Tankless Water Heater causing buzz in my monitors

[eq1]

^ Thanks for all the help, everyone. You've stuck it out longer than I would have expected... You'll have to forgive me if/when I just forge ahead and follow the process I'm on. At this point I'm not thinking about the noise in my monitors per se; I'm simply trying to fix the Rheem unit that I had installed prior to it failing, at which point I had replaced it with the Eco11 - which had a lot more noise/responsible for the noise in my monitors. Going back to the Rheem will at least reduce the noise; going back to my older Rheem might eliminate it, as the same model in use right by my studio has no noise.

The symptoms for my newer Rheem seemed more cut and dried than the older one, so I opted to try to fix that first by replacing the SCR. As I've now found out for sure, that didn't work, so I'm back to either fixing the circuit board or buying a replacement.

The other option at this point would be to buy a replacement 'something' for the older Rheem, but I'm not sure if it's the SCRs or the circuit board. Since the replacement SCR didn't fix the newer unit, makes me skeptical replacement SCRs would fix the older unit. If a fixed or replacement circuit board for the newer Rheem fixes it, then I'd be prone to think a replacement circuit board for the older unit would fix that one as well. But, I think I can just take the new circuit board and plop it into the old unit to test, before or in lieu of buying anything else; I think they're interchangeable. I emailed the company to verify that, but haven't heard back...

That's where things stand...

Originally Posted by eq1

Dammit, man, I just realized something: I think the thermostatic breaker on the tankless heat exchanger was tripped when I tested it after swapping in the new SCR...This means I have to install the unit again and test again - it still could have been the SCR...

So, as I mentioned above, I re-installed the newer Rheem, thermostat reset, and... it didn't fix it. So I'm back to the circuit board. I could see that indeed that thermostat/breaker was tripped when I tested it last time.

Everything 'works', it just seems to spaz-out when it comes time to turn off. There's two red LEDs that light solid when the unit is ON; when I turn the water off those LEDs flicker instead of turning OFF.

And, actually, I don't think it's ever OFF: I had to control it via the circuit breaker in the service panel. When I turned the breaker ON with the faucet OFF (no flow), didn't matter, as soon as I turned the breaker ON the unit turned ON. So I flipped the breaker OFF, went to the faucet to turn water ON, then went back to the panel and turned the breaker back ON. The unit worked, heated the water, temp could be adjusted, but when I turned the faucet OFF it kept heating, with the two red LEDs flickering instead of turning off... That ended my testing.

I'm pretty sure the mechanical flow actuator thing is fine, as is the magnetic 'reed' switch. It's a basic, robust mechanism: the 'flow actuator' is simply a plastic plunger in the water inlet, with a spring, when water flows it pushes the plunger, the plunger has a magnet attached, the inlet tube has a magnetic reed switch attached, when the plunger with magnet passes under the reed switch, the switch contacts engage and send the ON signal to the PCB. When the faucet is turned OFF, the spring on the plunger pushes it back in the tube, magnet moves away from the reed switch, switch contacts disengage...

Actually, I should probably double check this whole mechanism. I checked the old unit a long time ago, which was fine, this one I checked the reed switch and I can hear the plunger moving, it's not gummed-up or something. But I never actually tested everything in place. Maybe the magnet on the plunger 'fell off' or something?...

[eq1]

Well, checked the whole flow actuator/reed switch thing, it all seems to work fine...

[In case some stray reads this, it's pretty easy to test: use a needle to pry the screen out of the inlet tube so you can access the plastic plunger in the tube. Use a metal rod like a hex wrench to pull the plunger out (it will stick to the wrench as it's magnetic). Make sure it's all clean and that the plunger can move freely. I disconnected the leads from the reed switch at the PCB and attached an ohm meter, but you don't really even need to do that, as you can hear a very very faint 'tick' every time the reed switch contacts engage and disengage - as you move the plunger in and out of the tube and under the reed switch. That's it. Move that plunger in and out, listen for that faint 'tick', and if you hear it every time you move the plunger under the reed switch, it's fine.]

Based on my understanding of what we've covered re. the 'ICs', sounds like neither of them are programmed with proprietary stuff, that they're just off-the-shelf items. So, I decided to just order those, they're so cheap, and replacing should be easy, no de-soldering/soldering, etc. I ordered 3 of the 'opto-isolators' (MOC3063, they actually call it a "triac driver" at the digikey website), and 3 of the op amps (LM324N). Grand total with $5 shipping was ~$10. Should be here in a few days.

* * *

Originally Posted by cyrano

I'll bet you the controller is shot... These are cheap to get, but they come without the software that runs inside the controller... An opamp is easily replaced, as it has no programming. It's just one or more amplifiers in a chip...

Originally Posted by eq1

While I got this open again, for reference, the 'other chip' says:
L1503
MOC3063...

