WEAK-AM

The way you adjust the audio controls on an SSTran can have a major effect on the sound of your signal. The unit has considerable flexibility, but it can be a little confusing to set up if you don’t have test gear. Here are some ideas.

Before you do any audio tweaking, make sure you have tuned the RF section of the transmitter properly, or you won’t get satisfactory results. In addition, use the best radio you have available for monitoring the transmitter. Among portable radios, I still prefer the GE Superadio III. There are lots of good choices in ac powered sets. I am currently using a Cambridge Sound Works 735i as my monitor.

I would suggest starting with the Gain and Compression controls turned all the way down (fully CCW) and the Modulation control at 12 O’Clock (mid-range). Next, advance the Gain control clockwise until you hear audio from the radio. As you turn up the gain, you will probably notice a point beyond which the audio doesn’t really increase much. Set it at that point. If you can’t find that point even with the gain control all the way up, the level coming from your audio source might be too low. Next, turn the Modulation control down (if necessary) until you don’t hear any distortion on peaks. The modulation control adjusts the limiter threshold. It is very helpful to have a scope that you can use to sample the RF output, but lacking that, you will have to rely on your ears. On my unit, under these conditions the Gain control sits at about 9 O’Clock while the Modulation control sits at between 10 and 12 O’Clock, but your settings will depend on the audio level going into the transmitter. It is better to have a fairly hot level going to the SSTran because if the level is too low, you’re going to have to compensate for that by turning up the Compression, which you may not want to do.

If you’re using the transmitter for classical music or want the most accurate sound, you’re pretty much done at this point. With the Compression control set fully CCW, the transfer function is 1:1, and the noise gate does not engage. This is the way I prefer to operate my SSTran. On the other hand, if you want to “punch up” the audio, here are a few things to try.

The Compression control will allow you to adjust for 1:1 up to 5:1 compression. If you listen to speech, you will find that even 2:1 compression is quite a lot. There are no panel markings to indicate the 2:1 point, so you either have to infer it from the data sheet or be able to measure it. Based on my measurements, the 2:1 point occurs around 9 O’Clock on my unit. With quiet program material, you will start to hear the noise gate function at this setting. You may or may not like this– I didn’t. On the other hand, if your programming is “dense”, like rock music, you’ll most likely not notice it. I personally don’t see a need to use more than 2:1 compression, so I am planning to modify the circuit to “spread out” the adjustment range to make the Compression pot only cover from 1:1 to 2:1. If you are interested in doing this yourself, it’s not hard. According to the SSM2166 data sheet (the audio IC used in the SSTran), 2:1 compression is achieved with a resistance value of 12.5K, whereas the pot used in the SSTran is a 100K unit. This is why the 2:1 point occurs at such a low setting. In case you’re wondering, this compression slope is not “digital”– there’s nothing to stop you from choosing a modest 1.5:1 slope instead– if you can figure out how to set it.

If you want maximum range at the expense of audio quality, you may want to try the maximum compression of 5:1. What this sounds like will depend significantly on the type of program material. I tried it with dense, heavily processed hard rock, and honestly could not tell the difference between 2:1 and 5:1. On most other types of music, I did not like the way it sounded. On speech, it sounded like a Travelers’ Information Station (very compressed, but intelligible). In the end, it’s up to you.

Once you’ve adjusted the Compression the way you want, it’s time to go back and re-check the setting of the Modulation control. That’s because with heavy compression, you’re going to be hitting close to maximum a lot more often. If you’re watching on a scope, you’ll notice that the signal can be set to clip for short intervals without that being audible. But if it clips all the time, it will definitely be audible.

When you get done, you’ll probably find that your SSTran sounds a lot better than most commercial AM stations on the air– and it can actually sound about as good as FM if you’re listening on a wideband AM radio like the Superadio. It’s amazing how good AM can sound, and it’s too bad that most stations don’t deliver on this promise. The audio quality of many of the big 50 kW stations is especially bad, now that a lot of them have converted to IBOC transmission. These stations are generally limiting their analog audio bandwidth to 5 kHz or less. That’s part of the reason they sound so bad– the other is due to the assorted buzz, hiss, and screech sounds coming from the digital sidebands.

