Questions- Audio Transformers

CoolCat

Well-known member
I notice some get serious about transformers.
Others assume quickly whats "cheap" yet, others who upgrade post "honestly I cant hear a difference".....(I fall into the latter somewhat.)

This article piece states:
How does core composition affect the sound?
Different materials have different abilities to contain magnetic flux—this is called “permeability.” Core materials with higher permeability create higher primary inductance, and therefore better low-end response. However, more permeable core materials will also saturate faster than less permeable ones. Ah, nature, where everything's a tradeoff!

The most common core materials for audio transformers are M6 steel (steel with a bit of silicon) and nickel/iron alloys. Cores with high nickel content are more permeable and more expensive, with less hysteresis than steel cores.

In general, steel will have higher distortion at normal signal levels due to hysteresis, while nickel will have higher distortion at higher levels, due to saturation. For this reason, you'll often find high-nickel cores in high-quality transformers designed for lower signal levels, and steel cores in cheaper transformers or those designed for high signal levels.


I got a new preamp...it has 3qty transformers, from what Ive read its the same china factory making GAP, PELUSO and others TNC, Chameleon Labs and who knows who else has their stuff built in China. Seeing pictures of the factorys the place is high tech, immaculately clean and far above most old factories with magic dust. As a engineer type I get the physics of saturation and transformers and it can effect sound. The application and goal of design is a huge part...ok...

My question is more in line with different brands of Transformers. Why one might be magic and others are garbage?

Can a person really hear if the transformer is more steel or nickel or the % of mix?
To date I havent been able to hear a difference. Maybe I dont squint my ears enough to hear it?
 

Attachments

  • quality.png
    quality.png
    2.1 MB · Views: 9
First off CC, everything you have ever heard that was remotely commercially produced came via transformers. Mixers had them at the front end for both mic and line inputs and more for the outputs. Tape machines were invariably traff coupled in and out as were disc cutters. Much broadcast gear still uses transformers , nothing quite like them for keeping RFI out and they have a massive CMRR.

The coming of the transistor and especially the op amp meant designers could eliminate the transformer from most audio stages but an RFI proof, low noise mic pre amp proved a hard nut to crack.

Core material has to be judged in conjunction with size. A bigger core will give lower LF distortion than a smaller one at any particular level. But THE most important part of output transformer use is that of the drive amplifier's output resistance. A transformer with zero primary resistance and driven from a zero source Z will deliver zero distortion. Theoretical of course but in reality, keeping drive Z as low as possible should be a design aim. Clearly the latter is impossible when peeps can buy ad hoc traffs and connect them to gear of unknown drive characteristics. Some pieces of gear have OP Zs of several 100 Ohms, even 1k and that ain't good.

I always cringe when I see the word "saturation" when applied to transformers (and valves) In practice it is extremely difficult to 'saturate' a transformer with audio because the term really means the point where the core has been energized to and beyond its magnetic capabilty and no longer functions as a transformer.

This effect can be seen in MAINS transformers when zapped with storm damage. The pulse of current momentarily saturated the core and the impedance of the primary drops to zero (well, DC resistance) Thus the breaker trips often saving the rest of the equipment from further damage. This is why old kit with mains traffs often survive storms and new gear with SMPSUs just go up in smoke!

Dave (you know! That pedantic old fekker!)
 
thanks for the info,
I get the first part...

delving into the core, and drive Z I dont understand... so the saturation term seems to be used with the Neve designed 1073, input overdriving the output transformer.

however Im not understanding it.

I kind of understand the Mic signal> hits the Input transformer> then goes to the "preamp transistors" gain circuit and drives the Output transformer...is that right?

