Can loose connections overload a fuse?

Users who are viewing this thread

Alternety

Like an engineer
Messages
768
Reaction score
0
Points
16
Location
Washington
We should find out why the fuse is blowing. But, in a fixed resistive load like you have, eliminating the fuse should work just fine. The overall system is protected by the breaker connected to the thermostat. That breaker should be appropriate for the load on that circuit. That is easy if this is a dedicated circuit. The fuse on the SCR controller is only for fault protection of the SCR. It is not an overall safety issue. A length of resistive wire is unlikely to have a fault mode that fries the SCRs. Note that I edited my previous post to say a 25W bulb for a better safety margin for inrush current on a cold bulb. It should draw about 0.2 A hot and may got to 2 or 3 amps on turn-on. Sorry, I should have done the arithmetic.

The thermostat seems to be OK. From what I read it is a proportional control but its output is not proportional pulse modulated and uses a relay for discrete state load control. It seems to be the kind of thermostat you really want to use for this applicati0on. PID is the way to go. It learns.

One point - since you are not varying the voltage applied to the floor wire (you removed the potentiometer on the SCR controller), how about a transformer instead of the SCR thing.? A transformer will be a reactive load for the relay, but check with the thermostat manufacturer to see if it is protected from non-resistive loads. Arcing/wear of the relay contacts can result if it is not and you probably don't want to go in there and add that.

A note for the lawyers out there, advice is not guaranteed. It can be confusing to people who don't do this sort of thing.
 
Last edited:

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
We should find out why the fuse is blowing.

Yes, and I am hoping that is related to the loose output lugs.

But, in a fixed resistive load like you have, eliminating the fuse should work just fine.

At least at the moment, I do not know enough about all of this to be comfortable doing that or even how to do that.

The overall system is protected by the breaker connected to the thermostat. That breaker should be appropriate for the load on that circuit. That is easy if this is a dedicated circuit.

Understood, but I have never seen a 6-amp QO breaker. So, I must use a fuse or something else somewhere to protect my house from a heating-wire malfunction.

The fuse on the SCR controller is only for fault protection of the SCR. It is not an overall safety issue. A length of resistive wire is unlikely to have a fault mode that fries the SCRs.

Understood.

One point - since you are not varying the voltage applied to the floor wire (you removed the potentiometer on the SCR controller), how about a transformer instead of the SCR thing?

I cannot find the original discussion about all of that, but cost was an issue there. The SCR is much less expensive and I did not know the actual voltage needed until after I had some kind of controller to do some testing. So, I began with the variac and then got the SCR.

A note for the lawyers out there, advice is not guaranteed. It can be confusing to people who don't do this sort of thing.

At least you do not have to worry about me being a suer person!
 

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
Yet another point. For this use a properly sized light dimmer would work fine.

Wow! I never would have thought of that, and several months ago some of us were having such a good time talking about how bulbs-in-parallel and in series with the heating wire could reduce 120 volts to 70 volts that evidently nobody else did either! Maybe this SCR will end up on the garage shelf along with the variac ...

A 1000 Watt dimmer would work ...

I would put a stop on it, and probably a fuse, because putting 120 Volts on a heating circuit designed for 70 Volts would nearly triple the heat output.

Yes, or else as I recall being warned in large red type, I would end up burning down my house!

Also, just yesterday I began suspecting I have been putting just a bit too much voltage into my heating wire. Since it is a cut from a large mat, I had spent a lot of time measuring it and trying to figure its percentage of the whole ... and all I ever really needed to do there was to figure its percentage of total ohms:

14.4 (ohms) / 52.3 (ohms) X 208 (volts) = 57.27 volts (for this piece)

So whatever I end up doing here, I will go from 70 volts down to 60 even though the thermostat already keeps the wire from overheating itself by not allowing it to overheat the floor.
 

Alternety

Like an engineer
Messages
768
Reaction score
0
Points
16
Location
Washington
Do not assume the thermostat will protect the wire under all conditions. Even if the thermostat sensor was tightly coupled to the actual heating wire, the minimum cycle of the thermostat is 15 seconds. If you are controlling floor temp with an embedded sensor it stands a better chance of helping some. If you are measuring air temperature it will have no useful impact on controlling the wire temp.

From what I read about the thermostat, when you are referring to the % power it is talking the amount of on time vs off time, with the minimum interval of 15 seconds and max of 15 minutes.

