Guitar & Bass OT - wich is the bandwidth of a good and a very good OT ?

[kagliostro]

As you know I'm not a musician, but I know that a Guitar can go down till 82Hz and a Bass till 41Hz

what I would like to know is which is the bandwidth of a good and a very good OT for each application

At the moment, because I forgot what I've read some years ago,  I'm studing how to wind transformers (I would like to wind one for an SE amp)

so I must decide which will be the lower frequency that the OT must allow (on the two musician application not for HiFi were all is around 20Hz-20,000Hz)

Thanks

Franco

[PRR]

> Guitar can go down till 82Hz and a Bass till 41Hz

Some players "drop tune". However I have never heard of adapting the amp/speaker for that.

I think you want to be *both* musician and technician to design OTs for musical instruments.

You should distinguish low-power frequency response from Power Response. Many push-pull hi-fi OTs are flat to 4Hz at the 1 Watt level, but won't go below 50Hz at full power without making bent waves.

However on stage we are not interested in "small bass" because we can't hear it, between crowd and room rumble, band mates, and Fletcher Munson. Probably the Full Power bandwidth is most critical.

I think many-many guitar OTs have OTs that won't make full power below 150Hz. Others are clearly good to 40Hz even at high power.

[kagliostro]

OK, I know there are a lot of OT that has the bass response not lower than 150Hz and you are saying that some can manage 40Hz without problems

but don't know if this low frequency OT (those that manage 40Hz) are reserved to Bass players or also Guitar players can keep advantage

as per your explanation I can understand that a 40Hz response is not so useful for Guitar players but may be indispensible for Bass players

Changing a bit the question, if you personally have to plan an OT for guitar which is the lowest frequency you put on your project ?

Thanks

Franco

[jjasilli]

"Bandwidth" is more complicated than it should be.    Typically a component or circuit does not fully stop passing signal at the edges of its bandwidth; instead there is a drop-off at a certain rate of dB per octave.  So if an OT is rated, say, at 150Hz - 10,000Hz, what does this actually mean? Probably that it's response is flat in terms of voltage or dB output within that range.  But what is the actual output @ 80Hz?  At what lo/hi frequency is it 3dB down; 6dB down? Often we don't have this info, so we don't know until we test it.  But standard tests are into a nominal fixed impedance load.  Actual speakers will vary.

[kagliostro]

Ciao Jjasilli

Yes, that has sense

Effectively I told bandwidth but the concern of my question is about a value to work around with the calculation of the OT

and precisely the allowable lower frequency to have a good transformer

Franco

[trobbins]

The simplest example I have seen discussed is the very interesting history of the Sunn range and the general consensus amongst bass players who wanted clean full power response - which was provided by the Dynaco Mk3 origin.  I took the time to look through that history a few years ago when restoring a local McPherson prototype instrument amp that was a dead copy of the Dynaco Mk3.

[shooter]

Frequency response at least 20 Hz. to 20 Khz. at full rated power (+/- 1 db max., ref. 1 Khz.)

The above quote is from Hammonds classic tube, SE transformer section.  I have used both Hammond and edcor, I plotted a Hammond, it was indeed close at 20HZ, but there was a *knee* about 17.5K, no problem for me, I am completely deaf above ~14Khz

hope that helps

dave

[jjasilli]

Ditto to the above posts & see: 



http://tubelab.com/articles/output-transformers-explained/


http://www.victoriaamp.com/understanding-your-output-transformer-part-one-by-mark-baier/


https://peavey.com/monitor/pvpapers/Chapter7.pdf


http://www.hammondmfg.com/clsover.htm


The question is easy to answer for a hi-fi amp.  But for a guitar or bass amp, hi-fi perfection may not be the design goal.  The compression, saturation and other distortion offered by an OT with a smaller bandwidth may be desirable for tone production (as opposed to hi-fi REproduction).


I don't know if the frequency response specs of guitar & bass amp OT's is published or otherwise available. 

[kagliostro]

Thanks to All

Jjasilli I'll read the linked material with interest

at the moment I've read only a small part of those articles and they are very interestings

thanks for sharing

Franco

[xm52]

You might find this guide of value:
www.dissident-audio.com/Transfos/Papers/Wolpert_Audio_Xfmr_Design_Manual.pdf



It is important to keep in mind that the speaker cabinet is often the bandwidth limiting factor in the system. In terms of bass low end, fundamentals are not being reproduced, the harmonics play an important role and that is often what you are listening to when you listen to low notes. Most cabs can't reproduce down to 40Hz. Power amps often have low pass filters designed into them to reduce low frequency speaker excursion and subsequent damage.

{EDIT - URL link simplified - PRR}

[kagliostro]

Thanks


I'll go to read also that link


Franco

[HotBluePlates]

> Guitar can go down till 82Hz and a Bass till 41Hz
... You should distinguish low-power frequency response from Power Response. Many push-pull hi-fi OTs are flat to 4Hz at the 1 Watt level, but won't go below 50Hz at full power without making bent waves.  ...  I think many-many guitar OTs have OTs that won't make full power below 150Hz. Others are clearly good to 40Hz even at high power.

This thread needs pictures.



The two trapezoids in the picture are 2 different OT's. The horizontal axis is frequency, while the vertical axis is power through-put.

Any transformer used well below its rated power with have wide bandwidth. As power through-put is increased, bass response is constricted by the core size (and saturation at some power level below the full rated power), while treble is constricted by winding capacitance.