Why is the input transformer needed for ? it seems any gain circuit can over drive the output transformer?

in my mind I can think of frequencys and a mic signal hitting a input transformer and maybe the low freq has more energy, the highs less energy so a transformer can "tame" the cheap chineese shrill mics better, naturally...a good combination for the peaky-high freq capsules is tamed with a transformer at the input...(my interpretation of TedFletcher JoeMeek w/ transformer benefit)

but the discrete circuitry (transistors) was always looked at as lesser tone-love than tubes (per Geoff and George Harrison on the Abbey Road new board both prefering the old tube console supposedly)...

then why is discrete praised over IC, it seems, at least marketed that way usually. "DISCRETE COMPONENTS!!" right?

then if Im understanding this design thing, the output transformer adds some tone, often called saturation in preamp jargon, as kind of light distortion?

and then to make it more confusing the mentioning of actual metal materials in the core....and how does a HR gearhead even know which metal is in their gear?
 
Heh! Don't worry about the 'Z drive' thing, all to do with 'Rates of change of Flux' and similar swaddlin'
It is just a fact that the performance of a transformer cannot be taken in isolation, how it is driven and how it is loaded has an effect on distortion and frequency response (input transformers invariably need a series CR network across secondary to damp HF resonance)

Why are input transformers a good thing? It starts with valves. Noisy bstds! So you have a transformer at the front to step up the signal. 1:60 was common giving you 30dB or so of noiseless, 'free' gain that overcame some of the valve noise. Transistors with their much lower circuit impedances 'promised' lots of low noise gain but early attempts were not that good and often a 'hybrid' circuit was evolved using a 1:6 or so input traff. Nowadays with SMT it is pretty easy to make a very good mic pre for 5 bucks but don't take it close to any powerful radio masts!

Yes, you can drive a transformer and make it distort but that is NOT the same as "saturation"
Might be easier to understand for a valve? If saturated it means the valve is turned on SO hard that the cathode can no longer supply electrons,i.e. the valve has ceased to function. That will actually eff them up in no time but fortunately it is impossible to do for pre am stages and next to impossible for power stages, a fuse would blow first. OK! Yes I am being tekktpedantic here but this IS thread about transformers so I think in THIS instance we should get the terms right? And yes, I agree, it is a lost cause generally like 'rms' bloody watts and 'Class A', terms generally bandied by people who don't actually have 'king clue what they mean!

"Discrete Components. Oh ***k! How long have you got?! Yes, there are some advantages, mainly that you can run them off higher supply rails than op amps, sometimes much higher. That translates to much higher output levels (adpuffed as "Massive Headroom) but since the top pro A/D converters only need around +24dBu for 0dBfs what use is +36dBu or more?

The discrete transistor has perhaps one last place where op amps are inferior? Mic pres and summing amps. There are specialized very low noise transistors that, in the right circuit configuration be the best ICs by 3 of 4 dB for noise.Then you use op amps to provide massive feedback and vanishingly low THD.

How is the HR punter gonna know? He/she won't (and I can't tell them!) but this is all to do with very tiny differences in performance and a bucket of snake oil. It really isn't technically difficult these days to make a very good AI for $100 and there are plenty about. In fact I could not name a bad one!

Dave.
 
1) Mic Preamp Input Transformer....what is its purpose when placed before a discrete amp circuit is used for, like a 1073?

2) ok discrete amp circuit, headroom power rails...I can just accept that and not try to get too deep....and snake oil is ok, of course, famous records help sell gear these days.

3) Mic Preamp Output Transformer, purpose is ? what is the gooey magic of this one? some preamps have them others dont? hhmmm?

I stumbled on a old Neumann comment on microphone transformers and they state their TLM series are Transformer Less Microphones....and tend to have more HiFreq sizzle...marketed at a time "cleaner" was sought, (and saving cash on mic build parts).... HF pre-emphasis and de-emphasis yaddayaddayadda.....
the older mics had transformers and smoother High Freq...in a short regurgitation from their site.

as the Neve 1073 is so awkwardly and sometimes obnoxiously cloned and copied in design or with just snake oil blue and red knobs...what is the purpose then of the transformers of that Era? Was it more needed as you say for tube circuits and now its not required but has a tone-vibe? I suppose the discrete circuit is required by the more involved "cloner's"....

appreciate the transformer info....Ive high respect for those that know their craft but Im not going to be learning Transformer Math formulas..lol Im just curious of why Rupert stuck transformers in his channels...and maybe a little why, everyone copies that 73 design?
 
