flue temperature sensor interlock device

For this application I'd second what Yanche says ... use a temperature controller.

A temperature controller is more complex than a bimetallic thermostat, but gives a lot more flexibility. Once you buy a particular thermostat, you are stuck with it's characteristic trip temperature and hysteresis, but these are completely under your control with a temperature controller. Although a type J thermocouple will work I'd use a type K instead - iron in the type J TC is prone to corrosion in this application. It doesn't necessarily need to be in the flue gas stream; for this service it would do to simply clamp it to the outside of the flue pipe using a large diameter hose clamp.

AutomationDirect sells the 'Solo' series of controllers which combine a good feature set and low cost (the SL4848-RR at $99.00 would work). You may find equivalent controllers on eBay for less, but will need to do a bit more research to make sure they fit your needs.
This one also has MODBUS over RS485 serial communications capability, and opens the possibility of logging temperatures to a computer using any MODBUS capable client, or their free configuration/recording software (although you'll need an RS-232-to-485 converter as well). http://support.automationdirect.com/products/solo.html


Solo SL4848 Product Overview
http://web2.automationdirect.com/adc/Sh ... /SL4848-RR

Manual

Note: I can't get the product overview link to display correctly as a hyperlink, but you can copy the text between the bracketed URL indicators, and paste it into your brower's address bar.

In this type of circuit the controller output isn't used for temperature control, but rather simply as a limit indicator using a relay output that changes state when setpoint has been reached. The controller would be set up for on-off control (rather than PID). Typically you can get up to two alarm outputs as well on such controllers, although the SL4848-RR looks like it has only one.

Alarms can be set up in a variety of ways, but one that may be useful is 'high absolute limit', and use it to turn off the burner if the stack temperature gets uncomfortably high. In implemented, you'd want to set a rather wide hysteresis for this - a small hysteresis would allow short cycling. For instance, if the stack overtemp setpoint was 200°F, and you used a hysteresis of 1° then the burner would turn on again at 199° ... not a good idea. Here an alarm hysteresis of 50°F would make more sense.

I measured the temperature of my flue pipe 1/2 way between the boiler chimney outlet and damper, and got the following temperature curve - starting at 99°F, a rise of about 15°F per minute until it started to settle out and 'knee over' at 162°F about 4 minutes afterward.

Whether you use a thermostat or controller it seems you'd need a bit of timing and hold in logic, too, something along this line ... [if burner is on for more than x amount of time, and flue temperature is less than y degrees, then "fire out" - turn off burner, sound alarm, and lock out burner until reset]. In this case I'd probably use a 2 minute burner on time, and figure for a flue temp of 120 degrees. If the burner has been running for more than 2 minutes, and temperature hasn't reached 120 degrees then the "fire out" condition would be satisfied, and burner would be locked out until reset.

This would miss a fire out condition until a demand call occurs (i.e. - a 'keep lit' timed running of the burner wouldn't trigger it, since the burner wouldn't be running for long enough), but this won't waste too much unburned coal.

TechCurmudgeon wrote: Note: I can't get the product overview link to display correctly as a hyperlink, but you can copy the text between the bracketed URL indicators, and paste it into your brower's address bar.

That URL contained parentheses which pose a cross site scripting vulnerability, these are special characters and can be used to insert malicious code into a link. In this case perfectly harmless. The forum looks at URL's before it applies the link and any suspicious pattern will be rejected. In this case simply escaping them fixed it. Not sure why it differentiates between the two....

More here, note the example code given at the bottom with the escaped characters fails on his forum: http://www.symantec.com/enterprise/secu ... ripti.html

TechCurmudgeon wrote:Whether you use a thermostat or controller it seems you'd need a bit of timing and hold in logic, too, something along this line ...

All the coal fired boilers cry out for modern electronic controls. Something that integrates boiler water temperature sensing, outdoor temperature sensing, ash temperature, flue stack temperature, domestic hot water temperature,coal feed and blower motor control. Add to it room carbon monoxide monitoring and flue gas oxygen measurement and you would have a hell of a system. If I didn't have so much on my plate I'd design one.

