Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Why TWO Thermal Fuses?

1,800 views
Skip to first unread message

Jeff Wisnia

unread,
Nov 2, 2013, 8:19:26 PM11/2/13
to

SWMBO went to make a pot of coffee with our Bunn coffeemaker and
reported to me that the water which flowed through the ground coffee was
cold.

I checked for the obvious and then drained the unit and took it down to
my basement shop for a look see.

I quickly found the problem was an open 141 degree C thermal fuse
mounted on top of the heater tank. That fuse was the first thing in
series with the hot side of the power cord and fed everything else in
the coffeemaker.

But, I was surprised to find that there wasn't just one thermal fuse
there but two identical ones in series, clamped to the top of the tank
within a millimeter of each other.

I resisted the temptation to bypass the open thermal fuse and was
pleased to find that the Radio Shack store a quarter mile from our home
stocked them. Fifteen minutes later I was back home with a new $1.99
thermal fuse, installed it and the Bunn was back in business.

But why does Bunn use TWO identical thermal fuses in series? Are those
little suckers so unreliable that they felt the need to use two in case
one of them failed to open when the tank temperature soared because of a
stuck thermostat?

My curious mind wants to know.

Thanks guys.

Jeff

--
Jeffry Wisnia
(W1BSV + Brass Rat '57 EE)
The speed of light is 1.8*10^12 furlongs per fortnight.

DerbyDad03

unread,
Nov 2, 2013, 8:50:17 PM11/2/13
to
Could be a positioning strategy...

http://forum.allaboutcircuits.com/showthread.php?t=40819

bob haller

unread,
Nov 2, 2013, 9:36:54 PM11/2/13
to
the manufacturer of the coffe maaker is using a beltsand suspender approach to make certain they never cause a failure.

purchased in bulk the extra fuse likely cost under 25 cents

nestork

unread,
Nov 2, 2013, 10:47:49 PM11/2/13
to

'Jeff Wisnia[_8_ Wrote:
> ;3143616']
> But why does Bunn use TWO identical thermal fuses in series? Are those
> little suckers so unreliable that they felt the need to use two in case
>
> one of them failed to open when the tank temperature soared because of a
>
> stuck thermostat?
>

Most probably.

In order for Bunn to sell that Coffee Maker in the USA, it has to make
it past UL, or Underwriter's Laboratories. In Canada, it has to be
approved by CSA, or the Canada Standards Association.

If Bunn comes to them with a coffee maker that has a single thermal fuse
and says "We want to sell a million of these things in the USA", UL is
going to think: "If that thermal fuse doesn't open to shut off power to
the heating element if the thermostat sticks, then we have a potential
fire. Even if only one in 100,000 thermal fuses fails to blow, we're
still talking about 10 potential fires. For the few cents that a second
thermal fuse costs, it buys a lot more security that power will be shut
off to this thing if it gets too hot." After all, the chances of one
thermal fuse not blowing are low, but the chances both won't blow are
miniscule, and therefore worth the few cents that a second thermal fuse
is gonna cost.

So, most probably UL told Bunn they wanted a second thermal fuse in that
coffee maker just to be on the super safe side or Bunn designed it with
two thermal fuses just to streamline the approval process.




--
nestork

micky

unread,
Nov 3, 2013, 1:33:46 AM11/3/13
to
On Sat, 02 Nov 2013 20:19:26 -0400, Jeff Wisnia
<jwisnia...@conversent.net> wrote:

>
>I resisted the temptation to bypass the open thermal fuse and was
>pleased to find that the Radio Shack store a quarter mile from our home
>stocked them. Fifteen minutes later I was back home with a new $1.99
>thermal fuse, installed it and the Bunn was back in business.

How did you attach the thermal fuse? I'm afraid to solder them, and I
seem to have nothing good with which to crimp them.

larrymo...@my-deja.com

unread,
Nov 3, 2013, 2:58:37 AM11/3/13
to
On Saturday, November 2, 2013 5:19:26 PM UTC-7, Jeff Wisnia wrote:

> SWMBO went to make a pot of coffee with our Bunn coffeemaker and
> reported to me that the water which flowed through the ground coffee was
> cold.
>
> I quickly found the problem was an open 141 degree C thermal fuse
> mounted on top of the heater tank. That fuse was the first thing in
> series with the hot side of the power cord and fed everything else in
> the coffeemaker.
>
> But, I was surprised to find that there wasn't just one thermal fuse
> there but two identical ones in series, clamped to the top of the tank
> within a millimeter of each other.