To summarize, here are some recommended “starter” settings, based on program type:

Classical– Gain: source level dependent. Modulation: 10 – 12 O’clock. Compression: fully CCW (off), up to about 9 O’Clock (2:1) if you want to boost soft passages.

Talk Radio– Gain: source level dependent. Modulation: 11 – 1 O’Clock. On higher settings, you will start to hear some clipping. Compression: 9 O’Clock up to about 12 O’Clock.

Rock– Gain: source level dependent. Modulation: 10 – 12: O’Clock. Compression: 10 O’Clock up to full CW (5:1)

Test (tone modulation)– With a 1 kHz tone, set Gain to point at which level stops increasing, with Modulation at 2 O’Clock and Compression at full CCW (off). This will be close to 100% modulation. This information is based on my tests in the lab.

If the SSTran is working properly, there should be *no* audible distortion due to the transmitter. But I have found that under some circumstances, problems can occur. Here is what I learned.

If the SSTran is working properly, there should be *no* audible distortion due to the transmitter. But I have found that under some circumstances, problems can occur. Here is what I learned.

In the manual, there are two components that are affected, depending on your operating frequency and whether you are using the supplied wire antenna or a base-loaded antenna. The first is R18. The manual says that you should leave this out if you are operating above 1,000 kHz or using the base-loaded vertical antenna. The second part is C23. The manual says that C23 should be 820 pF if you are using the supplied wire antenna, but it should be 560 pF if you are using the base-loaded antenna.

I was planning to mount my transmitter outside and use the base-loaded vertical, so I left out R18 and used 560 pF for C23. I thought I could get by with this and operate temporarily using the indoor wire antenna. WRONG! Depending on the setting of C5, the output loading trimmer, I noticed distorted audio and low modulation level. I think the reason for this might be that the collector breakdown voltage of the 2N2222 output transistor is fairly low, and when this is exceeded it can cause distortion. It is also possible that the output may be going unstable; I have not looked into that.

Experimentally, I decided to tack solder an 820 Ohm resistor in for R18 and see what happens. The first thing I noticed was that the maximum DC voltage measured at T1 and T2 dropped almost in half, but the audio cleaned up completely. So the resistor is taking a little power, but it helps stabilize the output stage loading considerably. For best efficiency, it would probably be better to leave the resistor out and change out C23 for the correct value. But in my case, I was only broadcasting around the house and didn’t need the extra signal.

Bottom line: if you are planning to go to the base-loaded antenna eventually, but want to get on the air indoors using the wire antenna in the mean time, you’ll have to install the appropriate parts or you probably won’t be happy with the results. Don’t forget to make the necessary changes when you do change to the base-loaded antenna later on.

I have invested some effort in finding out what can cause hum on in-home broadcasts to ac powered radios when using the SSTran transmitter. I hope this information will be helpful. If you disagree with my findings, that is fine– I am just reporting on what I found.

First, I will deal with the topic of hum that is present on the carrier when there is no audio input and the Gain control is turned all the way down. There have been some complaints that the included ac transformer is responsible for hum on the carrier, and I have to say that after careful testing, I cannot agree with this. Below, I describe what I did.

I set up the SSTran in a typical “home broadcaster” fashion, with the transmitter indoors and the ac transformer plugged into a wall outlet along with other appliances. The standard wire antenna was run up along the wall, and the transmitter was grounded with a short wire. I experimented with various grounds available in the living room, and eventually found that grounding to the outlet screw produced the lowest hum level.

At this point, I could not hear any hum on a good portable radio (GE Superadio III), but I confirmed that I could hear a small amount of residual hum on a couple of ac powered radios around my home, including my new Cambridge Sound Works 735i. This radio has extended low frequency response and very good sound quality. For the rest of my tests, I used it as the reference. To put the hum level in perspective, it was well below the point that you would be able to hear it with any modulation present, but with no programming it was certainly audible.

I first disconnected the ac transformer and substituted various dc supplies that I had on hand. A few produced noticeably more hum than the supplied adapter, and I confirmed that with these, I could also hear hum on the portable radio. The reason turned out to be that the SSTran requires at least 20Vdc at its input to remain in regulation (if you want to know why, just ask). However, even the dc supplies that didn’t introduce additional hum still resulted in the same level of residual hum as the supplied ac transformer.

