Measurements of several different transfomers will be presented in this section.
The basic test setup is as follows:
Source impedance 100Ω
Load impedance 20KΩ
Frequency response measured at 1Vrms
Impedance is measured between 10Hz and 40Khz
Harmonic distortion measured from 10Hz to 50Khz at 1Vrms, .1Vrms, and .01Vrms.
The measurement system consists of an M-Audio Audiophile-192 sound card. The Audiophile-192 is connected to an interface box that provides buffering and protection. The interface box provides both attenuation and gain as needed for signal conditioning.
AudioTester V2.2 is used for the frequency/phase measurments. Arta
Steps is used for the distortion vs frequency measurements. Arta LIMP
is used for the impedance measurements.
With transformer measurements the driving impedance has great
influence on the measurements. With distortion, the lower the driving
impedance the lower the distortion. Frequency response is a different
matter. With low driving impedance the HF peak (if any) will be worse.
The HF peak that sevaral of the transformers tested show is influenced
by the drive impedance. If the drive impedance increases the HF peak
will decrease. The down side to increased source impedance is the
distortion rises.
To keep the number of measurements to a managable number, the
transformers are all measured with 100Ω source impedance and 20KΩ load
impedance. I may do one transformer at several source inpedances to
demonstrate the effects of source impedance on frequency response and
distortion.
Block diagram of input buffer interface used for the frequency response and distortion measurements:
Block diagram of interface used to measure impedance:
Note that in all the impedance measurements the baseline 20KΩ
measurement is set as an overlay and shows up in gold.
Here are the loopback measurements of the test system for frequency response at 1Vrms, impedance, and distortion at 3 voltage settings. Note that at the -40dB signal level noise is becoming an issue and limits the resolution of the measurements. Also note that the distortion of the measurement system starts to rise above ~3Khz.
The frequency/phase measurements use the second channel of the sound card plus buffer as a reference to correct for system errors. The distortion measurments do not have a reference as both channels of the buffer are used for attenuation and gain to be able to test at the 3 voltages while keeping the sound card in a linear measurement range.
With the M-Audio card the output voltage is 1Vrms when the digital output gain is set to -11dB.
Frequency response loopback test at 1Vrms:
Impedance into 20kΩ load
Distortion loopback test at 1Vrms:
Distortion loopback test at .1Vrms:
Distortion loopback test at .01Vrms:
Now that the house keeping is done lets look as some transformers...
Line level transformers
Lundahl 1676
1+1:2+2 amorphous core.
Sowter 3575
1:1ct mumetal core 10K:10K
UTC PC0-25
1+1:.5+.5 nickel core 2K/500:600/150
Low voltage interstage
Sowter 9100
1+1:1+1 100% mumetal core 20K:20K
Both O/Netics small interstage transformers and the Slagel small
nickel bifilar transformer were custom builds for the interstage
position in the solid state Tabor amplifier. Both primary and secondary
windings carry DC current. Max voltage of 25Vrms at 20Hz or lower.
O/Netics Small Nickel interstage
High voltage interstage
Lundahl 1635/5ma
1+1:1+1 125Vrms max @50Hz
Lundahl 1671
1+1+1+1:2+2 300Vrms max @50Hz