DST 100 Power Supply

With the refurbishment of any WW2 equipment I aim get the thing roughly into working order before spending a lot of time on the restoration, but as this example of a DST100 appears to be complete and with no obviously unserviceable parts I reckon its worth expending effort making a contemporary-looking power supply before powering it up. The two pictures below were taken by Brian, M1JLM of his Rectifier No.8 power supply which is the type used with the DST100. The dimensions of the case are 177mm wide x 160mm high x 300mm long. I've been building, upgrading and repairing computers for many years and I've collected lots of cases because they may come in useful and one old case measured exactly 180mm wide so I marked it out and cut it to shape. I won't bother cutting louvres but I'll cut the holes in the sides and cover these with steel perforated sheet.

Cutting top louvres will be too difficult so the top will be plain.

I need to find a suitable mains transformer, choke, capacitors, rectifier valve and a few other bits and pieces then make a simple chassis with front and rear panels. What size transformer? The spec for the Rectifier No.8 gives the transformer size as 92VA. The DST100 spec says the power consumption is 110mA @ 250 volts and 4.75A @ 6.3 volts. This is about 58VA. The 5U4 rectifier valve requires a heater current of 3A @ 5 volts and there's a dial lamp, say 300mA @ 6.3 volts adding an additional 17VA. That's a total of 75VA. I'll add a little more to deal with cable losses say 5A @ 0.5 volts bringing the total to 77.5VA (this is because an unloaded heater supply of 6.8 volts terminal voltage is likely). Transformer losses will be say 10% equating for example to heating due to winding resistance. The requirement is therefore around 86VA so the quoted figure of 92VA is perfectly reasonable. Some transformers are marked with their VA rating or their rated voltage and current outputs but not always so. The physical size and weight is therefore a reasonable way to select a transformer from one's junk box. Once a transformer has been selected it's a good idea to check the output voltage under a specific load. If there are extra outputs provided which are surplus to requirements, these can be used with some care to supplement say the HT voltage. For example a 20 volt winding can be added to an HT winding if a full wave bridge is contemplated as that type of circuit does not need a centre tapped winding. A typical example is a transformer I found yesterday. It's a toroidal type and is wound with three untapped secondary supplies: about 220 volts, 20 volts and 6.7 volts. I added the two higher voltage windings together and this gave 240 volts, ideal for 250 volt DC output. I noted that the low voltage winding supplied about 6 volts at 5 Amps so I added four extra turns of heavy copper wire in series with the 6.7 volt output which gave me under load about 6.3 volts at 5 Amps. This would be OK if I used semiconductor rectifiers rather than a 5U4 valve rectifier. Of course adding transformer windings must be done correctly. The windings need to be connected in-phase to get a higher voltage. Below for reference is a power versus approximate weight table for toroidal mains transformers. The one I tested weighed in at 1.8KG so would be about 180VA. I decided against using this toroid as its rating was double that for the job in hand.

Identify a toroidal transformer by weight  Left, a junk box toroid weighing 1,800 gram, and therefore about 180VA. Easy enough to add extra turns of wire to improve the heater voltage under load.    POWER IN VA  WEIGHT IN grams  POWER IN VA WEIGHT IN grams  15 325  160 1,690  30 475  225 2,300  50 675  250 2,450  60 765  300 2,900  80 1,000  500 4,800  100 1,160  800 7,000  120 1,360  1,000 8,000

I also checked some standard transformers. One was a WW2 example marked 38VA and weighed 1,800g giving an indication of the widely different power/weight ratios of the older style versus newer toroids. Below is a table showing power versus approximate weight for standard modern transformers. A second transformer labelled "Gardeners L425" marked with voltages and currents totalling 57VA weighed 1,900 grams. Assuming say 80% efficiency this would make the rating about 72VA. I suspect any ex-military transformers will be conservatively rated so a commercial rating for this of 100VA lines up with the chart below. I haven't differentiated between normal and C-core types and of course extra weight can come from the method of construction especially in mounting and terminal arrangements.

Identify a standard transformer by weight Left, a very heavy un-representative 38VA transformer dated 1945. It seems to have only three low voltage windings.    POWER IN VA  WEIGHT IN grams  POWER IN VA  WEIGHT IN grams  6 200  200 3,500 12  400 250 4,300 20 550 500 7,700 50 1,150 800 10,000 75 1,650 1,000 11,000 100 1,900 1,500 14,500 150 2,500 2,000 17,000  The VA rating basically is the product of the transformers primary voltage and current when running at its maximum design load. Roughly speaking this will be similar to the sum of the products of its secondary windings. The higher its efficiency the closer this will be and the higher the rating the thicker will be the copper wires used in the windings. Something to note is the fact that thin wire has more resistance and its voltage drop at the maximum design current will therefore be higher. Run any 6.3 volt transformer open circuit and the voltage will be higher than 6.3 volts. Generally speaking, the higher the VA rating the closer the voltage will be to 6.3 volts. This being so you'll need connecting wire which is heavy enough to maintain the transformer output voltage to the equipment at full load.