Sorry, I get ahead of myself typing, Greg, You are correct. If it goes out for extended periods (over an hour or two) THEN the power comes back on, it will just push the coal off the end....especially if we are not home (at work) for 8 hours or more....big mess. At least with an temp interlock on the flue pipe, the pusher would stop if the temp falls below a certain setpoint after the power comes back on, we would still have the furnance backup once the power did come back on and keep the house warm.

Yanche wrote:All the coal fired boilers cry out for modern electronic controls. Something that integrates boiler water temperature sensing, outdoor temperature sensing, ash temperature, flue stack temperature, domestic hot water temperature,coal feed and blower motor control. Add to it room carbon monoxide monitoring and flue gas oxygen measurement and you would have a hell of a system. If I didn't have so much on my plate I'd design one.

Been thinking along the same lines, but I give up on it after doing back-of-the-head calculations ... it always ends up costing more than I want to spend.

I'm very familiar with Allen-Bradley PLCs, and would do it around a SLC500 or Micrologix 1500 platform, except (even if I sit around, and lurk eBay for bargains) I'd need to use my work license for programming. This isn't an issue, but it makes the "dream system" less than transportable - for someone else to benefit from it they'd need a license of RSL500, and I think it goes for about a grand per seat these days.

Haven't used EZAutomation PLCs yet, but they are fairly cheap for what you get, and the demo ladder software didn't look too bad. I'd use one of the Koyo PLCs from AutomationDirect, but never warmed up to that software, especially in the analog realm.

It looks like Modbus TCP/IP is a good communications protocol for such a thing - a lot of people are making analog/digital I/O for it, and it'd be fairly easy to set up a 'scrap' computer running an OPC server for rudimentary HMI (setpoints, and such), and to do trending.

Keeping the cost down while still using mostly off-the-shelf automation hardware and software is the sticking point. Once prices drop to where I can get a 6" color touchscreen, at least 20 digital I/O, and 8 analog channels (12 bits or greater) for $600 or less ... hmm, we'll see.

I think you could get near that price point now (sans touchscreen) using something like a Rabbit, and roll-your-own analog input board and mux, but would have to put a lot of time into it.

Yea and it should e-mail or call you when it's low on coal. I'll set up script that will aotumatically alert you when viewing the forum if you need coal.

Seriously though I'm sure in some situations adding technology such as some of things mentioned may be beneficial to some but I think it's a bit overboard for most. You two may have very technical backgrounds but most people do not, they want to plug something in and have it work and as far as furnaces go have it work 20, 30 years. What are you going to do 10 or 15 years from now and the hardware isn't available or there's no one that knows anything about it. You may be able to fix it but your average homeowner is going to be stuck out in the cold. They'll be paying exorbitant labor fees for parts and labors or as a worse case scenario be stuck with a 1 ton paper weight.

NEPAForum Admin wrote:
Seriously though I'm sure in some situations adding technology such as some of things mentioned may be beneficial to some but I think it's a bit overboard for most..

I think that’s a good point. New boilers are already selling for as much as $6500 + and adding another one or two thousand dollars worth of electronics along with the added installation cost and set up costs would make a new coal heating system far to costly for most people not to mention most people burn coal because they are trying to save a buck to begin with. For those who have the knowledge and expertise in this area and experiment for their own satisfaction is one thing, but as far as mass manufacturing equipment with all these bells and whistles would be a wasted effort in my opinion.

A local plumber told me that he replaced many old Axeman-Andersons years ago when they would break down because the replacement parts were very expensive and a new oil burner was relatively cheap. Cost of equipment has to be taken into account. I am not going to shovel coal and carry ashes for just a few saved dollars, I want to see real savings, and likewise I also would like to see a return on my equipment cost in 10 years or preferably less.