Originally Posted by cyrano

The MOC3063 is a 5 kV opto isolator. LED + photo transistor. Probably for null detection. Looks like that thing is just a fancy thermostat with (hopefully) null detection. Of course, if the opto isolator fails, it would stop the output. Unsolder and check?

[Tone Deft]

crazy ride...

at this point I'd call a meeting with all the interested parties with a big ass whiteboard.

sketch out the system on a whiteboard.
define the blocks as defined inputs/outputs (a transfer function aka gain.)
focus on the interfaces between the blocks.
identify all the system variables.
--what's their range in units of measure?
verify all the constants.
--perform a test on constants
---test your test that it's guaranteed to only fail when it should.
put in a mechanical hat
etc
put on a electrical hat
- I think in the frequency domain. think of taking an FFT (frequency domain) of power over a spectrum of frequencies and looking for spikes and harmonics to find the noise sources, just like with audio Production. it's all the same basic physics, applied over and over in different ways. nature repeats itself. you wouldn't use an oscilloscope (time domain) to find hum in an audio file, you'd use a spectrum analyzer.

heck, you can do the above with the above list and so on and so on...

(I'm unemployed, maybe retired, I dunno. having fun learning new stuff. watching AXS-Tv who's showing Clapton all day.)

or you can just put some hair on your peaches and take cold showers.

edit - check out 'one over f' noise... aka 1/f noise:
https://en.wikipedia.org/wiki/Quantum_1/f_noise

1/f noise is an intrinsic and fundamental part of quantum mechanics. Fighter pilots, photographers, and scientists all appreciate the higher quality of images and signals resulting from the consideration of quantum 1/f noise. Engineers have battled unwanted 1/f noise since 1925, giving it poetic names (such as flicker noise, funkelrauschen, bruit de scintillation, etc.) due to its mysterious nature.

how would YOU filter out 1/f noise? it goes down to DC! this gets into the resonance of the earth as an electromagentic waveguide (giant chunk of Iron with electric/magnetic fields around it.) I read a paper that it also involves how puddles form into pools of water. I've had lots of time to read.

[Allybye]

Wow that's some post! Do you not think it is getting on going off topic and a bit complicated? However......

@eq. Your relay test is good, that's great, but it might be easier to just use your ohmeter on the disconnected PCB wires whilst turning the water on and off? Real flow all aspects go or no go!

The symptoms you describe do sound like something is not getting fully turned off -like the transistor driving the opto isolator (or it is getting some noise in it's drive that then turns on the SCR in part of it's supply cycle).

This could be from mains supply getting into the drive too much. Does it flicker at supply frequency/times two? Consider changing the capacitors too!
This goes along Cyrano's failed controller theme.

Pity we do not have a circuit diagram

[eq1]

Originally Posted by Allybye

@eq. Your relay test is good, that's great, but it might be easier to just use your ohmeter on the disconnected PCB wires whilst turning the water on and off? Real flow all aspects go or no go!

Well, my unit's on the bench, so there's no water flow... I've got two other people living here, and they need their hot water, so I can't leave the unit installed yet dysfunctional for testing over periods of time...

[Tone Deft]

Originally Posted by Allybye

Wow that's some post! Do you not think it is getting on going off topic and a bit complicated? However......

we've already been over this with you and cyrano. please chill.

[Allybye]

To use your language....I'm cool.

It was only a suggestion based on my opinion! I don't recollect a previous suggestion of off topic? Maybe there was......

[Allybye]

Good point @eq !

[Tone Deft]

Originally Posted by Allybye

To use your language....I'm cool.

It was only a suggestion based on my opinion! I don't recollect a previous suggestion of off topic? Maybe there was......

if it helps... I'm bored and unemployed. I'm prone to rambling and over-thinking everything.

noise is a fun topic. with inline water heaters, I'd love to buy a hottub ASAP.

this is like an episode of This Old House. there's something to learn from all this. we're certainly beating the hell out of a simple little device.


did another search, found this
http://www.gibsonsheating.com/blog/2...-here-are-som/

Tankless water heaters will all make some noise when starting up and heating water, but the noise levels are usually reasonable and not cause for concern. When the noise occurs when there is no hot water being used, it's likely there is a vacuum that's siphoning water away from the unit and causing loud noise and vibration. In most cases the solution is to have a plumber install a check valve in the water line to prevent other plumbing fixtures from interfering with the flow of water to the water heater.

Another common problem that can create noise in tankless water heaters is a dirty flow sensor which controls how much gas is sent to the unit. Other sources of noise include blocked or inadequate ventilation or a defective or improperly set pressure valve.

/shrug

[eq1]

Tried replacement opto-isolator and op-amp, no fixum...

