So, I have a transformer that has two primary windings that can be typically be hooked up in parallel for 115 VAC operation or in series for 230 VAC operation as well as a couple of identical secondary windings. It looks something like this:
Let's take the scenario where we hook up only one primary to the mains. My thinking is that I can treat the second primary winding as another "secondary" and should see the mains voltage on this winding and it would therefore act like an isolation transformer as well as a step down transformer for the secondary voltages.
Where I am confused is the following scenario. Let's say we hook the mains to 1 and 2 above and use 3 and 4 as the input voltage to a full wave bridge rectifier, filter and regulator for my B+ supply. The B+ only needs about 100 ma current capability.
So, what if the transformer is rated at 50 VA and we are drawing sufficient current from the combined secondaries to equal 48 VA including the current drawn from the second primary winding. This is obviously less than the 50 VA rating of the transformer. But, in this case does the manufacturer expect that the primaries will be always wired in parallel and actually used as primaries? Do I risk drawing too much current through the single primary wired to the mains in this scenario? Anyone have any ideas? Is it likely that each winding can handle the maximum rated VA?
If I understand this correctly, if the transformer is rated at 50VA, then each winding must handle the equivalent current (voltages chosed to make the math easy):
50VA
------ = .5 A primary current total or .25 A each winding
100V
50VA
------ = 2 A secondary current total or 1 A each winding.
25V
So, if I have this correct, I risk over-heating the single primary hooked to mains if I am running at close to the rated VA of the transformer unless the manufacturer over-engineered the windings wire size. It seems I should limit the total secondary current to 1/2 the VA rating of the transformer if only a single primary is hooked to mains.
Thoughts? Am I thinking about this correctly?
I managed to find a nice little period (sort-of) speaker for the SW-3 receiver on eBay. It is a nice little "tombstone" style cabinet with about an 8 inch speaker and impedance matching transformer. Presents a nice high impedance to the receiver and the pin connectors exactly match the connectors on the SW-3, so it plugs right in.
The cabinet is cast metal (likely "pot metal" (zinc)) and is quite brittle, though the exterior condition is quite good for something as old as this is. However, as is the case sometimes with eBay purchases, it was unfortunately damaged in shipment and one of the corners was broken off. The seller did right by me and refunded 1/2 of what I paid for it and so I kept it. Fortunately, the damage is on the back and relatively easy to hide from view.
The speaker works admirably with the SW-3 and I enjoyed imagining what it was like in 1930 to be listening to this little rig whilst running off batteries. There was a lot of activity on 40 metres as this was a contest weekend, so there was able listening material along with the 49 metre broadcast band.
Last evening, I changed out the frequency-determining coils in the SW-3 and tuned around the 80-75 metre band finally settling on 3885 khz where a group of AM enthusiasts was parked holding a round-table discussion on everything under the sun for hours on end.
Pretty fun.
So, today I finally sat down to see if the SW-3 is actually functional. The first order of business is to come up with 2.5 volts at 4 amps for powering the valve/tube filaments. I found a Triad transformer that provides 2.5 volts at 6 amps. I decided to try just using AC on the filaments even though it is likely to introduce some hum into the receiver. For now, I just lashed up the filament supply with a power cord and fuse.
For now, I am just clip leading the filament transformer to the power cord on the receiver. Sure as shooting, the valves/tubes all lit up as they should. The AC voltage drops to about 2.25 volts and is drawing just short of 4 amps. How did these guys ever run these puppies on batteries? Must have gone thru them like no tomorrow.
So, with all the valves/tubes lit up, it is time to deal with the audio output. Looking at the bottom of the receiver, I see that the B+ goes directly to one side of the audio output and the other goes to the plate of the audio amplifier. Bloody lovely... 135 volts across the headset... Looking at the valve/tube data sheet it looks like the output impedance will be on the order of 20K ohms. Obviously this 1930 receiver expects a magnetic, high impedance headset. One of my Bogen audio transformers to the rescue. I wired up one of them to the audio output and put a 3.5mm audio jack on the secondary to allow using a powered computer speaker.
Now for the B+ supply. I am using my Heathkit lab supply to provide 135 volts regulated. For now, just clip leading it in place.
So, I cranked up the high voltage supply a little at a time. I didn't want to have any 80 year old capacitors explode if I could help it. After an hour or so, I had the B+ up to 135 volts and no magic smoke was released into the atmosphere.
So, with everything up to temperature and up to voltage, I started trying to figure out how to tune this little gem.
The main tuning dial is very smooth with no backlash and very little yellowing of the plastics and dial card. Amazingly, the main dial also has the ability to change the turning rate from about 18 to 1 to about 4 to 1. Those guys at National did a nice job on this rig back in 1930. The dial left of the main dial is the antenna coupling. It operates very smoothly to peak the signal without introducing any frequency change. The horizontal dial below the main dial is an RF gain control. The regeneration control is the one to the right of the main tuning dial.