Almost everything in music comes around full circle and transformers are one such item. All of these things start with "I remember..."
There was a day when you couldn't give away a Hammond.
There was a day when you couldn't find any interest in modular synths.
There was a day when electronics manufacturers touted "transformerless designs!" and Golden Ears like Bob Clearmountain would sing its praises. I probably have a few issues of Mix magazine from the 90s with such ads.
To me, most of it reeks of cork sniffing.
When everyone was in pursuit of the most pure signal path ever, transformers were given the boot because they do indeed color the signal. Today, everyone is looking for the "right sound" which is, by definition, at variance with the original source. Transformers came back into vogue for that.
Even Walter/Wendy Carlos added a bank of (input) transformers to the console to restore organic vibe to the then-new digital synth recordings. She loved them and sang their praises. But then, synth players have that luxury because every component right up to the speaker can be considered "part of the synth", where a recording engineer has to preserve accuracy. Or, in today's world, just get "the right sound".
There is a bit of confusion. A transformer, by itself, feeding the inputs of SS circuits, can hardly saturate - unless it's grossly mismatched to the circuit. In a tube circuit, however, things are different.
Often a guitarist will plunk down some cash for different tubes because "these tubes distort better." This is not the case. The tubes and the output transformer act as a coupled loop, feeding one another, building a 2nd Harmonic distortion -- which is considered the 'most musical' because it's the most pleasant.
The distortion is seldom from the tubes but from the transformer reaching, yes, saturation. This is why the explanation of "it's a different circuit" is the only answer given someone when they try to move tubes around in similarly rated amps, but the results are either nothing or radically different. It's not the tubes doing the distorting, it's the windings of the transformer. But who wants to change out THAT thing? Especially when you buy another pair of tubes (matched pair!) hanging on a card, right?
This is why even in preamps, changing out a 12AX7 with a 12AU7 can result in new and interesting distortion patterns: different gain.
(Counter-intuitively, LESS gain tubes can have more distortion than greater gain tubes. Yes, it's a bit involved.)
But when you ask someone to plunk down a cool quarter mil on a console, you gotta have more going on inside than an incredibly accurate and simultaneously boring O2 preamp-per-channel. You gotta have tubes and transformers and tech straight out of the 1950s.
Because, after all, it has 'the sound'.
We KNOW it has 'the sound' because look at the great selling products that were recorded with it. Also look at the perennially stoned endorsers. Don't ask about promotion and distribution, though. Just look at that console!
What's that for you? Well, i really don't know except how I do it. I keep everything as clean and accurate as possible. Very sterile. Very boring. No wine club descriptions for me.
Then I introduce these various artifacts through the preamp chain no differently than I regard a chorus or phaser. I have 8 channels of good, British, EQ out of a good, British mixer that I patch in to give me that "British sound". I have another 12 channels of various tube circuitry when I think a good "tube sound" will help. Some guys do this sort of thing with software plug-ins, but I do not. Real circuits with real variances.
Now i can easily record 72 full digital channels simultaneously, so, no, I don't have enough to go around for every channel, nor do I WANT it.
Maybe I took Roger Nichols at his word. He mentioned working the Steely Dan album and how they'd travel all over to just that one room with just that one mixer to have just that one sound for just those couple of tracks on just those couple of songs. (In retrospect, i regard that as self-aggrandized hype.) He said that sticking to one console, whether Neve or SSL, or whoever, made your recordings always sound the same.
So I never pursued the "right sound" as anything other than part of the effects chain. That way, I could remove it as the winds changed direction.
But then, I got no superstar album cred, either. Maybe I'm totally wrong.
 