Transformers rated for full 20Hz-20kHz at full power (like Hammond transformers) are much bigger than 60's guitar amp transformers rated for the same power, which themselves are much bigger than 50's Fender transformers which seemingly carry the same power (e.g., for a 2x 6L6 amp).

The net effect is that the modern Hammond OT seems very solid and hi-fi compared to a typical guitar amp transformer. It might also seem "harsh" especially under distortion compared with a tweed OT which is both restricting bass and rolling off some of the treble hash of a distorted signal.

I imagine designing a transformer will not be a "once and done" chore. Especially for distorted guitar, it may be "build, evaluate, modify, revaluate, build, ..."


FWIW, the same type of trapezoid exists with amplifier gain & frequency response. When cutting amplifier gain with feedback, frequency response is broadened. You could look at the upper and lower slopes as the dB per octave roll-off Jjasilli mentioned, which stays fixed. But when you cut gain, the new lower-gain circuit has a wider bandwidth than it had at full gain. Either way, the graph is the same (see below).

[jjasilli]

A picture is worth a thousand words!  But here's a way to spend a thousand dollars.  Actually lot's more, but it's for a good cause.   


KAGLIOSTO:  you'll want one of these:  http://el34world.com/Forum/index.php?topic=20491.0 & http://www.roomeqwizard.com/ & room test eq.


Then acquire a bunch of venerable bass amps and speaker cabs.  Do spectral analyses of the amps, and the amps through the cabs.  Publish the results.  Get rich.  Cut us in for 10%.

[HotBluePlates]

A picture is worth a thousand words!  ...

In (I believe) a Desktop Reference to Hip Vintage Guitar Amps, Gerald Weber made a comment to the effect of modern transformer designers being too clever and making output transformers that weren't tone-ful (or maybe too hi-fi). He then went on to glorify paper-insulated transformers for amps.

I don't think the paper insulation was the key so much as small cores. GW was looking to exactly replicate tweed amps, and seemed to attempt to duplicate every detail in absence of understanding why things gave the sounds they did. Or maybe he did know, but chose to obscure the real reasons (I don't know)...

Anyway, the Victoria Amps link earlier points out the undersized OT cores in tweed amps contributed to the sound (and apparently kept the speakers safe, which I didn't know before!).

[shooter]

same type of trapezoid exists with amplifier gain & frequency response

That factoid snuck up and bit me when I was spec'n the amp I mentioned, couldn't figure out why at max my tranny was not producing the expected response, then the webs parted

[PRR]

> Any transformer used well below its rated power with have wide bandwidth.

But the reasons are different, low and high.

Low freqs, the iron can only hold so much for so long. Past that it is too distorted even for NFB. The maximum useful _voltage_ is inverse to frequency. If it can hold 10V almost-clean at 100Hz, it will only hold 5V almost-clean at 50Hz.

The transformer has no such built in distortion for high freqs. Instead stray reactances divert some of the amplifier's output. Capacitance steals current, and tube amps are often not over-eager to deliver more current. But also leakage inductance steals voltage.

Capacitance is almost all about the bulk size of the transformer. Not much you can do to reduce it without hurting other factors.

Leakage inductance is when the primary flux is not intimate with every turn of the secondary. Interleaving reduces leakage inductance. At the extreme is bi/many-filar winding, but this does raise capacitance in interesting(?) ways.

Which to worry about? Speaking very broadly: in an audio transformer winding under 1K, C does not matter, LL will be your problem. Over 10K, C will eat your sizzle more than LL. In an intermediate range around 10K impedance, *both* C and LL will get together and ring.

Which is why a Williamson OT has such a complicated winding. A 4-Ohm to 70V PA line output can be thrown together any old way. 

[PRR]

> the undersized OT cores in tweed amps contributed to the sound (and apparently kept the speakers safe

These tweedy amps were, in many ways, "good parlor radios and record players". There was 20-30 years experience making bigger radios sound loud yet nice (and without too much rude sound or early burn-ups). The Tweeds used many of the same parts, though toward the more robust end of the selections.

It was the rise of players like Dick Dale who exposed the difference between a parlor radio (even one big enough to fill a VFW ballroom or a Grange) and playing Rock-n-Roll out on the beach. The move from Eminence radio speakers to theater chuffers (JBL), and the shift to bottles and OTs more commonly used for shaker-tables.

[kagliostro]

Thanks HotBluePlates & PRR

Very very good explanations about such a difficult matter

---

I think that what PRR say is very important

There was 20-30 years experience making bigger radios sound loud yet nice

Experience is what permitted to achieve such results and I think no homebuilder can get it if not after a long practice of years

Franco

[drgonzonm]

At least one boutique amp builder, Zimmerman, recommends a greater frequency response for bass amps when compared to guitar amps.  If you have ever heard a standup bass and an electric bass, side by side, the standup bass sounds a lot deeper, even though they are tuned to same notes.  E0, for the low note.  Remember many bass players are using the B0, (about 30 Hz) as their low string. 

While there are pedals for electric basses, they are quite a few less number of brands. As, many pedals use harmonics to provide their "flavor". 

Regarding, designing SE O/T's for bass, expect to use a lot of iron.  Norman Crowhurst in his acoustic basics books, provides some design information for SE O/Ts.  I also believe RDH 4th Ed also has a fair amount of information. 

I would seriously consider using HiFi technology to select bass amp O/T's.  Any O/T that does justice to tubas timpani, and bassoons, will do justice to a bass. 

As stated, speaker design is also very important.