Lochinvar Noble data logging

In_New_England

Does anyone have information about how I could tap into the control board of the Lochinvar and extract data that the Noble knows about such as

1. Firing rate
2. Incoming water temp
3. Outgoing water temp
4. Ambient temp

The Noble's processor must track all these variables for its control system.

ratio

If it doesn't intentionally export them via some method, you're not going to get them from it. Pulling an analog signal off a μC input pin or using JTAG to grab it (what comes to my mind when you say "tap into the control board") will be setting yourself up for heartache if anything goes wrong, ever, at all.

FWIW, three of the four are easily available independent of the boiler control. The last one (firing rate) can be determined by monitoring the gas flow into the unit, but there's a larger buy-in than with the temps.

In_New_England

ratio said:

If it doesn't intentionally export them via some method, you're not going to get them from it. Pulling an analog signal off a μC input pin or using JTAG to grab it (what comes to my mind when you say "tap into the control board") will be setting yourself up for heartache if anything goes wrong, ever, at all.

FWIW, three of the four are easily available independent of the boiler control. The last one (firing rate) can be determined by monitoring the gas flow into the unit, but there's a larger buy-in than with the temps.

I agree: if the control board fries for example, there is no way to know if my tinkering caused it.

Ok, I will move forward with my ESP32 based boiler monitoring project. As you say, the temps are easy (and safe) to monitor. For gas, what do you suggest? I wonder if I could monitor something else, like current to a blower fan or something as a proxy for firing rate?

In_New_England

Actually, the temperature difference between the incoming and outgoing water will be a proxy for the firing rate, since this is the amount of heat the boiler is putting into the water.

This ignores changing efficiency and variations in circulation rate, but I believe that the pumps will run at a fixed rate, so that approximation will be good enough for my use case. Looking at the output temp I will also be able to infer when the boiler stopped firing by looking at a drop in output temp, since this will be a low mass system and the heating ability will stop as soon as it turns off.

Another input I forgot to mention is the thermostat. That's a simple on/off signal which I think I can intercept harmlessly before it gets to the boiler.

cc: @fentonc who made the original boilertron.

ratio

IIRC the inducer motor is directly proportional to the firing rate. Maybe a snap-on CT will give you a usable signal?

Dave Carpentier

A pair of thermal sensors in the intake and exhaust streams should be able to extrapolate the fire rate ?

In_New_England

Dave Carpentier said:

A pair of thermal sensors in the intake and exhaust streams should be able to extrapolate the fire rate ?

It took me a while to understand this suggestion, but this idea is genius! Is the exhaust under 125 C (257F)? Then I can use these temp probes I got to monitor the hot water. I'm quite excited by this idea, TBH!

Edit:

Actually, not sure about firing rate. The flue temp will indirectly measure water temperature, since the exhaust gas temp is related to water temp. That said, ignoring inefficiencies, and assuming a fixed circulation rate, change in water temp as it goes through the boiler will capture firing rate. But I'll now measure flue temp as well, because what the heck.

Dave Carpentier

In_New_England said:

The flue temp will indirectly measure water temperature, since the exhaust gas temp is related to water temp.

You're right. Thats the whole point of "condensing", so that would be an unreliable data point. Mybad.

Does the Nobel display screen stay on all of the time ? Ive heard of people using camera OCR to read hard displays.

Zman

Anything that measures kWh should do the trick. Note the power consumption at high and low fire and then you can figure out the rest

In_New_England

Zman said:

Anything that measures kWh should do the trick. Note the power consumption at high and low fire and then you can figure out the rest

Assuming that there is a component of electric consumption decently related to firing rate and this component is measurably more than constant load, this is a genius idea.

A clamp on ammeter is what I thought about, but this is a bit invasive.

God, this instrument will be bristling with sensors.

@Dave%20Carpentier OCR is an idea. It will up the cost but has the potential to record everything needed without any other sensors. I have to look if the Lochinvar has a stay on display.

Hot_water_fan

What kind of gas meter do you have? A smart meter would at least give you hourly visibility into usage.

In_New_England

@Hot_water_fan I'm having a gas line installed. I hope next month, but who knows. I'll be supplied by National Grid. Do I have to do something special to have a smart meter, or do you think all new meters are smart meters?

edit: I looks like gas meters will be AMR, as I guess is my electric and for sure water. Now doing web searches of how to use microcontrollers to read these gadgets ...

hot_rod

are are you looking to log this data, or watch in real time?  Depending on points and how often you sample it ends up being a huge amount of data to collect and store. We used to log 4 points on solar thermal systems at 1 minute intervals to get useful graphs 

In_New_England

@hot_rod you are correct about easily being deluged with data.

My current intention is to log a set of temperatures and the thermostat status for the new boiler (possibly the old one too, if I finish early).

I will log the data at 1 min intervals. This comes out to 7 bytes/row x 1440 rows/day. I can fit the temperature data into one byte because I'll use Fahrenheit and offset it by 20F. In this scheme -20F = 0, 0F = 20, 235F = 255. This will be sufficient for all measurements. I might saturate for the flue intake and outlet, but that is ok, and will probably be rare.

If I get firing rate somehow, I'll fit that into one byte too, probably just convert it to a percentage.

I'm not so concerned about gas directly, though several people have assumed I'll be measuring this, so, maybe I should. It'll be painful enough to pay the company EOM, let alone watch the values in real time .