Output transformers and resistor types

[leewongfei]

Hi everyone! Over the last couple of weeks, while planning my last 5E3 built from spare parts, I decided to test out some unicorn/unobtanium/sacrilegious voodoo and essentially see if my untrained ear could hear a difference between metal film vs carbon film, and coupling caps. **Spoiler, it did make a difference**.
Im not a reviewer, or a researcher. I am cursed in the sense that I love to see how little changes can impact the big picture (for lack of a better term). I am very much a try everything until my curiosity is 1,000% satisfied kind of a person.

Both 5e3's we very close to each other spec wise. Both amps have 22uF filter caps. The only variation was the resistor types, coupling caps, and transformers. Bothe amps had very similar power transformers spec wise (one Classictone and one Heyboer). The orange drop amp had a Classictone Deluxe Reverb Output transformer, while the Mallory amp had the Mojo 5E3 transformer. Both amps share the same tubes.

The Mallory amp sounded SO MUCH better than the Orange Drop amp (OD amp).
The Mallory broke up very nicely when pushed, but stayed very sparkly when the volume was lowered (it cleaned up well). The Mallory amp actually had a decent amount of clean headroom, but not as much as the OD with its Deluxe Reverb transformer. The note articulation seemed to hold up significantly better on the Mallory amp with its stock 5e3 Output Transformer. The amp was very responsive and overall, a dream to play.

The OD amp didnt have the soul that the Mallory amp had. The high notes got mushy and overly saturated in a way that was very unpleasing/shrill/stiff, while the low end stayed very stout/ pronounced, almost boomy. The OD stayed clean a little longer, but the amp didnt feel like the Mallory amp did. Overall, I wasnt feeling it. This lead to the first change. After playing the OD amp for a while, the mailman finally dropped off my package of amp parts (capacitors and another output transformer). I popped the OD's out and put in a set of Mojo Dijons, turned on the amp and played it again. I couldnt believe it, I heard a difference! The overly saturated sound of the OD's were gone! It did sound warmer, but it still didnt sound as good as the Mallory amp. So after an hour of playing, I pulled the chassis again, and replaced the output transformer with the one that came with the Mojo Dijons, another Mojo 5E3 output transformer. I plugged it back in, started playing, and couldnt believe it. It sounded just as good as the Mallory amp, even with the use of metal film resistors.

What I learned from this is that resistor type doesnt matter. While changing the coupling caps smoothed out the amp, it didnt change the actual sound of the amp. Both amps still sounded like a tweed 5e3, but the OD's sure did sound stiff/cold. The output transformer had the biggest impact on the sound outside of the speaker. What im not sure about though is why the change in feel and sound was so dramatic from the deluxe reverb transformer to the 5e3 transformer. Is it the impedance? The 5e3 transformer has an 8k primary while the deluxe reverb is a 6.4k (or 7k with how mojotone specs it).

Sorry for any grammatical errors. I am trying to do this on a mobile device. I will try edit it to flow better later in the evening.

My next and final experiment for a while is something im going to try on my Deluxe Reverb clone. I may mic that one up too since I wasnt able to find a clear answer. The Soursound OT vs the mojotone deluxe reverb transformer.

 

[__HKGuns__]

I have a mojo tranny (Heyboer) in my 5E3, with orange drops and it sounds great to my ears. I have Mallories with a Webber tranny in my 5E8A and it sounds good as well, but a tad bit siffer and not as dynamic. I guess you really need the side by side to feel the difference and I think it is more feel than sound as your fingers are responsible for half of your tone in my view.

 

[Lowerleftcoast]

The output transformer had the biggest impact on the sound outside of the speaker. What im not sure about though is why the change in feel and sound was so dramatic from the deluxe reverb transformer to the 5e3 transformer. Is it the impedance?

Chances are the impedance is the determining factor between these two OTs.