They probably used 2 thermal fuses in series to avoid the problem GE
had with many of its drip coffee makers produced from 1976-1984
causing house fires when their single thermal fuse failed to melt:

http://www.cpsc.gov/en/Recalls/1991/General-Electric-Voluntarily-Recalls-Certain-Drip-Coffeemakers-That-May-Pose-A-Fire-Hazard/



Stormin Mormon

unread,
Nov 3, 2013, 2:58:59 AM11/3/13
to
On 11/2/2013 9:36 PM, bob haller wrote:
> the manufacturer of the coffe maaker is using a

belts and suspender approach to make certain they

never cause a failure.
>
> purchased in bulk the extra fuse likely cost

under 25 cents
>
Cheaper than fire liability claims?

--
.
Christopher A. Young
Learn about Jesus
www.lds.org
.

larrymo...@my-deja.com

unread,
Nov 3, 2013, 3:40:24 AM11/3/13
to
On Saturday, November 2, 2013 7:47:49 PM UTC-7, nestork wrote:
> 'Jeff Wisnia[_8_ Wrote:
>
>> But why does Bunn use TWO identical thermal fuses in series? Are those
>> little suckers so unreliable that they felt the need to use two in case
>
>> one of them failed to open when the tank temperature soared because of a
>> stuck thermostat?

> Most probably.
>
> In order for Bunn to sell that Coffee Maker in the USA, it has to make
> it past UL, or Underwriter's Laboratories. In Canada, it has to be
> approved by CSA, or the Canada Standards Association.

In the US there's no requirement for products to be safety approved, but
I had a $4 no-name Walmart drip coffee maker that was ETL approved that
had only one thermal fuse in series with the bimetal thermostat.

Here are some UL approved and non approved power supplies for external
hard disk enclosures. The one on the bottom is really bad and has a
low voltage circuit board (swung to the left) that normally sits right
above the high voltage board, separated only by a thin but tough piece of
plastic. The dealer claimed it was UL approved, but when I pointed out
it wasn't, they corrected their website in 30 minutes. Another customer's
hard disk blew out in such an enclosure, and the dealer not only refunded
his money but also paid for his hard disk, which had been bought elsewhere.
The AC adapter in the middle pictures, included with Bytecc and some other
brand USB hard disk enclosures, is also not safety approved, and despite
the fact it takes a 3-wire AC cord, the ground connection is fake. The
photos at the top are for a UL approved AC adapter that replaced it.

http://www1.picturepush.com/photo/a/11504419/1024/Anonymous/USB-enclosures-power-supplies.jpg

k...@attt.bizz

unread,
Nov 3, 2013, 11:42:47 AM11/3/13
to
On Sat, 2 Nov 2013 18:36:54 -0700 (PDT), bob haller <hal...@aol.com>
wrote:

>the manufacturer of the coffe maaker is using a beltsand suspender approach to make certain they never cause a failure.
>
>purchased in bulk the extra fuse likely cost under 25 cents

Probably more (I've been trying to something that cheap).

They're probably in series because they aren't very reliably and
lawyers are expensive.

k...@attt.bizz

unread,
Nov 3, 2013, 11:47:10 AM11/3/13
to
On Sun, 3 Nov 2013 01:40:24 -0700 (PDT), larrymo...@my-deja.com
wrote:

>On Saturday, November 2, 2013 7:47:49 PM UTC-7, nestork wrote:
>> 'Jeff Wisnia[_8_ Wrote:
>>
>>> But why does Bunn use TWO identical thermal fuses in series? Are those
>>> little suckers so unreliable that they felt the need to use two in case
>>
>>> one of them failed to open when the tank temperature soared because of a
>>> stuck thermostat?
>
>> Most probably.
>>
>> In order for Bunn to sell that Coffee Maker in the USA, it has to make
>> it past UL, or Underwriter's Laboratories. In Canada, it has to be
>> approved by CSA, or the Canada Standards Association.
>
>In the US there's no requirement for products to be safety approved, but
>I had a $4 no-name Walmart drip coffee maker that was ETL approved that
>had only one thermal fuse in series with the bimetal thermostat.

Correct. There is no requirement at all to have products safety
approved but it does leave one wide open to lawsuits. The first
products a former employer designed weren't approved at all. We then
went to UL for a contract, and ETL for everything else.

>Here are some UL approved and non approved power supplies for external
>hard disk enclosures. The one on the bottom is really bad and has a
>low voltage circuit board (swung to the left) that normally sits right
>above the high voltage board, separated only by a thin but tough piece of
>plastic. The dealer claimed it was UL approved, but when I pointed out
>it wasn't, they corrected their website in 30 minutes.

..or UL's lawyers would have been all over them. They're as nasty as
ASCAP's.