Sometimes you have to just sit back for a bit and think about what you were trying to do, rather than how to fix what has been done. That is a generic you; not directed at you. That is what I finally did. It is actually harder than you might think for people to do that. It is kind of counter to human nature. Over the years I have periodically had to relearn that.
 

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
Sometimes you have to just sit back for a bit and think about what you were trying to do, rather than how to fix what has been done.

Yes, understood! And personally, I do not care how many times I might even have to "start from scratch", if necessary, to get something done well, properly, safely, best, rightly or whatever.

Do not assume the thermostat will protect the wire under all conditions.

Sure, and so: The bottom line there is that the wire must not ever receive any more voltage than is completely safe when on 100% of the time.

Even if the thermostat sensor was tightly coupled to the actual heating wire, the minimum cycle of the thermostat is 15 seconds ...
From what I read about the thermostat, when you are referring to the % power it is talking the amount of on time vs off time, with the minimum interval of 15 seconds and max of 15 minutes.

This particular model has a 15-minute, air-heating cycle at the end of which it evaluates what it needs to do next. All the while, however, it is monitoring the floor temperature and will immediately (or maybe within fifteen seconds) shut the power off whenever the floor arrives at its preset limit. Because I have the desired room temperature set high, the thermostat never gets to room-temperature setting even though it is always on at either 40% or 60% unless the floor is at its limit. Its output is proportional and not simply either on or off, and it learns how to stretch over fifteen minutes the total power it will actually send out.

If you are controlling floor temp with an embedded sensor it stands a better chance of helping some. If you are measuring air temperature it will have no useful impact on controlling the wire temp.

Floor temperature is limited, not controlled, and the air temperature determines whether the wire will get any power at all. In other words, the thermostat tries to use the floor to heat the air, but will always stop when the floor gets to its limit even if the air is still cold.

Here is how Aube describes that:

AF [Air/Floor] mode controls the ambient air temperature and maintains the floor temperature within desired limits using an external temperature sensor.

Also, this particular model will not work at all without a properly-functioning sensor connected.

I should have some more fuses tomorrow, but I might not be able to again try the SCR until the day after.
 
Last edited:

Alternety

Like an engineer
Messages
768
Reaction score
0
Points
16
Location
Washington
I went back and read the documentation of the thermostat. The only place I could find that they talk about the output is where they talk about percentages. The words are ambiguous. I sent an email to the manufacturer asking a question about just what they are doing.

When you are heating a floor you do not necessarily need an actual pulse width controlled output to control the heater because of the high mass. By PWM I am talking milliseconds/pulse vs every few minutes. For instance, a normal hydronic heating control system (which is the wet version of what you are doing) uses PID learning controls but they do not attempt to modulate the water flow in real time. The controls are slow acting; many seconds to open or close. So the thermostat you have may do that. Their words can, I believe, describe either way.

OK, why have I said all this. If the thermostat is a PWM output, I WOULD NOT try connecting it to your SCR box. This could be one of the problems you hare having with your fuse. The SCR box is designed to be supplied by a 60 Hz sine wave. All of the controls for the SCR switching is based on not firing the individual SCR until a specific time has elapsed after each zero crossing of each half of each sine wave. At 60 Hz. At the very least it should fail to work properly. If the thermostat switches complete half cycles rather than PWM, it could work. The fact that you got a stable 70 VAC leads me to lean toward it not being a PWM output from the thermostat. Anyway. I would suggest waiting until I get a response from Aubetech. Some of the things I am talking about probably don't make a whole lot of sense to you. But I think they are correct and it would be safer for you to be sure.

This is wearing my brain out, but I live to confuse. Have you thought of a kerosene heater :)
 

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
I went back and read the documentation of the thermostat. The only place I could find that they talk about the output is where they talk about percentages. The words are ambiguous ...

Yes, they are. So, and admittedly, much of what I am saying is based on my best recollection of a long-ago phone conversation with a technician that helped me select the model of thermostat I have.

If the thermostat switches complete half cycles rather than PWM, it could work. The fact that you got a stable 70 VAC leads me to lean toward it not being a PWM output from the thermostat. Anyway. I would suggest waiting until I get a response from Aubetech. Some of the things I am talking about probably don't make a whole lot of sense to you. But I think they are correct and it would be safer for you to be sure.

Okay, and yes, much of this is "greek" to me! When I first got the SCR, I powered it directly and slowly brought the voltage up over time while using a thermometer to observe the surface temperature of the floor. After holding that temperature at about 100* for a couple of days (using about 65 volts, as I recall), I added the thermostat and everything worked fine. When the SCR's OEM fuse blew a few weeks ago, the system had been running continually and "perfectly" for a couple of months. But, maybe your suspicions relate to some kind of accumulated stress on the SCR?

