AC Power Supply Discussion for Ham Shack

[CharlesHarris]

The following is provided courtesy of Arlington County, (VA) RACES and is a discussion between Steve Cuccio, NB3O of Oakwood Electronics and one of their RACES radio operators:

Q - I want to run a Kenwood TS-480 HX HF-SSB radio (25 Amps at 13.8VDC) and a Kenwood TMV71A (about 13 Amps) dual-band 2m/70cm rig for continuous monitoring in the shack. How large a powder supply do I need to power both rigs at full power? Should it be a switching supply or linear? What does a 50A power supply draw from a household 120V supply line? Should it be on its own breaker? Is a 15A household circuit enough?

A- Running the HF and VHF rigs in the example totals 38 amps. An argument could be made for the limited duty cycle on HF SSB or CW, but it is better to have a little extra overhead. Of the two choices between switcher and linear, I hate to be a Luddite, but chose the linear..but only if it is in a fixed situation (non portable).

The following example compares the Astron SS-50 switcher and the RS-50 linear.

Pros:

A switcher is lighter and runs 80 to 90% efficient when at half to full load, great for economy when running off a gen set.

The voltage regulation is pretty good on lousy cheap-gen set waveforms.

It is also much lighter in weight (5 versus 46 pounds). Very important in a go-kit or portable setup.

A bit cheaper ($200 versus $280).

The linear has open heatsink fins (no fan), with just some minor transformer hum, which is tolerable in a quiet room.

It has a crow-bar circuit to short the output in the event of an over-voltage failure (if the series pass transistors short).

The linear can be easily repaired (assuming the power transformer is OK) and the schematics exist on the web.

The Astron linears use the old 723 regulator and 2N3771 transistors, which are available.

Cons:

The switcher usually has thermostatically controlled fans which can be noisy.

It does not have a separate crow-bar circuit apart from the internal controller IC, in the event of an over-voltage failure (open control loop).

Forget about trying to repair a switcher, unless you have the schematic (I can't find one for the Astron). If the failure is more than a dried out cap, an oscilloscope and some familiarity with the design is important, as there are a lot of different controller ICs and variations in architecture.

It appears Astron made at least two variants for their SS-30M product. The later (>2005 vintage) has more HF switching noise on 80 and 40 meters.

The linear is a true "boat anchor" with the heavy iron. The fins have sharp edges and the box does not have a handle.

The linear runs closer to 50% efficiency (or less), dependent on load and will dissipate the rest as heat.

It does get warm, even when lightly loaded, since the series pass transistors are dropping the raw 22VDC down to 13.8V, which results in 8 volts times the load current worth of dissipation.

The transformer magnetizing current and core loss probably eat over 50 watts when there is no load, so a 24/7 electric bill will be affected.

Sometimes the 15 amp breaker will trip when the unit is first turned on, since Astron did not provide any in-rush current limiting (to charge up the big cap).

Do not hook up the output of a linear-with-crow bar directly to a battery without a fuse and/or steering diode.

Summary recommendation is to buy is a linear for the shack, a switcher for the EMAC Deployment/Field Day bag.

A 50A supply is a good match with your Kenwood TS-480 HX.

50 amp Astrons are typically de-rated by approximately 25% for continuous duty (40A for the SS-50 switcher and 38A for the RS-50 linear).

You may still need to watch the ambient temperature and provide good air circulation, especially if you run RTTY at full power.

Max current ratings are generally based on a +25C degree shack and not an outside table on a +35C degree Field Day (we found out the hard way).

The Daiwa SS-505 is 55A peak / 50A continuous rated, and the MFJ 45 MightyLite is 45A peak / 40A continuous rated, however I don't know anyone that owns these, so cannot speak from hands-on experience. Sometimes manufacturers play fast and loose with continuous duty ratings, which is why I'm be hesitant to recommend other brands without having any direct input.