>Another customer's
>hard disk blew out in such an enclosure, and the dealer not only refunded
>his money but also paid for his hard disk, which had been bought elsewhere.
>The AC adapter in the middle pictures, included with Bytecc and some other
>brand USB hard disk enclosures, is also not safety approved, and despite
>the fact it takes a 3-wire AC cord, the ground connection is fake. The
>photos at the top are for a UL approved AC adapter that replaced it.
>
> http://www1.picturepush.com/photo/a/11504419/1024/Anonymous/USB-enclosures-power-supplies.jpg

A disk drive is cheaper than its contents, or a lawyer.

k...@attt.bizz

unread,
Nov 3, 2013, 11:47:51 AM11/3/13
to
On Sun, 03 Nov 2013 01:33:46 -0400, micky <NONONO...@bigfoot.com>
wrote:
Soldering is no problem. Be quick, with a hot iron.

Jeff Wisnia

unread,
Nov 3, 2013, 12:00:42 PM11/3/13
to
Yep, that's just what I did.

I thought about using a pair of long nosed pliers with a rubber band
around the handles as a heat sink between the solder point and the fuse
body, but didn't bother.

bob haller

unread,
Nov 3, 2013, 12:13:29 PM11/3/13
to
soldering thermal fuses leads to premature failure. it stresses the fuses.

I service roll laminators for a living, one manufacturer had mass thermal fuse failures, traced back to soldering them.

there are crimp connectors made just for thermal fuses

k...@attt.bizz

unread,
Nov 3, 2013, 12:20:07 PM11/3/13
to
On Sun, 03 Nov 2013 12:00:42 -0500, Jeff Wisnia
<jwisnia...@conversent.net> wrote:

>k...@attt.bizz wrote:
>> On Sun, 03 Nov 2013 01:33:46 -0400, micky <NONONO...@bigfoot.com>
>> wrote:
>>
>>> On Sat, 02 Nov 2013 20:19:26 -0400, Jeff Wisnia
>>> <jwisnia...@conversent.net> wrote:
>>>
>>>>
>>>> I resisted the temptation to bypass the open thermal fuse and was
>>>> pleased to find that the Radio Shack store a quarter mile from our home
>>>> stocked them. Fifteen minutes later I was back home with a new $1.99
>>>> thermal fuse, installed it and the Bunn was back in business.
>>>
>>> How did you attach the thermal fuse? I'm afraid to solder them, and I
>>> seem to have nothing good with which to crimp them.
>>
>> Soldering is no problem. Be quick, with a hot iron.
>>
>Yep, that's just what I did.
>
>I thought about using a pair of long nosed pliers with a rubber band
>around the handles as a heat sink between the solder point and the fuse
>body, but didn't bother.

If you've got the space to do that, they work fine. Forceps are the
perfect tool for this. For $5 at Harborfrieght, can't go wrong.
Harborfrieght is good for some things.

micky

unread,
Nov 3, 2013, 12:32:03 PM11/3/13
to
On Sun, 03 Nov 2013 12:00:42 -0500, Jeff Wisnia
<jwisnia...@conversent.net> wrote:

>k...@attt.bizz wrote:
>> On Sun, 03 Nov 2013 01:33:46 -0400, micky <NONONO...@bigfoot.com>
>> wrote:
>>
>>> On Sat, 02 Nov 2013 20:19:26 -0400, Jeff Wisnia
>>> <jwisnia...@conversent.net> wrote:
>>>
>>>>
>>>> I resisted the temptation to bypass the open thermal fuse and was
>>>> pleased to find that the Radio Shack store a quarter mile from our home
>>>> stocked them. Fifteen minutes later I was back home with a new $1.99
>>>> thermal fuse, installed it and the Bunn was back in business.
>>>
>>> How did you attach the thermal fuse? I'm afraid to solder them, and I
>>> seem to have nothing good with which to crimp them.
>>
>> Soldering is no problem. Be quick, with a hot iron.

Thanks.
>>
>Yep, that's just what I did.
>
>I thought about using a pair of long nosed pliers with a rubber band
>around the handles as a heat sink between the solder point and the fuse
>body, but didn't bother.

