Ever had a Toroidal Secondary Go Open Circuit?

[thermionic]

Hi,

A project I have (runs in Class A, so no dynamic loading) drains 1.35A @ 18VDC per +/- rail. I used a toroidal with a 2.2A secondary rating. After around 100 hrs use one of the secondaries has gone open circuit.

Description: each secondary (2 x 18 VAC) feeds its own bridge rectifier, which feeds LM338K Vreg (creating 2 isolated supplies). These in turn feed sub-PSUs. The total reservoir capacitance for the LM338K circuit is  only about 4,700uF - but the following sub-PSUs in total have around 60,000uF of reservoir capacitance prior to their 317 / 337 Vregs.

What would cause a secondary to fail open circuit? Usually an overloaded or overheated toroidal will go open circuit on the primary, right?

Could the substantial amount of capacitance coming after the LM338K pre-regulated circuit create such an inrush current spike as to damage the secondary? The secondary failed after being switched on for 4 hours...but the venue is renowned for voltage spikes...

Toroidal - 18VAC secondary - Pre-reg LM338K supply @ 21.5v - several sub-PSUs (each with substantial reservoir caps prior to Vreg) all feeding 18v out.

Thanks in advance.

 

[PRR]

The AC RMS of a cap-input DC load is 1.6 to 1.8 times the DC current.

So you are right on the edge.

Were they running HOT?

Still, the most likely reason for a single failure at 100 Hours "at" rating is a badly made joint. The low-pay workers make good joints 99% of the time. You got un-lucky.

 

[thermionic]

Thanks. The toroidal was totally cool - never above room temperature. It was a quality toroid, made in Ireland (I'm told decent toroidals are usually conservatively rated?). What strikes me as odd is that the secondary went open circuit - not the primary, which is where the damage occurs when a toroid is overheated (and where the thermal fuse is, right?); and that the damage was assymetrical, i.e. to one secondary but not other.

I've just fixed another item this week which had an overheated toroidal (18VAC x 2 - a commercial grade type). I didn't have the inclination to plug it into the mains as it'd blown a packet of the client's fuses, so instead I hooked the primary to a sig gen @ 50 Hz (50R o/p Z). When I put in 2.4v I got 180mV x 2 - but I noticed that, in order to get 2.4v at the primary, the gen had to be set for around 4.5v - thus suggesting the primary had fused a few windings together??

In the 25+ yrs I've been playing with toroidals I can't remember seeing them fail - yet I've seen 2 fail in the last week... I think it was Randy Sloane who said in his amp cookbook that your transformer is probably the most reliable component you'd ever come across, and that he had no problem using salvaged transformers that were 30 or 40 yrs old. The t/former that failed was old stock that I had in a drawer - from 2002. It was made by the highly-respected Antrim Transformers - when I recently sold one at Diyaudio (a 500VA - I regret selling it) my PM box was full. They were the mechanically quietest toroids I've ever used. When I first tested it I thought it had a fault as it made no noise whatsoever.

Cheers 

 

[mitsos]

Like PRR said, maybe an intermittent joint on the toroid.  This thread was first in the "related topics":

http://groupdiy.com/index.php?topic=22598.0

 

[thermionic]

Ha! The 'related topics' actually delivers a link which is relevant :- ) Thanks.

PRR didn't mention in-rush current from the combined reservoir capacitance (8 x 6,800uF) of the Regulated PSUs which come after the pre-reg LM338K-based power supply which is fed from the toroidal. My suspicion has always been that this may have weakened the secondary. However, when it failed, it did so after being switched on for 4 hours - not on start up. However, the venue is notorious for power fluctuations. One member of staff said the fridges started playing up if they cranked the house system beyond a certain level...

 

[PRR]

The in-rush current is seconds at most. Transformer is a heavy lump which takes a minute or more to get hot.

The power in secondary is nearly the same as the primary; I don't see why primary failure would be more common.

The random bad joint fits the clues so far.

 

[JohnRoberts]

Transformer primary open failures are probably more common because most thermal fuses are inserted in primary.