I made an image of the circuit board with components labeled, possibly on my way to some kind of circuit diagram (though I doubt I'll get there). I tried to trace the path/s of anything that might prevent the unit from turning off - but I just get lost, go in circles, don't know enough...

Everything works - it just doesn't turn off, and it's not due to a bad SCR, a bad 'ON/OFF' flow switch, or bad opto-isolator or op-amp...

All I can gather is that one or both of the gates on the dual-SCR should not be conducting, there shouldn't be a 'trigger voltage/current' or something like that. In other words, something on the board should be stopping the current from flowing to 'SCR G1' and/or 'SCR G2'. I just can't wrap my brain around it all, seems like it could be anything (though I'm kind of leaning towards one of the capacitors)...

Here's that labeled image:



"D1" and "D2" are red leds that are supposed to light when the unit is active, those don't turn off when the water's turned off, but instead they flicker. "D7" is a green LED that lights when the unit has power; that seems to work fine.

Most of the components are horizontal and the labels are situated between the lead ends. A few are vertical, such as "C2" and "J2," the horizontal label is situated between the lead ends. J2 is just a jumper.

"ED1" I think is just a trimmer pot for initial calibration, it's hot glued.

"POT" is the pot used for adjusting the temperature.

From my haphazard reading and research, I think this unit tries to control power by adjusting the 'phase angle' of blah blah blah, it's a 'phase-fire' control...

"L1 120V" is one leg of the house power. "L2 120V" is the other. From 'the house', one leg splits and goes to one side of each of two heating elements, and also a small wire branches off from one of those splits and goes to "L1 120V."

The other 120V leg ('from house') splits, with one branch going to one terminal on the 'dual-SCR' and the other going to "L2 120V." The remaining terminal on the dual-SCR goes to the other side of one of the heating elements, which is wired to the other heating element...

[cyrano]

I see no picture?

[eq1]

hmm, should be there, I see it...

Try this:

[cyrano]

Thanks!

This one is visible.

[cyrano]

Been staring at the pic for a while.

Suppose one of the diodes (D4, D5, D6, D8) is open, it could produce what you're seeing. I mean switching on, but not off.

Have you checked these diodes?

From what I think I see, the mains is rectified by the diode bridge to pulsing DC. A bit unusual.

A test I would consider, is moving the "noiseless" Rheem to the position the noisy Rheem is in. If the noise moves with the unit, it's the device itself. If the noise stays in place, I'd reckon both units to be noisy and the wiring just not injecting it into the studio mains.

[Tubeguy]

If monitors still buzz when they're off, they can't be off so thaty doesn't make sense to me.
But are you using ballaanced cables for the monitors? I have to use them for my DBX monitors otherwise they pickup even an ant under the carpet.

[drumphil]

What are your presonus monitors connected to, even when they're turned off, where you hear the noise? Does the noise go away if they're only connected to mains power with no audio input cables?

Is it a balanced audio cable connection? Does it make any difference if the shield on the balanced cables is disconnected?

[eq1]

Originally Posted by cyrano

Suppose one of the diodes (D4, D5, D6, D8) is open, it could produce what you're seeing. I mean switching on, but not off. Have you checked these diodes?

Have not checked the diodes, yet... I think I've been waffling in a state between wanting it fixed, on the one hand, and wanting to understand what's wrong/how it works, on the other - all while trying to minimize the time/$ I spend on it.

I'll probably check the diodes today. I haven't done a ton of soldering, though, and there always seems to be a risk of messing stuff up when I desolder!

I think I might be able to buy a completely different board for this unit - it's a newer one, mostly digital...

Originally Posted by cyrano

A test I would consider, is moving the "noiseless" Rheem to the position the noisy Rheem is in. If the noise moves with the unit, it's the device itself. If the noise stays in place, I'd reckon both units to be noisy and the wiring just not injecting it into the studio mains.

I might try that. I have assumed that the 'noiselss' Rheem by my studio is indeed noiseless, partly because when I did the AM radio test the radio didn't pick up any noise like it did with both of the ECO11s (one in garage laundry room, one 'downstairs', which is the one I'm trying to replace with the seemingly lower noise Rheem). I know, this is all confusing, hard to follow.

What you're calling the "noisy Rheem" is actually not hooked-up anywhere at the moment, it's the one I'm trying to fix to replace the really noisy ECO11. I'm not exactly sure how much noise the 'noisy Rheem' emits; for a long while I was hearing a faint buzz around my computer and monitors - when the 'noisy Rheem' was installed 'downstairs' - but I didn't put 2 n' 2 together until I installed the really noisy ECO11, realizing that the faint buzz must've been from the 'noisy Rheem'...

[eq1]

Originally Posted by Tubeguy

If monitors still buzz when they're off, they can't be off so that doesn't make sense to me.

Yeah, that doesn't make a lot of sense to me, either. But their switches are in the OFF position.

