Water heater Stiebel Eltron DHE - resetting after heating...

[Kromag]

Hi,

I have a problem with my ~5 years old water heater Stiebel Eltron DHE (http://www.stiebel.com.au/images/brochure/DHE-Brochure-9987.pdf) which is constantly resetting itself after heating.
At the beginning (around a year ago) the same issue was happening, but only after the heater was working with 100% power consumption - removing the valve that slows the water flow fixed that issue. Now it resets every time after heating.

The service guys said that the main board "is most likely broken and it would be better to replace this heater with another one" - blah, blah, blah...
It is just too expensive to be replaced after 4-5 years of service...

Does anyone have any clue if there are any "magic" caps inside there (as in the Samsung monitors or Yamaha amplifier that Dave fixed) or do you have any schematic I could use?
I have been looking for such information for a long time, but uncle Google gives me nothing

Any help appreciated...!

[niflheimer]

I didn't had the opportunity to open one , but usually this type of units are equipped with an overheat protection that trips out if the water or heating element exceeds a certain temperature. Given what you said it's quite probable this is the cause - either because the water flow is insuficient , causing the element to overheat or from a fault in either the heater unit itself ( partially shorted ) or the detection circuit which should be a simple termistor or thermostat relay.

I very much doubt it's the actual motherboard at fault , since that usually present itself with either a totally crapped unit ( doesn't start ) or random resets.

What you can try is :

1: Check the resistance of the heating element with a cheap multimeter- google around for what the value could be , but anything below 10-16 ohms could mean its faulty. There should be at least three of them , one for each phase.
2: Check the thermostat(s) or thermistors , since they often fail and are cheap to replace . The one used for the protection circuit is most probably separated from the one controlling the normal temperature selection. Do not bypass it though for any long term operation - you don't want to burn your house down if something happens.
3: Check the mainboard for obvious signs of damage - corroded connections , cracked traces , power contacts that wiggle .

4: Given what you said about the flow control valve and the age of the unit it's likely there are deposits inside the pipes and on the heating elements. If so the water flow decreases , causing the heat transfer to worsen - meaning the elements overheat . Depending on what needs to be cleaned it could be an easy fix or it might not be - the unit looks pretty hard to access .

And if everything else fails you can try calling another local tech guy - it's in their interest to convince you to replace it altogether and the "oh , it's the mainboard , chuck it in the bin" is one of the old excuses for "I'm not paid enough / I don't have the time to see what's actually wrong"

[Kromag]

I should make it more clear on the very beginning - it does not reset when heating, but right at the moment of closing the tap...

It can work on 100% for more than 20 - 30 minutes without a problem (until you close the tap), while it also resets when you just use it for 30% in couple seconds (and close the tap)...

So, it does not really look as a problem with the heater itself...

Too bad I cannot find any schematic for it and it is really a pain in the a** to take it apart and not break it at the same time Especially that this heater is the only one in my house :/

Anyway, I will need to check the board as you suggested - thanks!

[niflheimer]

I found some photos of it and it does indeed look like a pain to tear down.

From what you say it does suggest something wrong with either the thermistors/thermostats/overcurrent protection or loose cable connections. The cables on your unit look beefy and given the 40 amps it must be drawing an imperfect contact might drive the unit into protection/reset.

Might be worth looking if it's using relays to switch the heaters and not some sort of electronic control . If it's the former the relay contactors might be toast - cutting the power from time to time or even arcing over. If the later check the board for bad solder on the beefier cables . The mosfets might be sound since at that current if they failed you'd know ( magic smoke ! )

[Alana]

Closing tap means sudden change of pressure - maybe overpressure sensor [if exist] is acting up?

[niflheimer]

Good point - I missed the tap closing comment.

[Kromag]

Yeah, but I would not expect resetting if the pressure is too high...

Anyway, it looks like l will need to take it apart...

[Niklas]

That amount of heating power will need a large area to transfer all the heat over to the water. There is also a limit of Watts/area that you must stay under to get a long and reliable service life for the product. With a constant wall thickness of the heater tube, the total volume of heated metal is proportional to the length of the tube.