Technology for a lot of things has come a long way. The issue is it's sometimes used not because its needed but because they can or instead of. Take car keys for example, a lot of them now have chips embedded into them to prevent theft. That's great right up until the point you need to go buy one for $200 because you lost a set or bugs in the design such as your cell phone erasing the chip. Add to that its rip-off because those keys probably only cost a few dollars at most to produce but now they got you stuck because its a proprietary piece of equipment.

http://www.latimes.com/business/la-fi-g ... s-business

I agree any add on modern electronics needs to be cost effective, just like any other consumer product. If it doesn't have a two or three year payback in coal savings it's not going to sell. I believe it's possible. But it would need to be designed from the ground up. Adapting industrial control sub-systems would be way to expensive. That's fine for turning a single coal boiler installation into a home science project but not for a consumer product. Given the high markup margins on existing boiler controls distributed through the plumbing supply houses it may be possible to sell something direct for the same dollars that's much more capable. For example the Honeywell indoor/outdoor electronic aquastat sells for about $200 trade price at supply houses. It's got a simple two layer PC board and a 8 bit microcontroller in it. Total electronics component cost of $20-25 tops. The rest is mark up. Typically 100% at each level of distribution. Right now the residential coal heating market is to small for any manufacturer to get his engineering cost back. The component costs for an electronic device that would measure boiler water temperature, outdoor temperature, flue gas temperature and then use a microcontroller to control the combustion blower and coal feed motors would be less than $100. Modern electronics is incredibly cheap.

Those are good points, which is why only a hardcore tech nut would want to play around with a full blown integrated system at this time. To make it tasty for general consumption I'm thinking we'll need to wait one or two equipment generations (call it 5 to 15 years) for the costs to come down enough.

Already, your garden variety electronic thermostat that can be had for $50 to $80 - and some of them are down into the $30 range; same pricing as a mercury switch thermostat. All of these electronic thermostats are packing a computer of some sort, and under the hood are fairly complex gizmos, but, so long as that complexity is transparent to the user its six of one/half dozen of the other. I bought one of these early on (back when Heathkit was still in business; early 1990s or late 80s, back when they were still in the $130 range), and haven't had a problem.

The challenge is to do an integrated system that is similarly transparent to a typical homeowner, is cheap enough to pay for itself within 10 years, has a reasonable design life (say, 25 to 30 years), and which isn't so integrated that it becomes a maintenance nightmare, or can't be replaced with standard components. A fairly large challenge, to be sure, but do-able.

I'm guessing my Van Wert Anthratherm is well beyond any reasonable design life (built somewhere in the mid 50's), and main problems with obsolescence so far have been with the mechanical system. For instance, it uses a 600:1 gearbox that nobody makes anymore, but the boiler itself still runs like a champ.

I believe it's possible. But it would need to be designed from the ground up. Adapting industrial control sub-systems would be way to expensive. That's fine for turning a single coal boiler installation into a home science project but not for a consumer product.

Gotta agree with you there ... to make it cheap enough to be a viable consumer product building from scratch is the only way to go. Don't want to veer this topic off course any more than I have (and I've just fallen into a spate of breakdown mess, projects, and such, so don't want to start another topic now), but I'd like to spitball this a bit more down the road.

Maybe this'll be the impetus for me to throw some instrumentation at the Anthratherm, and collect various run time data ... use it as a test bed to see what changes in control actions yield favorable operational changes (i.e. - less run time/coal usage but still giving a good comfort level, hot water on demand, etc.).

What kind of sensors would you be thinking of for ambient and flue temperatures? Guessing the DIP and SMD temp transmitter chip lines that Maxim and others make would be OK for lower temps, but (for circuit simplicity) would you be thinking RTDs vs. thermocouples for the hot stuff?

TechCurmudgeon wrote:What kind of sensors would you be thinking of for ambient and flue temperatures? Guessing the DIP and SMD temp transmitter chip lines that Maxim and others make would be OK for lower temps, but (for circuit simplicity) would you be thinking RTDs vs. thermocouples for the hot stuff?