Originally Posted by Tubeguy

But are you using balanced cables for the monitors?

Yes, balanced cables...

[eq1]

Originally Posted by drumphil

What are your presonus monitors connected to, even when they're turned off, where you hear the noise?

They're each plugged-in to a power strip, power strip plugged-in to a single wall receptacle, receptacle cable (wires) go back to the service entrance 20 amp breaker that serves this 'studio' receptacle (there's a couple more receptacles on this circuit, but they're on the same wall, in the studio, with nothing connected).

Balanced cables connect the monitors to balanced 1/4" jacks on my audio interface, audio interface set to 'balanced mode'... I don't know how to disconnect the shield of the balanced cables.

Originally Posted by drumphil

Does the noise go away if they're only connected to mains power with no audio input cables?

I'm pretty sure I tested that at some point - I pulled the audio cables out, or at least one, and the noise remained. I'll have to double check that. When I unplug them from mains power the noise goes away...

edit: OK, double checked that -- noise the same even with audio cables disconnected from monitors.

[eq1]

Originally Posted by cyrano

Been staring at the pic for a while....

I think I'm about ready to desolder those diodes. I noticed that resistors R10 and R11 are discolored quite a bit - all the other resistors are nice and clean and bright, while these two are lightly charred-looking in the middle, brownish, like they've heated up quite a bit and/or frequently.

Can you tell if failure of those resistors might cause turn-OFF problems? R11 is connected to the two red LEDs - that flicker upon water shut-off rather than turning off...

* * *

For reference, I also tried to get the part numbers off the diodes. The three big black ones, D3,6,8, are 1N4004. I can't quite make out the small, clear ones. One of them, D5, I think is 1N52xxx? with a 5T underneath [edit: 1N5233B 5T]. The other, D4, I think is 1N4742..also with a 5T underneath [edit: 1N4742A 5T]. I'm likely missing characters for D5, might be missing a character at the end of D4.


* * *

OK, that didn't go so well. Try as I might, I could not get a leg of those diodes off the board in a neat and tidy fashion. I ended up cutting them off, checked them, and they all check out fine (I used the diode function on my multimeter and forward voltages looked about right, and they didn't conduct when I reversed the probes). I checked the resistors R10 and 11, in place, and they both measured about what they should be (100k Ω for brown black black orange)...

I think it's about time to buy a new circuit board...


Just want to add one more thing, again for 'reference', in case someone else stumbles upon this thread, looking for technical details, etc...

I noticed that the heating elements on this unit are rated at 208V, one's 5000W and the other is 5500W. Looking at various similar units available today (mostly the 'Titan' brand), and reading about them, I'm pretty sure that, all else being equal, these devices can utilize a range of heating elements -- 208V, 220V, and 240V, without a different control PCB. A similar Titan unit I'm pretty sure mentioned it -- that you could choose among power levels depending on your home power available. I guess some places only have 208V?? Not sure, but I have 240V output, so it seems like I could bump these heating elements up to 240V. I think that would result in faster heating at a lower current, but not necessarily more heating...


* * *

Now that I've ordered a 'digital' replacement control board, it just occurred to me that maybe my testing of the diodes wasn't specific enough? I was just looking for major failure - they have forward voltages that look about right, and didn't conduct in reverse direction. But maybe functionality of the tankless depends on something more specific? For example, two of the diodes are Zeners:

"A Zener diode is a special type of diode designed to reliably allow current to flow 'backwards' when a certain set reverse voltage, known as the Zener voltage, is reached."

I would imagine the tankless design requires the Zener functionality - that's why they're there. Obviously I didn't check what this reverse voltage is supposed to be, and I didn't check what it actually is...

* * *

I looked up info on those zeners, and I also looked for 'How to test' them - and re-tested. According to the instructions, measure resistance forward and reverse, I should see low-ish resistance forward, and high to infinite resistance in reverse.

For 'D5' I measured 11.35MΩ forward and 'OL' in reverse. The instructions say 'modest resistance' forward, "a few hundreds of thousands of ohms," and uses 450KΩ as an example of a good reading. 11.35MΩ doesn't seem to fit the bill.

When I use my multimeter's diode test, it produces a voltage of 0.606V.

For 'D4', I measured 'OL' forward and reverse, and 0.694V using the meter's diode test...

These resistance tests seem to say both of these zener diodes are bunk.

There's another voltage test I could do, by I'm too hungry to continue with this.

Here's a link to the webpage with instructions I followed on how to test a zener diode: http://www.learningaboutelectronics....-a-zener-diode

[Allybye]

Been looking to understand the circuit for you.
Can you tell me what the resistor is between the pot and ed1 -cannot read what it says!
Any idea where R2 might be? Cannot see it but it might just be the pot or ed1....

