www.wavebourn.com

Once being in an audio recording studio I connected my bass guitar to a really High-End amplifier. It was used to monitor records. It had very flat and wide frequency response characteristics, and very low THD. Sounds from the big loudspeaker with huge woofer and 18 tweeters mounted on a carpeted base under tiny black net were heard as if it was a window to the studio... A technician inserted a jack in his pult, moved a control up...

But I was really disappointed. My guitar sounded flat, dim, colorless, and tasteless!

That time I understood that the guitar alone does not sound. What does sound, is the whole system, which includes a guitar, an amplifier, and a speaker with a cabinet.

Each component has own "character", it means each component distort sounds. The same amp with different speakers gives another sound. The same speakers with different amplifier give another sound.

Jensen speakers are not good speakers, Fender amps give big THD if to compare to modern amplifiers. But people like them... Why?

I think, the answer is simple. They like the music recorded by musicians who used such rigs. It would sound different, if they had modern rigs. And probably, those old famous rigs would be considered sounding ugly?

Significant part of a tube amplifier is output transformer. It is needed to match high impedance of tubes to low impedance of speakers. It carries significant powers in a wide frequency band. For guitar, the band must be from 80 Hz to 10 kHz at least, for bass guitar it must start from 40 Hz, for classic organ from 30 Hz, for Hi-End, which means "all band which an ear hears and wide" at least from 16 Hz.

However, high-end transformer may be used for guitar amplifier, if to consider a frequency band only. But it will give fewer distortions, so a sound will differ. It will be "poor", not so "rich" of specific harmonics and resonances.

Is it possible to use guitar transformer for Hi - End amplifier?
Probably, yes. But transformed made for guitar will give less power in Hi-End applications. And probably, more distortions, because cheap iron sometimes was considered as good for "rich" sound.

A first, it may be made on a core of thick plates, in such case eddy currents on higher frequencies will suck energy and turn it into a heat, pulling down an amplitude - frequency curve on higher frequences. A second, it may have lower permeability and lower induction of saturation. It means, more turns of windings will be required to work on lower frequencies. Also, higher coercive forces may cause more distortions.

I recently bought some surplus transformers made for guitar amplifiers and want to find optimal conditions for them to work in hi-fi tube amplifier. However, it will transfer less power because of smaller inductance. And, probably, will require deeper feedback to get good linearity and flat frequency response. The iron itself looks quite hi-fi, the main concern is inductivity. So, I am going to find what maximum power will be transferred through it on the frequency of 30 Hz (it is low enough for hi-fi) without significant distortions. To do so I am going to connect a resistive load to the output of amplifier made with the transformer, connect the audio generator to the input of the amplifier, and to find the highest possible voltage which will show saturation of the transformer's core.

Another way is to connect some light bulb (in our case 25 or 50W) in series with the secondary winding of the transformer, plug it into a power outlet through variac, increase voltage while waveform is still sinusoidal, measure amplitude of 60 Hz AC on the coil, and find the maximum power of the output signal, using the formula:


Pmax=A^2/( R*8 )

where A - amplitude of square pulses, R - impedance of a speaker that will be connected to the coil.
Maximum AC on 30 Hz will be half of maximum on 60 Hz. (1/2)^2=1/4
Also, maximum sinusoidal AC will be 1/sqrt (2) of the maximal amplitude. (1/sqrt (2))^2=1/2.
1/4*1/2=1/8
Os simpler, to measure size of pulses on the oscilloscope's screen, power them 2, and divide by 16.
It will be the absolute maximum power the transformer is capable of on the 30 Hz frequency, limited by unacceptable distortions.

Tomorrow I will experiment...

After that, I will figure out what tubes is better to use with transformers, what plate currents and voltage are optimal for them.

Today I tried it, connecting 16 Ohm secondary coil in series with 15W electric bulb.
I connected 360 Ohm 20 W trimmer in parallel with the the coil, and playing with trimmer's resistance found that the maximum voltage on the 16 Ohms coil of the transformer on which it starts clipping on was 26V RMS. It means, the maximum power on 60 Hz limited by distortions is about 40 W.
However, on 30 Hz frequency it will be 4 times less, i.e. 10 W, while for the electric guitar it may be as big as 76 W.

Maximum "clean" voltage on a half of the primary winding on 60 Hz is 208 V RMS. However, on 30 Hz it will be twice lower, i.e. 104V.

Conclusion: it needs four 6BQ5 (or probably six) tubes to be driven with 200V power supply in Hi-Fi applications. It will give 10W of clean power on 30 Hz. And it is the maximum power, not nominal one. The same transformer, but with 6 KOhm of the primary instead of 4 KOhm, may be driven by par of tubes, with 375 V plate voltage, for the same results.

Later I will try it in the amplifier I built recently using hand wound transformer, with 2x6L6 and 300V plate voltage. I will replace the transformer and measure maximum power on different frequencies.

Tests on the real amplifier shows that on 30 Hz maximum voltage on 8 Ohm coil with 8 W resistive load is 12.5 V RMS. Power supply is 300V, tubes - 6L6WGC (6P3S-E). Strictly speaking, the characteristic is nearly flat from 30 HZ up to 30 kHz. Below 30 Hz and above 30 KHz distortions start to grow rapidly. It means, the transformer is ideal exactly for 300V plate and 20W output in hi-fi applications.

I listen it now and enjoy...

However, the transformer is absolutely cold, so it may be connected to 4 Ohm load from 8 Ohm coil, and probably couple of tubes per one shoulder will be needed, so maximum output power will be 40W instead of 20W. I'll try it later.

Bulb, resistor, 120V AC

30 Hz input signal

Form of a signal, 30 Hz, maximum output power just before symmetrical saturation. Bunch of parallel resistors, 8.001 Ohm total.

Output voltage is 12.5 V RMS on 8 Ohm on 30 Hz

I have already the design in mind. There will be metal construction with wooden sides covered by black vinyl with metal angles, transformers on the back of a chassis covered by a single metal box, tubes in front of them covered by a metal grill. Before tubes, near the front edge, will be horisontally mounted input jacks, volume knob (or pair of knobs), power switch, and couple of fuses. Power input socket with filter and output terminals for speakers will be mounted on the back panel, under transformers.

I have already power transformer from tube radio station. It was made by Hammond for Spilsbyry & Tindall back in 1960'th. The rtansformers gives 375 Volt of rectified anode voltage, bias, and filament voltages.

I've found also a peace of wood which satisfy my requirements:

Couple of peaces of vinyl to wrap wooden parts...

Some soft plastic foam, and a glue spray...

...and heavy weight. Voila!

Here is the amplifier I used to test the transformer. The amplifier had been made for a shortwave receiver.

How it looks today.
Convection from hot toobs will suck an air from the transformer's compartment cooling the power trans. Narrow aluminum mirror will reflect heat radiation from toobs.

Preliminary mechanical test of the prototype. Indeed, the subconscious mind is the best computer ever made... But anyway I miss our vibrostand with thermostation we had in the Laboratory of the Equipment Tests in our TIASUR I graduated...

How it will look with tubes and controls...