I don't think the tread has be hi-jacked. It's just expanded into all the issues considered in any product development. The problem to solve, how to solve it, product features and product cost. There are some very, very exciting (at least to an electronic engineer) developments going on now in the industry. Every wonder how all these new electronics stuff can sell so cheaply? Well it's not just because much of it is manufactured overseas. The really hard engineering part is done here in the USA. The dream of any designer is "System on Chip". That means all the electronics is on that tiny little integrated circuit which is made in 1-3 billion dollar semiconductor plant. Yes, that is 1-3 billion. The piece parts that come out of that plant only cost tens of cents each or for really complex parts a dollar or two. In the past the engineering costs to design some would cost hundreds of thousands to millions of dollars. Well no more. Enter Cypress Semiconductor. They are selling programmable "System on Chip" parts. It idea is this, gather up all the little things electronic engineers use, amplifiers, digital to analog, analog to digital, filters, input circuit, output circuits etc. and put them all on one chip. They are not connected together, they are just there. Now use software to tell how they should be connected to do something useful. Further store the program as to how they are connected together in a flash memory on the chip. Now the same part, made in huge volume, can be used for many, many things. These parts are cheap. Single piece parts are less than ten bucks. Much less in volume. They have more than enough capability to do the simple things to control a coal appliance. That's why I said in an earlier post any controller must be designed from scratch. The technology has changed so much. BTY that 1-3 billion dollar plant is no good after 4-5 years. It obsolete and needs to be scraped and rebuilt to the latest technology. How would you like to be the one making those business decisions?

Getting back to your specific questions. I would measure all temperatures except flue gas and coal burning with thermistors. I'd measure flue gas and coal burning temperatures with type K thermocouples. Motor controls would be multi-phase power FETs. This would allow motor speed control electronically. The micoprocesor would use outside temperature and domestic hot water demand to adjust the coal appliance firing rate. I'd have an option for flue gas O2 measurement. This would use a automotive lean burn engine sensor to adjust burning for minimum CO emissions. The selling cost goal would be to have it pay for itself in 2-3 years of coal use savings.

multi-phase power FETs

Would I be correct to assume you are looking at building a PWM motor controller inverter using a fixed DC bus supply, power FETs in the output bridge, and driving a standard three phase squirrel cage motor? Most such inverters I've seen in commercial designs use IGBTs, and I'm wondering why FETs?

I haven't kept up with what is happening at the device level for awhile, but figured I'd have seen more FET-based controller designs by now, seeing that they don't need as complicated a drive logic as IGBTs, don't need snubber networks, and, especially at higher switching frequencies, have lower losses.

Maybe my question should have been, "why are IGBTs so much more common"?

I'm thinking it would be interesting to drive a switched reluctance motor - it seems we're are finally nearing the point where such motors become practical (I've seen them in fan applications already), and I know one of the target markets is for automotive applications.

This suggests a rather low DC voltage supply, although probably for the next automotive generation; I've heard 24 or 48 VDC batted around. It would be nice to be able to back up the burner logic and power controls using conventional deep cycle lead-acid batteries to float across the drive bridge supply in trickle charge mode when AC line is available, and then throw them directly across the bus when AC is lost.

The batteries would be sized to provide a decent amount of run time, but that could be extended by entering a 'power saver' mode which, when AC power is lost, only runs the burner enough to keep the fire from going out, and preventing room air temperature from falling below a truly uncomfortable level. That would depend on the person, but somewhere around 60 and 65°F isn't too unreasonable for longer blackouts, if it means not having the fire go out, and getting some amount of heat versus no fire/no heat.

TechCurmudgeon wrote:For this application I'd second what Yanche says ... use a temperature controller.

A temperature controller is more complex than a bimetallic thermostat, but gives a lot more flexibility. Once you buy a particular thermostat, you are stuck with it's characteristic trip temperature and hysteresis, but these are completely under your control with a temperature controller. Although a type J thermocouple will work I'd use a type K instead - iron in the type J TC is prone to corrosion in this application. It doesn't necessarily need to be in the flue gas stream; for this service it would do to simply clamp it to the outside of the flue pipe using a large diameter hose clamp.

