After school I started working for the Philips company. We build machines for the production of parts for consumer electronics. Color TV parts in particular. (Deflection units and HV transformers.)
Later I started as trouble shooter electronic maintenance on these production machines. And slowly I entered the digital world.
Production machines got computerized and I studied 'computing & control' learning assembly languages. I was 27 when Philips gave me the opportunity for a one-year retraining to IT/ICT (TU-Eindhoven). Later they offered me a job as systems programmer IBM mainframe computers at the Philips computer center in Eindhoven. (VM/SP, VM/SP HPO, VM/MA, VM/XA & VM/ESA). Great time!
Later, at 38, I got a same kind of job with the Ericsson company and moved later to SAP, Unix, Java and mainframe / Unix networking.
I remember the moment I got the first TCP/IP connection between our mainframe(IBM) and Unix up and running. Hi.:-)
Internet did not exist jet!
Great time, but I kept my interest in electronics and combined it with my software skills.
There is so much that I have build (and programmed) over the years.. a fraction of it on this web site.
Let's talk about a stabilized(?) power supply for HAM use. 13.8 Volt / 0-30 amps.
A very popular power supply is the EP-925. It is sold under several 'makes' but they all look the same.
I had one for many years but I was not impressed by its stability. I saw the voltage needle swing during my SSB QSO.
A drop of 0.8V or more was normal. An instable power supply on a transmitter results in a distorted signal (SSB/CW).
I took the diagram from the web and studied it. Also did some measurements.
Below the original diagram.
Diagram of redesigned power supply.
Weaknesses and what I've changed:
The 723 voltage regulator is powered from a separate secondary tap on the transformer. Purpose: a stable source.
Due to the inferior quality of the transformer the voltage on this tap varies considerably under the influence of the load on the main secondary tap and the current through the primary tap.
To solve this, I added an extra stabilizer uA7824 (IC3) to give the 723 a stable source. In my EP-925 C17 was missing. I added C17, 100nF.
The 7824 should be mounted on the heat-sink.
The circuit around the 723 could be improved on some points.
TR6, NPN TIP31C is replaced by a PNP TIP42C. This design makes use of the Hfe of TR6 to add more stabilization.
(Print track interrupted under 723. Pnt.12 = V+ on 7824, pnt.11 = Vc to base TIP42C and pnt.10 = Vout to collector of TIP42C.
R6 = 330 Ohm and R7 = 680 Ohm.)
The inputs of the 723 were set up for lower output voltages (up to 9 volts).
I chose 13.8 volts and adjusted the input setting accordingly. R14 to 680 ohms. Parts removed: R15, R16, D13 and D? on pnt.1.
Input pnt.4 changed (R16 bridged) and additional resistor (R13.8) parallel VR3 (output voltage) in order to limit the maximum output voltage to 14.0 volts.
The circuit around the fan I never understood!
A thermostat, two op-amps and a transistor to switch a fan On and Off ?!
The thermostat used is a NC (Normal Closed) type. May be that's why they came up with this strange circuit.
I dropped it.
The thermostat is replaced by a NO-type (Normal Open) which directly switches the fan.
Parallel on the thermostat a 220 Ohm / 5 W resistor lets the fan run silently at half speed.
The heat-sink is continuously cooled and I have not yet noticed that the thermostat set the fan at full speed.
Fan replaced by a PAPST 8414NG mounted blowing outwards. This reduced the noise a lot. On idle (half speed) even not noticeable.
I closed the air gaps at the top of the cabinet with tape so that the cool air from the side gaps is sucked in along the heat-sink to the back.
Over-voltage protection. (*)
The maximum output voltage is limited to 14.0 volts.
I added a surge protector of +/- 16 volts.
Thyristor BTV24 / 1400R connected in parallel with the bridge rectifier (replaced by a 50A / 120V type) is controlled
by ZD2 and R 100 Ohm sensing the output.
When over-voltage occurs, the thyristor will short circuit the rectifier DC output which will cause the fuse to blow.
(*) This is a simple but not so neat protector.
Damage on the rectifier is not unthinkable.
In one occasion the fuse broke when turning on the power, but after replacing the fuse all was OK.
The fuse is a S-type (slow) because it has to handle the switch-on current surge.
I'm thinking about a 'slow start' update so the fuse can be replaced by a lower amp F-type (fast).
This would prevent an accidental activation of the over-voltage protection and spares the rectifier when there is an over-voltage situation.
!!! For this overhaul several board tracks have to be cut, parts replaced, added or removed. It's not a novice job!
Results. Measured direct at the output, the voltage varies less than 5 mV between minimum and maximum load. No noise.
Now I can use my K3 with full power without being distracted by a dancing voltage needle. :-)
My 'workshop'.... Eimac 3-500Z