Welcome! Here are the website rules, as well as some tips for using this forum.
Need to contact us? Visit https://heatinghelp.com/contact-us/.
Click here to Find a Contractor in your area.
InfinityLab i8124-A Boiler Triple Limit
Bob Tonner
Member Posts: 64
With regard to the i8124-A (which, by the way, is being renamed 8124i to keep it in sync with our other product offerings) there was a problem and we are offering a free software upgrade. Last winter we started to get some tech calls about the control locking out and giving a "sensor trouble" error code. Upon investigation we found that in certain regions of the USA electrical noise was being induced onto the sensor wires.
Now, this is where those who are not interested in the fine technical details should skip down to the bottom of the page. For the others who care about why things are the way they are I am going to give some detail here.
One of the great advantages of microprocessor designs is that a manufacturer has a new opportunity for safety that was not present in the old electro-mechanical designs. That is, in the older style limit controls the safety of the design was based upon the robustness. Specifically, in order to make sure that the contacts on a limit don't weld together a manufacturer chooses a good quality relay and then tests it for 100,000 operations. Once the product is built and installed there is no backup system to prevent an unsafe condition in the event that the relay does, in fact, weld shut. Similarly, there is no back up in case the capillary-type sensors fail. If they develop a leak, or if the tube is damaged in a certain way they operate in an unsafe manner.
Microprocessor designs, on the other hand, offer a method of increased safety called "supervisory systems". Supervisory systems are circuits and software that monitor certain critical functions on the control and then, in the case of failure (or perceived failure), will shut the control down to prevent an ugly situation.
In our case we have many supervisory systems. One method that we employ to ensure that the B1 terminal is always under control is to provide two (2) solid-state relays (SSR) in series. These are monitored continuously by the microprocessor for proper functionality at all times possible.
But the area that was problematic for us last winter was that the sensor design uses two electronic temperature sensors (thermistors) side-by-side encapsulated in the tip of the sensor probe. There is a common 5-volt supply to them and two return signals to the microprocessor. The software was programmed to:
1. Take a measurement of each one
2. Compare the results to each other and ensure that they are within a certain tolerance of error.
3. If the sensors are within tolerance, the average of the two is taken.
4. This average is fed into a software module that "smoothes out" the results.
Now, in step 2 above we said that the software checks to see if the signals are within a certain tolerance of error. If they are not, which can happen in the case of electrical noise from various sources, the software would discard the signal and record the occurrence. If there were more than 256 occurrences of bad signals in one burn cycle the software concluded that something was wrong and would shut down the control, and report a problem with the sensor. What we did not expect was the level of consistent electrical noise in some areas of the USA. It was definitely regional, in that there were a lot of problems in Hartford CT, but virtually none a half hour up the interstate in Springfield MA.
We uncovered this problem in February 2004 but were not aware of how many times it was occurring. I put my phone number (direct to my office) on every single control but the number of calls to us was quite small. The method mentioned above, of counting the errors and then shutting down is called a "threshold limit". We made the decision in February to change to a "Majority Detect" method. This means that rather than counting the errors we compare the number good signals to bad, and if the majority of signals are good, then the bad ones are rejected. Once we completed testing the upgraded software was put into production in May 2004.
The decision to make public this situation was put back into my hands in November when marketing of our products in the USA was taken on directly by InfinityLab Inc.. So, in November our representatives in the USA were given a technical bulletin outlining the situation and offering a "free software upgrade" to all wholesalers that were carrying old stock.
Furthermore, we have upgraded our Warranty from 3 years to 5 years, and have a bonus replacement guarantee. That is, if any of our controls fails within the first 12 months of the installation not only will we replace it according to our warranty, but we will send out a voucher for a free 5100i Oil Burner Primary Control for the inconvenience.
To those who have tried our product and were unfortunate enough to have a problem; my sincerest apologies. We work tirelessly here to make new, exciting, cost-effective and reliable product and this was a big blow to our hopes for the US market.
Here is the Technical Bulletin that we have distributed to wholesalers recently:
______________________________________________________________________________________________________________________
Technical Bulletin
#1104-03
18 November 2004
i8124-A Free Software Upgrade
Affected Units: i8124-A
Date Codes: 0903, 1203, 0104, 0204, and 0404
InfinityLab Inc. has become aware that there have been some reliability concerns with some early production units of the i8124-A Boiler Triple Limit control. There has been less than 100 incidents reported to InfinityLab Inc. Our investigation has isolated the following:
Nuisance Lockouts: Some units produced and shipped before May 2004 exhibit a tendency to lockout and display a diagnostic code indicating a faulty sensor. This diagnostic code was displayed through the red LED, which would flash three times. Simultaneously the control permanently shuts the boiler down. Field reports showed that when power was removed and then restored the problem might go away permanently, or could return at random intervals. InfinityLab found that the problem was related to the inability of the software to distinguish (during prolonged burn cycles) between excessive electrical noise from random external sources and a failure of the electronic temperature sensors embedded in the sensor tip.
