R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: How Transformers Fail  (Read 1657 times)

klaus

  • Moderator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2203
How Transformers Fail
« on: June 02, 2023, 03:00:05 PM »

The author of this post, Bryan Sours, is a highly regarded, sought-after transformer specialist. He designed a PT, OT and choke for the 18W Marshall 1974x, transforming (!) the little brother of the mythical JTM 45 "Bluesbreaker" combo into a tone monster, without incurring hearing damage K.H.
--------------------------------------------------------------------------------------------------------------------------------------------------

Hey all,
Klaus asked me to pop in and kick some thoughts around about how that KM84 transformer might have failed. (https://repforums.prosoundweb.com/index.php/topic,37648.0.html)

At my shop we predominantly do paper layer winding with the exception of a few bobbin wound designs, but I've torn down enough BV-type transformers (as used in Neumann mics) to have a good feel for what makes them special and how they're likely to fail.

Phantom power generally shouldn't be an issue with them as Klaus mentioned.  One of the functions of the floating secondary is to isolate external dc voltage systems from the primary side preamp circuit, power supply, and capsule. 

DC Voltage on the secondary winding (with no leakage to ground and no tap/CT to ground) is effectively unseen by the transformer.  The DC is on the winding, but with no path to ground. 

But DC current is a different story. If one side of the secondary sees any path to ground, the P48 voltage present will draw DC current through the secondary winding. It’s current limited by the 6800Ω usually present in the preamp (not always 6800Ω unfortunately), but because the wire in the secondary is so small (doesn't need to be big - its AC current spec is minuscule), even small levels of DC current can cause the winding to go open. 

Plug in a cable with a connector pinned incorrectly, or frayed wire shorting a pin, and it is possible to eventually cause an open coil in the transformer secondary with Phantom Power.  It won’t happen right away, and it probably won’t happen with a properly designed preamp, but it is possible given all the wrong circumstances. 

You would most likely hear that issue happening as well: Current limited P48 (6800Ω) is something like 7mA DCI which would draw through the secondary of the OT, which would bump the flux density in the core up, and with the high permeability Permalloy laminations in there, your saturation point is very very low. That means you would surely hear your low end turn to garbage if there was standing current in the secondary.

I agree also that voltage spikes are to be avoided (with all microphones).
These types of transformers (high permeability core, high inductance, low leakage inductance) couple very low and very high frequency information between the primary and secondary, so although the primary-secondary isolation aspect of the design is there, the spike transient can reflect back to the primary via the turns ratio, which is relatively high in these types of transformers. 

For this reason there's usually an RC ramp on the Phantom Power switch of mic preamps. The failure mode in this sort of case is most likely going to be primary-to-secondary breakdown. Tesafilm, Mylar, or polyester (tape mostly - haven't seen anyone use sheet yet) was generally used as inter-winding insulation in these, so the voltage breakdown rating is really high (that was their intention along with low LL I'm sure), but repeated spikes could breakdown the material (and the wire insulation). This becomes exasperated by the primary to secondary capacitance, since the interwinding insulation is so thin there’s a lot of C between the two sections.  May not cause a failure the first time, or even the 20th time, but eventually is could break down.  Klaus- that’s basically how your Marshall 1974X OT failed (corona burst through the polyester tape).  Poly insulation isn’t perfect and it can (and does) fail.

In my opinion, the biggest source of failure in this type of transformer is physical. You're random-winding insanely tiny wire (42-45 awg) on a bobbin usually made of PET or (modern builds) nylon. If you've ever grabbed a thicker sheet of PET it's really sharp on the edge - and brittle.  Even nylon bobbins can have sharp edges, and with how small that wire is, even the slightest burr on the bobbin can cause a break eventually. 

With that smaller wire, winding tension is absolutely critical as well. If your entrance angle to the bobbin isn't just right, tension will increase more than expected as the winding builds up, and by the end the strain on the wire is much higher than when you started. If you've got a hard edge under the wire at the starting layer, you've now got even more downward pressure on it. Then add a secondary on top of that primary winding. Problem compounded.  The output transformer has nested leads in it, so wire build up will be irregular and this decreases production consistency in the transformers, in addition to the random winding. That means some transformers might eventually have a problem. 

When it gets installed at the factory the leads get moved around and probably yanked a bit, this likely filters out most of the output transformers with immediate manufacturing defects, so any that will exhibit physical problems will be long out of a warranty period. 
Additionally these transformers typically aren't vacuum pressure impregnated (there's exceptions to that, I believe), so wire does move internally despite a very tight wind. Combined with the potential issues involved with winding it in the first place, a wire break can easily happen when the microphone is of an older age (they're just like us!).

Long post barely scratching the surface on this stuff, but fun to wonder how the thing broke.  Glad you fixed it and can get back to making music with it!
-Bryan Sours // Soursound
Logged
Klaus Heyne
German Masterworks®
www.GermanMasterworks.com
Pages: [1]   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.204 seconds with 17 queries.