As is the case with most regens, they can be very sensitive. While it is tempting to turn the RF gain all the way up and crank the regen control all the way up, it is counter productive and counter intuitive. In this case, you want to peak the antenna coupling for maximum noise and then advance the regen only as far as necessary to obtain a rushing sound in the speaker. This indicates the regenerative detector has started oscillating. Strong signals will overload the receiver and really complicate trying to tune things in, so back off the RF gain nearly all the way and only turn it up sufficiently to be able to hear the signal you wish to hear. With the antenna coupler peaked, you now back off the regen until just before it starts oscillating if you want to tune an AM station or just after it starts oscillating if you want to tune a CW or SSB signal. This is the point of maximum sensitivity. Back off the RF gain until the signal is clear and free of distortion. I don't seem to have any AC hum issues, even though the filaments are running on AC. Sweet!
So, tuning around the band, it appears that the tuning coil I have covers about 5 to 9 Mhz. I was able to hear WWV at the bottom of the dial, a couple of numbers stations, an upper sideband (USB) weather station (National Weather Service - Miami FL), CW and LSB signals on 40 metres, CBCS in Vancouver BC, etc. Nice little receiver! Next, I need to package up the power supply bits and audio output transformer into a tidy external package.
After discussing with my buddy Eldon, I have decided not to use the voltage divider approach to power the filaments. They are all wired in parallel and the risk from one of them opening up and thereby taking out the entire set of tubes is too great, not to mention the royal pain in the tush finding a 1.15 ohm 15 watt resistor will be. I decided to order a 2.5 volt 6 amp transformer from Mouser for $13 today and will use it to power the tubes. The other option would have been to rewire them all in series and use a voltage divider to come up with 7.5 volts from a 12 volt source. At least if a filament opens up we power everything down, plus the voltage divider would be easier to construct without having to locate 15 watt resistors.
I love a few select old pieces of equipment, but powering them can be a pain in the empennage. I have a couple sources for the B+ 135 volt supply, so we should be good to go once the transformer arrives.
Meanwhile, I am on the lookout for an old National "doghouse" power supply for the SW-3.
A while back, I posted a note about having found an old 1930's National SW-3 receiver. The rig came without a power supply. I need to have 2.5 volts at about 3 amps to power the tube filaments. I have a 6.3 volt supply at 4 amps, but need to bring this down to 2.5 volts at 3 amps. So, I thought I would just use a voltage divider.
o--------------------+-----------------o +6.3V
|
R1 3.8 volts across R1
|
+-----------------o +2.5V
|
R2 Load = 3 amp @ 2.5 volt
|
o--------------------+-----------------o 0V
So, how to design a simple two resistor divider? Assuming that there is no load current:
Vout = Vin * R2 / (R1+R2)
The problem with this is that our load DOES draw current (about 3 amps) and therefore this will not work because the load can be considered another resistance in parallel with R2.
The 10 percent rule is a standard method for selecting R1 and R2 that takes into account the load and minimizes unnecessary power losses in the divider.
So, the first thing to do is select R2 so that I2 is 10 percent of the desired load current. This resistance and current is called the bleeder resistance and bleeder current. In this example, the bleeder current is:
Ibleed = I2 = 0.1 * 3 A = 0.3 A = 300 mA
Using Ohm's law to calculate the bleeder resistance:
Rbleed = R2 = 2.5V /0.3 A = 8.3333 ohm
Now, we need to select R1 so that the output is maintained at 2.5V. To do this, we simply calculate the total current through the resistor and use Ohm's law:
I1 = I2 + Iload = 0.3 A + 3 A = 3.3 A
R1 = (6.3 V - 2.5 V) / 3.3 A = 1.1515 ohm
Now considering standard resistor tolerances and value, this would indicate the need for R1 = 8.1 ohm for a 5% resistor and R2 = 1.15 ohm.
In terms of power ratings:
P1 = V1^2 / R1 = 3.8^2 / 1.15 = 12.55 W
P2 = V2^2 / R2 = 2.5^2 / 8.1 = 0.77 W
Did I do this right?
I am not sure what happened, but the first time I did this exercise, I (somehow) came up with R1 = 7.5 ohm 3 watt and R2 = 5.6 ohm 5 watt...
More coffee...
After a couple years of looking, I have located a National SW-3 receiver that is in good shape.
It came with a couple sets of coils, a bandset coil set for 80 metres and an unknown coil set that was hand-wound by a previous owner as well as two sets of empty coil forms for my own hand-wound coils. Subsequently, I have obtained a bandset coil set for 160 metres.
I am now looking for a period power supply, preferably one of the National "Dog-House" power supplies.
My SW-3 is one of the earlier version 2 receivers. It has a pair of 58 valves (tubes) and a 50 valve (tube). The filaments are 2.5 volt and the B+ is anything upwards of 350 volts though the National power supplies typically supply around 135 volts. Meanwhile, I plan to locate a nice high voltage lab supply to be able to test this little beast out.
