How much high level exists below 40Hz?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ruffrecords

Well-known member
Joined
Nov 10, 2006
Messages
16,158
Location
Norfolk - UK
As part of the Mark 3 development I have been looking at alternative transformers and in particular I have been looking at the low frequency distortion at high signal levels because this is one area where transformer performance can be limited. The spur for this is the desire to find an output transformer that will fit in the new 35mm wide modules.

Input transformers have little difficulty with this, probably because the power involved is very small. However, for output transformers some real power is involved. I have tested a range of output transformers from 1KHz down to 20Hz and below driving +16dBm into a 600 ohm load. Types tested range from the big old Carnhill VTB2291 used in the EZTubeMixer project through medium sized ones like the Cinemag CM2820 to as small as the Sowter 1461. As a result of this work I have come to the conclusion that all the transformers tested have low enough distortion down to 40Hz but there is great variability from 40Hz down to 20Hz. Size does play a part in this but core material is also important.

Bottom line is the transformers that fit into the new module I judge to be OK down to 40Hz but not to 20Hz. So my question is does it really matter? Is there ever any content down there? Given the lowest note on a bass guitar is just over 40Hz perhaps it doesn't.

Thoughts/opinions?

Cheers

Ian
 
Cleanliness on lowest piano or organ notes, it'll matter.  Yet, everyone seems to want to overdrive absolutely everything these days, so maybe it doesn't. 

Plosives, wind noise, and other accidentals that are allowed to pass, while still errors, will create even more errors if low distortion is added to it. 

What is your benchmark for judging them ok to 40, but not 20? 
 
As usual it depends... in real acoustic music not much (any?), in dance/trance/sound effects could be a bunch.

Back in the day when I had to deal with power output transformers for constant voltage install market that was a concern.  Saturating audio output transformers could damage amplifiers etc. There is generally not much low bass in background music, but give the customers loudness EQ, and tone control, and.... they will boost bass to sound better, often challenging magnetics.

I got a patent back in 1996 for a dedicated bass clipper (US05509080 Roberts), expired now, that clipped just the bass, while the unclipped HF masked the sound of clipped bass.

Probably not a problem, but worst case, protect your driver electronics against transformer saturation (low impedance). 

JR

 
As always, "it depends." I recall that most all pop material was intentionally high-passed at 40Hz, especially back in the LP area when the lowest octave couldn't be reproduced well.

Four-string bass guitars' low E is 40Hz. The increasingly popular 5-string basses have the B below that, which is at 30Hz. If it's got a low B (and bassists play 5-string so that they can choose to use it, and they usually do), then you'd (IMHO) need good response to 30 Hz.

Pianos go down to the A below that, which is 27.5Hz, but there's very little energy in the fundamentals of piano notes that low, so for a piano you don't need "full power response" below 40Hz or so. It's good to have flat response to 25Hz, but the signal power at those frequencies  shouldn't be a problem for a smaller transformer.

Pipe organs can go loud at sub-20Hz, and the listeners want to hear (or feel) it.

If a mix has substantial signal below 40 Hz and the transformer can't quite handle it, you could cheat by doing some high-pass around 40Hz to drop down  the level of the really low stuff. This might be okay on a lot of pop music, but probably not on classical such as pipe organ.
 
+1 on "Depends"

Remember that none of our beloved 2" tape machines reproduced much below 40Hz at standard 30ips.

And that frequency-limited distortion like that does not always subjectively translate to "distorted", but rather often to "warm"

Also, you rarely if ever put high-level 30hz sine waves through a mixer alone. Any simultaneous presence of higher-frequency content can help bias low-frequency through your iron, or masking the phenomenon.

Be aware that theory is not all: The only real way of deciding on a transformer is having someone who'se ears you really trust evaluating under real-life conditions.

Jakob E.
 
Thank you for all the replies. To add a little more context, the frequency response of all the transformers down 20Hz is fine. For comparative distortion at 40H and 20Hz  I measured with +22dBu fed into the transformers and a 600 ohm load on the secondary. The results for the four transformers I tested, with the 1KHz figure for comparison,  are:


Transformer                                1KHz            40Hz                      20Hz                  Core
Carnhill VTB2291                    0.031            0.035                    0.16                    Steel
Cinemag CM2820                    0.011            0.041                      0.39                    Steel
OEP K30A11C                          0.007              0.22                        4.16                    Nickel/Steel
Sowter 1461                                0.0051          0.21                        15.7                  Mumetal/Nickel/Steel

To my mind both the Carnhill and Cinemag are acceptable down to 20Hz but both are too big and the price of good low frequency distortion is higher mid band distortion. Both the OEP and Sowter fit in the module and have reasonably low distortion at 40Hz but not at 20Hz.. However, their mid band distortion is better than the two steel cored types.

My front runner right now is the OEP . What would probably be ideal would be a steel cored transformer slightly smaller than the Cinemag.

Cheers

Ian
 
ruffrecords said:
Thank you for all the replies. To add a little more context, the frequency response of all the transformers down 20Hz is fine. For comparative distortion at 40H and 20Hz  I measured with +22dBu fed into the transformers and a 600 ohm load on the secondary. The results for the four transformers I tested, with the 1KHz figure for comparison,  are:


Transformer                                1KHz            40Hz                      20Hz                  Core
Carnhill VTB2291                    0.031            0.035                    0.16                    Steel
Cinemag CM2820                    0.011            0.041                      0.39                    Steel
OEP K30A11C                          0.007              0.22                        4.16                    Nickel/Steel
Sowter 1461                                0.0051          0.21                        15.7                  Mumetal/Nickel/Steel

Just out of curiosity, how did you measure them? (I suppose those are THD figures, 15,7 for the Sowter sounds a lot to me)

But then again you're using tubes aren't you?
 