After lots of rummaging in various boxes I've selected the components for the new PSU so can proceed with the metalwork for the case.. below.

The power supply chassis (177mm x 300mm x 25mm) resting on top of the half-finished cover.          The base plate will be screwed to the five studs. Two metal brackets will support the front panel and these, although smaller, like the original PSU will be mounted slightly inboard from the chassis edge. My chassis is the full width of the case unlike the original because it makes construction easier. The front panel will carry an indicator lamp, two fuses (mains and HT) and a toggle switch. It will not include a mains selection panel. There will be two transformers, one for 6.3 volts and the second for HT, a small choke and a chassis mounted electrolytic capacitor. Because the HT winding of my preferred transformer has no centre tap I'll be using a silicon full-wave bridge rectifier. A rear plate will carry an IEC mains connector and a connector for LT and HT.

Above on the left a step-down transformer according to its NATO number and right a low voltage transformer carrying three 6.3 volt secondary windings. Not far to go now. Fit mesh to the outer case and fit mounting screws ...below, a trial fit.

Then make and fit a rear panel with brackets.. below. I've used an IEC mains connector and a beefy Belling Lee connector for power out. It needs painting and of course wiring...

Having finished the mechanical construction of the power unit, wiring up proceeded. Like my collection of computer cases saved for metal bashing, I have a collection of old computer power supplies saved for parts and wire. The wire used in these is generally designed for use in mains circuitry so has a decent insulation factor and comes in a nice selection of colours. Like computer power supplies, I can increase current handling by paralleling wires. Because I'd already bench tested the mains transformers to check their outputs everything should have proceeded smoothly, but this was not to be the case. After I'd fitted fuses into the mains and HT feeds and carried out a visual check plus tests with a ohm-meter, I plugged in and switched on. The workshop lights dimmed momentarily and the HT voltage measured a mere 30 volts and was slowly dropping. The 5 amp fuse in the PSU mains lead was open. I decided to fit a new fuse and to use my variac but switching on changed the variac hum to silence. The variac was set to zero output.. very odd. After a quick check I found the PSU HT bridge rectifier had a short circuit between its negative output and an AC input. Why was this? I switched on the workshop computer to check the ratings of the rectifier... no Internet... the fuse blown by the variac was shared with my ethernet switch. I fitted a new fuse and recovered the Internet and found the rectifier was rated at 600 volt and 4 amp. The transformer output was 220 volts so maybe marginal? I fitted another.. this time rated at 800 volts and 2 amps. This also went pop and the same bridge diode had failed. From distant memories with TV sets I recalled that this used to happen if a leg of the mains was grounded through a bridge rectifier, but maybe it's just a surge current which is damaging the bridge rectifier as I hadn't fitted surge limiters.. the output windings totalled something like 12 ohms so a significant surge current is possible. On the other hand it could be a faulty transformer because, unlike initial tests the two transformers were now bolted to an earthed chassis. I even wondered about the on/off switch grounding through its securing nut. I disconnected the HT leads feeding the bridge rectifier and after fitting yet another fuse, switched on. The 6.3 volt dial lamp came on and the LT measured 7.2 volts AC. The HT measured 220 volts AC so the problem lay within the HT rectifier and smoothing circuitry. I weighed up the options.. first a leak from mains to the HT output.. second a high surge which is destroying the bridge rectifier and third, something completely different. An obscure option maybe, because each time the fuse blew I was able to measure around 30 volts at the smoothing capacitors? One possibility is the HT choke. This is the primary winding (2 Henries) of a 2 amp 12 volt transformer screwed down very close to the chassis using brackets that came with the transformer so maybe a leak from the choke to chassis which happens only after the HT has risen to say a couple of hundred volts? Then again.. my junk box HT transformer is a complete unknown and may have been removed from its original equipment because it was faulty? From its markings I'm pretty sure the thing came from a KG40A crypto so I'd imagined the windings would have been well isolated to suit Tempest requirements. Maybe it's just an initial surge which is blowing the bridge rectifier? A thermistor will sort this out and also reduce any HT surge. I'm going to carry out two tests. One is to wire a mains lamp across the 220 volt output. Then, I'll connect the mains lamp from one side of the HT wires, then the other to chassis. This should prove the winding output and isolation to mains ground. A 40 watt lamp should be OK as that will draw about 200mA. What worries me is the fact that my variac blew a fuse when the output was set to zero volts. Sure enough the lamp lit between one of the HT leads and ground. What I'd assumed to be a 200 volt output turned out to be another primary winding coupled to the others so back to the drawing board. Peering under the top of the KG40A transformer you can clearly see a link from the mains primary winding to what I thought was a secondary tag. After all said and done the tags for this winding are on the opposite side of the pair of screen tags where the secondary windings are terminated. The transformer seems to be designed for 120 or 240 volt mains and I mistakenly ran a 120 volt winding at 240 volts and measured the other half of the primary thinking it was a secondary (effectively an auto-transformer). Below is the circuit diagram of the new power supply