Safety Risk: None
Other Risk: Unoccupied or otherwise unsupervised buildings may have heating systems shut down for long enough periods of time to cause damage to the heating system or the building contents.
Solution: In units with date codes 0405, 0407, 0408, and 0409 and later the software has been upgraded to filter out spurious electrical noise. InfinityLab has confirmed this solution by upgrading the software in known failed units, and then re-installing these units in the same residential application wherein they first failed. All units are continuing to work without incident.
Customer Action:
Although the number of failed units is miniscule in relation to the number of units installed, InfinityLab prides itself in the quality of such products and wishes to both retain and enhance its reputation. Accordingly, units bearing the affected date codes should be dealt with as follows:
· Installers: Return such units to wholesaler for exchange.
· Wholesalers: Contact your InfinityLab Sales Representative to have such inventory returned to InfinityLab for free software upgrade.
InfinityLab Inc sincerely regrets any inconvenience that may have been caused to our customers. Please be assured that we work diligently toward continuous improvement in our designs to ensure that the industry is offered a viable and cost-effective alternative to traditional sources.
If you have any questions that cannot be answered by your local representative or on our web site (www.infinitylab.com) please feel free to contact me directly at (905) 619-0837
Robert B. Tonner
President
InfinityLab Inc.
1590 Greenmount Street
Pickering, Ontario,
Canada
L1X 2H9
Now, this is where those who are not interested in the fine technical details should skip down to the bottom of the page. For the others who care about why things are the way they are I am going to give some detail here.
One of the great advantages of microprocessor designs is that a manufacturer has a new opportunity for safety that was not present in the old electro-mechanical designs. That is, in the older style limit controls the safety of the design was based upon the robustness. Specifically, in order to make sure that the contacts on a limit don't weld together a manufacturer chooses a good quality relay and then tests it for 100,000 operations. Once the product is built and installed there is no backup system to prevent an unsafe condition in the event that the relay does, in fact, weld shut. Similarly, there is no back up in case the capillary-type sensors fail. If they develop a leak, or if the tube is damaged in a certain way they operate in an unsafe manner.
Microprocessor designs, on the other hand, offer a method of increased safety called "supervisory systems". Supervisory systems are circuits and software that monitor certain critical functions on the control and then, in the case of failure (or perceived failure), will shut the control down to prevent an ugly situation.
In our case we have many supervisory systems. One method that we employ to ensure that the B1 terminal is always under control is to provide two (2) solid-state relays (SSR) in series. These are monitored continuously by the microprocessor for proper functionality at all times possible.
But the area that was problematic for us last winter was that the sensor design uses two electronic temperature sensors (thermistors) side-by-side encapsulated in the tip of the sensor probe. There is a common 5-volt supply to them and two return signals to the microprocessor. The software was programmed to:
1. Take a measurement of each one
2. Compare the results to each other and ensure that they are within a certain tolerance of error.
3. If the sensors are within tolerance, the average of the two is taken.
4. This average is fed into a software module that "smoothes out" the results.
Now, in step 2 above we said that the software checks to see if the signals are within a certain tolerance of error. If they are not, which can happen in the case of electrical noise from various sources, the software would discard the signal and record the occurrence. If there were more than 256 occurrences of bad signals in one burn cycle the software concluded that something was wrong and would shut down the control, and report a problem with the sensor. What we did not expect was the level of consistent electrical noise in some areas of the USA. It was definitely regional, in that there were a lot of problems in Hartford CT, but virtually none a half hour up the interstate in Springfield MA.
We uncovered this problem in February 2004 but were not aware of how many times it was occurring. I put my phone number (direct to my office) on every single control but the number of calls to us was quite small. The method mentioned above, of counting the errors and then shutting down is called a "threshold limit". We made the decision in February to change to a "Majority Detect" method. This means that rather than counting the errors we compare the number good signals to bad, and if the majority of signals are good, then the bad ones are rejected. Once we completed testing the upgraded software was put into production in May 2004.
The decision to make public this situation was put back into my hands in November when marketing of our products in the USA was taken on directly by InfinityLab Inc.. So, in November our representatives in the USA were given a technical bulletin outlining the situation and offering a "free software upgrade" to all wholesalers that were carrying old stock.
