Pleased with the seemingly functional Beocord 5000 (4715/4716) tape mechanism, an initial test recording was conducted using a waveform generator. Playback revealed a 3 dB channel imbalance on the Peak Program Meter (PPM), indicating a calibration issue. This was unsurprising given the deck’s age and highlighted the fortunate performance of a previously rebuilt Beocord 5000 in a Beolab 6000 setup, which operated well without calibration. To establish a solid calibration baseline, a reliable frequency response measurement was deemed essential. Unfortunately, the available bench equipment lacked this capability. A search for suitable freeware compatible with a computer ensued, as the audio range posed no challenge to modern PC sound cards.

Finding open-source software for hassle-free measurements proved difficult. However, “audioTester,” developed by German enthusiast Ulrich Müller, emerged as a suitable option. The evaluation shareware version, permitting 10-minute measurement intervals before requiring a restart, was downloaded onto an older MacBook Pro running Windows 7 via Bootcamp. Installation proceeded smoothly, with the MacBook’s integrated sound card proving sufficient. Following some experimentation, a non-synchronized sweep-measurement was successfully implemented to determine the frequency response. This method was necessary due to the 2-head Beocord’s lack of a monitoring function for real-time recording feedback. The software’s ability to determine and plot signal level relative to frequency facilitated this. A 60-second logarithmic sweep from 10-20kHz was recorded, rewound, and then measured during playback. Testing the mechanism by directly inputting the sweep from the waveform generator into audioTester resulted in a consistent level across the spectrum, confirming the linearity of the MacBook’s sound card signal path.

A baseline measurement of the Beocord at 0dB followed. This involved adjusting the recording signal level to achieve a 740mV(RMS) output at pin 7 of the Dolby B ICs. This was accomplished by setting the waveform generator to approximately 50mV(pp) and using the recording level potentiometer to fine-tune the output to 740mV(RMS) at pin 7 for both channels. Importantly, Dolby should be deactivated during calibration.

Visual confirmation of the 0dB signal at pin 7 at 333kHz was provided through an oscilloscope capture. The display indicated an RMS voltage of 740mV, though the peak-to-peak voltage measured 2.3V, slightly exceeding the ideal 2.1V due to noise. A separate image highlighted the oscilloscope probe’s connection to pin 7 of the right channel Dolby B IC.

After recording the sweep, playback revealed a frequency response curve illustrating a noticeable divergence between channels at higher frequencies. This discrepancy explained the observed PPM imbalance during playback.

Following the service manual’s calibration procedure, the recording current was adjusted to ensure equal playback and recording signals at pin 7 at 333Hz. Using a TDK SA 90 tape, this adjustment, targeting a medium signal level of 200mV(RMS) at the Dolby ICs’ pin 7, involved manipulating trimmers 1R99/47, specific to CrO2 tapes.

Adjusting the Dolby B output to 220mV(RMS) marked the initial step. Fine-tuning involved recording, playback, and monitoring the signal at pin 7. Clockwise trimmer adjustments addressed excessive signal strength, while counterclockwise adjustments compensated for weakness. This iterative process continued until approximately 200mV was achieved during playback. Small adjustments (around 5 degrees) were recommended due to the trimmer’s sensitivity. An image documented the left channel adjustment during Fe2O3 tape calibration.

With the recording current optimized, attention shifted to bias adjustments. This entailed setting the 92kHz bias signal fed to the erase head to an amplitude that ensured optimal high-frequency recording. Trimmer capacitors, responsible for tuning the resonance frequency of a pickup coil that received the bias oscillator signal, were used for this purpose.

The procedure involved setting the input signal to 333Hz and 22mV(RMS), then adjusting the record level potentiometers to achieve 740mV(RMS) at the Dolby ICs’ pin 7. Reducing the input signal amplitude by a factor of 20 (-26dB) to 1.1mV (RMS) followed. With accurate PPM calibration, only the lowest lamp on each channel should be illuminated. Setting the input signal to 15kHz allowed for bias trimmer adjustment for both channels (C70/72 for CrO2, C71/73 for Fe2O3) to achieve a consistent -26 dB signal (37mV RMS) at pin 7 during both recording and playback. An image captured the adjustment of the right CrO2 trimmer.

Adjustments necessitated lifting the preamplifier PCB due to the lack of a specialized tool that could access the trimmers from the solder side, a potential future 3D printing project. This process proved tedious, as recording required a properly grounded PCB, necessitating repeated board removal and reinstallation. Additionally, the trimmer’s sensitivity demanded multiple adjustment cycles for accurate bias calibration.

Following calibration for both tape types, the frequency response was re-evaluated using audioTester. The results, compiled into Excel graphs for both tape types, revealed minimal difference between them. This highlighted the quality of modern Fe2O3 tape formulations. As expected, the low-level recordings (-20dB for CrO2, -26dB for Fe2O2) exhibited better high-frequency response compared to the 0dB recordings, smoothly extending to 15kHz. This disparity between low and high levels is a characteristic of tape recorders. The measurements also indicated that the Dolby B system introduced negligible distortion, with curves closely mirroring the non-Dolby results.

Comparing these curves to publicly available data for high-quality 3-head decks revealed a wide range of responses, including some with oscillatory characteristics. The Beocord 5000 demonstrated respectable performance, especially considering the comparative modernity and inherent advantages of the 3-head decks, which benefited from monitoring capabilities enabling real-time response measurement during recording.