Power Transformers With Internal Thermal Protection Fuse

[oc disorder]

I know best practice is to replace the transformer as it's safety may be compromised but when its encased in metal on a metal chassis with the correct value fuses in the device it powers I consider it to be safer than a lot of the domestic cheap heaters (of the "warm the house" variety).

The main culprits I gather were the under rated wall wart plug packs powering answer phones faxes etc that were left on all the time.

It seems in order to be competitive and keep the price down the less copper
the better.. they always seem to get remarkably hot and when on skirting boards and around curtains well it's um curtains for a lot of them causing house/office fires.

I guess this is why switchmode are leglislated in by their nature ..and yes RG 1 spot adapters are now in kangaroo land !
Deluxe Guitars - South Melbourne VIC Australia.

HTH's post re Fender Stage 1600's £105 ($166) transformer
and thermal fuse issues http://music-electronics-forum.com/t9804/
+ some pics of typical therm fuses and symbol

[jrfrond]

All of the aforementioned points are extremely valid. If you own or operate a repair shop, here's one more: replacing the thermal fuse in a transformer voids all safety ratings (UL, CE, CSA) and puts you in the direct line of fire (no pun intended) should an accident occur as a result of the transformer tampering. And rightly so. Severely overheated transformers should NOT be put back into service.

[J M Fahey]

I have had 3 classic 30's with blown therm fuse transformers they pop at about 144C....

144ºC? Wow!!
That's too much!!
Well into the plastic bobbin melting area, which by definition renders the transformer useless, even if (still) not shorted.
Not forgetting that temperature inside the transformer is not uniform, there may very well be some "hot spots" 20 or 30ºC above the average.

[Mark Black]

Let's not forget that these fuses also blow at a certain amperage, so a quick over-current event might be to blame. That fact is what actually bugs me the most about these embedded fuses - if they were purely temperature-sensitive that's all well and good, but in fact they can blow due to current draw so in my opinion should be more accessible for replacement.

[J M Fahey]

Wasn't sure about this, so I checked it up:
Wiki says (in "Thermal_Cutoff"):

Unlike electrical fuses or circuit breakers, thermal fuses only react to excessive temperature, not excessive current, unless the excessive current is sufficient to cause the thermal fuse itself to heat up to the trigger temperature.

Cantherm Thermal Fuses from Cantherm says:

In order to repair the circuit, the complete thermal fuse must be replaced. Thermal fuses have a solid, dust and dirt- tight housing. They react to ambient temperature and are generally insensitive to current at rated levels.

So in general they state that although in an indirect way they may have some current sensitivity, it's only a byproduct of the thermal issue.
EDIT: just checked Maplin Thermal Fuses : Thermal Fuses : Maplin : the ones they sell blow at 40A or 20A , independent from temperature.
That's *way* above any reasonable fuse specified in a regular guitar amp.

[R.G.]

Yep. They're designed to be as insensitive to current as possible. However, there is some current that will melt *anything*. The self heating in a thermal fuses is pretty negligible.

Another issue is that to work well, a thermal fuse has to be in contact with the thing it's protecting, and ideally imbedded inside so that air cooling does not keep it cooler than what it protects. There are ways around this, of course, but they're expensive and require more care - and money - in making the transformer, so they don't as a rule get done. The thermal cutouts are put in because they're *less undesirable and expensive* than other ways to meet the letter of the specifications on fire safety.

Fundamental to the idea of protection devices is the concept that they in effect decide to stop things before damage actually occurs. So if a protective device operates, it is likely that things aren't damaged. It would be a reasonable attitude to say "Whew! I'm glad that operated and kept me from having to pay to get actual damage repaired.". But that's not what happens, with musicians at least. Their reaction is "Hey! It just quit. Nothing is damaged, so it should not have quit. Must be that $#&*() protector keeping me from doing my music."

Back in my power supply design days I once had a field problem with an installation where a 5V, 25A supply was intermittently going down, but not having any permanent damage. The customer was very frustrated that these "false trips" were giving them problems. One of our power supply group came up with a good solution. We'd put a 50A relay in there and wire it to the shutdown signal so when that was detected, the relay would reverse 5V and ground and burn out 100% of the logic in the thing. This would convert an intermittent into a solid failure and make the customer happy. Right? >8->

Took replacing the whole machine, and when we finally found the bug, there was a logic flaw turning on all the logic gates opposing each other at times, tripping the overcurrent. They never let us use the relay.

If nothing is damaged, customers think that all protection is worthless, not saving their butts before damage happens.

[km6xz]

A hobbyist who uses the gear only himself might thin bypassing the thermal protection is a reasonable risk....but is it? By not having UL CE or other compliance, your home owner's insurance is also voided for damage caused by your disabling it. Any shop that does it, should be closed down as a threat to the community. Look back through archives of newspapers before UL fire certification and you will see that house fires were very common in the early days of electric service to homes, it was a major cause of death. It still is in countries that overlook regulations. More people where I live die each year from electrical fires even now than any other form of home event. In the US, there are 310 deaths average a year according to FEMA, only 1/100th the shooting deaths. The introduction of strict safety protection regulations has dropped the electrical fire death rate by a factor of 100.
The thermal switches are not current fuses, they open when the transformer has gotten so hot that it can no longer be safely put back into operation. The transformer is heat damaged. You can replace the thermal switch if for your own unit but not for a customer.