thats bizarre but true, reading the era fad where transformers were outcast... clean was the new impyoved thing!...haha long before my recording gearhead time.

cant argue a U47 or 87 and NEve 1073 type and LA1176 standard gets plastered for eternity it seems...probably no different than guitar heads still searching for the Marshall amp and a Fender amp or Ampeg etc... but I do wonder what was in the head of the inventors and why they put certain parts in their design, when they didnt have to... was it luck? I mean Rupert invented many many preamps and for some reason the 1073 is the "famous one" and no one is cloning his others.

great post ponder5.. yeah, Im no keys man, but hauling a Hammond around doesnt sound fun....so long Hammond organ...
 
Sorry Ponder5 just not so with regard to output transformers and valves.

Yes, transformers do produce distortion but it is predominately ODD, mainly third harmonic (look at the BH loop, it is symmetrical about zero flux) . A very well balanced, push-pull amplifier will also produce virtually no even harmonics but guitar amps are never that well balanced so the VALVES produce some 2nd H.

I have already said that the term "saturation" has gone into the audio lexicon and is impossible to remove now but, for the purposes of discussion about transformers in a 'proper' sense, you cannot saturate them in conventional circuits* Neither can you valves. Drive a transformer hard and it distorts, IF it were to saturate it would stop passing signal. But! We will never stop the adpuff men when they have latched onto such a mellifluous word!
The technical director at the amplifier Co I worked for told me they had TRIED to saturate an OPT but it could not be done with conventional circuitry.

One job I was given was to built a 'Traff Switcher" This potentially lethal rig allowed them to play a guitar into a given amp (a 40 watts and a 50W) and compare transformers made by different suppliers. At all power ranges from dead clean to manic metal no one could tell which transformer was which.

Yes, the transformers were made to the same specification, ratios and such but each company has it own little trade secrets and it would have been reasonable to expect SOME audible difference. None was evident.

However the traffs WERE different. This was shown when one of them failed in a 60W (2x EL34) amplifier whilst on 'torture' test. We had never had a transformer fail in that amp before, either in the lab or in the field.

*Mains (aka "power")transformers can saturate if the supply goes even a little bit higher than nominal maximum. If you check the magnetizing current for traffs around 100VA it is about 70-80 mA. Wind the volts up past 260 (here) and Imag starts to climb alarmingly!

Dave.
 
So a preamp, like Neve 73 running on 24V probably isnt going to blow any transformers at the output , not even close on the input it seems. Is that safe to say?

Thats another off topic question but strange....why the hell does the BAE clone use that big bulky power supply box that only puts out 24V to the little 19" 1073 rack copy1073? Its a big huge power supply box with an output of 24v? ..... Neve corp has 24v standard wall ps, on small desktop units and the 500 series stuff even less...so why the big bulky box and all the cost for 24v??

"no one could hear a difference" :eek:
and thats probably with close listening. inside a mix with a ambitious drummer and a bass going and some strings, can the transformer on the vocal be heard between a transformer with all settings the same?

still Ive got a nice clean loop and am getting a cloner-setup 1073ish,LA2Aish going ..for Tracking. Transformers & Tubes...
seems I actually enjoy gear and tv as much these days and more often.
 
The BAE PSU will run two rackmount clones and each one draws 28 watts. A 56W supply is going to need a 100VA transformer to be totally reliable and keep cool. Your will NEVER hear me complain about OTT BSH engineering ESPECIALLY power supplies!

To their credit BAE make the pre amp available sans PSU and this used to be typical of companies selling into the true professional market, they expect the studio people to be able to knock up a simple 24V 1.5A regulated supply (Nagra used to not only include a schematic in their TRs but a wee bag of spare transistors and suchwhich!)

No, it would be a very small, cheap and nasty input transformer that could be overloaded by a microphone.

