Do output transformers "saturate"

[Mike K]

https://victoriaamp.com/understandin...by-mark-baier/


Here's something I ran across the other day, not even really looking, but this guy is well respected in the amp industry. I'm thinking perhaps the OTs he's talking about here were just primary inductance limited and reducing low frequencies in that way. This "saturation" business seems to be not very well defined. Again, in an attempt not to hurt anyone's feelings here, I'll say that Mark is perhaps just not using the correct words... There's probably some truth in what he is saying, it's just not being conveyed properly.

[teemuk]

Why do we have to turn the bass to "0" on those vintage Fender amps to get rid of the flub at overdrive then...?

Smells like a thinly veiled ad. Seen it all before: write something technical that surprisingly puts on the frame the products you happen to sell. I treat this sort of stuff with the same grain of salt as Sozo's capacitor "break in" or Mercury Magnetics glorifying the wonders of rusted transformer laminations. Read between the lines and it's all just an advertisement.

[Mike K]

Of course there is a lot of that.


I think, in some instances, the transformer has nothing to do with flub, it's all coupling capacitors. But there are certainly amps in which the stock transformers were bass limited - I only know about the 5C1 champ and the AC4, personally. They still have too much bass for heavy overdrive, IMO. And this can be remedied with coupling capacitors even without touching the OT. WTBS, there are times when increasing the OT size actually makes the amp a bit too "woofy" on the low end. The combination of both a larger core OT and large coupling caps is not always good for overdrive. Generally, I think, when you tune the preamp of the amp to cut the bass, you wind up with very little or light bass distortion, which can come through a larger OT as being tighter and clearer.


It also can come down to the speaker you are driving. An 8" speaker doesn't typically produce a strong bass (although some can), and most tend to be more midrangey that a similar 10 or 12". If you're designing to drive said speaker, it can be used as part of your tone shaping and overall presence.

[Helmholtz]

What else ? How about primary impedance ? On the charts I looked at speaker impedance doubled at 5k . If the load line goes through the knee the load line for 5k would rotate counter clockwise and hit the shin at about Vgk equals -2 giving less gain . Same for bass resonance . This would result in a midrange emphasis . Does this make any sense ? I could be wrong , my understanding of magnatics is quite thin .

Pentode gain is given by gm times Ra. So higher load impedance increases gain (but may lower available power) and means a higher voltage across speaker terminals.
In fact measured sonic frequency response shows higher SPL where speaker impedance is high.

Global NFB tends to equalize this effect, so a speaker typically sounds brighter without NFB.

[Helmholtz]

If the amp uses global NFB, bad sound with a certain OT might be due to signal-triggered oscillation.
Different OTs sometimes require different frequency compensation for stability.

[Helmholtz]

I can also comment that there is some leakage inductance energy exhibited (the inherent reason for the test) at the time of turn-off of the conducting half, which then forces the transient voltage overshoot on that winding (and hence the interest in the MOV management of that transient) as well as transfer of that leakage energy over to the other half-winding to force a transient voltage undershoot.

I assume "no-load" means worst case regarding the spikes.
What happens with a load connected?

The other observation is that during cut-off, and after the other half-primary voltage has collapsed back to B+, the plots show the 'off' half-winding voltage noticeably changing (albeit to a minor level) and during the time that the other conducting winding has its current ramping down from a max level. To me that suggests the core BH operating point is starting to move back from its peak excursion level into saturation, and the loadline per se has some low impedance level.

Yes, I saw.
Maybe the core had some remanent B at the first cycle.
Also looks as if there's a bit of duty cycle imbalance.

[Mike K]

Mercury Magnetics glorifying the wonders of rusted transformer laminations. Read between the lines and it's all just an advertisement.

https://www.mercurymagnetics.com/tra...friend-or-foe/

I assume you mean this? I'm not sure what is incorrect about that or what they are trying to sell me here. If anything, they are encouraging you to leave your transformers alone. They clone and restore vintage stuff, so they could have easily turned this around into a way to boost that part of their business.


The one thing I DO recall them saying that was something I don't readily know, but I assume there is some basis in metallurgy, is that "good magnetic" iron rusts. This was not elaborated on but there are certainly changes that can be made to alloys to reduce rust, such as additional Ni and Cr. Of course at some point you wind up with a "stainless" alloy, which may or may not be magnetic, depending on the actual mircrostructure - austenitic vs. martensitic vs. ferritic but I don't think that applies to transformers.


Perhaps completely unrelated to transformer iron, but I know even with the same "specifications" there was always a vast difference in bearing quality between certain "low-cost" manufactures and those that were not. They'd give you the same outline drawing, but I could tell you the performance was not the same. It's the stuff they didn't tell you on the drawing that was making a big difference.

[Helmholtz]

The MM article looks like an excuse for rusty cores.
Superficial rust has no measurable effect on Eddy currents and certainly doesn't improve transformer performance.

[Mike K]

The MM article looks like an excuse for rusty cores.
Superficial rust has no measurable effect on Eddy currents and certainly doesn't improve transformer performance.

Do you have any sources for your claims?


So I’m still not sure what the actual excuse is. You have a vintage transformer wound by Schumacher or Triad or whoever and they say leave it alone. At the end they say if you insist they will laquer them for you but they aren’t promoting that. Again, I’m not following your logic here.

[Mike K]

There also appears to be empirical evidence that (specific) types of capacitors change after use. So I’m not sure how anyone is profiting from this. I never do this, they’ll get there eventually but I tend to wonder if this is a property of self-healing capacitors?


