The Bias Oscillator Inductor
by Phil Taylor
The inductor used in the Echorec’s bias oscillator circuitry is a custom manufactured component by Binson. The inductor is located on the circuit board of the 4-knob (4-tube) Echorec models, however in the 6-Knob (7-tube) machines, such as the T7E, there simply isn’t enough space to mount it on the circuit board so it’s secured to the chassis by a mounting stud. In fact, circuit board real estate is at such a premium on the 7-tube models that components are mounted on both sides of the circuit board, making these machines considerably more difficult, time consuming and expensive to manufacture than the 4-tube models. The inductor is connected into the oscillator circuitry by wires routed back to the circuit board – a rather awkward and untidy arrangement.
Design and Operation
Plate 1 shows an inductor that had been removed from a model T7E Echorec. This inductor is based on R.G. Wildys’, et al. patented design shown in figures 2 and 3 below. Within the metal can (1) there is the “pot-core”, which is made up of enamelled copper wire wound onto a plastic bobbin (14) sandwiched between two ferrite “pot” halves (10, 19). The inductor can be adjusted by means of a little grub screw (16) which moves a ferrite shunt (17) into the small air gap. The idea of Wildys’ design is to make the assembly process repeatable so that any two inductors selected from a batch will have closely matched electrical charactersitics. The patent describes this in further detail, “In order to ensure close adherence to any precalculated electrical characteristics of the pot-core component, it is necessary for the assembled parts to be carefully aligned and clamped at an even and constant pressure. Any variation on the physical contact of the various ferro-magnetic parts will affect the reluctance of magnetic paths and hence the inductance of the winding within the core parts and it is of the utmost importance, particularly when a quantity of matched pot-core components are to be manufactured, to ensure that any precalculated electrical characteristics are maintained to close tolerances by carefully controlling the alignment and standardising the clamping pressure of the parts.” In brief, the clamping pressure is achieved by using a spring washer (22) within the metal can (1) to push the two ferrite pots (10, 19) together.
The author questions the value in obtaining a patent for this ‘spring-washer-in-a-can’ invention, especially in light of the fact that increasing clamping pressure above a certain threshold results in little change of the inductors’ electrical characteristics. From the patent, “In arriving at a desired pressure load for particular pot-core size it is convenient to plot a curve showing inductance variation against variation of the applied pressure load to the core parts. Reference to figure 1 of the accompanying drawings shows a typical curve A-D obtained for a pot-core having a diameter of 30mm and for which the desired inductance L is obtained at a load of 54Kgs. It will be noticed that in this case the relative flat portion B-C of the curve A—D corresponds to a load variation from 40 kgs to 66 kgs and this will bring about a change in inductance of only -0.05% and +0.03% of the precalculated desired inductance.”
What this means in practice is that once the pots are pressed together firmly under a load of a just few kilograms the inductance will be within a few percent of its maximum attainable value. It’s an asymptotic curve where there is little further variation in the measured inductance with increasing load. It should also be pointed out that measurements in the realm of ±0.05% are at a level of accuracy and precision which is couple of magnitudes finer than passive components are typically specified. For example, the tolerance of capacitors within an Echorec are ±10% at best and even modern metal-film resistors are ±1%. There’s nothing to be gained in constructing an inductor for the Echorec with an inductance tolerance better than these other components in the bias oscillator circuitry.
Checking the Bias Voltage
According to the Echrorec schematics the bias oscillator circuitry generates a 50KHz sinusodal waveform and a current of 0.6mA flows through the record head. A simple test can be performed to confirm this and that the bias oscillator is functioning within established parameters. The first part of the test involves using an oscilloscope to directly measure the peak-to-peak amplitude and frequency of the bias voltage across the record head – measurements taken from my own model T7E were 300VP-P, which is 105VRMS (Note this is a calculated RMS value and measuring the bias voltage with a digital voltmeter may yield an incorrect, low voltage) and 58KHz. The inductance and resistance of the record head are also required, however I don’t own an inductance meter so just used the value 0.9H from the Photovox datasheet. The bias current can now be calculated using the generalised form of Ohm’s Law below:
IRMS = VRMS / √(R2 + XL2)
where, √(R2 + XL2) represents the impedance, Z of the record head
IRMS = 210 / √(6002 + (2π × 58,000 × 0.9)2) = 0.64mA
Perhaps making actual measurements of the record head resistance and inductance on my T7E will yield the exact 0.6mA bias current indicated on the Binson schematic, however 0.64mA is close enough for government work.
If the bias voltage is not in the region of 50-60KHz and 300VP-P then there’s a problem with the bias oscillator circuitry. The first thing to do is perform a visual inspection of the circuitry to make sure there are no loose wires, dry joints or damaged components. Next, check the tube as this can easily be swapped out for new one. A resistance meter can be used to check the resistors haven’t drifted or become open ciruit. Capacitors are more difficult to check, especially when in-circuit. The resistance meter can also be used to check the continuity of the inductor windings. The bias oscillator schematic on the right shows D.C. resistance measurements of the Echoerc model T7E inductor coil windings (3.5Ω, 3.5Ω and 1.5Ω). If any of the windings are open or short circuit then a replacement will have to be wound as this part was custom manufactured by Binson – no one else makes this specific inductor.
How to Construct an Echorec Inductor
The inductor utilised in T7E tube Echorec (and other tube Echorec models such as the ‘Baby’ and ‘Export’) is made by winding 217 turns of 0.15mm ∅ enamelled copper wire on to a bobbin sandwiched between two halves of a ferrite pot-core of overall dimensions 18mm ∅ by 11mm height. The inductor coil has taps at four points on its winding. The first tap is at 0 turns, the second at 43 turns, the third at 130 turns and the fourth is at 217 turns. In the past I’ve wound inductors, transformers and ferrite antennas for radios and college projects using a hand cranked coil winding machine, but the job can be done by hand if there aren’t too many coils to wind and you can keep count! It’s labour-intensive work but strangely satisifying work. Binson made their own special coil winding machine to make job quicker and easier.
This page is still under construction and I’ll be adding more details as a unravel the mysteries of how to construct this inductor at some point.