Furthermore, we have upgraded our Warranty from 3 years to 5 years, and have a bonus replacement guarantee. That is, if any of our controls fails within the first 12 months of the installation not only will we replace it according to our warranty, but we will send out a voucher for a free 5100i Oil Burner Primary Control for the inconvenience.
To those who have tried our product and were unfortunate enough to have a problem; my sincerest apologies. We work tirelessly here to make new, exciting, cost-effective and reliable product and this was a big blow to our hopes for the US market.
Here is the Technical Bulletin that we have distributed to wholesalers recently:
______________________________________________________________________________________________________________________
Technical Bulletin
#1104-03
18 November 2004
i8124-A Free Software Upgrade
Affected Units: i8124-A
Date Codes: 0903, 1203, 0104, 0204, and 0404
InfinityLab Inc. has become aware that there have been some reliability concerns with some early production units of the i8124-A Boiler Triple Limit control. There has been less than 100 incidents reported to InfinityLab Inc. Our investigation has isolated the following:
Nuisance Lockouts: Some units produced and shipped before May 2004 exhibit a tendency to lockout and display a diagnostic code indicating a faulty sensor. This diagnostic code was displayed through the red LED, which would flash three times. Simultaneously the control permanently shuts the boiler down. Field reports showed that when power was removed and then restored the problem might go away permanently, or could return at random intervals. InfinityLab found that the problem was related to the inability of the software to distinguish (during prolonged burn cycles) between excessive electrical noise from random external sources and a failure of the electronic temperature sensors embedded in the sensor tip.
Safety Risk: None
Other Risk: Unoccupied or otherwise unsupervised buildings may have heating systems shut down for long enough periods of time to cause damage to the heating system or the building contents.
Solution: In units with date codes 0405, 0407, 0408, and 0409 and later the software has been upgraded to filter out spurious electrical noise. InfinityLab has confirmed this solution by upgrading the software in known failed units, and then re-installing these units in the same residential application wherein they first failed. All units are continuing to work without incident.
Customer Action:
Although the number of failed units is miniscule in relation to the number of units installed, InfinityLab prides itself in the quality of such products and wishes to both retain and enhance its reputation. Accordingly, units bearing the affected date codes should be dealt with as follows:
· Installers: Return such units to wholesaler for exchange.
· Wholesalers: Contact your InfinityLab Sales Representative to have such inventory returned to InfinityLab for free software upgrade.
InfinityLab Inc sincerely regrets any inconvenience that may have been caused to our customers. Please be assured that we work diligently toward continuous improvement in our designs to ensure that the industry is offered a viable and cost-effective alternative to traditional sources.
If you have any questions that cannot be answered by your local representative or on our web site (www.infinitylab.com) please feel free to contact me directly at (905) 619-0837
Robert B. Tonner
President
InfinityLab Inc.
1590 Greenmount Street
Pickering, Ontario,
Canada
L1X 2H9
0
Comments
-
Excellent read. That is good that you caught this minor ....
technicality. best of luck in the future ....it is difficult to catch every corner some experiences we would all rather not have...catching them before they are a nusciance speaks well of a company...0 -
Hey Bob,
Posting this shows how committed you are to the industry too. Stepping up to the plate and laying the "facts" on the line....What a concept!
Wish I could say I've heard the same from some other companies in the 20 years I've spent in this fine field.Any admission of a problem, has never been as forthright as your post anyway!
Thanks for the detailed explanation. Some may be bored by the tech. speak spewed forth, but the guys that really want to know why, are sitting on the edge of their seats waiting for an explanation. Great find and the backing of any "problem units" is an even higher accolade to your belief in your product line.
Happy New Year my friend. Chris
0 -
Bravo Bob...
you have cajones.
It takes cajones to stay in business.
ME0 -
Another reason why
Chris thinks so highly of you Bob.
Marks right on the money, you got cajones !
Scott
To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"0
This discussion has been closed.
Categories
- All Categories
- 86.3K THE MAIN WALL
- 3.1K A-C, Heat Pumps & Refrigeration
- 53 Biomass
- 422 Carbon Monoxide Awareness
- 90 Chimneys & Flues
- 2K Domestic Hot Water
- 5.4K Gas Heating
- 100 Geothermal
- 156 Indoor-Air Quality
- 3.4K Oil Heating
- 63 Pipe Deterioration
- 916 Plumbing
- 6K Radiant Heating
- 381 Solar
- 14.9K Strictly Steam
- 3.3K Thermostats and Controls
- 54 Water Quality
- 41 Industry Classes
- 47 Job Opportunities
- 17 Recall Announcements