06-12-2011, 03:02 PM
Thank you gentlemen for your encouraging responses. And so to the details . . . .
First, a simple diagram explaining the initial concept:
[attachment=3484]
Then an expansion of this:
[attachment=3485]
And finally to the detailed drawings:
[attachment=3486]
RL is the variable load resistance. This compromises a number of individual resistors and 3 'resistor modules', as detailed in drawings 2 and 3.
[attachment=3487]
The rotary switch for RL - which for convenience is drawn as a straight line - must be a MBB type. At each end of the switch there is one no-load position. C1 is included simply to reduce the arcing at the switch contacts.
[attachment=3488]
The 'resistor modules'
[attachment=3489]
Rectifier unit & its PSU; mains input arrangements
[attachment=3490] [attachment=3491]
Two sketches illustrating component layout. (Useful as a maintenance aid).
Comments.
Metering: I did consider having separate meters for voltage and current. However the size of the available case made this not possible.
The scaling of the milliammeter, ideally, would have been 100 mA and 200 mA. But this would have meant awkward and confusing scales on the one meter. For most transformers, 100 mA FSD is adequate, with the occasional need for 200 mA. On the grounds of maximum allowable dissipation of heat in various resistors in RL, the continuous current rating becomes 100 mA, but on test, currents up to 200 - 250 mA were found to be 'acceptable' without overload-heating of those resistors. This makes the unit suitable for 'soak testing' certain transformers.
The chassis, which is plated brass, is simply a salvaged sub-chassis from a scrap TF144G signal generator.
There is no provision for monitoring the temperature of the transformer under test - apart from using the 'finger on the laminations' approach!
I did contemplate fitting additional circuitry (based on an analogue multiplier) that would calculate the total power dissipated and use that result as another meter-displayable function. Lack of chassis space prevented this. (Maybe a feature in a later Mk. II)
Calibration: Two pre-set pots. are provided: one for FSD voltage, one for FSD current, with 'select-on-test resistors' to give appropriate scaling.
Usage: Basically, obvious: by adjusting the value of RL and switching between voltage and current, the drop in voltage as the current is increased can be observed. From that, a graph could be drawn from which the transformer regulation characteristic can be derived.
Application: For selection of a transformer for a specific application, the particular rectification circuit and the necessary filter circuit needs to be known, since the mean current demand on any given transformer depends on that configuration: usually bi-phase or bridge rectification; capacitor or choke input filter.
Finally, my 'get-out' clause! I'm not a professional designer of electronic equipment, so there may well be short-comings or over-sights in the design. As such, any comments in those directions will be appreciated. At present, I do not regard this unit as a 'finished & proven' unit: 'input' from others and time will lead to that result.
Al.
First, a simple diagram explaining the initial concept:
[attachment=3484]
Then an expansion of this:
[attachment=3485]
And finally to the detailed drawings:
[attachment=3486]
RL is the variable load resistance. This compromises a number of individual resistors and 3 'resistor modules', as detailed in drawings 2 and 3.
[attachment=3487]
The rotary switch for RL - which for convenience is drawn as a straight line - must be a MBB type. At each end of the switch there is one no-load position. C1 is included simply to reduce the arcing at the switch contacts.
[attachment=3488]
The 'resistor modules'
[attachment=3489]
Rectifier unit & its PSU; mains input arrangements
[attachment=3490] [attachment=3491]
Two sketches illustrating component layout. (Useful as a maintenance aid).
Comments.
Metering: I did consider having separate meters for voltage and current. However the size of the available case made this not possible.
The scaling of the milliammeter, ideally, would have been 100 mA and 200 mA. But this would have meant awkward and confusing scales on the one meter. For most transformers, 100 mA FSD is adequate, with the occasional need for 200 mA. On the grounds of maximum allowable dissipation of heat in various resistors in RL, the continuous current rating becomes 100 mA, but on test, currents up to 200 - 250 mA were found to be 'acceptable' without overload-heating of those resistors. This makes the unit suitable for 'soak testing' certain transformers.
The chassis, which is plated brass, is simply a salvaged sub-chassis from a scrap TF144G signal generator.
There is no provision for monitoring the temperature of the transformer under test - apart from using the 'finger on the laminations' approach!
I did contemplate fitting additional circuitry (based on an analogue multiplier) that would calculate the total power dissipated and use that result as another meter-displayable function. Lack of chassis space prevented this. (Maybe a feature in a later Mk. II)
Calibration: Two pre-set pots. are provided: one for FSD voltage, one for FSD current, with 'select-on-test resistors' to give appropriate scaling.
Usage: Basically, obvious: by adjusting the value of RL and switching between voltage and current, the drop in voltage as the current is increased can be observed. From that, a graph could be drawn from which the transformer regulation characteristic can be derived.
Application: For selection of a transformer for a specific application, the particular rectification circuit and the necessary filter circuit needs to be known, since the mean current demand on any given transformer depends on that configuration: usually bi-phase or bridge rectification; capacitor or choke input filter.
Finally, my 'get-out' clause! I'm not a professional designer of electronic equipment, so there may well be short-comings or over-sights in the design. As such, any comments in those directions will be appreciated. At present, I do not regard this unit as a 'finished & proven' unit: 'input' from others and time will lead to that result.
Al.