Once I'd carried out a simple test and discovered the problem, I needed another HT transformer so I selected an isolation transformer that I'd salvaged from an old Sony TV set about 25 years ago. I guess it had been too difficult to meet new safety regulations and Sony had given up and just fitted an isolation transformer. I fitted it with the higher resistance winding conncted to the mains supply and this gave about 230 volts out from 240 volts in. I added a small thermistor of about 10 ohms in the primary connection to reduce switch-on surge and the rectified output was 305 volts off-load and 205 volts with a 1.8kohm dummy load. I swapped the original choke which had a resistance of 330 ohms and an inductance of 2H for another with 80 ohms resistance and 10H inductance to reduce voltage losses and I got 235 volts output under the same load. Again it's a 12 volt transformer not a real LFC but should be OK. These changes meant moving the LT transformer slightly and redrilling the chassis. If you look closely at the front and rear panels below you'll see I left a gap so that the outer case flanges slide into place between the brackets and the panels to provide extra rigidity. The blue wires below carry outputs from the three 6.3 volt secondaries and all three terminate at the same Belling Lee connector pin. Paralleling windings is OK if they have roughly equal voltage outputs. I checked this and each measured exactly 7.1 volts across their connection tags. The exact voltage will be determined by the original primary winding rating together with the mains voltage at the time of measurement. The reading will be slightly higher than the marked 6.3 volts because allowance has been made for the winding resistance at the rated current. At 3 amps a secondary resistance of about 0.26 ohms will reduce the output voltage from 7.1 volts to 6.3 volts.

Above you'll see I've used an IEC mains connector and a Belling Lee socket for output power. The former for convenience and the latter to handle the 4.75 Amp heater current of the DST100. I haven't included a power rheostat, as is fitted to the Rectifier No.8 unit, for altering the HT voltage. Because my PSU will be used for a specific DST100 I can make a fixed modification to the HT voltage if this is required. Later I'll tidy things up and maybe paint the case. See below.

Maybe it's a symptom of advancing age, but yet another glitch in proceedings. I needed to make a new cable to connect the completed PSU to the DST100 and after about an hour of searching came across a 4-pin WW2 connector (a match to that used in the DST100) . I then found I'd mislaid the Belling Lee mating connector. Eventually I'd got both and also some heavy duty mains cable of the maximum diameter to fit the pair of connectors. Another hour passed and the cable was completed. I plugged the thing into the new PSU and checked the resistance of the leads. There were three.. ground, LT and HT. I checked ground and it was open circuit. Very odd... visually it was perfect.. but wait... the LT and HT sockets in the plug fitted to the cable were the same size and fitted perfectly but the ground socket was 25% larger in diameter than the ground pin in the chassis connector. Although the junk box had yielded a mated pair of connectors, it seems the two were incompatible. I realised the problem fairly quickly because it happens a lot with those small DC power connectors used on laptop computers. You need a mating connector with exactly the right size of centre hole otherwise it will be intermittent. After a rummage I found a WW2 5-way chassis mounted socket plus three mating plugs. At least these would allow me to make extra leads for other equipments... maybe my R1132? Fortunately the hole in the chassis was a perfect size for the new connector. I fitted this but had no end of trouble soldering the heavily tarnished plug pins to the cable. I used two each for ground and LT and the other one for HT. Plugging the PSU into the receiver lit the front panel lamp and I could see heaters warming up. I'd removed the HT fuse temporarily so I could connect my variable HT supply to save trouble from and unexpected leaks or shorts.

Above the new PSU with a new cable capable of carrying the LT current of nearly 5 amps without significant voltage drop. The cable is a length of 3-core heavy duty mains cable and terminated in a 4-pin DST100 style power connector (which was also used on the R206, R107 and other WW2 sets) and a 5-pin connector with an AM logo and part code retrieved from my junk box. The mains connector is a standard IEC type with a leading earth pin and the mains cable an ordinary computer type.

Above are pictures of the original and its modern cousin. Exactly the same physical size and power ratings.

Return to the DST100 Refurbishment