I would imagine.
You may also have a open external triac.
http://www.fairchildsemi.com/ds/MO/MOC3063-M.pdf
In a pinch you could replace the SMD with miniature axial resistors, although it won't be pretty!
Al.

[ger21]

You may also have a open external triac.

Yes, that's what I was talking about. There are 4 that turn on the heating elements. I believe those are what caused the resistors to go bad.

Yes, I thought about just using regular resistors, as those SMD's are so small I can barely see them

[Al_The_Man]

If the resistors have also been stressed then you may also have to replace the opto's as they only take 60ma on the output triac, and the resistors are in series with them.
If you need to test the outboard Triacs, you should be able to turn them on by connecting a resistor of the same value as the overheated ones placed across the two output terminals.
They will be turned full on however.
Al.

[ViperTX]

Gerry, how many heating coils does you heater have? I would say that you would have one triac per coil in which case you should be able to tell which heating element is not turned on which would point to the triac or the control circuitry.

Thanks,
Paul

[ger21]

Edit: I already know that at least the resistors are bad.

4 coils

Yes, there is one per coil. Originally, one Triac burned up, and they sent me a new one with the new board. Although I think the board was still good.

A few months ago, I was seeing similar issues to when it burned up before. The unit uses a flow switch to turn on and off. What started to happen, was it didn't shut off when the water stopped. You could hear the water stat to boil until the thermal overload tripped. (Each coil has one). One of the triacs shows signs of getting very hot. (brown wires). They're only about $9 each, so I was going to replace them all and see if it works if I swap out the resistors. I was poking around in there, and when I turned up the temperature (pot connected to the board), I saw the resistors burn with a quick bright flash. Changed the board, and saw it happen again. It's been running for a few months now at about 50% power or less. I'm a little unsure how to check the triac, as the terminals have heat shrink covering most of them, and I was afraid of the probes touching and having 50 amps of 220 at my fingertips. Actually, there are 2 50 amp breakers, so 50 amps for two elements.

I bought the heater 8 years ago, and it had a lifetime warranty. Until they went out of business. A new one is $800

Do you guys think that replacing the resistors, optos, and triacs should do it?

[Al_The_Man]

I would also investigate fitting a higher rated triac and a decent heat sink, what is the part number?
If you change the resistors, you should not damage anything by testing to see if the opto's are OK.
If they are open, you will have no control, if they are shorted it will just turn on the heater triac with no control.
It sounds like they may not be phase angle controlled for heat adjustment, the unit probably switches each heater full on in turn.
Al.

[ger21]

I would also investigate fitting a higher rated triac and a decent heat sink, what is the part number?

I'll open it tomorrow and check the #, and post a pic. They are mounted to the cold water inlet pipe, so they get plenty of cooling.

[Al_The_Man]

I am surprised they they would heat up if they are water cooled?
Especially with a resistive load.
Al.

[ger21]

Al, they are Q6040J7.

These are the right ones, right?
http://www.mouser.com/ProductDetail/...J%252bQNeZk%3d

[Al_The_Man]

That looks like the ones.
If you want to test them you can usually use a 12v/24vac transformer and connect the triac in series with a lamp, and use a 75/150ohm resistor from MT2 to gate to turn it on and off.
Maybe also check the heater resistance to make sure they are all the same.
Al.

[ger21]

Al, is there a way to determine the wattage and voltage ratings of the resistors?

[Al_The_Man]

If these are 1k resistors, they should be ample, these are zero crossing opto's there is normally very little current and when the power triac conducts there is very small volt drop across it and the subsequent opto, I suspect the problem most likely is if or when one power triac goes open, this is placing a higher than normal voltage across the opto and its series resistor.
Did you managage to calculate the wattage rating of the elements, these are 40 amp triacs.
Al.

[ger21]

The heating elements? They have the wattage rating on them.

I mean for the resistors, do I get 1watt? 1/4 watt? ???

[Mariss Freimanis]

Gerry,

The resistors appear to be 0805s, making them 0.125W parts. The markings on the unburned ones seem to read "102" making them 1K 5% resistors.

I assume you don't have a hot-air rework station so the best bet on removing them is to "tweezer" them off use two small pencil-point soldering irons. If you have only one iron, a trick you can try is to use some rosin-core solder and melt a decent size solder ball on top of the resistor. The solder ball must be large enough to cover both ends of the resistor. Lay the iron flat across the resistor so you heat both ends of it simultaneously. The resistor should stick to the iron when you lift it away from the board.

Clean the pads with solder wick afterward and place the new volunteers with a pair of tweezers. Solder one end, then the other.

Avoid removing the resistor by wicking the solder off first. The resistor will still be firmly held in place by the solder you cannot wick off. Do not try to lever the resistor off or remove it with needle-nose pliers. You are almost guaranteed that you will delaminate the solder pads from the board should you try. If you insist on trying anyway, make sure you only apply a torsional shearing force to the resistor. Do not attempt to lift it off of the board.

Mariss

[ger21]

Thanks Mariss. I would have never gotten them off with the plan I had.

[Mariss Freimanis]

Gerry,

Just for fun, another method is to use Bismuth or Indium solder. Both metals will lower the melting point of ordinary solder to below 100C when alloyed with these metals. Both are used to de-solder difficult parts at very low temperatures, sometimes as low as 60C. Indium solder costs $300 for a 3-foot length of 0.03" diameter wire solder. I have some in my lab but we use it very sparingly, like almost never.:-)

Mariss

[Al_The_Man]

Gerry, One method I have used in the past where boards combine surface mount and through-hole is to if the board has double sided pads for the through hole components is to snip the SMT res off and solder the ends of a normal axial resistor, in this case either 1/8w or 1/4w directly to pin 4 of the opto pad and the other end to the terminal pad on the Reverse side of the board.
This essentially bypasses the trace on the top side, the resistor is mounted to the underside of the board.
I am assuming the other end of the resistor goes to the screw terminal?
I mentioned the heater wattage as it appeared strange that 40amp triacs were failing on a resistive load.
Al.

[amplexus]

Gerry,
there are several sites on the web with instructions for a cheap diy hot air rework iron I think hack a day may have been one of them. if you do much with smt they will save you a lot of grief.