Ps no one has still provided proof or not that output transformer bandwidth changes without saturation. And if it does, what the effect is. Other than that it’s all been “words” distinguishing what actual saturation is. But I’m willing to bet this not an abrupt, linear process.

[trobbins]

I assume "no-load" means worst case regarding the spikes.
What happens with a load connected?

Yes the 'no load' scenario reduces the available paths for energy in leakage inductance to dissipate more benignly. With a secondary side load connected then some of the leakage inductance energy can transfer into that load, and hence suppress the level of primary winding voltage overshoot (and hence stress on the output transformer insulation). Of course substantial energy in leakage inductance only becomes a problem when a transient di/dt condition occurs.

I'm not keen to do more testing (due to other projects of more interest to me), but perhaps testing could be more honed to the topic of OT saturation by using direct coupling to the output stage (to avoid coupling cap blocking issues), and perhaps LEM style current sensing (eg. LTS 6-np) of the primary winding, and perhaps an RC network as an alternative scope voltage input for core flux density 'B'.

[trobbins]

Ps no one has still provided proof or not that output transformer bandwidth changes without saturation.

It's worth appreciating that the low frequency bandwidth of an output transformer in an amp can be dynamically influenced by the signal voltage level across the primary winding. The primary winding inductance is significantly influenced by signal voltage, with inductance typically rising with voltage, and depending on the design of the transformer the inductance may start falling as signal voltage increases further if the amp's output stage allows. So the amp may have a signal level influenced high pass filter characteristic.

Another aspect of PP amps is that the distortion level from the output stage can noticeably increase when the output stage valves have increasing imbalance (eg. idle bias current imbalance, but also signal gain imbalance), with harmonic distortion level rising as signal frequency reduces. A good example of this effect was provided in the Heathkit W5-M hifi amp manual, albeit at low signal levels and relatively low distortion levels compared to a guitar amp.

With respect to output transformer superficial rust on laminations, there has been detailed assessment and simulation of the influence of peripheral 'issues' such as where mounting holes are located in lamination stacks, and where welding has been used to clamp core stacks, and the general outcome is that due to the bulk of the core flux and the bulk of any eddy-current related losses being 'internal' to the core, then any changed circumstance to the peripheral surface is really just an aesthetic concern.

[Mike K]

It's worth appreciating that the low frequency bandwidth of an output transformer in an amp can be dynamically influenced by the signal voltage level across the primary winding. The primary winding inductance is significantly influenced by signal voltage, with inductance typically rising with voltage, and depending on the design of the transformer the inductance may start falling as signal voltage increases further if the amp's output stage allows. So the amp may have a signal level influenced high pass filter characteristic.

Another aspect of PP amps is that the distortion level from the output stage can noticeably increase when the output stage valves have increasing imbalance (eg. idle bias current imbalance, but also signal gain imbalance), with harmonic distortion level rising as signal frequency reduces. A good example of this effect was provided in the Heathkit W5-M hifi amp manual, albeit at low signal levels and relatively low distortion levels compared to a guitar amp.

With respect to output transformer superficial rust on laminations, there has been detailed assessment and simulation of the influence of peripheral 'issues' such as where mounting holes are located in lamination stacks, and where welding has been used to clamp core stacks, and the general outcome is that due to the bulk of the core flux and the bulk of any eddy-current related losses being 'internal' to the core, then any changed circumstance to the peripheral surface is really just an aesthetic concern.

Thanks, that makes sense - what about on the high end?

As far as the laminations - that was exactly my thought, but I also can't find a counterargument why it would even be a concern. That is, intuitively, I definitely think that eddy current losses will not be very prevalent on the surface, where there is very little contact for current to flow. But I can't see what harm it does either, and I can't figure out how that would actually promote anything positive to any transformer winder (almost all of them rust). The only thing I could figure is it was somehow indicative of specific material properties, as I alluded to above, but also that perhaps scraping and clawing at it may open up pits in the laminations which could promote further decay. A light oxide coating without some mechanical or further chemical means to promote it is actually a deterrent for further oxidation. That is to say simply, just given contact with moist air, it won't rust further if you don't disturb it.


The part that was a bit confusing to me, which I read in another article, is that they don't disturb the rust during restorations. Maybe that's just belt and suspenders kind of thinking, or maybe they are truly concerned with damaging a vintage core lamination which is perfectly usable given some minor cosmetic issue.

[g1]

As far as the Victoria link (post #46) about OT size in Fender amps, I thought the amps he talks about having smaller OT's also had smaller PT's (or lower B+) and put out less power. So smaller OT would make sense.
He makes no mention about that. Rather, I get the impression he is implying that Fender somehow reduced the output power by using a smaller OT. If I am reading that right, it seems to be a misunderstanding of what determines power output?

I'm not so sure putting a Vibrolux OT in a Super is such a good idea.

(edit: my schematics show 360 vs 315 VAC PT's, so memory serves. However, Dave Hunter makes the same kind of claim ignoring the PT difference here: https://www.guitarplayer.com/gear/cl...ll-round-combo)

[teemuk]

I assume you mean this? I'm not sure what is incorrect about that or what they are trying to sell me here.

Well, despite that Mercury Magnetics does not deliberately rust their transformers themselves, or that oxide layer of a lamination isn't exactly the same thing as rust, so the article doesn't actually say anything, the name Mercury Magnetics and their transformers has now been imprinted to your subcobsciousness. Mission of subliminal marketing achieved.