AFAIK it was only to eliminate radio interference, especially for the AM radio modes.
But how I can check the flawless operating when I have such a tape recorder for service ?
But how I can check the flawless operating when I have such a tape recorder for service ?
Absolutely, 19Khz pilot tone is used to sync the 38Khz decoder for the stereo difference signal. As tape recorder bias is often set between 30-40Khz when recoding from FM radio it could produce a beat tone. The beat switch drops in a filter but also affects the HF response of the recoding.
In the US there is a 10KHz beat between adjacent stations. Most AM radios don't approach 10kHz. One I had did, and had a 10KHz notch filter, not switched.
Not very likely. If there was a bias frequency that didn't cause problems why wouldn't you use it all the time?
Because there are different sources of tone that might beat against the biass oscillator, for instance nearby radio channels putting ultrasonics into the output of the radio - not every receiver has a crystal filter in its IF chain so nearby channels leak into the output, which if ultrasonic are unnoticed till mixed with a tape head bias oscillator.
It should be possible to find out who is right by looking at the schematic, if available. Or measure if the frequency of the signal at the erasing head changes with the switch setting, but if it doesn't, it can either mean that Mark's hypothesis is incorrect or that the circuit doesn't work as it should.
Last edited:
Possible, but quite unlikely. That sort of leakage would cause problems within the radio. Without crystal filters the IF strip was just tuned for best selectivity. If anything were to leak out, it would be the IF itself, but at 10.7MHz for FM and 455kHz for AM, not likely it would make it very far into a tape recorder.
The beat switch had the purpose of reducing the 19kHz pilot leak which would definitely beat with the bias oscillator. I had this exact problem with a Heathkit tuner and an old tube-type tape machine. The method of reducing 19kHz leak in the tuner was to sum in a sample of it, then adjust phase and level to null it. Unfortunately, getting a good null was nearly impossible, and it would drift and not stay put anyway, so in stereo mode the tuner squirted 19kHz out the audio jacks. In mono the issue went away. The switch on the tape recorder filtered out more pilot at the expense of some high end, which in those days wasn't there much anyway. A good 19kHz filter was impractical in the Tube era. Think 6 or 7 pole Chebychev or Elliptic. So we lost some response above 10kHz to kill the bias beat tone, which IIR was in the 5-10kHz range. Nasty.
That's the most compact way to say it correctly. It was mainly used to avoid beating through bleeding of the bias oscillator when recording AM stations. In most cases there was just a capacitor that was switched to or from the bias oscillator in order to switch the bias frequency by around 10 kHz. It was more often found in what we called radio recorders (combination of tape recorder and radio) than bare tape recorders. Its functioning can easily be tested with a scope for instance.
Regards
Charles
No, it was for recording FM stations with a stereo pilot. The typical IF and resulting audio bandwidth of an AM receiver was barely 5kHz to begin with, often less, and directly tied to the tuner's ability to reject adjacent channels. Nothing really outside of that came out of the output jacks. FM, entirely different problem, they leaked 19kHz like crazy.
You could stagger tune the IF strip in an AM radio and get response out to 10kHz or higher, but the amount of noise and poor selectivity, and reduced sensitivity usually made that kind of tuning useless outside of major markets where the dial was populated with big 50kW signals, spaced much farther than 10kHz apart. The McKay Dymek AM tuner was the one exception, it had a wide-band mode and a narrow band mode, prior to the adoption of the current NRSC mask, which changed all that.
50kW clear channels are protected against first-adjacent interference completely (that's another station 10kHz away, close enough to interfere), and Class B and C stations (lower power regionals) had to consider daytime first-adjacent channels and night skywave propegation, which to this day causes many to operate at reduced power an/or a directional pattern. Today in the US, the FCC stringently limits the bandwidth of all AM stations to a special "mask".
In short, bias-beat is an FM stereo problem with older tuners that leaked 19khz.
That sounds logical, at least it explains why there isn't one fixed setting that always works.
I think I was precise enough in pointing out that it isn't the output of a tuner that is interfering with the tape recorder but the recorder bias oscillator interfering with the radio stations on the RF side.
Radio interference from a bias oscillator harmonic seems more likely when you are tuned to a weak station.
Not possible. AM band is about .5mHz to 1.5mHz. Bias was 25kHz to 35kHz. No audible beats, even at 3X bias. The bias signal at the head is relatively high voltage, the field strength of even 50kW a 1mHz at 1 mile is very low. You might get demodulated audio, you won't ever get a beat note that way.
You do understand that in FM, the 19kHz pilot is -20dB below 100% modulation, right? In the days of those switches, average modulation was in the -10 range, before aggressive audio processing became available.
I have never encountered a tape recorder where the bias frequency was that low. And if there really was a problem with the pilot tone why not choose a fixed bias frequency that doesn't interfere with it to begin with ?
Regards
Charles
Regards
Charles
Yeah, I know. You have to look at consumer machines in the few years between FM Stereo's start (1961) and the general exit of tubes. Early recorders hasd low bias frequencies. Once FM tuners became SS they could do better pilot cancellation, and it was no longer as much of an issue, though the MPX switch did appear on stereo cassette decks through to the early 1980s.
Not that simple. Bias frequency is rolled up in the whole design including the heads and equalizers. The consumer stuff had cost issues, and lower bias frequencies saved cost initially.
You don't get phase_accurate's point at all. The erasing head in a radio-tape recorder is not that far from its ferrite antenna. When you add a harmonic of the erasing frequency to a weak AM signal, you get a tone at the difference between the AM carrier and the erasing frequency harmonic at the output of the AM demodulator. Besides, I'm quite sure the AM medium wave band is in the megahertz rather than the millihertz range.
The leakage from bias harmonics into the antenna of a (specific) AM radio-tape recorder would be a design flaw, and has nothing to do with the bias-beat switch. If bias were at 35KHz and the station was 1000KHz, the beat would be at 1000KHz +/- 35KHz. Not an audible beat. Even 3rd harmonic of the bias frequency, 105KHz, still cannot produce an audible beat. I don't believe any harmonic of the bias oscillator at a frequency that differs form the AM station by an audible beat would be present.
Look at how beats work. Two signals mix nonlinearly. Two sets of intermodulation products are produced, one pair around the upper frequency, one pair around the lower. The products are the sum and difference of the two intermodulating frequencies. So in the example, you'll have 1000kHz plus 35KHz (1035 Khz), 1000KHz minus 35KHz (965KHz) because the difference is so great, you won't see the new products around the 35KHz bias. Also, there's nothing there to cause an audible beat. Assuming "audible" has to be less than 15KHz, we'd need a bias and AM carrier that were either separated by 15KHz or less, or very harmonic-rich bias oscillator. There's a problem there. At 35KHz, to get a harmonic at less that 15KHz away from 1000KHz, it would have to be the 29th harmonic of the bias oscillator. Pretty much not going to be there. I could see a distorted bias oscillator, but it's never a square wave, which is what it would take to get the 29th harmonic out of it. Change AM frequencies, and the picture changes. But there's another problem. The intermod product would not be at the same frequency for any station. The filter would be severe to catch them all.
AM is .55 megaherts to 1605 megaherts. There. Written out so we know what I'm talking about.