[Dark Fader]

[QUOTE=eq1;2341044]Yeah, that doesn't make a lot of sense to me, either. But their switches are in the OFF position.

It may have been covered before, but rather than trawl through all the posts I'll ask for clarification to save time, because I'm a bit confused here.
Apologies if I have completely misunderstood, just trying to help as an electronics engineer.

Are you saying the speakers are still emitting noise when they are switched off?

Does the noise go away when they are completely unplugged from the mains?

If both of those are true it sounds like the noise is a mechanical noise coming from a component in the power supply (such as a transformer or an inductor) before the switch, rather than coming from the speaker cones themselves due to an electrical noise.

[cyrano]

The picture sent me off in the wrong direction...

You can't test a zener with a typical DMM.

In this case, it's a 6V zener (from memory). So you'd need a >8V power supply and a resistor to limit current. Then you measure voltage over the zener.

I watched Big Clive explain a halogen load yesterday that was quite a bit funkier than what I could imagine. I can't "picture" this one either, sorry.

[eq1]

Originally Posted by Allybye

Been looking to understand the circuit for you. Can you tell me what the resistor is between the pot and ed1 - cannot read what it says!

"R18"...

Originally Posted by Allybye

Any idea where R2 might be? Cannot see it but it might just be the pot or ed1....

It's not labeled in the image, but I just found it on the board - it's the big pot, it's labeled "Radj2."

I think that's everything, but it seems like there's a couple missing, I don't see an R5 or an R17...

Does the numbering mean something in terms of functionality?

[Allybye]

Not at all! It's just for reference in the circuit.

Thanks.

[eq1]

Originally Posted by Dark Fader

Originally Posted by Dark Fader

Are you saying the speakers are still emitting noise when they are switched off?

Yes.

Originally Posted by Dark Fader

Does the noise go away when they are completely unplugged from the mains?

Yes.

Originally Posted by Dark Fader

If both of those are true it sounds like the noise is a mechanical noise coming from a component in the power supply (such as a transformer or an inductor) before the switch, rather than coming from the speaker cones themselves due to an electrical noise.

I think that's right. I don't hear the noise from the cones, rather, I hear it more clearly when I put my ear to the front port, or even around to the back panel...

[eq1]

Originally Posted by cyrano

You can't test a zener with a typical DMM...In this case, it's a 6V zener (from memory). So you'd need a >8V power supply and a resistor to limit current. Then you measure voltage over the zener.

OK, that was the 'other' test method described at that linked webpage. I might try this today... Even though I ordered a replacement board, I'd still like to know why this board failed. If I can test the zener that I already pulled off the board and find that it's bad, well, I guess it'd be peace of mind, at this point.

One of the zeners is 6V (D5), D4 is 12V.


* * *

I tested both zener diodes, D4 and 5, and they both look OK. I applied ~14V in reverse, D4 measured 11.98V and D5 measured 6.15V...

I also removed C1, the big cap, and went through the 3 tests listed at that webpage. The 'resistance' and 'capacitance' tests -- no clue, I measured 'OL' with resistance, and capacitance was all over the place. I don't have any confidence in these. Then I tried to charge it up with the ~14V battery power supply - and it charged up. So, C1 looks OK too.

So much for my 'peace of mind'. I don't know what's wrong with this thing. The only thing left is the miscellaneous resistors, and C2. I tried to pull C2 off the board but couldn't get it. Maybe I'll give that one more try...

OK, got that off, tested it, looks OK too. C2 says "104J" on it. I measured capacitance with my DMM and the reading was 103.8 nF. I tried to charge it with my 14V supply and it charges...

C1, BTW, is a 25V 220 microF electrolytic.

[eq1]

"Just when I thought I was out..."

I just spent 4 or 5 hours trying to find the right connector for the replacement digital control board that has arrived. As you may recall, I decided to buy a replacement board, and as far as I could tell a newer 'digital' version of the analog unit should work. Well it arrived, but the location where all the hardware wires attach just has some pins, that looked like typical header pins on a PCB when I bought it, looking at the image.

But they're bigger than typical, with a 5mm pitch (spacing). I could not find or figure out what was supposed to plug-in to that... The analog board has what looked like a more or less typical screw block terminal, where you insert the wire ends into lugs and screw them down (blue block in earlier image). I thought it was soldered to the board.

It turns out that that 'blue block' actually pulls off - and underneath it are these larger-than-normal header pins! So all I have to do (had to do) is pull that off and plug it into the new board...

What a headache! I've been searching, looking around for alternatives, thinking I'm gonna have to try to desolder the new, desolder the old, re-solder the old onto the new, etc etc...

For reference, the connections are made by NTE. The header pin portion is what they call a "wire protector" from the "E1500 series" ('terminal block header, male pins, unshrouded'), while the plug/block portion (also called "wire protector") is from the "E1600 series" ('terminal block plug for unshrouded header'). Both of these can be found within Digikey's 'connectors/ Terminal blocks - headers, plugs and sockets' category...