JR

 

[CJ]

a tape wound core has sharp edges, and no bobbin is used,

so those first layers can get cut like a knife, which was a good song by some canadian,

dead xfmr = take-a-part

 

[thermionic]

dead xfmr = take-a-part

Is this an offer? I've got another toroid (2 in one week - never killed one before in my life!) that died this week as well. You can do a post-mortem on that one as well if you fancy. It'll be fun when I fill in the customs declaration...'dead transformers - for dissection purposes'.

The one that failed on the secondary is a much-fabled Antrim transformer - said to be the Rolls Royce of toroidals.  The other is a non-descript one that seems to work ratio-wise, but the primary draws a large load (suggesting overheating on the primary methinks).

 

[CJ]

why not do it yourself?  it's fun!  and usually non violent, 

yours is just tape over wire, right?  maybe a little varnish?

just heat it up in an oven to 120 or get out a heat gun or blow dryer,

 

[ricardo]

..  The toroidal was totally cool - never above room temperature ...
...
but the primary draws a large load (suggesting overheating on the primary methinks).

I don't believe this.

Transformers are designed so both primary & secondary have about equal dissipation.  If the secondary is overheating, the primary will be doing so too.

And the transformer on your Class A amp should be one of the things that are stinking hot.

 

[mobyd]

I have some 100s of Kgs of useless toroidal transformers in the back of my garage. The only sin many have committed is being 110v primary in a 230v country, but a goodly number are kaput - and the failure mode is consistent across the lot (several dozen) of them : Shorted primary turns. These are out of low end Crown, QSC and JBL (EON) products and originate from China. I wish I could devise a time and energy efficient method of extracting the copper from these as the scrappy only pays $1 a kg for them, but pays $10 for copper (oh for the hippy days where many of my cohort had pottery kilns in their backyards). O/C Secondary ??? - never seen it.
M

 

[merlin]

A project I have (runs in Class A, so no dynamic loading) drains 1.35A @ 18VDC per +/- rail. I used a toroidal with a 2.2A secondary rating. After around 100 hrs use one of the secondaries has gone open circuit.

If this is a +/-15V 2.2A transformer then it going to have some pretty hefty secondary wire- there's no way you're going to melt or break that in a hurry. The primary would always break first.  I would suspect metal fatigue in the lead out wires. Does this unit get moved around or vibrated much?

 

[thermionic]

A project I have (runs in Class A, so no dynamic loading) drains 1.35A @ 18VDC per +/- rail. I used a toroidal with a 2.2A secondary rating. After around 100 hrs use one of the secondaries has gone open circuit.

If this is a +/-15V 2.2A transformer then it going to have some pretty hefty secondary wire- there's no way you're going to melt or break that in a hurry. The primary would always break first.  I would suspect metal fatigue in the lead out wires. Does this unit get moved around or vibrated much?

No. To my knowledge the t/former had never been used before. I bought it around 10 years ago, but it had lived in a drawer since. I did use it briefly for a project, but that was pulling less than 100mA, so no chance of heating it up. I've yet had the time to dissect it, but until I know conclusively I can only go on what PRR has suggested so far.

BTW - the reason it lived in a drawer was that it was too large for most projects. Bearing in mind its illustrious make, I wanted to save it for a special project...whereupon it failed after less than 20 hours of use.

 

[andy91350]

Hi everyone,

from my experience, thermal fuses are inserted into secondaries quite often, so this may be the case with this toroidal transformer.

 

[s2udio]

from my experience, thermal fuses are inserted into secondaries quite often, so this may be the case with this toroidal transformer.

???
Not in my experience.......as the purpose of said thermal fuse is to disconnect the primary ,removing the  possibility of fire or shock from potentially lethal voltages
that statement makes no sense .
And in fact is nonsense.

 

[andy91350]

Maybe, fuses in secondarieas are not so common thing,  but they do exist, and I personnally had an experience with such a transformers, so what do you mean by "nonsense"?

 

[s2udio]

So to verify ,you have seen thermal fuses embedded in a transformer in series with its secondary windings ?

 

[andy91350]

Exactly. I had similar situation, when one of the secondaries went open circuit, so I decided to remove secondaries, and found that the reason was the blown fuse. Interestingly, the fuse was provided only for one of the secondary windings.
Before that my colleague told me similar story from his experience. Maybe, we have some local anomalies here...