"15,7 for the Sowter sounds a lot to me).."    that is the core saturating due to the nickel content.

Mu Metal good for 5 k Gauss, steel is good for 18 k Gauss.

What kind of circuit is this? Neve, API? how much current? How mush primary voltage? What ratio?



are you using the CM 2820 with PC pins or flying leads? looks like the flying leads model will mount sideways with room to spare>

 

Attachments

  • 2820.png
    2820.png
    16.7 KB · Views: 14
I would take the problem from the other end, how many speakers are able to reproduce the 20-40Hz range ? From my experience only main speaker like Tad, Augspurger and of course those array of sub found in live show event.
So I would say that if your are not planning to use it for live show, most people won't ever notice
Also I have seen mixing console channel that included frequency response tuning. I was implemented by tuning the corner frequency and gain of 2nd order Hpf and Lpf.
Cheers,
G.
 
the price of good low frequency distortion is higher mid band distortion

You're thinking this way too theoretically, imo.

What you have is two different flavors of transformer sound. Find out what your core customer likes best, or listen (music!) and take the decision yourself. Numbers does not correlate with subjective sound.

Jakob E.
 
efinque said:
Just out of curiosity, how did you measure them? (I suppose those are THD figures, 15,7 for the Sowter sounds a lot to me)
I am using my standard Twin Line Amp tube output stage from the EZTubeMixer project to drive the transformer. The signal source is my Lindos MS20 audio test set. The transformer is loaded by a 600 ohm resistor and fed into a Scarlett 2i2 and then into the RTA of REW. This has a THD function but it does also list the harmonics. As you would expect with a transformer it is mostly third harmonic.

Cheers

Ian
 
CJ said:
"15,7 for the Sowter sounds a lot to me).."    that is the core saturating due to the nickel content.

Mu Metal good for 5 k Gauss, steel is good for 18 k Gauss.
You are right. Not sure how these combined Nickel/Steel cores work. does the Nickel saturate first followed by the steel?
What kind of circuit is this? Neve, API? how much current? How mush primary voltage? What ratio?
It is a tube circuit. My TwinLine Amp circuit. Basically an SRPP with NFB. Idle current is around 6mA. Output impedance is about 150 ohms. Transformer ratio is 2:1. In these tests primary voltage was about +22dBu (enough to get +16dBm in a 600 ohm resistors on the sceondary} so about 9.75V rms on primary.
are you using the CM 2820 with PC pins or flying leads? looks like the flying leads model will mount sideways with room to spare>

It is close. The space available between the PCB and the top steel screen is 27mm so we are about 5mm short.

Cheers

Ian
 
I think the ear is way less sensitive to low frequency distortion , even very good monitors probably show high distortion at very low frequencies . If you dont have much low end content in your source it looks like the smaller transformers do great .

Low end was curtailed a lot back in the days of vinyl ,as large low frequencies tend to cause tracking problems , even still as Jackob said the roll off in analog tape and vinyl is quite gradual ,where digital below a certain frequency everything is cut off .I did also hear of many producers prefering 15ips bass end  to 30ips ,probably less ultra low end with 15 ips ,but more saturation which rounded things out nicely , Bass tends to sound slightly stronger on 15 ips in my experience

Maybe a HPF with another setting in between the usual 75/80 hz and thru might be enough. I agree also that classical music with large low end will probably be the only place it could really matter . 

Maybe having  a few options for transformers could allow usage of better quality small transformers in vocal channels and the bigger steel ones on say 'bass drum' channel or in places where large low end was likely  . I always felt providing full lowend bandwidth on all channels was over the top  in transformer coupled stuff ,the vast majority of sources wont go near those
depths . I think in any case flexibillity for the console owner to choose either smaller transformer with excellent performance in the mid/high range or one with very good low end at the expense of distortion higher up  would be good thing .
Looking at a typical  band ,bass drum, bass guitar and keyboards are the places most likely to need the extra bass headroom, all the other sources the low end isnt required or desirable really .

Its looks a bit frustrating with the transformer selection nothing seems to fall into the middle ground . I also see your loading the outputs for worst case 600 ohms ,maybe the same test with more real world loads might be instructive .
 
For a premium audio path I'd try to deliver bandwidth down to 20 Hz... the nominal LF skirt for most spec sheets. Then confirm the stage driving the transformer doesn't get hurt by outlier saturation events... 

If there are audible saturation artifacts, the operator can reduce level or compensate as long as the gear keeps on keeping on.

JR
 
benb said:
As always, "it depends."
+1

I recall that most all pop material was intentionally high-passed at 40Hz, especially back in the LP area when the lowest octave couldn't be reproduced well.
But for a mixer, you can't prejudge what teh user may want to do.

Pianos go down to the A below that, which is 27.5Hz, but there's very little energy in the fundamentals of piano notes that low, so for a piano you don't need "full power response"
That is also true with 5-string bass; players are used to hearing their B string via their harmonics. Many of them use 10" loaded cabs!

Pipe organs can go loud at sub-20Hz, and the listeners want to hear (or feel) it.
But again, almost no existing system can reproduce 20Hz flat.
 
ruffrecords said:
Bottom line is the transformers that fit into the new module I judge to be OK down to 40Hz but not to 20Hz. So my question is does it really matter?
What matters IMO is that the xfmr should not be allowed to distort grossly when driven with sub-40Hz. The usual practice is to band-limit before, usin active filtering for optimized response. Are you ready to sacrifice one tube section for that? Or maybe you could use some negative-impedance drive? I used negative impedance to decrease THD from 2% to 0.1% at 20Hz.
 

Latest posts

Back
Top