AutomationDirect sells the 'Solo' series of controllers which combine a good feature set and low cost (the SL4848-RR at $99.00 would work). You may find equivalent controllers on eBay for less, but will need to do a bit more research to make sure they fit your needs.
This one also has MODBUS over RS485 serial communications capability, and opens the possibility of logging temperatures to a computer using any MODBUS capable client, or their free configuration/recording software (although you'll need an RS-232-to-485 converter as well). http://support.automationdirect.com/products/solo.html


Solo SL4848 Product Overview
http://web2.automationdirect.com/adc/Sh ... /SL4848-RR

Manual

Note: I can't get the product overview link to display correctly as a hyperlink, but you can copy the text between the bracketed URL indicators, and paste it into your brower's address bar.

In this type of circuit the controller output isn't used for temperature control, but rather simply as a limit indicator using a relay output that changes state when setpoint has been reached. The controller would be set up for on-off control (rather than PID). Typically you can get up to two alarm outputs as well on such controllers, although the SL4848-RR looks like it has only one.

Alarms can be set up in a variety of ways, but one that may be useful is 'high absolute limit', and use it to turn off the burner if the stack temperature gets uncomfortably high. In implemented, you'd want to set a rather wide hysteresis for this - a small hysteresis would allow short cycling. For instance, if the stack overtemp setpoint was 200°F, and you used a hysteresis of 1° then the burner would turn on again at 199° ... not a good idea. Here an alarm hysteresis of 50°F would make more sense.

I measured the temperature of my flue pipe 1/2 way between the boiler chimney outlet and damper, and got the following temperature curve - starting at 99°F, a rise of about 15°F per minute until it started to settle out and 'knee over' at 162°F about 4 minutes afterward.

Whether you use a thermostat or controller it seems you'd need a bit of timing and hold in logic, too, something along this line ... [if burner is on for more than x amount of time, and flue temperature is less than y degrees, then "fire out" - turn off burner, sound alarm, and lock out burner until reset]. In this case I'd probably use a 2 minute burner on time, and figure for a flue temp of 120 degrees. If the burner has been running for more than 2 minutes, and temperature hasn't reached 120 degrees then the "fire out" condition would be satisfied, and burner would be locked out until reset.

This would miss a fire out condition until a demand call occurs (i.e. - a 'keep lit' timed running of the burner wouldn't trigger it, since the burner wouldn't be running for long enough), but this won't waste too much unburned coal.

well i think im gonna check my flue temperature sensor interlock device, thanks for the couple test..


_________________
4396841

While there seems to be a number of semi-complex to full-blown solutions put forth in this thread, I'm trying to find some simple 'plug & play' type protection equipment, in the $20-30 range each, to do simple tasks.

1. Something simple as: if the fire goes out, (temp too low in stove), cut off the 110v electricity. Manual reset only. What I envision is a simple 110v outlet style little box with a lead to a sensor that is in the stove (1/4" hole drilled to mount the sensor). Plug it in, set the 'low temp' limit, and that's that. This would also handle power outages as when the power comes back on, the fire is out, so unburned coal would not be pushed over the edge.

2. Variation of device #1 - If the chimney gets too hot, cut off the 110v electricity. Perhaps an optional 2nd set of leads from box #1.

3. Variation #3 - if no airflow, cut off the electricity. This would sense a convection fan failure. Such a failure would ultimately cause my stove to overheat and become damaged.

4. Variation #4 - Carbon Monoxide detector. If some limit exceeded for more than 5 minutes (30ppm?), shut off the electricity.

Each of these devices, except, perhaps #4, exists in some form in my side-vent gas water heater and gas furnace. They both sense chimney air-flow for 30 seconds or so before they 'turn on the fire'. There's also a no-fire-cut-off-the-gas sensor in them, too. I highly doubt there's any fancy computer boards, ASCII buses, or any high-tech stuff that does these functions.

For novices like myself, selecting one of a zillion sensor devices and figuring out some wiring scheme without creating a short circuit is beyond my skills and desire-to-do-so level. Drill a hole, insert sensor, tighten down, run sensor wires to device, and plug stove into it is about the limit of what I want to do. If anybody knows where I could find such critters, I'd be ready to buy right now!