The next experiment (Deluxe Reverb clone) with the Soursound OT vs...
The two OTs may have a different winding pattern but it might just be impedance differences. Please measure the impedance of each transformer so the observation has some empirical evidence. (Measure the static DC resistances of the leads as well to compare the OTs wire size.) So far the reviews, I have read, of the expensive OT have no empirical evidence of impedance or winding pattern so it is unclear why the expensive one sounds different.

 

[King Fan]

Oh, dear, @leewongfei , your experimental spirit is admirable, but I'm afraid my experimental design prof would be rolling in his grave. Too many variables is the least of our problems. The resistor noise difference is likely measurable but *very* subtle and would be expected to go unnoticed in listening. The cap difference, OTOH, is not (easily) measurable but unfortunately is also a matter of fervent belief *and* bitter doubt. The OT difference is at a third extreme -- large, well-recognized, expected to make a difference. Though down to simple engineering for the most part, it has the added spice of a few mystery ingredients (winding patterns, metallurgy, reverse engineering, etc.) As for blinding, analysis of interactions, confounding, multivariate analysis, yada yada, just tell the prof this isn't science, this is art.

Let's emphasize the upside for a minute. You had fun; you heard what you heard. You can apply your findings and your understanding (you know what changed when) to your builds in the future. And at that point, your experiment was a success.

 

[leewongfei]

Oh, dear, @leewongfei , your experimental spirit is admirable, but I'm afraid my experimental design prof would be rolling in his grave. Too many variables is the least of our problems. The resistor noise difference is likely measurable but *very* subtle and would be expected to go unnoticed in listening. The cap difference, OTOH, is not (easily) measurable but unfortunately is also a matter of fervent belief, and bitter doubt. The OT difference is at a third extreme -- large, well-recognized, expected to make a difference. Though down to simple engineering for the most part, it has the added spice of a few mystery ingredients (winding patterns, metallurgy, reverse engineering, etc.) As for blinding, analysis of interactions, confounding, multivariate analysis, yada yada, just tell the prof this isn't science, this is art.

Let's emphasize the upside for a minute. You had fun; you heard what you heard. You can apply your findings and your understanding (you know what changed when) to your builds in the future. And at that point, your experiment was a success.

My applied science professor would be kicking me in the @$$ again if he knew I was back to my old tricks again lol. All schooling aside, what all of my useless experimentation taught me is keep an amp as stock as you can, proper impedances do matter, and the 2 largest contributors to sound are the speaker and the output transformer . I also had a blast trying different things out. My total cost out of pocket was negligible since I traded a TV repair for the transformers, which took the fear out of the project.

 

[echuta13]

I've monkeyed a bit on different types of caps, but I think that cap value makes more of a difference than the brand. That being said... once you get your values where you like them, you can start tweaking the other stuff to you hearts content!

Resistors. I only use metal film. Not because they sound good, but because I want a low noise floor in my builds.

Output transformers.. I likes mine big n' beefy! A little overkill in this area is a good thing IMO.

 

[Peegoo]

and essentially see if my untrained ear could hear a difference between metal film vs carbon film

Yes it sure makes a difference.

Used to be, I would save a few dollars on my builds by using carbon comp/carbon film resistors in non-critical (non-signal) paths. They're cheap as chips.

Several years back I build an amp using only metal-film type, and it made quite a reduction in the amount of hiss in the signal. My builds these days are super quiet.

If you're playing live with the amp it's probably a wash--but if you're recording with it you want the quietest performance possible.

I also bump all the resistors up at least 1/4 watt from what's specified on the schematic. Yeah, it costs a little more to use all metal-film resistors and over-spec them, but it's worth it to me.

 

[King Fan]

IIRC a builder here a few years ago went to the trouble of building two identical 5E3s, one with red Jupiters, one with yellow. He noted he could hear a difference. Sadly, this info was wasted on me — I hadn’t even known there were red ones. But it was useful to him.

 

[sds1]

Which Orange Drops? It is important to note which as they have both polyester and polypropylene film caps.