Bottom line is: Input and output transformers USED to be mandatory (as did 600 Ohm 'image' matching!) They are no longer needed but people seem to like the way they 'sound' hoss feathers in their heads or nay!

Dave.
 
Sorry Ponder5 just not so with regard to output transformers and valves.

Yes, transformers do produce distortion but it is predominately ODD, mainly third harmonic (look at the BH loop, it is symmetrical about zero flux) . A very well balanced, push-pull amplifier will also produce virtually no even harmonics but guitar amps are never that well balanced so the VALVES produce some 2nd H.

I have already said that the term "saturation" has gone into the audio lexicon and is impossible to remove now but, for the purposes of discussion about transformers in a 'proper' sense, you cannot saturate them in conventional circuits* Neither can you valves. Drive a transformer hard and it distorts, IF it were to saturate it would stop passing signal. But! We will never stop the adpuff men when they have latched onto such a mellifluous word!
The technical director at the amplifier Co I worked for told me they had TRIED to saturate an OPT but it could not be done with conventional circuitry.
Dave.

I know better than to argue with the Queen's Own Geezer on this, so I'll just say you've neither corrected nor explained anything.

Transformers cycle in and out of saturation in microseconds, so yes, while conductivity is truncated during that time (duh), it reappears equally quick, thus creating a clipping and slope effect where some of the waveform becomes distorted.

And any "technical director" incapable of such can stop by my bench. Luckily, I have such abilities.
 
thats bizarre but true, reading the era fad where transformers were outcast... clean was the new impyoved thing!...haha long before my recording gearhead time.

cant argue a U47 or 87 and NEve 1073 type and LA1176 standard gets plastered for eternity it seems...probably no different than guitar heads still searching for the Marshall amp and a Fender amp or Ampeg etc... but I do wonder what was in the head of the inventors and why they put certain parts in their design, when they didnt have to... was it luck? I mean Rupert invented many many preamps and for some reason the 1073 is the "famous one" and no one is cloning his others.

great post ponder5.. yeah, Im no keys man, but hauling a Hammond around doesnt sound fun....so long Hammond organ...

There was an article where they took a technical tour through a Neve console describing all the superfluous circuitry compared to SSL where Rupert would color code connectors for gain staging within the channel, correcting to unity as the signal went through the chain. Why? only Rupert knew. No one else could explain the reasoning.Patent infringement is a minefield, especially in a highly competitive, time-to-market industry. But that was only one theory.

The article concluded with the somewhat random stumble of "whatever he did right or wrong, he got the SOUND right".
So it's back to that, again.
SSL was far more accurate, as shown in repeated analysis. No one doubts it.
Neve was inaccurate, but the preferred sound. It would evoke an emotive response where everyone liked or loved "the sound".
I don't operate in such rarefied air, so I can largely ignore these problematic comparisons.
 
I know better than to argue with the Queen's Own Geezer on this, so I'll just say you've neither corrected nor explained anything.

Transformers cycle in and out of saturation in microseconds, so yes, while conductivity is truncated during that time (duh), it reappears equally quick, thus creating a clipping and slope effect where some of the waveform becomes distorted.

And any "technical director" incapable of such can stop by my bench. Luckily, I have such abilities.

"Microsecond" events in a ~20kHz bandwidth? Unfortunately I no longer have access to scopes and such so I have to rely on what I was told and my (75yr old!) memory but if you have any scope shots that demonstrate the effect I would like to see them.

"Queen"?? Lot of freeloading parasites in my opinion.

Dave.
 
Well, I'm sure that some folk will thinnk that I am far off base here; but I don't worry much about the kind of transformers included in any component I buy. My primary concern is how well does the unit sound - how clean and free from distortion is that sound? I understand a little about the physics related to transformers; but I see no need to concern myself with that stuff. I figure that for my purposes, a moderately-priced component from one of the better equipment makers is all I need; I'll trust the company to put in the best stuff it can for the price it is charging.
 
Back
Top