Cheapest power supply transformers?

[Enzo]

Those things may be true, but when they were building the 5E3 new, they were not shopping around buying various replacements. I thought the premise here was different 5E3s sounded different, and you were thinking about why. I remember going to the stores and having them demo several of an amp so we could pick the one we thought sounded best - of the same model. It was expected that all 5E3s (or whatever model) would not sound alike.

Got an old Fender with more than one speaker tap? Listen to it on the "right" tap. Then move the speaker to the other tap. That will demonstrate how much difference impedance matching makes, or doesn;t make.

[woodyc]

I was using the Fender replacement series from CE Dist for about 3 years. I saved about $15 or $20 each compared to Hammonds, which I passed on to customers. For a while it seemed like a good idea. Then they started coming back. I've had to replace/upgrade ALL of them and it hasn't been cheap.

For example here's a pic of a failed vibrolux transformer P-TF22848. The short is visible on the left where the "layers" are overlapping. If you have never seen an output transformer before I can assure you that this was manufactured by people who have absolutely no clue how to make one.

My point is, if you buy cheap stuff you may regret it.

[Steve Conner]

The blue solderable magnet wire says it all.

I think there is quite a wide range of transformer specs that will make a good sounding guitar amp. It doesn't need to be "just right".

[J M Fahey]

I can't believe Fender uses such a piece of cr*p.
That transformer is scatter wound, even looks like the wire was hand guided, can't believe it.
The winding piles up in the center, wires are not parallel, and criss cross everywhere ... so you have turns with widely different voltage touching each other.
Couple that to an atom-thin insulating enamel *designed* to evaporate leaving no solids behind at less than soldering temperature, and you have a recipe for disaster.
Still can't believe it.
Just curious, roughly what's the price and that of a Hammond equivalent?
And a Fender original?

[Enzo]

I don't think he said that was a Fender part, I think he said it was one of the Fender Replacement types from AES/CEDist.

[tubeswell]

I don't think he said that was a Fender part, I think he said it was one of the Fender Replacement types from AES/CEDist.

So you found a portal on the train huh? (Or didn't you leave yet?)

[woodyc]

...Just curious, roughly what's the price and that of a Hammond equivalent?
And a Fender original?

Sorry if I wasn't clear Juan, but the part in the photo is a REPLACEMENT output transformer for a Fender vibrolux. In this particular case, CE Dist sells an excellent Hammond P-T1750J for $43.30. But they also sell the cheaper P-TF22848 in the photo for $28.50.

The cheaper one has the correct turns ratio and dc resistance, and it works at low power... but as you can see it is an utter piece of crap that will eventually fail. I get neater winds on my fly rod, and I don't even use it at high voltage.

[Tom Phillips]

...The interesting things I here about the 50s amps is it might be the 1 in 100 that really does have the magic...

One thing about 5E3 era amps that is untrue today is that they were built with loose tolerance parts...

In addition to Enzo’s point we must consider that the performance consistencies and inconsistencies of the old Fender amps are obscured by the passage of time. The original Fender 5E3 Deluxe amps are all over 50 years old. The storage conditions, service history and old age all combine with the original mix of parts tolerances to determine the present performance. This makes it impractical, if not impossible, to do a statistical listening test that will tell us the percentage of original amps that had the “magic” when they were new. There have always been especially sweet sounding examples and lemons. However, back in the day, most people didn’t pay any attention to the small details that are often discussed today. If an amp malfunctioned it was usually returned to the music store or taken to a radio repair shop to be fixed. Virtually no one did detailed tune ups, requested special brands of parts or even used matched tubes. I have never seen discussions or magazine articles of statistical listening tests that were done 50 years ago.

I’d venture to say that most 5E3s could be overhauled and tuned up to sound great today. That is, if the work was done by a competent tech using the common sense musical instrument amp knowledge available today (rather than the internet lore). When you go through an amp from stem to stern and fix all the anomalies the owner is usually surprised with the result. Comments are usually something like “Wow it never sounded so good” or “I didn’t know it could sound so good.” Sometimes they ask if special mods were installed. My answer is usually “I just made it work as well as it did the day it left the factory.” (or as it was intended / designed to sound) Of course, it’s easier to achieve consistency with modern close tolerance parts but I don’t find the need to replace the magnetics or shotgun replace all the caps with{insert favorite magic type here}. Sometimes an original factory mistake is discovered that, when corrected, turns a lemon into one of the really sweet amps. Such mistakes can be incorrect component values, wiring goofs, unsoldered connections etc.

I still see amps that are in almost original condition. They are much easier to restore if they have not suffered through multiple repair / upgrade events. Attached for fun is a photo taken in 2012 of a 1963 Model 6G2 Brown Face Princeton showing the condition as received at my shop. To me it’s a real blast from the past to see such a well preserved example of a 50 year old amp.

Cheers,
Tom

[tubeswell]

Attached for fun is a photo taken in 2012 of a 1963 Model 6G2 Brown Face Princeton showing the condition as received at my shop. To me it’s a real blast from the past to see such a well preserved example of a 50 year old amp.

Cheers,
Tom

Thanks for the amp porn Tom. That is indeed one fine example of workwomanship. Merry xmas.

[R.G.]

