Now that my Kenwood X-3WX tape deck is working again I’d like to make a back-up copy of my old tape before it decays to dust. This sounds like a simple exercise of taking the tape deck stereo outputs and connecting them to the microphone input of a computer. The digitization and any editing can then be carried out with my favourite software package in slow time.
Wait just a minute before you go connecting your tape deck to your computer microphone input.
The two ports are not necessarily electrically compatible. The signal drive level at the output of my Kenwood tape is rated at 270 mV from a 3,900 Ohms source impedance. The microphone input to my Vaio laptop computer is made for an electret microphone. There are no technical specifications for this in the computer’s manual but these microphones typically have a signal level of just a few mV and an output impedance of about 2,200 Ohms. They also require DC bias of between 1 and 9 Volts for the FET amplifier integrated within the microphone which draws around 0.5 mA from the microphone input socket.
Note that on older computers the microphone input may be mono as opposed to stereo.
I don’t want to overload the computer’s microphone input or feed DC back into the tape drive pre-amp because this might damage something and in any case the audio will be distorted.
Reducing the tape output voltage signal by a factor of more than 100 seems counterintuitive. Why would we want to get rid of the tape’s preamplifier signal gain and risk adding noise to the audio? While I can certainly identify a suitable signal level in the tape preamplifier chain for takeoff, this will negate a whole lot of audio filtering (such as NAB equalization and Dolby) which is very important to the audio fidelity from tape. It will also require flying leads soldered to the pre-amp printed circuit board.
All that I need is a passive network (no active component or power supply noise contribution please) with an input impedance of 3,900 Ohms or higher, DC blocking, attenuation to about 2 mV, an output impedance of around 2,200 Ohms, and an essentially flat audio frequency response from about 10 Hz to 15 kHz. Apparently some computers sense the presence of an external microphone by monitoring for a drop in the DC bias level. Best I implement this too.
With no meaningful specifications for the Vaio laptop’s microphone input available I measured the open circuit voltage at 2.04 V (on both channels). With a 2,200 Ohm resistor between the signal pin and ground this dropped to 0.45 V. So the DC bias was being sourced through a 7,800 Ohm resistor. I figure I want the bias voltage to drop to around half the open circuit voltage for microphone detection by the computer and to put the computer’s input amplifier DC level input in the middle of the bias range. This means a 7,800 Ohm DC load. The current from the microphone socket will be a little over 0.1 mA which I am completely comfortable with. I still have no idea of the input impedance of the microphone socket but it must be less than the DC bias resistance and is probably very close to it.
My basic circuit for one channel (you will need two for stereo) is shown below. Both channels of the prototype were built using discrete leaded components on Veroboard in less time than it took to find some RCA connectors. The only polarity-critical component is the electrolytic capacitor (C1) where the positive terminal is towards the microphone. Alternatively you can use a bipolar capacitor. I have also designed some enhancements including increased frequency roll-off at the high end, a mono output, and a monitoring output.
Figure 1. Schematic of One Channel
With the basic circuit my Kenwood tape deck sees a 3,900 Ohm load impedance. The Vaio microphone socket sees a 2,200 Ohm 2.2 mV AC signal source and a 7.8K Ohm DC load which reduces the bias voltage at the input by exactly half. The –3dB response is from just under 10 Hz to just over 20 kHz.
Figure 2. Frequency Response
The frequency response exceeds the specifications for the tape deck and in any case you won’t actually hear anything outside this range (but you may hear tonal harmonics). Frequencies above about 20 kHz may also cause aliasing problems with the computer’s analogue to digital converter which typically operate at 44,000 samples per second for CD quality recording.
Each channel has the inputs and outputs physically and electrically separated and the prototype circuit was mounted rigidly in a grounded metal enclosure to prevent noise pickup and microphonics (signals induced by physical movement of components). The grounded metal case is really important. Not only can I see the positive influence that it has on the no-signal noise with an oscilloscope, I can hear it.
Figure 3. The Prototype Tape to Computer Interface
I plugged everything together and it worked perfectly first time. The computer detected the microphone, and the signal levels on the tape deck display and Microsoft Windows Sound Recorder even had similar amplitudes. If I crank the Vaio computer’s volume way up I can just hear the tape noise in the silence between tracks. This noise is inherent with magnetic tape and any filtering that attenuates this will also attenuate the original sound. Your favourite digital audio editor software will allow you to experiment with a number of filters and adjust the audio to your liking.
Personally I have a dislike of tone controls and equalizers. Why would I want to mutilate the frequency spectrum that the recording studio spent weeks getting right? You will find that the tone controls on my Kenwood amplifier are set smack in the middle - no treble or bass emphasis or de-emphasis.
After an hour of play time I have a digital back-up of my precious tape.
If you’d like one of these interfaces and don’t have the time, tools, instruments or inclination to make one yourself then I can make one for you for about US$75.00 + postage and packing. Please contact me.
Your device will be housed in a robust grounded die-cast aluminium case with a matt black finish, complete with quality stereo RCA connectors. It will contain all SMD components made with RoHS compliance. If you want to save a few dollars with a cheap ungrounded plastic case and low quality connectors then I won’t be making one for you - best you dust off your soldering iron.
If you need a custom design specifically to match your audio equipment and computer, maybe with enhanced frequency response, gold plated connectors to your specification, perhaps a mono-output, or with a case colour and finish of your choosing then I can do this for you too. Price is subject to application and you will need to provide at least the model of you tape deck and computer. Please contact me.
Finally if you need audio leads I can either source these for you, or make custom leads with connectors, cable and length to your specification. Just let me know what you need.