Next, I brought home my 20+ pound lab dc supply from work. This supply produces dc power that is arguably as clean as a battery (maybe cleaner). I set it for 20V and hooked it up to the SSTran and guess what? The residual hum level was no different! While I was at it, I measured the current drain. In the absence of modulation, my unit draws 50.7 mA. So if you are thinking of running yours on batteries, plan accordingly.

At this point, I am of the opinion that there is no intrinsic source of hum within the SSTran due to the power supply. If you disagree, that is fine, but this is my conclusion. HOWEVER, please read on for some additional very important qualifications that may help you if you are having hum problems.

There are three chokes inside the SSTran with associated jumpers. Jumper S1 shorts L1, which is in series with the audio input ground. I cannot condone operating without this jumper installed. I have never yet found a situation in which you can obtain hum-free audio if you leave it open. The reason is that with L1 in the circuit, the ground return for your audio input will be primarily through the power ground, which is bound to be dirty. You are better off to operate with S1 installed and leave S2 and S3 open. The manual says that you can jumper S2 and S3 to take advantage of the ground path provided through the ac adapter and your house wiring. Now this is probably OK for casual operation, but it is really not a very good RF ground, and it is definitely going to superimpose RF on your house wiring. When this RF enters the power supply of an ac line powered radio, some of it is going to be rectified in that radio’s power supply, which will almost certainly add hum. The reason is that unlike the diodes in your SSTran, the diodes in the radio are probably not bypassed. This is the problem that carrier current stations have to fight all the time. So if you can avoid it, you should try to isolate the transmitter ground as much as possible from the ac line and ground it to a good earth ground if one is available. If not, I think it is better to ground it to the ac outlet screw, which should be attached to the conduit, than to drive the hot and neutral lines in common mode fashion by grounding to the ac transformer.

Upon further testing, I have found that a significant amount of RF energy still gets injected into the ac line through your audio equipment when you have the audio input cable connected and have jumper S1 in place (as previously recommended). If this is causing a problem, I would suggest trying a common mode choke made by winding the audio input cable around a #77 ferrite toroid for as many turns as possible (you will probably want at least a 1.4″ core). This will attenuate the RF without affecting the audio.

Next, I am going to further discuss the issue of hum that is present when the Gain control is turned up. The SSTran has a relatively high impedance input circuit that makes it vulnerable to hum pickup, particularly since it is in a plastic case. If you are driving the transmitter from a low impedance source like a computer sound card or an Ipod, you could benefit by lowering the input impedance. This can be done by bridging the audio inputs with a 1K resistor. This should not drop the audio level significantly if you have a low-Z source. If it does, you can experiment with a higher value (e.g. 5 or 10K). You should notice the difference right away if you turn up the gain with the inputs disconnected– the hum should be a lot lower. If you do have a high impedance source, you will benefit by keeping the level coming into the SSTran as high as possible, and controlling it with the input gain control. In other words, to minimize hum pickup, you want to run the input gain as low as possible, consistent with obtaining the proper modulation level.

Finally, as has been discussed elsewhere, if you are feeding audio to the transmitter over a long cable, as for example when it is mounted outside, you should get an audio transformer and drive the transmitter with a balanced audio source. Mount the transformer at the transmitter input. This will pretty much totally eliminate any hum pickup on the cable.

In the future, I will devote some time to explaining more about what I have learned about setting the Gain, Compression, and “Modulation” controls. My format is classical, and I don’t like a lot of compression, but a moderate amount can be helpful on AM. However, what particularly bothered me was the low level “noise gate”. I have found that if you drive the transmitter input hard enough, you can turn down the compression control all the way and pretty much eliminate all of the noise gating. You can still engage the “Modulation” (limiter) circuit and benefit from some top end limiting, which is helpful.