I never measured the Astron's AC line side current, but I can give an estimate. Unfortunately the schematic is missing a few key voltages, however I'll try to remember how to get through it. This is the long version of the story, which might be useful when troubleshooting:

The supply's transformer, full-wave center tap rectifiers and filter cap provide approx 22VDC unregulated, which probably sags to 20VDC under full load, somewhat dependent on AC line voltage. I am guessing a little, as it has been over decade when I had one open for servicing.

Since the transistor regulator section drops only the voltage (down to 13.8VDC), the load current can be multiplied with the unregulated voltage to obtain the total DC power, minus the rectifier voltage drop (each diode drops 0.7V on the alternating half wave) and transformer efficiency (maybe 90%). 20VDC plus 0.7VDC, times the full 50 amp load = 1035 watts. A 90% efficient transformer will increase the DC equivalent power to 1150 watts total.

However the 1150 watts cannot be directly divided by 117VAC to yield the actual AC current (almost 10 amps). What is not included in this equation is the power factor due to the transformer shunt inductance (unknown) and the actual waveform of the AC current. Even if the transformer has enough windings and core permeability to make its' inductance insignificant, the AC current waveform looks like a sharply clipped sine wave with high peak values, rendering a straight DC-equivalent calculation under-estimating this effect. The current peaks occur when the rectifier diodes start conducting as the AC voltage on the secondary side of the transformer exceeds 0.7V of the filter capacitor value.

These short-duration current peaks can be fairly massive (5 to 10 times), as only the transformer winding resistance and the filter cap series resistance play a major role in reducing this. This is one of the main reasons to observe the peak repetitive current rating on rectifier diodes when finding replacements. Below this voltage threshold during the rest of the sine wave, the rectifiers fall out of conduction and only the current to magnetize the transformer core is present (less than 1 amp). This large peak to average current ratio will make some circuit breakers hum or false-trip. It will also cause some of the cheap gen sets to introduce harmonics (ringing) during the off-peak current time, which themselves can produce high peak voltages (beating the heck out of the MOV surge protectors).

The only way of a truly accurate current measurement for a power calculation is to use a true-RMS current meter which can integrate at least a ten to one crest factor to average out the peaks. Or use the older thermocouple current meters. This is true for most supplies these days, as even the switchers have the same issue; sometimes even worse since the diodes hang directly on the AC line, unless some additional filtering is provided (FCC type accepted versions).

A dedicated 15 amp circuit would be a VERY good idea. A 20 amp circuit would be even better, as its AWG#12 wiring will also provide a lower (stiffer) impedance during the high current peaks. The stiffer the AC line, the less voltage sag during the spiked current waveform (less aggravating to other equipment). I've had a few rare occurrences of false-tripping the old 15 amp breaker in the house when the supply is initially turned on, due to the in-rush current when charging the big 100,000uF filter cap from zero. Probably not enough to consider adding an inrush circuit, but enough to keep other big ticket items of the circuit during the initial power up.

Surfing the eham.net reviews and saw some recent posts regarding the construction quality of newly purchased Astrons. It appears that Astron now uses the same caliber folks as MFJ, which means you would want to open the case and check for cold solder joints and loose hardware before power up (some assembly required). Maybe even put a headlight on the output and measure the voltage, before committing to a nice expensive rig.

[Lightspeed]

an A) from a European perspective:

I’d say its unlikely that the operator would ever be drawing full load from both radios at the same time. As he said in the OP the VHF/UHF radio would be used for permanent monitoring, in which state its current draw would be minimal, typically 1 to 2 A at 13v.

So, lets say his absolute max current draw would round up to 50amps at 13v = 650w

That’s 2.7amps current draw from the UK AC which is delivered at 240v

Assuming a switch mode supply is 80% efficient, that would see it drawing a rounded up (again) 4 amps. From the AC circuit.

So is it essential to run the shack on its own supply line and contact breaker? In terms of current draw for the described operation, no, its not essential. But I’d personally follow best practice and still run a dedicated circuit protected with its own 6a or 10a breaker to the shack.