[nashvillebill]

Yet (with my limited experience at least) these thermal fuses DO seem to pop with inrush current. Case in point: our kitchen blender, cost well over $100. Worked great for about 2 months then went dead on turn-on. Motor winding with thermal fuse tested open. Returned under warranty. Replacement worked about a month...then same thing. Died when turned on, motor winding tested open. Returned AGAIN under warranty...guess what! SAME THING. The 4th unit has lasted a few months now. (Internet searches showed a number of complaints on same model blender with similar symptoms)

My guess is that the motor manufacturer bought a batch of thermal fuses made under poor quality control. So maybe in theory they should not have been current-sensitive, but certainly the inrush current did all of ours in, and the blender in each case was cold, it had not even run yet.

[J M Fahey]

Maybe, why not ?
Although there is also another possibility: all thermal fuse manufacturers *insist* on not soldering them, for obvious reasons, but crimping or similar "cold" mounting methods.
A poorly crimped one, coupled to a strongly vibrating environment (blender motor) is not a recipe for reliability.
Just as a side thought, it might very well have failed from any other cause, including a purely electrical one.
Unfortunately, pouring $100 into a blender is not a guarantee of quality anymore.

[R.G.]

Case in point: our kitchen blender, cost well over $100. Worked great for about 2 months then went dead on turn-on. Motor winding with thermal fuse tested open ... (Internet searches showed a number of complaints on same model blender with similar symptoms)

My guess is that the motor manufacturer bought a batch of thermal fuses made under poor quality control. So maybe in theory they should not have been current-sensitive, but certainly the inrush current did all of ours in, and the blender in each case was cold, it had not even run yet.

I don't see any reason for "and yet".

It should be self evident that there are cases where an overcurrent inrush will melt anything. I've melted solid steel rods about 1/4" thick accidentally. I was using a welder set to limit current at 120A. The locked-rotor current of a faulty motor can get to that in the milliseconds before the wall-breaker trips. In the absence of more information, it would be just as reasonable to say that the motors were faulty and were killing properly made thermal fuses.

Of course they're current sensitive. Everything is. But that's not what to focus on. If I were responsible for a design that was burning out thermal fuses on an inrush to a cold transformer or motor, I would carefully examine what the inrush current was, and whether the thermal fuse even had a chance to survive.

Hmmm. A little thought leads me to most devices needing some way to prevent primary overcurrent with a current sensitive fuse, and then possibly a thermal fuse to protect against overtemp, that being a similar but slightly different issue. Both thermal and current fuses rely on the melting point of a low-melting alloy. For current fuses, the idea is to make it melt at a specific current under self-heating. For temp, you make it bigger to carry more current to minimize the self heating and be sensitive mostly to the heat coming in from outside.

Nothing says that somebody would not try to get clever and combine both of those into the same device, and still call it a "thermal fuse". As an engineer, it occurs to me that if it's tricky designing an overcurrent fuse right, and it's tricky designing a thermal fuse right, it's tricky-squared to design an overcurrent/thermal fuse right.

That's the problem with anecdotal evidence, no matter how many times it's repeated. It can offer a place to start digging, but there isn't really a gold nugget in one's hand until the digging is done.

[jrfrond]

As a tangent to the whole safety ratings issue, performing a voltage conversion mod on a unit by swapping a domestic or export transformer ALSO voids the safety ratings. The only "good" voltage conversion is one where a multi-tap transformer exists.

When customers ask us to do voltage conversions, we sell them step-up/down transformers. We will not modify any internal power supplies. If they piss and moan about it, we tell them to go elsewhere. Particularly in this litigious society we live in, one good lawsuit can close your doors for good. It's not worth the money that is turned down initially.

[Prairie Dawg]

RG, ain't that what a properly rated thermistor is supposed to do?

[Enzo]

A thermistor starts out with some rated resistance when cold (room temperature), then as it heats uo, the resistance goes DOWN. That is why they are useful as inrush limiters in the amplifier mains circuit. The couple amps of current flowing through it heats it up over a few seconds so it drops to a low resistance that essentially vanishes from the circuit.

[R.G.]

Thermistors to reduce inrush current? Sure. That's what a properly rated thermistor is supposed to do.

The problem is that all of these - power line fuses, secondary fuses, thermal cutoffs, inrush limiters, etc. - are there because someone predicts something bad will happen, based on either pessimism or experience, and the problem is then to make the vast majority of all of the units work correctly for their useful lives. It's absolutely certain that they all won't work right; some of them *will* fail. So the designer is in the business of divining the future, and putting in enough to keep the customers happy without losing his job because the boss' nephew who runs accounting thinks he put in too much.

If all this could be calculated, it would be no problem. Because of the real nature of the Real World, it's always a prediction, and there is a penalty for guessing either too high or too low. In fact, in some companies, there is a penalty for guessing exactly correctly, because there's always some newly-minted hotshot who'll second guess it and do the analysis that they could have had more profit by allowing more of them to fail.

Things that are certain to work forever cost too much for the economics we live under.

[R.G.]

A postscript to my last observations. One of the personal joys I get out of living like a technology monk, spending a preponderance of my adult life learning technologies, is that I can put the little niceties into my stuff that makes it more bulletproof and resistant to trouble. I recently got two dozen 75C resettable thermal cutouts from a surplus outlet for a buck apiece. That happens to be the perfect temperature cutout to put on a power amp heat sink to shut things down safely. I can do that because I can find them, know which ones to buy, wire them up safely, and know what's happening when one trips.

But learning all that took a long time. I wonder sometimes if I should have just practiced playing the guitar.