Alastair Couper Design Desulfator
For this documentation effort, pictures of the waveforms observed on the scope CRT were followed by pictures of the a) horizontal timebase 'sweep' (Time/Div control) as well as the b) vertical attenuator 'gain' (Volts/Div control) settings.
The performance of Unit #3 as a pulser (in the same circuit configuration as Unit #2) is documented on another webpage, i.e., the performance of Unit #3 with the L1 power inductor and the Diode operated into a 10 Ohm resistor (as Unit #2 is shown on this web page) is shown here
Schematic of an Alastair Couper style Desulfator - notice: 1) the slight value changes from 'classic' Alastair Couper values and 2) that correction to C2's value has already been made.
This schematic better shows the minor circuit changes made when 'bench testing' (and not battery pulsing/desulfating) to observe the pulse output by Desulfator into a fixed resistor (instead of a battery).
For bench testing only (with inductor L1 and diode D1 installed) the following minor temporary changes were made to the power circuitry of Unit #2 as described here:
- L2 is jumpered across (dotted line - this allows C4 to function across the + and - inputs)
- The right side of D1 is connected to a 10 Ohm 1 Watt non-inductive Carbon composition resistor
(shown in dotted line format) instead of to the (-) side and the other end of the resistor is connected
to the negative (-) or -12 V rail.
For the rise/fall time testing (no inductors) of Unit #3 shown further below:
- L1 and D1 are removed
- L2 is jumpered across (this allows C4 to function)
- A single 10 Ohm 1/4 Watt resistor is placed from Q1 Drain (bottom terminal of the FET shown above)
to ground, the (-) side of the circuit.
Layout of board.
Note the low-level LM555 'driver' circuit is on the left and the high-power high-energy components are on the right.
Layout of the power 'strip'; this is a diagram showing how a piece of PC board stock will be cut up to create several individual circuit 'islands' where the power components are soldered.
The high-power/high-energy components mount here in a manner that makes changing them a snap for various tests during the several phases of construction.
This wide strip with short interconnections also works to preserve the high-frequency characteristics of this circuity; these are tricks I've learned from designing various demanding, high-frequncy, high-performance RF devices.
The strip before cutting and tinning, This picture shows where the power components will be physically placed. Diode D1 is not shown and only one lead of inductor L1 is shown.
This picture shows the small strip of PC board stock that the power components will be soldered to. This strip has been cut and tinned with solder effectively creating several different islands of copper or large area 'circuit board pads'.
Close up shot of the strip mounted at the edge of the board; this can be a single or double sided piece of fiberglass FR4 board stock. Several cuts have been made in the copper foil creating several individual circuit 'areas' where the power components are soldered to.
Missing in this photo is the small 1 mH inductor, L2, which should be on the far end of the power 'strip'. The yellow thing in the middle is capacitor C4 (100 uF at 16 V).
1
This is a picture of an Alastair Couper-style Desulfator bearing the designation: Unit #2.
o Unit #2 uses a slower P-channel HEXFET (an IRF5210) than the HEXFET in Unit #3 (described further below)
o Inductor is a 145 uH air-core coil for L1
o The output DIODE is going to 1 Watt carbon composition Resistor which acts as the load for the pulser; this is done in lieu of connecting the pulser to a battery and allows various characteristics the pulser to be observed without the variability of battery entering into the equation.
2
o Pulse as viewed across 10 Ohm 1 Watt resistor connected from diode D1's anode to (-) rail.
o Vert: 10 V/Div
o Zero volt reference is 1 major division down from top
o Horz: .1 ms (.1 millisecond or 100 us) per division
Pulse peak amplitude is approx. 42 V
3
Photo showing scope sweep time per division:
.1 ms (.1 millisecond or 100 us) per division
4
Expanded view of pulse
o Vert: 10 V/Div
o Zero volt reference is 1 major division down from top
o Horz: 2 us (microsecond) per division
5
Photo showing scope sweep time per division:
2 us (2 microseconds) per division
6
Close up -
o Pulse as viewed across 10 Ohm 1 Watt resistor
o Vert: 10 V/Div
o Zero volt reference is 1 major division down from top
o Horz: 100 ns (100 nanoseconds) per major division; this translates to 20 ns per minor hash mark
Pulse Rise time appears to be about 100 ns (Nanoseconds)
7
Photo showing scope sweep time per division:
.1 us (.1 microseconds) or 100 ns (100 nanoseconds) per division
8
Scope vertical setting, in conjunction with the use of a X10 scope probe yields:
o 10 V/Division on the vertical graticule
9
Unit #3 Tests - Examine the switching times on a faster HEXFET
10
This a picture of Unit #3.
o Unit #3 uses a faster P-channel HEXFET, an IRF9540
o The FET drain is connected to a 10 Ohm 1/4 Watt carbon film resistor acting as the load for the pulse.
o No Inductors (L1 removed, L2 jumpered across to allow C4 to function)
o No Diode
This configuration produces a nice positive-going rectangular pulse; it is easy to observe the FET on and off times by observing the rise and fall times of the pulse at the HEXFET's Drain.
11
This another picture of Unit #3 with power and 'instrumentation' (the scope probe) connected.
Notice the scope probe is connected to the end of the 10 Ohm resistor that is connected to the P-channel HEXFET's Drain; the Source of that same HEXFET is conncected to the +10 V (or so) supply for these tests.
12
Pulse waveform observed at Drain of HEXFET
13
14
15
16
o Pulse as viewed across 10 Ohm 1 Watt resistor
o Vert: 5 V/Div
o Zero volt reference is the center line (3 major divisions from top and bottom)
o Horz: 100 ns (nanosecond) per division
Pulse peak amplitude is approx. 10 V
Pulse rise time is approx. 30 ns.
17
18
o Pulse as viewed across 10 Ohm 1 Watt resistor
o Vert: 5 V/Div
o Zero volt reference is the center line (3 major divisions from top and bottom)
o Horz: 100 ns (nanosecond) per division
Pulse fall time is approx. 20 ns.
19
20
21
22
Close up view of the fast HEXFET, the IRF9540.