Also, as far as I can tell, the wire connections should parallel the wire connections on the analog board. The actual board has no labels/indicators of where the wires go, but the image at the website where I bought it did, and they compare to the analog. Unfortunately, too, I noticed that I was sold an old version 'Rev 2.2', whereas the website image shows 'rev 2.3'; there's quite a few visual differences...

I couldn't find a single image of this board online, I'll put one here... I actually can't believe the dearth of info online about these things.



Anyway, not quite there yet in terms of trying this out...

[eq1]

Got a chance to try it out with the new board. It seemed to be working, I even got a chance to see if it caused noise in my monitors - and it didn't. But it looks like the emphasis is on the word seemed, as 10 minutes or so after declaring success it stopped working - and then a bit later with some sleuthing I discovered that it actually wasn't turning fully OFF, despite the LED lights indicating otherwise...

I just checked the resistance of the heating elements in this unit vs. another older one that failed in a different way. One element in this one reads ~4.5MΩ, while the others (there's two per unit) read around 8Ω. What's more, I read something other than infinite resistance between the element's terminals, one at a time, and the copper tube/heating chamber in which the element is installed. The same measurements on the other elements/heating chambers show infinite resistance...

So, it looks like one of the heating elements is shorted and faulty. I still have to open it up and check... This unit is only about 3 1/2 years old, I generally just assumed that the heating elements would be the last thing to go - so never did any checks on them...

[eq1]

Now that I'm looking into replacing a heating element and thinking more about those, I'm wondering about what I wrote earlier, about 208V vs. 220V vs. 240V. I've pasted that at the bottom.

I don't think the gist of what I wrote is correct, the idea that it might be advantageous to "bump-up" to 240V. If anyone knows either way - I'm all ears...

If an element's rated for say 5000W at 208V, I think that would actually put out more heat if/when the voltage is higher; i.e. at input=208V power consumption is about 5000W, at say 240V power consumption would be, I think about 15% higher, so about 5800W.

On the flip side, an element rated for 5000W at say 240V would put out less heat than the 208V-rated element, it will consume about 5000W at 240V.

Originally Posted by eq1

I noticed that the heating elements on this unit are rated at 208V, one's 5000W and the other is 5500W. Looking at various similar units available today (mostly the 'Titan' brand), and reading about them, I'm pretty sure that, all else being equal, these devices can utilize a range of heating elements -- 208V, 220V, and 240V, without a different control PCB. A similar Titan unit I'm pretty sure mentioned it -- that you could choose among power levels depending on your home power available. I guess some places only have 208V?? Not sure, but I have 240V output, so it seems like I could bump these heating elements up to 240V. I think that would result in faster heating at a lower current, but not necessarily more heating...

[cyrano]

FWIW, I don't think going from 208 to 240V would make a difference.

Mains over here is supposed to be 230 VAC. People with 250V are around. Ours is 235. I don't see much lower than 225, unless it's a very old install.

Heating elements aren't linear either. I don't know by how much. Probably not a lot. It's just that I recently noticed running a 12V relay on 24VDC doesn't mean immediate burnout. It gets warmer, but not twice as warm. I measured the current and it was about 15% higher. Not double, what one might expect.

[Allybye]

Just back after several days 'offline' so I have made no progress with your circuit that is significant to report....still working on it again!

With regards to the different heater elements.

I would be careful about just mixing and matching.
The elements temperature is restricted by the water and the controller. So say a 40 or 50 degrees C range. Non linearity for resistance owing to the temperature change is not a huge affect.
Relay coils are different and the temperature rise can be considerable.

What is my concern is the relationship between the applied voltage and the power given a consistent resistance when moving between 208v and 240v and the consequential increase in current.
The Power is V squared devided the resistance.
At 208v and 5kW the nominal resistance is just over 8 ohms and a current of 24 Amps.

Use that same heater element at 240v and the current rises to about 30 amps.

So in total the current rises from 48amps to 60amps! That might be a issue? ...power being about 6kW.

Going from a 5kW heater nominal 11.5 ohms at 240v down to 208 volts would be safe but a significant power reduction (to 4kW) and thus poorer heating effect.

[eq1]

Originally Posted by Allybye

Just back after several days 'offline' so I have made no progress with your circuit that is significant to report....still working on it again!

Don't worry too much about it on my behalf - I'm still curious about the circuit and how it works, so if you're curious too and want to explore please do. But practically speaking we've moved on - now that I've found out the element is the real problem.

Originally Posted by Allybye

With regards to the different heater elements. I would be careful about just mixing and matching....What is my concern is the relationship between the applied voltage and the power given a consistent resistance when moving between 208v and 240v and the consequential increase in current. The Power is V squared divided by the resistance.

