Was sup fellas, I have a tankless electric water heater and this thing is killing my wallet. It's a Rheem rte18 and it's 2 months old and I've replaced the circuit board numerous times. It would be a blessing to have some assistance with this thing. I have pics http://s1274.photobucket.com/user/jophillips350/library/water heater

 

And the question is?

 

Numerous times in 2 months and no warranty?
Can you identify the area it is failing in?
Max.

 

Sorry. The circuit board fails. I would like to build one that will last.

 

I can't board looks normal. Maybe a design flaw?

 

OK. This is the safety lecture.
You better be real sure what you're doing,
Shock hazard, scald hazard, tempering valve, etc. etc.

However, I would really like to see you get your money back. Other manufacturers have done this very well. You shouldn't have to put up with this.

 

. Maybe a design flaw?

I would think so!
I would have contacted the manuf by now.
Rheem is reputable make.
Max.

 

I have called them all I get is the run around. I bought it on eBay. I saw a post on another guy having trouble but mine is a Lil bigger unit. Right now I'm jumping it out to get hot water. Has to be safer than me doing that. I just don't have the extra money to replace. Rheem just wants to avoid a refund.

 

I am thinking that is an older model with the through-hole components?
Many now are surface mount.
At least it is easier to fix or reverse engineer if you wanted to.
Are there any symptoms?
Or just stops working?
Max.

 

Looking at the photos, it seems there is a lot of corrosion on the solder, and it looks like the board may have been wet. I would remove the three ICs, and clean the pins. I would try as best I could to clean the connectors in the sockets, and re-seat the ICs in the sockets.

Sometimes, just removing and re-seating ICs in their sockets can clean enough corrosion off the pins to restore function. Be careful to put the ICs back in the sockets in the same orientation as they were before removal.

If you have a replacement LM324, that might be worth a try.

 

The S6065 SCR are now no longer Teccor but Littlefuse so that gives an indication of the age, if they are the culprit, they could be upped in size.
As power devices I would suspect those and any power supply regulator IC's that may be under rated.
Max.

 

Did you see this, smaller model but may apply.
Max.

 

i can't post new thread , please help white screen appears after i click submit . do i have problems with this forum?

 

I am thinking that is an older model with the through-hole components?
Many now are surface mount.
At least it is easier to fix or reverse engineer if you wanted to.
Are there any symptoms?
Or just stops working?
Max.

It just stops. It has a green led for standby mode and red in use led. When it stops working it just stays in standby mode.

 

Did you see this, smaller model but may apply.
Max.

Yes I went through it and all measurements on elements flow switch and voltage is correct. It seems to have a issue with the control board

 

If it was me I would first start out by touching up every single solder joint with new solder.

The simple trick for doing that quick and clean is to melt it from the backside and blast it though from the front with caned air and the tiny straw that comes with the can.

If that fixes it I would clear coat the whole board in enamel or acrylic spray paint to make it moisture resistant. That's basically all they do to circuit boards that have to survive in extremely damp and dirty conditions and it works well.

 

I did pick up a Lm324 at Radio Shack but it didn't fix it

 

I did pick up a Lm324 at Radio Shack but it didn't fix it

There really aren't many components to go wrong. Apparently, it's not the op-amp. There are two optocouplers, but they are fairly durable. What about the sensor itself? I presume it's a thermistor or maybe a thermocouple. Have you googled the part number to see if there are others reporting trouble?

 

I did pick up a Lm324 at Radio Shack but it didn't fix it

You need to try an identify the failed components.
First check all the voltage points for correct voltage, e.g. the 7806 is a 6v regulator etc.
Then check individual components, either by resistance test or bread board test.
You are fortunate that most are socketed.
Max.

 

Looking at the pictures I would say that R1 and C3 power the board. The silkscreen on the PCB for the zener diode is incorrect. A 1n4742a is a 12v part, not a 1n4740a 10v part. Might have something to do with it. The 100k resistor next to R1 looks to be getting really warm... enough to change the color of the paint on it.

Other than the zener diode, the two MOC3063 optocouplers, the LM324 op-amp, 78L06 voltage regulator, and it's 220uf filter capacitor there isn't too much stuff on that board that should blow up. One thing though is the electrolytic capacitor. That is a Chang, one of the cheapest of the cheap caps. Seeing it on that board would indicate that it is made by some low cost fly by night PCB manufacturer in China wherein the origin of a lot of components is rather dubious. I actually wonder if those MOC3063 opto's are in fact real.

If it were me I would rebuild one of my dead boards by removing the LM78L06 and zener diode and replacing them with machined pin headers. This way you can easily drop in all new semiconductors, or as I call it "take the shotgun approach". Also get rid of the Chang electrolytic capacitor and replace it with a 105C rated Nichicon VZ series or equivalent. Should cost $3-5 for parts minus shipping costs. If replacing "everything" on the board fixes it, then you can start putting back old components until you figure out what has died. Not really the most efficient way of troubleshooting but saves you from having to poke around on a live board connected to mains.

If the repaired board dies again in another month or so, and it is the same component that originally let out the magic smoke then you have a design flaw.