The Mallory 150's are polyester.

Anecdotally, the OD polypropylene film are said to be more sterile/harsh than the polyester.

 

[King Fan]

Which Orange Drops? It is important to note which as they have both polyester and polypropylene film caps.

The Mallory 150's are polyester.

Anecdotally, the OD polypropylene film are said to be more sterile/harsh than the polyester.

And FWIW, the Mojo Dijon are polypropylene but with a different construction than the ODs...

A few more easy, evidence-based experiments to try at home:

For a fun, easy resistor experiment, swap carbon *comps* with metal films for the input grid stoppers (Fender's 68Ks). The MF have made an audible reduction in hiss in every amp where I've done this. Reference: Google "Valve Wizard grid-stoppers."

For a subtle but noble resistor experiment, swap carbon comps (1/2W for most effect) with MF (1/2W for comparison) in the few selected spots where Mr. Keen notes CC **may** produce their legendary 'sweet distortion.' Reference: Google "Geofex carbon composition." The exact slots depend on whether your amp has NFB around the PI: "The place to use CC's is where there's big signal - plate resistors, and ideally the stage just before the phase inverter. The phase inverter would otherwise be ideal, with plate resistors carrying the highest signal voltage in the amp, but phase inverters are often enclosed in a feedback loop." A stock 5E3 without NFB might be a great test bed. (Also note CF don't have this distortion, despite having 'carbon' in their name -- it's the 'composition' that creates the faults and virtues of CCs.)

If you have a 'real' (not artificial) heater CT, it's super easy (in a cathode-biased amp) to try it elevated and non-elevated. This is just a bit more complex with an artificial heater CT; for one method in a cathode-biased amp, see Rob's encyclopedic 5E3 mods page. And it can be done in fixed/adjustable bias too; for both reasoning and methodology, Google "Valve Wizard Heaters" and see the nice short discussion of heater elevation.

Easiest of all: If you have discrete filter caps (not a cap can) try attaching the high-voltage CT *directly* to the first filter negative pole instead of whatever nearby ground location you might have started with. This can reduce ripple return hum up to 5-fold. For reasoning, Google "Valve Wizard Ground" -- wow, long and detailed, so go straight to Sec. 15.3.

Warning: All of these are based on evidence and objective differences, but your results will be subjective, unblinded, and subject to several forms of unconscious bias. So wear safety glasses.

 

[leewongfei]

Which Orange Drops? It is important to note which as they have both polyester and polypropylene film caps.

The Mallory 150's are polyester.

Anecdotally, the OD polypropylene film are said to be more sterile/harsh than the polyester.

Orange Drop 715's/Polypropylene are the ODs used. Mojo Dijons are also polypropylene, but they sounded a a pinch warmer.

I found the linked video below on this thread

 

[King Fan]

Yep, that video gets a lot closer to ideal experimental design. The results, if nothing else, show why folks in cap-war threads can't seem to agree...

 

[gabasa]

The output transformer had the biggest impact on the sound outside of the speaker.

In my experience the output transformer is the heart and soul of a great guitar amp.

The problem with transformers is that a lot of considerations in their manufacture are never included on any spec sheets. The quality of the wire and laminations can make a huge difference, as well as winding patterns. You would hope that the more expensive transformers use better materials, as well as more accurately reproduce complex winding patterns, but it isn't always the case. For example, I've compared Mojotone's 50246 with Soursound's SST1839, which are both 8k/8 Ohm 5E3 OTs, and supposed to be built to exactly the same specs. Side-by-side, one sounded flat, lifeless and boxy, while the other sounded three dimensional, holographic, godly, and invoked a visceral listening experience. Four other guitar players were unanimously as amazed as I was.

With exactly the same specs on both OTs, I attribute the differences I hear to materials and workmanship. I'm open to being wrong about this, but I don't really believe that something like, "a few less turns on the primary" could possibly transform the sound in such a profound and impactful way. The first time I tried Soursound, I felt like I found the missing link.