There are a lot of useful calculators online these days. Here is one that I liked from Compliance Engineering: http://www.compeng.com.au/emc_conversion_tables.asp

There are a lot of useful calculators online these days. Here is one that I liked from Compliance Engineering: http://www.compeng.com.au/emc_conversion_tables.asp

This calculator will allow you to enter measured field strength and determine EIRP, or the other way around. For example, data from the test report for the new ChezRadio transmitter shows a field strength of 75 dBuV/m at 10 M. You can enter this value in the calculator to determine the EIRP in mW. Try it– You might be surprised! One caveat: this was apparently a near-field measurement, so accuracy might not be very good. I am a little surprised they didn’t perform this test outdoors, on an Open Air Test Site (OATS). Of course, you do have to be concerned about the ambient RF environment in that case. From this calculation, the predicted field strength at 1 km would be 40 dB lower, or about 56 uV/m. Not earth-shattering, but certainly audible on a good car radio.

Here’s another page with lots more interesting calculators: http://www.rfcafe.com/references/calculator_links.htm

There are basically two issues with the Ramsey AM25C. The first is that the included Wal-Wart ™ power supply is unregulated and delivers WAY too much voltage to the unit. This is because it is designed to supply 15 Vdc at 1 Amp, per the nameplate label (on my unit, anyway). I have read criticisms of Ramsey elsewhere that the hum problem many users encounter is caused by a “cheap, underpowered supply”, when actually the opposite is true!

The result is that the power supply delivers too much voltage to the transmitter, causing several bad things to happen. First, the on-board modulator transistor runs extremely hot because the unit draws excessive current when operated at this abnormally high input voltage. As a matter of fact, I think that if you ran it continuously it would eventually burn the PCB (I admit I did not try this). Second, the RF power input is WAY too high. Besides being out of compliance with Part 15, this screws up the modulation. Third, since the power supply is unregulated, it introduces hum on the carrier. The amplitude of the hum is exacerbated by the high current drain.

So the first thing you need to do is power the transmitter from a filtered, regulated 12V power supply. This will completely eliminate the hum, stop the regulator from overheating, and drop the power input into the proper range for 100 mW operation. I really think that for this transmitter, a regulated supply is an absolute necessity. Otherwise, you will not be able to get rid of the hum completely. I noticed that on my unit, the power input was still a little higher than 100 mW, and Q5 was overheating (actually it was too hot to touch). Therefore I needed to reduce the RF drive level slightly. I padded down the drive into Q5 by putting a 1000 pF capacitor across R16. You will probably need to determine the exact value experimentally, because it will depend on your power supply voltage and also to some extent on your operating frequency.

Once you make these changes, you should be able to achieve reasonably clean modulation at the proper 100 mW power input level, with no hum and no overheating (of anything on the board). This design does not support positive modulation greater than 100%, but you can still have a very “loud” and “punchy” sound if you use a good audio processing chain with it.

One thing I do like about the Ramsey is that the RF output is well filtered and that it is designed for a nominal 50 Ohm load. So, you can easily match this to a base loaded shortened monopole antenna– no tricks required. This should work fine with the “Antenna Guy” coil antenna. You can also easily measure the RF output power.

I’d like to call your attention to a minor indiscretion on the part of Ramsey. In the manual, they seem to confuse input power with output power. Note that FCC rules impose a limit of 100 mW input power to the final amplifier– not 100 mW output power.

While driving home from work, I happened to be monitoring 1670, listening to a weak signal from WTDY (Madison, WI). Suddenly I picked up a “talking house” transmission. It was from a realtor (Baird and Warner) advertising a house for sale along my route. He went through the house description, and at the end, he put in a plug for the “talking house” and invited people to contact him about selling their home “the talking house way”.

That’s fairly unremarkable, but something else I noticed might be of interest here… the “woop woop woop” effect! Now, what is that? Well, as the talking house signal got weaker, it was beating with the background signal from WTDY. As the message played, it would cause this low frequency “woop woop” sound, with about a 1 second periodicity. I noticed that the “wooping” effect was strongest between words, and it would eventually die out at the end of sentences or during the quiet period before the message repeated.

Now, what exactly is this? Here’s my theory. I think there are two things going on. The power supply apparently is not well enough regulated, and the PLL is slightly unstable. On modulation peaks, the extra power pulls the supply voltage, affecting the carrier frequency. The PLL corrects it, but it overshoots, causing the “woop” effect. When there’s no modulation, the thing stabilizes. This could probably be corrected fairly easily by adding some additional regulation.

So OK, all you talking house transmitter owners, have you noticed this? I was curious about these units, but it sounds like the audio quality might not be as good as the SSTran or the Rangemaster. At least not without some work!