This is wearing my brain out, but I live to confuse ...

Sometimes that is a good way to learn, or at least to get outside the box.

Have you thought of a kerosene heater :)

My wife and I are not yet using this unfinished bathroom, but the cat dearly loves the warm floor and I would not want to have to clean soot from the tile if she curled and fell asleep too close to a flame.
 

Alternety

Like an engineer
Messages
768
Reaction score
0
Points
16
Location
Washington
I got a response from the thermostat tech support:

"Yes the cycle for the TH115 is 15 minutes and it is compatible with baseboards or Fan-forced heaters.

When, the temperature at the thermostat is close to the setting program, the proportional controller start to determine how much time, on the 15 minutes cycle it as to turn on the heating system to maintain the temperature.

It will repeat that four times an hour. The result of that ON/OFF cycles will gives you a temperature regulation."

So it is fine to use it.
 

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
Well, my fuses arrived and appear to be good, and my afternoon and evening plans got changed. So, I am presently using one of the fuses to power the floor wire through the variac while simultaneously using the thermostat to only check the floor temperature. Even though there is no load on the thermostat, I can still adjust its floor-temperature limit and watch to see whether it attempts to send power. One purpose there is to see what voltages and floor temperatures go together, but I am also watching the current draw to see whether the wire does anything strange as it gets warmer.

I will probably bring the SCR into the picture tomorrow evening and report back as to what does or does not happen then!
 

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
I am baffled. Here are the tests I have run:

power > SCR with fuse > different heating wire = okay
(This test wire has a higher resistance but I left the voltage alone.)

power > thermostat > fuse > variac > floor wire = okay
power > thermostat > variac > fuse > floor wire = okay

power > thermostat > SCR with fuse > floor wire = 1-minute blow
power > SCR with fuse > heating wire = immediate blow

If the wire in the floor was bad, I would think the fuse would blow whether the power is being supplied by the variac or by the SCR. But, could the SCR somehow make the fuse more sensitive to some kind of very-slight fault?
 

Attachments

  • 18d.jpg
    18d.jpg
    92.3 KB · Views: 373

Bob NH

In the Trades
Messages
3,310
Reaction score
9
Points
0
Location
New Hampshire
Circuit diagrams don't always show all of the connections because the connections are implied to people who understand the complete circuit. Sometimes there are subcircuit diagrams that may not have been furnished to you. The fact that some points are not shown as connected to anything proves that there are connections that are not shown.

What are the yellow and magenta wires shown in the diagram of the trigger of the SCR connected to? Can you draw a diagram of the complete circuit, especially any connection of those "unconnected points" to the return line through other components.

If they are connected to the return in some way, perhaps through other components that may have failed, a failure of the SCR could cause a short through them to the return line that would carry enough current to overload the fuse.

I continue to believe that the fuse blowing is related to a failure of the controller.
 

Alternety

Like an engineer
Messages
768
Reaction score
0
Points
16
Location
Washington
Have you thought of sacrificing a small mammal behind the controller?

Can you measure the resistance of both pieces of wire again. Am I right that the extra piece in the floor plus the extra one together would be a single heating mat for 240 V (I know you said 208, but that is kind of unusual for a residence)?

The different results with and without thermostat may be a function of a programmed delay in the thermostat. Do you know if the SCR powered up immediately with the thermostat?

One thought about why this might be happening. You have to think about how an SCR works. There are two SCRs and each conducts on alternate half cycles of the power line sine wave. An SCR blocks current when when the input voltage is reversed or in the forward voltage until the trigger is pulsed. Once the trigger is pulsed the SCR latches until the line voltage returns to zero at the end of the half cycle, when it turns itself off. Voltage output is controlled by delaying the turn-on until the right time within the half cycle. What happens if the the voltage output is 100% of line input, the SCR turns on at the beginning of the half cycle and the current increase follows the sine wave. If you have the SCR set to reduce voltage, and it is triggered say half way through the cycle, the load suddenly sees the line with the full half wave voltage across it. This peak voltage/current would be the same as if you connected the load to the voltage controlled source turned full on. Generally this does not bother anything. But that damn fuse's reaction time may be killing it when the resistance is as low as the floor wire. I suppose it is done, but I have never seen anyone trying to protect SCRs this way. A half wave cycle is only about 8 milliseconds. It should still be an amp or so below the fuse rating, but it might be enough. The fuses are old and we have seen that they deteriorate with age.