At 208v and 5kW the nominal resistance is just over 8 ohms and a current of 24 Amps. Use that same heater element at 240v and the current rises to about 30 amps.

So in total the current rises from 48amps to 60amps! That might be a issue?

This is the way it's supposed to be - that's the stock configuration, more or less. I'm just starting to wrap my brain around these things...

The stock elements are 208V 5000W and 208V 5500W. The device is rated or 'called' a 13kW unit, the breaker is 60 amp... The actual input voltage for my house power is around 240V.

My understanding is that the elements and the actual input voltage dictate the max power and current possible. I think the controller board 'works' with whatever the max is. For example, when the device turns on, the water is cool, max power is allowed/applied, but then the controller tapers that down by changing the 'phase angle' or whatever, as the water heats up... Basically, my sense is that the controller doesn't do a whole lot, if anything, to curb max power, it simply ends up limiting power based on the temperature it senses...

I actually think that might be part of my problem: this unit is heavily used and it's probably operating too often, too much near the max. I personally am aware of this and try to limit the water flow, to try to keep the unit from maxing-out all the time, especially in winter when the input water is colder. But my housemates...can't be bothered by such technical concerns...

It's actually quite interesting. I have an older Rheem RTE13 for 'my' bathroom. It's probably almost 10 years old now - and it still works. The difference between that and the one downstairs - where I've now seen the demise of two of these Rheems and have moved to that noisy Eco11 - is 2 more people using it. Plus it serves both kitchen and bathroom. My 'mother in-law' takes horrendously long showers...

Another thing to note: The Eco11 - they call it an "11" because the elements are each rated at 5500W - and I think the controller must enforce that, too. It heats the water up slower than the RTE13 - because of that extra power the 208V rated elements of the RTE13 allow, I think...

Originally Posted by Allybye

Going from a 5kW heater nominal 11.5 ohms at 240v down to 208 volts would be safe but a significant power reduction (to 4kW) and thus poorer heating effect.

Yeah, I actually ordered the stock 208V 5kW and 5.5kW elements, but now I'm starting to wonder whether reducing the max output by going to different elements would have been more prudent. As I mentioned earlier, the 208V elements actually have higher max heating potential than 220V or 240V rated elements. Given the heavy use of this water heater and the demise of the element, seems like the lower max power output could potentially increase their longevity...

The 208V 5500W rated element should see a max power of 7385 watts at 240V input, 30.8 amps.

A 220V 5500W rated element (one I've seen widely available, such as for Eco11) would only see a max power of 6545 watts, 27.3 amps.


Here's my calculations:

P=V^2/r, P*r=V^2, r=V^2/P

208V^2/7.8ohms=5547W (7.8ohms actual measure of 5500W element)

240V^2/7.8ohms=7385W (actual voltage input vis-a-vis this element)

V=ir, i=V/r, 240V/7.8ohms=30.8 amps

at 208V input: 208V/7.8ohms=26.7 amps


Now, for a 220V 5500W rated element:

r=V^2/P, 220V^2/5500W=8.8 ohms

at 240V actual input: P=V^2/r, =240V^2/8.8ohms=6545W
i=V/r, =240V/8.8ohms=27.3 amps

at 208V actual input: 208V^2/8.8ohms=4916W
i=V/r, =208V/8.8ohms=23.6 amps


Here's an image of the 'good', working heating element from the RTE13. Even this one has some corrosion spots on it. The bad one is ripped open around the same area. These areas are at the bottom of the heating chamber... There's a water port at the bottom of the chamber, right around where these units have failed/are failing...




Here are a couple images of the failed heating element, below:





Also, I removed the heating elements from my older unit and checked them out - they looked fine, no pits, no major corrosion, etc. The 208V 5000W element measured 9.3 ohms, the 208V 5500W measured 8 ohms. Note that the resistance values are little higher initially then fall a bit, and they also vary a little bit while holding the probes on the terminals, by like + or - about 0.3 ohms or so.

[eq1]

Here's a basic diagram of the wiring that goes with what I wrote in this post, https://forum.cockos.com/showpost.ph...3&postcount=90, mainly the part I've quoted below.

Originally Posted by eq1

"L1 120V" is one leg of the house power. "L2 120V" is the other. From 'the house', one leg splits and goes to one side of each of two heating elements, and also a small wire branches off from one of those splits and goes to "L1 120V."

The other 120V leg ('from house') splits, with one branch going to one terminal on the 'dual-SCR' and the other going to "L2 120V." The remaining terminal on the dual-SCR goes to the other side of one of the heating elements, which is wired to the other heating element...

[eq1]

Just in case anyone's still following along, I replaced the heating elements and reinstalled the unit. It sort of worked, but the temperature control is all over the place, it just doesn't control the temp right. I had to go back to the noisy ECO11. When I had the RTE13 installed I did check for noise in my monitors again - no noise. And of course it's back with the ECO 11...

