Direct drive from anode to tape heads (part 7)

At your risk. Take all safety precautions. A suitable fuse should always be used in series with a battery for thermal injury protection as well as fire hazard protection.


Signal path:


Recording:
Fame soundtrack CD - Sony CD Walkman at line out - Pleides V6, ECC83 - Tascam Porta 03 recording head - TDK SAX 60 cassette


Playback:
TDK SAX 60 - Sony TC-D5 Pro II - Sennheiser HD580


Referring to Pleiades V6 schematic.
Cc coupling capacitor to ECC83 grids is 2.7nF. Anode to grid bias resistor for each ECC83 section is 10MΩ. Supply voltage is 48V. Anode current is 400μA per each head channel.


This time ECC83, both sections were connected, one for each channel, so it is stereo. Best sound so far.


At playback VU needle goes up to 0VU. There is treble and a very nice sound. DC bias broadband fffffuuu hiss does not disturb. It is masked by music like listening to vinyl records.


So what is the problem?


There were some s problems.


Then Vb=36V giving a DC anode bias to heads of 200μA was tried. Much less sensitivity, more treble. Playback much less than 0VU. Hiss could be heard along music.


Vb=60V was tried too. Ia is 500μA. Unacceptable sound. Sounds like an expander.


Back to 48V and 400μA. Nice sound again. But track 01 Fame, fails. It sounds very distorted.


So next the Sennheiser HD580 headphones were connected to anodes before and after connecting the heads in order to monitor what quality of signal goes to the magnetic recording heads.


At the Fame track, with only HD580 at anodes, the sound is with very nice treble and deficient bass with some bass distortion. When connecting the heads in parallel, bass drops as expected since heads are inductors. Bass distortion dramatically increases.


Conclusion:


ECC83 with its high output impedance is great for premphasis, or constant current with respect to frequency head drive. But cannot direct drive such a low impedance load at low frequencies. The impedance of a head at very low frequencies goes too low.


So what seems best is to do what had been done all those decades.


Introduce a series resistor between amp and head for constant current drive.


(So it is again the same problem as with L only Pleiades microphone filters. Bass may distort due to source or load. An R,L filter is needed ie R in series with L. Bass will drop at R while the source of music signal, be it a mic or an anode, will see a high impedance). On microphones an (R series with L) high pass filter can be connected in parallel with voice coil and signal taken from voice coil. This creates a gentle slope less than 6dB per octave low cut. On tape heads a series resistor is usually feeding the head. Output of the filter is the inductor or head itself, so the low cut filtering is 6dB per octave, series R reduces stress to inductor and signal source and increases low cut frequency. Is this frequency 20KHz? Then due to the head gap losses etc, at reproduction high cut (integration) must stop somewhere and we have the well known playback deemphasis time constant.


What could be done in an effort to keep Vb low while solving the problem?


A clean low voltage ECC82, (ECC83?), E88CC could still be tried with other Vb or anode to grid bias resistors but direct anode coupling may not be the way. Impedance coupling may be the way.


If a clean higher supply voltage ECC83, ECC82?, E88CC circuit may be needed... Impedance again coupling may be the way. A high inductance inductor at anode producing flat voltage frequency response with large voltage excursions. Then capacitor coupling and series resistor. As it is done on the wise Uher 4000, 4200, 4400 report series using a transistor. It typical uses a 10-13 KΩ resistor with or without a shunt 1200pF capacitor (for extra treble boost) depending on model, see Uher 4000 report schematics.


Since the Uher uses small voltages, low currents, 1mA?, it may be possible to use electron tubes with Pleiades bias at small anode potentials, E88CC? but the anode would have to be coupled to head through a series resistor. Uher report typically uses 10-12 KΩ from transistor collector to head.


What else was gained?


These experiments provided an opportunity to see how brilliant the sound of many electron tubes is when Sennheiser HD580 headphones are connected directly to anodes (Pleiadss Electra 3). Pleiades bias immediately reduces distortion. Anode potentials from 12V to 24V seem ideal but attention to hearing safety is required. Sound Presure Levels must be always kept low for protection of the hearing mechanism.


DC tape bias may not that bad after all. Distortion took place because the signal going to heads was distorted in the first place. But perhaps easier magnetic saturation of tape may had made things worse? Not likely as DC bias seems to had been near the centre of the limear part of the H-B tape curve. DC tape bias hiss sounds louder, nice and thick, ie not treble heavy. It is nice for experimentation. If AC bias is needed an oscillator can be added later on.


Next day addition. After trying E88CC, RC coupled and then reries resistor, ECC83 direct to anode still sounds best and it seems worth carting on with direct to anode tape head drive sxperiments. Different values of series R to E88CC need also to be tried. See part 8.