I am suspecting this is the fault mechanism, but I can not immediately explain why it worked for a couple of months and not now. Your test with the higher resistance supports this theory. In any modern power device at this power level I don't think that sort of fuse is used. It is just too fast. If the manufacturer insists it will damage the controller, you may wind up trying without the fuse and/or try a dimmer. I have run lots of SCRs and TRIACS at these kinds of levels without any protection except the breaker in the power panel. I have never had one fail when running them with incandescent lights (a worse case than what you are doing). I have an SCR based controller sort of like what you have, in the basement running a 1.5 HP DC motor from a 240 VAC feed. There is about a 12 A fuse in the output line, but I do not believe it is one with a couple of milliseconds response. I will try to look at it tomorrow and see just what kind it is.

Incidentally, your volt and ammeters are unlikely to show you this happening. A scope will.
 

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
Circuit diagrams don't always show all of the connections because the connections are implied to people who understand the complete circuit ...

Yes, I see that, and I am certainly not one of those people!

What are the yellow and magenta wires shown in the diagram of the trigger of the SCR connected to?

The first picture below shows their beginning and end points, but I do not know what to call the components found there. However, I have shown the printed markings on the small (1/2") components located at one end of those wires and mounted on the heat sink.

Can you draw a diagram of the complete circuit, especially any connection of those "unconnected points" to the return line through other components.

The yellow and magenta (red) wires and their attachments to the second board both commonly and exclusively connect the load end of the fuse to output terminal 4. It seems everything the controller does happens there.

One of the brown wires taps off that same circuit on the lower board, runs up to the second board and comes back to the overheat sensor where the second brown wire then completes a Y-connection to input terminal 2 and output terminal 3. I assume that is the system's electrical "common".

The black wire only and exclusively connects input terminal 1 to the powered end of the fuse, and nothing else is connected to that circuit.

Drawing that out would be a bit difficult for me, but I will do so if my words have not presented a clear picture.

I continue to believe that the fuse blowing is related to a failure of the controller.

Same here, yet there *might* be one other possibility.

Since the heating wire being used is a cut piece, I had to make the feed connection on one end. Before doing so, I carefully cut open the set-aside "factory connection" to see whether there was anything all that special about it and found what I believe to be a simple crimp. The second picture below shows the temporary connection I used with one of my other pieces of heating wire during yesterday's testing, and the connection I made and buried in the floor is similar (but to stranded wire, not solid). As best I can recall, the feed and heating wires are in contact with each other inside a crimp I believe I soldered and definitely insulated well prior to embedding everything in the floor. Is there any possibility some kind of arcing might have since developed there?
 

Attachments

  • Im602.jpg
    Im602.jpg
    58.2 KB · Views: 346
  • Im603.jpg
    Im603.jpg
    17.6 KB · Views: 345

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
Can you measure the resistance of both pieces of wire again. Am I right that the extra piece in the floor plus the extra one together would be a single heating mat for 240 V (I know you said 208, but that is kind of unusual for a residence)?

The wire I have was originally configured as a commercial-grade 208-volt mat for casting in concrete. It was/is not the typical "home version" easily purchased online even though I did happen to find it at my favorite auction site. It had been retrieved from a job site after some kind of change order.

I initially removed the mat's lacing and cut the wire in half for use in two bathroom floors. At that time, I knew nothing about resistance and voltages ... therefore assuming I could simply connect it to any power source up to 208 volts. But, and gratefully said, some of the great folks here soon straightened me out about all of that!

The original piece of wire had a resistance of 50.1 ohms (not 52.3 as I had earlier posted) hand-written on its shipping box, and I did confirm that prior to cutting the wire. Half of that wire proved to be more than I needed for this first bathroom floor, so I now have 3 pieces: 25; 14; 9. The 14 is the one the controller has been powering, and I still have plans for the other two pieces. Overall, however, we are presently still working on the best way to rightly and safely power any of those pieces.

The different results with and without thermostat may be a function of a programmed delay in the thermostat. Do you know if the SCR powered up immediately with the thermostat?

When initially brought out of "standby", the thermostat always takes a few seconds before providing power. But as soon as the thermostat clicks on, and relatively speaking, my meter shows power coming out of the controller (variac or SCR). As best I can tell, there is no soft- or step-start.