Not sure where I'm going from here. Not sure why it's not working. The only thing I can think of that I didn't 'do right' to a T, that maybe would mess with the temp control, though I doubt it, is I didn't use any new thermal paste where the temp probe inserts. There's a tiny tube into which the thermistor is inserted. In stock form there's a white thermal paste in the tube and the thermistor is hot glued to the tube, so it doesn't pull out. Last time I changed the board I used new thermal paste (Arctic Silver, stuff for CPUs). But when I pulled it out and re-inserted the thermistor this time, I didn't put any new stuff in there, I just shoved the thermistor deep into the tube and put a dab of silicone on the end... Hard to believe that would make the difference I experienced...

So, I could try again with 'new thermal paste'... Or, I could fix the previous control board and try that again. Or maybe I could do something with the older RTE13 and try that one. I guess with that one though I had wanted to try the new digital control board and one of the newer all-in-one dual SCR units, since I now have an extra. But since the digital board doesn't seem to work right in the other unit, I'm reluctant to try it in the older unit, there shouldn't be any reason I'd see anything different...

One thought that occurred to me is that maybe these units are calibrated somehow - and since I don't know whether or how that's done, I haven't done it. There is an adjustment pot on the older analog boards - and on one of the digital boards - but there isn't one on my digital board. So not sure what could possibly be 'adjusted'/calibrated on this one... Somewhere along the line I got the impression that the adjustable trimmer pot was for the reed switch though, something to do with adjusting the water flow ON/OFF signal. Can't remember why I got that impression, nor why/how that would be adjustable...

[Allybye]

Hopefully I can help with your circuit soon...just have some other priorities I am sorry to say.

Would comment though that the heat conducting compound should be used to ensure that the water temperature is monitored correctly. New general purpose tubes are very cheap.
If you do not get good heat conduct than your probe will not measure correctly and you could over heat the water initially (as it will not 'see' the temperature rise) but the water might not get heated enough afterwards when the temp in the probe vicinity has risen. It will affect the response time and may cycle hot and cold! ....or if it is not getting enought heat from the water versus cooling of it then may be on too much!

Lack of compound should not prevent switching on as the probe may always be cooler (the locations ambient temperature) than the water set temperature.

[eq1]

Here's an image of the front side of the analog circuit board:

[Allybye]

Thanks, that pic was at my request...in case anybody wonders....

[eq1]

My tube of thermal paste finally arrived a few days ago. As you (probably don't) recall, after replacing the heating elements on the RTE13 - the last and only thing left that I hadn't replaced, and one of which was obviously bad - I installed the unit, but it still didn't seem to work well enough. It didn't regulate the temp well at all. I hadn't applied additional/new thermal paste/grease in the tiny tube where the thermistor inserts.

So now I have. And it seems to work pretty well now. I've been using it for a couple days. Hard to believe the thermal paste is what did it, I'm not totally convinced, as it didn't seem to regulate very well when I initially installed it. It took me some time to set the temperature output with the control buttons, and after maybe 10-15 minutes of fiddling, it seemed to settle in... I wonder if there's some additional, automatic calibration that takes some time to begin to work properly? Or maybe all the 'new parts' need some break-in - maybe initially their properties, such as the metal of the heating elements, maybe the silicon of the SCR? slightly changes, making temp regulation difficult??

I don't know. Maybe it was just the thermal paste... The thing is, when I had another unit apart ages ago, its thermistor wasn't seated all that well, no new thermal paste - and it didn't have a problem regulating temp...

Water temp still seems to fluctuate a bit more than it used to in the stock, old configuration, but it's not too bad.

The RTE13 is a lot faster at heating the water than the ECO11. And, of course, the main point: NO MORE NOISE IN MY MONITORS.


fyi: Major pain trying to get the thermal paste into the tiny tube where the thermistor inserts. It's really weird. There's a tiny tube - inside diameter about the size of one of those red plastic spray extension straws you get with spray cans of stuff, that fit into the spray nozzle on aerosol cans (like silicone lubricant, brake cleaner, etc.) - that tiny tube intersects a larger tube where the water flows from one heating chamber to the other, at the bottom of the two heating chambers. It must be recessed into the larger tube in some manner. Hard to explain. But, there must be a little pocket of space at the base of the tiny tube...

Anyway, I had to get a syringe and fabricate a little nozzle that fit into the tiny tube (out of one of those red nozzle straw things) to get the silicone thermal grease into the tube. Was not easy.

[cyrano]

Well, good that it's fixed!

I'm guessing that HVAC pro's probably replace the thermistor with a unit that already has the thermal compound in place. Saves time...

DIY is usually a tiny bit harder.