The heating is almost instant and the control electronics can not predict when the user will turn it off. When the user turns the water flow off, then the heater must be shut off immediately to not overheat. There is still a lot of energy stored as heat in the heater and its metal tube housing and without a cooling water flow, the water temperature will rise. The maximum water temperature to exit from the heater is probably about 50 degrees Centigrade so there is a headroom up to 70-100 degrees where a fire protection thermostat would cutout.

[G7PSK]

The element is hot from heating the water turn of the tap and you have no more cold water over the element so instant steam the element is covered in limescale.

[cosmos]

Could it be a build-up of deposits (from the water) on the inside of the heater ?
With water flowing it would still be cooled enough to not trip but the monitored temperature could jump a bit more than designed for when you close due to the insulation.

[Kromag]

Every office is normally a very busy place. With a water dispenser around, the staff can enjoy either cold or hot drinks anytime without refilling a pitcher in the refrigerator afterwards or boiling water in a kettle or thermo pot for a hot drink. This article provides the different types of water dispensers that are suited for a busy day at the office.

What are you talking about and how is that connected to my problem?

Could it be a build-up of deposits (from the water) on the inside of the heater ?
With water flowing it would still be cooled enough to not trip but the monitored temperature could jump a bit more than designed for when you close due to the insulation.

Nope, it has nothing to do with overheating. It is enough that you use it just for a second and you can see that problem...

Actually, I was able to take it apart (and put it back ) and I made some pictures - I will create some "review" post soon, as I need some help with identifying some of the components and they best way to measure its condition... I will publish it in couple days...

[Kromag]

OK, I can finally post some pictures...

The funny thing now is that each time I just change the flow or temperature - it resets itself... Have I messed it up? I don't know... 


Another funny thing is that it all started some time ago when this heater was fully functional and it had a "motor" attached (see Motor_circuit.jpg). That motor is (well, was) responsible for limiting the water flow in case the heater cannot produce requested temperature and was already working on 100% load. The result was that the flow was smaller, but the requested temperature was achieved. After reaching 100% load the motor starts working and at the end of the heating process - you could hear that it stops working (just like you would stop a drill with your hands after you end up drilling).
It was operating normally for 4-5 years and at some point it caused resetting the unit (only when you reached 100% load and the motor was activated) - at that point you could hear some noise from the motor and you could see that the unit resets. A workaround for that was to disconnect the motor from the board (motor impedance is around 60 ohms). Now - it is enough that I connect the motor back in and turn the power on (just the main power) - it resets like crazy...

I created a simple drawing of the circuits (see Motor_drawing.jpg - I put all the text on the small transistors to better identify them on the left side of the drawing) and also measured all the bipolar transistors - they seem to work OK, but I believe there is one MOSFET there (between the motor rails) - maybe that one is faulty? How to test it in this configuration without desoldering?

[Niklas]

The component marked with pink/violet is a BAW56, a dual diode with common anode. The other packages forms an H-bridge with red transistors as the lower legs (NPN) and the blue transistors as the upper legs (PNP). The green transistors are level shifters for the upper legs.

One way to test the bridge is to add a pullup or pulldown resistor to the motor connections and measure the voltages. Check that test points 1, 2, 3 and 4 are all low and then do the tests with the pullup/pulldown resistors. All the transistors should be off, thus a pullup resistor should result in a high voltage and vice versa with a pulldown resistor. If not so, then you have found a shorted transistor or diode.

[Kromag]

The component marked with pink/violet is a BAW56, a dual diode with common anode.

I tested the "pink" (BAW56) with a multimeter just like the transistors, so it should give me some similar results (~0.65V drop voltage on one configuration, "break" on another) right? If so - that may be the (one of many) faulty thing, since I could not get any reasonable measures from it (all the other transistors had reasonable readings)...

Anyway, since that circuit is in use only after the heater reaches 100% power, it does not say why the resetting takes place on lower heating values (unless it does some tricky things to the IC)...

Another thing there is an optotriac which is driving the main triacs directly (I believe) from the IC.
I need to find my notes in that matter, but I am pretty sure there is nothing else there...
Since the temperature seems to be correct - the triacs are working OK, but maybe there is one which is broken and once it will heat up a bit - it goes crazy...?