Thanks. I have a little set of 5 heat sinks, spring clips in two
sizes, one with the jaws bent 90^, and a fifth one with a metal clad
magnet connected to it. I also have forceps that I got at a
hamfest. But it sounds like I can do without all that if I'm quick,
and I'm quick if I try to be.
>Jeff

micky

unread,
Nov 3, 2013, 1:20:52 PM11/3/13
to
On Sun, 3 Nov 2013 09:13:29 -0800 (PST), bob haller <hal...@aol.com>
wrote:

>soldering thermal fuses leads to premature failure. it stresses the fuses.
>
>I service roll laminators for a living, one manufacturer had mass thermal fuse failures, traced back to soldering them.
>
>there are crimp connectors made just for thermal fuses

Thanks!

cl...@snyder.on.ca

unread,
Nov 3, 2013, 2:01:27 PM11/3/13
to
On Sun, 03 Nov 2013 02:58:59 -0500, Stormin Mormon
<cayo...@hotmail.com> wrote:

>On 11/2/2013 9:36 PM, bob haller wrote:
>> the manufacturer of the coffe maaker is using a
>
>belts and suspender approach to make certain they
>
>never cause a failure.
>>
>> purchased in bulk the extra fuse likely cost
>
>under 25 cents
>>
>Cheaper than fire liability claims?
At the factory door in China, closer to $0.025.

edc...@gmail.com

unread,
Apr 15, 2017, 5:52:06 PM4/15/17
to
Yes, if one in one-hundred thousand thermal fuses fails to break circuit (on an average), and just one is in series to the circuit, and they sell a million units, They can expect to burn down ten homes by calculated chance, could be more or less. Or P(fire)= 0.00001x1,000,000=10

But with two TF... P(F)=.00001x.00001x1,000,000=.0001
Or, having sold a million, there would be a 1:10,000 chance of any fires at all. They'd have the probability that they could build ten-billion units before burning one house. Life safety vs $0.99...worth it.

ncdav...@gmail.com

unread,
Mar 20, 2018, 7:52:11 PM3/20/18
to
On Saturday, April 15, 2017 at 5:52:06 PM UTC-4, edc...@gmail.com wrote:
>
> ...But with two TF... P(F)=.00001x.00001x1,000,000=.0001

That's assuming statistical independence. If they use two thermal fuses from the same batch, from the same factory, statistical independence is a very bad assumption.

If the coffeemaker manufacturers were smart they'd use thermal fuses from two different manufacturers. But, of course, they don't do that.

To a lawyer, two identical thermal fuses from the same batch probably sounds like the manufacturer was being extra careful. To an engineer, two identical thermal fuses from the same batch is probably no better than just one.

I have a Mr. Coffee sitting here with two identical thermal fuses which have BOTH failed "open," at the SAME time, after years of use. That surprised me, because for them to BOTH fail "open" in this Mr. Coffee, they must have failed at EXACTLY the same time. But it would surprise me even more if they weren't identical.

Note that if either or both thermal fuses fail "closed" then the coffeemaker will continue to work. So even if they don't fail "closed" at exactly the same time you have the same problem of two identical fuses being no better than one.

The thing that saves these manufacturers in spite of themselves is that is very rare for the thermal fuse to ever NEED to work.

People make the same mistake with RAIDed hard disk drives. They use identical drives, with the same date of manufacture, from the same factory, and subject them to identical usage patterns, and then they're surprised at how "unlucky" they are when two of the drives fail on the same day, and they loose their data from their "redundant" RAID array.

Davej

unread,
Mar 20, 2018, 8:06:21 PM3/20/18
to
On Saturday, November 2, 2013 at 7:19:26 PM UTC-5, Jeff Wisnia wrote:
> [...]
> But why does Bunn use TWO identical thermal fuses in series? Are those
> little suckers so unreliable that they felt the need to use two in case
> one of them failed to open when the tank temperature soared because of a
> stuck thermostat?
>
> My curious mind wants to know.
>

I suspect they are so cheap that it makes sense to use several
since they each detect heat only in one very small location. Plastic
appliances are quite a fire risk and can turn into balls of flame.

J. Jones

unread,
Mar 20, 2018, 11:44:06 PM3/20/18
to
replying to ncdave4life, J. Jones wrote:
> I have a Mr. Coffee sitting here with two identical thermal fuses which have
BOTH failed "open," at the SAME time, after years of use. That surprised me,
because for them to BOTH fail "open" in this Mr. Coffee, they must have failed
at EXACTLY the same time

Perhaps they both failed at the same time because they're both in proximity to
a heat source that exceeded their rated temperature -- say, when the
thermostat failed closed and didn't shut off the warmer? I'd guess both
failing open has less to do with some mystery defect in the batch of fuses and
more to do with there being a legitimate failure event that tripped them both.
Thermal fuses don't blow instantly (it takes a few seconds for the fused bit
to melt) so it's quite possible one failed shortly after the other within a
few seconds of hitting melting temperature; cutting power doesn't instantly
reduce the temperature below that point, so after the first popped and killed
the element, the second proceeded to melt anyways and boom, two open thermal
fuses.

--
for full context, visit https://www.homeownershub.com/maintenance/why-two-thermal-fuses-769088-.htm


0 new messages