 

[mountainhick]

Orange Drop 715's/Polypropylene are the ODs used. Mojo Dijons are also polypropylene, but they sounded a a pinch warmer.

I found the linked video below on this thread

Video is irrelevant. No testing of values. There can be as much as 20% difference in values due to tolerances.

Mallory yellow tolerance +/- 10 %

Mojo dijon tolerance +/-10%

Jupiter Red tolerance +/-10%

Sprague 716P Tolerance: ±3% to ±10%(I am sure depends on cap value)

Jupiter Yellow: Tolerance: +/-10%

So comparing each to each other there is potential for as much as 20% difference in value. For these listening tests to have any real validity, the values have to be tested accurately and test the same.

 

[King Fan]

Video is irrelevant. No testing of values. There can be as much as 20% difference in values due to tolerances.

Mallory yellow tolerance +/- 10 %

Mojo dijon tolerance +/-10%

Jupiter Red tolerance +/-10%

Sprague 716P Tolerance: ±3% to ±10%(I am sure depends on cap value)

Jupiter Yellow: Tolerance: +/-10%

So comparing each to each other there is potential for as much as 20% difference in value. For these lids of listening tests to have any real validity, the values have to be the same.

Interesting question. You're right, if one of the slots that really shapes frequencies got, say, an OD that was 10% off in one direction, and a Mallory that was 10% off in the other, we *might* hear a difference. A minor 'yeah, but' is that there will be some averaging of values across the slots. A moderate 'and also' is that you and I don't get to pick the ideal value of each cap within their tolerances, so our results will be randomized across all those tolerances. Major 'not only that' is that everyone from Leo to the modern cap factories thinks that +/– 10% tolerance, is, well, tolerable; it may well be inaudible. And my final note is that even the guys in person, much less us hearing YouTube audio, have a hard time hearing *much* difference. If we heard big obvious differences, then the lack of precision/uniformity might be interesting. If we hear small, almost imaginary differences, it seems less like + or – 10% could be having a big effect.

Sorry, long way to go and a short way to get there. My nerd self once studied statistics. One time I told a witty young woman that; she said, "On purpose?"

 

[printer2]

I think the biggest variable in an amp is in its transformers after speakers. The different steel they used for laminations and the winding pattern to extend the frequency response can have a big impact on the amp as compared to the different capacitors and carbon resistors.

 

[gabasa]

Capacitor listening tests have been done a lot in the stereo/hifi world, and they've been pretty scientific about it, too. Caps can certainly make a difference, but when you compare high-production metallized to hand-made tin foil/mylar, both of which are considered pretty bad for audio, it's understandable that one isn't much better than the other. Comparing really good copper foil caps vs. metallized or tin foil/mylar side-by-side ... there's a much bigger difference to be heard IME.

I think that Gord Rankin of Wavelength Audio in Ohio has the credibility to make such comparisons. He build really high end hifi gear that's well-respected, but also builds guitar gear. Regarding great copper foil caps, he states:

"I built two identical guitar amps (my new LunchBox series combo) and I was building the second when Chris sent samples of the new ToneCaps. I was like "ok I am game". Threw in the 0.01uF Tone Cap as a coupling cap in the amplifier. Fired it up and the other one sounded broken. These caps give an extension that is perfect for any guitar or guitar amp. I just ordered a bunch and will be swapping out all the tone caps in my guitars this week!"

I applaud the OP for going through this effort and sharing the experience with us. Even though there were multiple variables, good learning can still come from making these efforts. I'm sure he'll zero in on things better next time because of all the good advice offered here. Cheers.

 

[ievans]

Unless the tests are double blind where the only variable is the capacitor type/transformer/resistor type, I will remain skeptical of even the most expert opinion on hearing differences in sound. Maybe it’s there, but people hear what they expect to hear.

And anyway, all that ultimately matters is how it sounds to you. If you like it, great!