Dream tuner updates are posted regularly on the Alpha 4040 Yahoo Group - check in there for more frequent updates.  Here is a summary of the past month......

 

Software development continues, with some very nice enhancements taking place in the realm of user interface friendliness and functionality. We have fixed a few bugs, and continue to test every solution. We have ran into a couple of hardware issues, and as such we are looking at the best way to resolve these in a timely manner. None of these are delaying our beta test.

 

The first one deals with field upgrade of code to the motor controller. Solutions have been found, and the hardware is being reworked to fix the issue.  This will not impact the ability to continue with Beta testing, but must be resolved for production shipment. To that end, a spin of the motor controller PCB is underway.

 

The second issue deals with the stepper motors.  Again, we have solutions for beta test, and will be making a hardware change for the production build.

 

Neither of these are serious, but keeping in line with our commitment to ship a rock solid piece of hardware, these are being addressed for the production build. 

 

Software Updates:

 

1. The error logging facility is fully operational and is being tested for robustness. This facility is used to log any error conditions that may occur and stores this data for further analysis. Along with the error logging facility, the tuner will be able to "phone home" - allowing the error conditions to be uploaded to the engineers at RF Concepts for evaluation and possible corrective action.

 

2. The tuner also has the ability to execute "auto software update".  We have discovered a few bugs here, and those are in the process of getting resolved.  This function allows the tuner, when connected to the web, to look for the latest code versions, and automatically update itself if so enabled by the owner of the tuner.

 

3. The tuning algorithm is still not quite up to speed, and is being addressed.  Still needs a bit of "tuning."...Sorry for the pun folks....

 

All in all, we are progressing nicely. Still not where we want it to be for release, but becoming more solid by the day.

 

Thanks again for being so patient. I know everyone is looking for a tuner to be delivered just as soon as possible, and we are striving to achieve that expectation.  We continue to make progress at getting bugs out of the hardware and software.  

 

-Joe Jurneke, 

Letter from Gordon Hardman, WØRUN  

Mains wiring and where does all that power go?

Last month this article was inadvertently truncated, so we're reprinting it in its entirety this month - so here, in full,  is the rest of the story......

Many newsletters ago (or so it seems) I addressed the topic of the current drawn by an amplifier, and how this affects mains wiring. A recent experience with a customer who was convinced his amplifier was defective made me think it might be time to revisit this. After extensive telephone support, it turns out he had a very long run of 16awg wire bringing his 240V to the amplifier. After replacing it with 10awg wire everything was perfect. 10 gauge wire might sound a little on the thick side, especially if you use simply the total power drawn by an amplifier. After all, we fit the amplifiers with 20 amp line fuses, so obviously the rms current must be less than 20A. With an amplifier of typical efficiency, the rms current at 240v rms ac is around 12 amps. If you go to typical wire gage table, it might look like this:

It looks like maybe 16 gauge ought to do the job. But, as the saying goes, what about the rest of the story? It turns out you can make a pretty good run at looking at this using a cool piece of free software called LT Spice. See http://www.linear.com/designtools/software/ for more details. Linear Technology has made available what was a high-end analysis tool only a few years ago.

I developed a very simple model of what one of our amplifiers would look like, using LT Spice. This assumes that most of the power goes to the big HV power supply, which is represented by a transformer, bridge rectifier and capacitor bank. I represented the tube by a fixed resistor, it uses most of the power. It is obviously a highly simplified model, but it fairly accurately captures where most of the power is going.

R3 represents various secondary-side losses, including the transformer wire and two series resistors. R1 is the line resistance, and I varied its value to simulate various lengths and gauges of wire. The "ideal transformer" is a little tricky to set up, but is modeled as two coupled inductors, with the mutual coupling between them being very tight. The peak value of the voltage source is appropriate for a "240V" circuit at 60 Hz- the voltage across R2 is close to 3kV "key up" i.e. if you make R2 very large. The first notable thing is that the current waveform is nothing like you might expect. Using the "probe" provided by LT Spice, and setting R1 to zero, the following waveform results through R1 and V2.

Bizarre! The current flows in pulses. This is because the 30uF capacitor, C1, is only "topped up" when the voltage out of the bridge rectifier is higher than the voltage currently across C1. Notice that the peak current is now greater than 40 amps. The exact value of this depends on the fidelity of the model, but we have put a current probe on amplifiers, and connected this to an oscilloscope, and confirmed the general outline of this. In some cases, when the amp is driven hard, the current pulses are greater than 60 amps! (we are on 208V ac, so the current must be higher for the same power).

This results in what is called "poor power factor". Power factor is the ratio of real power to apparent power. In this case, the apparent power is much higher than the real power, so the power factor is low, around 20%. But this is not the worst of it! It turns out that the voltage drop across the mains wiring is due to the peak current. To explore this, I varied R1 to represent various real lengths of copper wire, according to various gauges of wire. I then recorded the voltage across the 2900 ohm resistor, R2, which represents the approximate plate voltage of the tube in an AP8410. Normally this sits around 3kV "key up". I looked at the plate voltage for four different wire gauges- 10 through 16. Here are the results.

On the left hand side, obviously, if the ac line length is zero, then it does not matter what gauge wire is used. The voltage comes out a tad over 2700 volts. But as the line lengthens, the voltage droop increases. And likewise, as the wire gauge grows, so too does the droop. In all these cases, the wire is probably not going to get excessively hot. Indeed, for the thicker gauges, the temperature rise might be all but unnoticeable.

To go back to our poor customer who had close to 150 feet of 16 gauge wire, it is clear why his amplifier was struggling. In actual fact, the graph above may be optimistic- in most cases, the ac line leading to your shack will also have some resistance, which will make the droop even worse. Our customer replaced his wire with 10 gauge, and all was well.

It is tough to explain all of this to an average electrician. They look at the total load, and do the standard calculations, and come out with a wire size that is way too small. If you are having some work done, you may want to have your electrician take a quick read through this article to see why much thicker wire is required. This is especially true if you are running at less than 240V, which is what the graph was prepared for. It also shows why using a big amplifier off 115V is a tricky proposition. The peak current pulses will be in 100 amp range, so the wire drop becomes an even bigger problem.

Finally, although heavier gauge wire may cost a fair bit more, it is still only going to be a fraction of the money you have spent on getting the finest amplifier in the world into your shack. It is money well spent.

73, Gordon, WØRUN