So thinking about it why is there only 1 in 100 5e3's with the magic, I suspect its because of poor QC in transformer production. A few turns up or a few turns down changes everything,

I'm with Enzo on this one. The transformers are the things most likely to be consistent. They were usually wound on 'sticks' of core tube several feet long. This was put in the winder machine, and six or a dozen coils wound at the same time, one full layer, with the machine counting. Then a whole sheet of interlayer insulation was fed in, covering all the coil-sections at once, and another machine-counted layer started; then another sheet, another layer, etc. until the last layer was done and covered. The coils were taped, then the whole stick was cut into sections for each transformer. The counting of layers was very consistent, as was the winding tension, etc. Pretty much there was at most one turn error, and that was probably zero with a skilled machine operator.

I happened to be looking at tubedata for a 6v6 if you increase or decrease the load resistance from the optimum on the tube the 2nd harmonics rise alot or the the 3rd harmonics rise increasing the overall total harmonic distortion. But if you look at it in respect to the plate current a higher load resistance it drops the plate current reducing the load on the power transformer, less p/t current less sag better amp.

There is a similar curve of output power versus load impedance. What is interesting is that the peak of the power curve is not the same as the minimum of the distortion curve. Plus, lots of people really like sag, and seek it out. In general, power transformers have a broad maximum of output power for a given frequency, core volume (stack, for the same lamination form) and window area (also same for same lamination form). Running the plate load higher does decrease the loading on the PT, but it does it by putting out less power to the speaker, not just from lower plate currents. So once again, it's the player's preference about what "better" means. Players generally vote for whatever their personal ears say is good sound and let the amplifier's longevity, internal temperature, etc. go hang. If they want more power and more sag, as many of them do, they'll say that less power and less sag is worse, not better. And they're right - for their ears. All ears vary.

I think we need to start a output transformer investigation thread, even a simple weight verse voltage ratio from simply pluging it into the mains, maybe even some dc resistance readings.

Not a bad idea. But I'd do it this way: There are many repairmen on this forum. That is where dead OTs and PTs are found. If there were volunteer repairment to find dead transformers and volunteer un-winders and documenters to do the autopsies, the pool of public information would accumulate. I believe that the Magical Mojo Magnetic Winders probably quietly went off and researched both paper and iron/copper data for transformers to accumulate their recipes so they could replicate OTs. Once you can replicate the originals in new materials, it's a short step to find various schemes to gild the lillies and advertise the gilding.

[jim1976]

One thing I do know about is sensory evaluation, I'm a winemaker and I can tell you it's not accurate its affected by so much mood to how hungry you are. So after thinking about it some more and doing some research harmonics only come into it once your amp is clipping if its not clipping its as good as a clean transistor amp.

So who uses a spectrum analyser? where are things clipping? hope its not all witchcraft.

[tubeswell]

...So after thinking about it some more and doing some research harmonics only come into it once your amp is clipping if its not clipping its as good as a clean transistor amp.

Harmonic distortion is not about clipping, its about how the plate's +ve and -ve voltage swings are assymetrical compared to what you're putting onto the grid. This is dependent on the HT supply voltage, the plate resistor load and the bias point of the tube (as well as on the tube type). Read this chapter to get how this works. http://www.valvewizard.co.uk/Common_Gain_Stage.pdf

[jim1976]

It isn't, watch this HP 3580A Spectrum Analyzer and Harmonic Distortion on the McIntosh MC275.MOV - YouTube
Plate resistor load and bias point only move it down the scale, do you have test gear? I really think its the only way to tell. That particulate spectro isn't an expensive one, you can look at every stage and you can see exactly whats happening.

Your a kiwi cool! have you had a play with any Jansens by chance?

[Tom Phillips]

... after thinking about it some more and doing some research harmonics only come into it once your amp is clipping if its not clipping its as good as a clean transistor amp.
So who uses a spectrum analyser? where are things clipping? hope its not all witchcraft.

I'll have to disagree with that. I do have a spectrum analyzer and it shows that there is significant harmonic content present before clipping in a typical tube based guitar amp. Of course the harmonics do increase dramatically as you drive the amp into clipping. Edit: As you see from the YouTube video using a McIntosh MC275 your statement holds true for a high quality HiFi amp.

The attached screen shot shows a 5E3 putting out just 4 Watts with a pure 400 Hz input signal. The first peak on the left side is a phantom signal known as the zero response of the analog spectrum analyzer. The highest signal is the 400 Hz output into a resistive dummy load. The other signals are harmonics and intermodulation distortion components. The intermodulation distortion is caused by the 120 Hz power supply ripple present in the 5E3 amp. A log scale was selected for the vertical sensitivity in order to get the lower amplitude signals to show on the same screen as the peak signals.

It may be best to start a new thread if this line of discussion continues.

Regards,
Tom

[Steve Conner]

I have a spectrum analyser and a distortion analyser. I can tell you for a fact that tube guitar amps produce audible levels of distortion before they start to clip. The subjective effect is a "warming" or "sweetening" of clean tones.

Many tube hi-fi amps do too. The McIntoshes are somewhat of an exception, they are about the lowest distortion tube circuit ever made.

Of course the real fun starts when clipping begins. You ought to read Russell O. Hamm's paper, "Why do tubes sound different to transistors" or whatever it's called.