Unfortunately, tearing it apart means get it out from its place, so testing it on a bench is not that easy as I do not want to burn the heater...
I need to do a second try to take it apart

Last thing: Thanks mate! I really appreciate any help I can get since it starts to drive me crazy!

[Alana]

How about changing caps and protection devices on that board and checking if any of current paths from psu [i see litle switchmode] to motor driver and cpu have bad solder joints [re-soldering them even if they look good]. From what you described it may be an issue with cpu power sagging below its reset treshold.

[Kromag]

How about changing caps (...) From what you described it may be an issue with cpu power sagging below its reset treshold.

Yep - you were right, it was the power supply - i measured the voltage and I've seen that the reset pin is going low on the CPU, later on I found the same action on the +5V voltage regulator.
Finally I've replaced two caps which looked OK, but were off by 60% and could not deliver enough power to the circuit...
That would also explain why it got worse and worse in time and the little tricks like removing motor (which consumes power) helped for a short period of time...

I'm glad I could fix that, since it cost me around $3 while the guy in the workshop said $600 for the new board...

Thanks for helping!

[gspiliot]

Hi,

Sorry for resurrecting an old thread but it happens to have exactly the same water heater with exactly the same symptoms. At first I thought that the flow limiting motor was shortening the board which I found out that wasn't the case. Service here in Greece is nowhere to be found and I was instructed to buy a new one!

The finding of this thread and your solution gave me hope. Can you elaborate a bit on what capacitors you had to change? Did you by any chance took photos of them or know their rating/values? Do I need to dismantle the pipes and the heating element to get access to the board? It would be extremely helpful if you describe the difficulties in dismantling/remounting the unit or the process in general.

Thank you again for opening this thread!
George

[Kromag]

Hi George,

I will try to dig out some pictures and will also try to describe the process you need to follow to safely remove all parts...
Give me couple days

Regards,
Radek

[gspiliot]

Thank you very much in advance!

George.

[kunze]

Hi,

sorry to dig out this thread again.
I also have a faulty device and would be very happy about some pictures that show how to do the repair.

Thanks,
Michael

[Seekonk]

Bet those were 85C capacitors too.  If I remember the temperature rating is for 4,000 hours at that temperature.  We had a legacy product that a customer would only get two years of operation before it failed.  No one else had complained before because at 60hz the relay would stay closed.  This customer operated at 50 hz and the relay would drop out.  Took internal temp of unit and did the calculation out of the manufacturers catalog.  Expected failure time was exactly what the customer was seeing.

[PahanV]

Hi Radek,

I have the same problem on my devices.
LCD display flashes crazy wenn motor cable is connected.
When motor cable is disconnected, deviceis ready for use but resets every 15 seconds
Could you describe which components must be tested and replaced?
Many thanks for your help.

Regards
Paul

[Kromag]

Hi Paul,

The faulty parts were not the motor-related items (as I thought on the very beginning), but the biggest "blue" caps, which are used to power the board - they are directly connected between the main power lines in some kind of a star or triangle way (so you have some lower voltage between some certain points).

Please, look at the attached picture and examine the biggest caps (I don't remember the values - it was long time ago).
In my case - two of those were lacking in capacity and that caused some current limitations, so when the motor started drawing more power - the voltage dropped significantly and it caused the main CPU to reset itself.

You may be able to find those two by examining the board in a way: <MAIN LINE 1>--|CAP|--<point>--|CAP|--<MAIN LINE 2>.
Since these caps are always connected to the main lines - they are used all the time - not only when you use heater.

Hope it helps.

[Dr_Welt]

Hi!

thanks for the pictures and the hint on the caps. Last Weekend I had the same issue on my "B158183 Stiebel Eltron DEL 18/21/24 SL" after nearly exact 5 years of service. I checked all the MKP caps and found the two on the picture had a huge loss in capacity.
Resulting in a reset of the logic board every time the big relay in the main supply path switched after heating.

Replacement type is MKP-X2 1,0µ (reichelt germany) pin pitch 22.5mm 1.0uF/305V

Thanks, Ole.