One thought about why this might be happening. You have to think about how an SCR works ...
What happens if the the voltage output is 100% of line input ... peak voltage/current would be the same as if you connected the load to the voltage controlled source turned full on.

I think I at least somewhat understand what you are describing there, but even at 120 volts my wire would still be well under the controller's 10-amp maximum while likely burning into two or more pieces and setting my floor on fire. So for what I think you are saying to happen, the controller is really neither as dependable or safe as I had been thinking.

In any modern power device at this power level I don't think that sort of fuse is used. It is just too fast. If the manufacturer insists it will damage the controller, you may wind up trying without the fuse and/or try a dimmer ...

If I continue using this controller, I will likely use a regular 5- or 6-amp fuse within so no voltage higher than the wire can safely bear can possibly get through ... and if a non-fast (NOS?) fuse causes the controller to destroy itself, then let it be done and gone!

Incidentally, your volt and ammeters are unlikely to show you this happening. A scope will.

Understood, and I again thank you and Bob.
 
Last edited:

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
Can you draw a diagram of the complete circuit, especially any connection of those "unconnected points" to the return line through other components.

Here you go, Bob. This was not quite as mind-numbing as I had anticipated.

The black components I had mentioned being attached to the heat sink are thyristors, whatever than means, and they each have three connections at that lower circuit board. Their center connections get power from the fuse, then their other connections go to and from their respective connections at the trigger circuit board. In other words, each thyristor goes there and back separately even though I have drawn only one thyristor circuit.

I also now have a scan of the trigger circuit, and I can e-mail that scan if I cannot find a place to put it with a url to post here.

If I do not end up getting this controller to work properly and safely, I would gladly mail it to you for further investigation or manipulation or whatever!

Oh, and the unused holes in the circuit board are for the control transformer I neither need nor have and they are not part of any circuit going anywhere.
 

Attachments

  • 18d.jpg
    18d.jpg
    24.2 KB · Views: 327
Last edited:

Alternety

Like an engineer
Messages
768
Reaction score
0
Points
16
Location
Washington
The heating cable is designed for 4 A RMS. That should be your target from the controller. If 120 V were put across the 14 ohm wire it would draw 8.6 A. You should be shooting for about 56 V to keep the wire within spec.

If you were to put the floor and the 24 ohm piece in series, the fuse would probably be fine. Probably even with the 9 ohm in series it might work. There is probably a minimum resistive load before the fuse reacts to the current of a switching SCR. It is less than 24 ohms and greater than 14 ohms.

With what you know now, perhaps you should try calling the controller manufacturer tech support. You have troubleshooting data now. Not just calling and saying my fuse breaks. They should know if what I have suggested can happen. Or give a better idea of what may be wrong.
 

Leejosepho

DIY scratch-pad engineer
Messages
2,483
Reaction score
0
Points
36
Location
200 miles south of Little Rock
Website
www.nonameyet.org
You should be shooting for about 56 V to keep the wire within spec.

Understood.

If you were to put the floor and the 24 ohm piece in series, the fuse would probably be fine. Probably even with the 9 ohm in series it might work. There is probably a minimum resistive load before the fuse reacts to the current of a switching SCR. It is less than 24 ohms and greater than 14 ohms.

I do not understand how that kind of thing works, but I do trust what you say. The papers that came with the controller say the RL (little L) must be greater than 12 ohms ... but I believe that is what the controller needs for itself regardless of how the fuse acts.

With what you know now, perhaps you should try calling the controller manufacturer tech support. You have troubleshooting data now. Not just calling and saying my fuse breaks. They should know if what I have suggested can happen. Or give a better idea of what may be wrong.

Yes, thank you!

Also, my last salvage fuse just blew a little while ago between the variac and the wire at 66 volts, but the wire is still open to ground and its resistance is where it should be. So, I am going to get a couple of regular and *new* 5-amp fuses to see what they do. I need to see a right-sized fuse be okay here before I can be confident about continuing to send power into my floor.

Bob: I got this small enough to post:
 

Attachments

  • scan11--.JPG
    scan11--.JPG
    97.4 KB · Views: 344
Top
Hey, wait a minute.

This is awkward, but...

It looks like you're using an ad blocker. We get it, but (1) terrylove.com can't live without ads, and (2) ad blockers can cause issues with videos and comments. If you'd like to support the site, please allow ads.

If any particular ad is your REASON for blocking ads, please let us know. We might be able to do something about it. Thanks.
I've Disabled AdBlock    No Thanks