"I did not realize ground conditions had such an affect."
Oh yeah, AM in the mid and low bands operates on ground wave during daylight, skywave at night ... but at this weak power level, our Part 15 signals never get there. Means we get interference caused by skywave from just about everything at night.
(snip)
"My antenna hangs from 2nd floor window on the back of my vinyl sided house."
Are you using the exact same antenna wire which came with the unit? Length is critical. Also, having it right up against a wall means any wiring in that wall might cause poor SWR, phase cancellation or doubling, RFI, etc. I'd think about getting it either above the wall or at least 2' away from it.
Also be sure your house wiring is grounded properly.
Ken N. - Yes I have no delusion about my wire antenna hanging from the 2nd floor. It is crude but affective. At least both the window frame and siding are vinyl. What it does give me is full coverage of the house for my tube AM radios and no HUM! That was a big issue, getting no hum operations. I have AC/DC powered smoke alarms, fluorescent lights and so on.
As far as wire length, I am going to say it's the stock 9' 10" length, but I went to some TEFL insulated stranded silver plated aircraft wire. The insulation color light as is my house siding, thus not as tacky. The antenna is tuned by the TH automatic internal tuning coils. I have full intent to make an outdoor antenna, either a mid loaded monopole or helical continuous wound, just for fun.
However I don't like the idea of mounting the TH in a box on the lawn with the antenna. May be I can get a deal on a Talking House ATU (Aux Transmitting Unit), aka, base loaded antenna. At least I can feed the TH ATU with coaxial and keep the TH unit indoors, legally. What is the rule on that? Seems that TH got an exception that does not apply the hobbyist folks. I'm not worried.
Lee - Thanks for the input. Yes they sell them new for $99, but their modified units are $500! I don't think they improve the PLL. I think shielding and may be better power supply might help? Before Radio Systems bought them out, the original Talking House folks sold these things for big bucks.
Adding a padding capacitor across the crystals leads really doesnt change the crystal frequency, but rather changes the circuit impedance for the reference oscillator and its crystal. To actually "bend" the frequency of a crystal, you add the trimmer capacitor in series with the crystal, or if the crystal circuit is of the series circuit type, it will have both leads isolated from ground, and you place the pad capacitor from one lead over to ground. However this is a very tricky trick because in series resonant crystal circuits, there is very little room for bending the frequency, and you end up killing the resonance and the oscillator simply stops oscillating.
The most likely cause here to all this "weeble warble" is the use of cheap crystals that are not exactly precise in their cut. They could have a variance away from the actual frequency just enough to work, but not be precisely on the frequency...but just enough to let the unit pass quality control at the time of manufacture. These cheap crystals will be ok when new..but as they age or are bumped around in shipping or handling along the way or during use..they can be bumped even further off the intended cut and cause problems like this.
Given the units were sold new at such a low price, though I am not well versed in the price range of TH transmitters because I never bought one..I thought they were (no disrespect to their owners), cheap and as such would be made cheap. However their low cost means the use of very low end and very wide tolerance components throughout the design.
As others stated, identical units have presented different power strengths and so forth, different power supply issues etc. These things are not intended for the use that members here are using them for. They were meant to be selling a house with a simple voice talking to potential real estate buyers driving by.
But these sound like good units to experiment with and potentially improve upon such as the PLL weeble warble.
I dont have a schematic or actual unit to see what type of components they actually have, but my first guess is that they are using standard ceramic capacitors in and around that PLL circuit, which in PLL design is "BAD MOJO". Silver Mica or tantalum capacitors are the preferred choice in PLL circuits as well as high end crystals. If I were to design such a unit, I would use a high frequency crystal say 4Mhz, then divide that down through an N-divider chip to have frequency selectivity. Then run that final frequency through a balanced modulator chip, then to a buffer amp and then final. Such a design is found in the famous Panaxis AM-100 transmitter, which I have one proudly sitting on my collection shelf along with its companion AM-5000 amplifier.
Its amazing some of these TH units sold for 99 bucks new...as the cost of precise high end, tight tolerance crystals usually go for around 60 to 80 dollars a piece and even higher for wide temperature range types.
Be careful about placing a ground shield around the PLL circuitry because obviously if you put your hand close to it and it begins to shift or eliminate the weeble warble, that means there is a huge amount of stray capacitance around the circuit..and placing a ground shield in there may actually throw the whole thing out of resonance and not work at all. Experiment in this area..however I think that there is not really much that can be done to improve it except replace any ceramic capacitors in the PLL circuit with the silver mica types or tantalum types as well as replace the crystal with a more precise cut and better tolerance crystal.
RFB
Well,RFB,the crystal stability is fine as I monitored it with my service monitor,rock solid but 580 hz off frequency until I added the 25pf varicap. This is done in most all crystal oscillators I work with in ham and commercial radio. These little reference crystals are of excellent quality asnd not expensive when you make them all on the same frequency,about $1 each for a 5ppm tolerance.It is when you special order a single crystal it becomes expensive.
Yes,there are a boatload of ceramic caps in the pll circuit,like you say-BAD MOJO. I think some of the output rf from the tuner is also bleeding back into that circuit. The proximity effect is so severe,you can change the warble by simply placing your hand anywhere on or near the unit or the antenna. I wish someone would come up with a fix,otherwise a pretty nice unit. The auto tuner is awesome!
Yes you can say it was good enough for a drive past a house for sale with a voice loop. These were first sold for big bux,$399.95 I think. For that price,it should have been more stable for sure. When i can buy a dual band commercial/ham radio handheld with all the best wanted features,a 1700mah lith/ion battery pack and drop in charger that is type approved for part 90 commercial and part 97 amateur service that works as well as and better than most major brands for less than half the price,I think the TH should have been better quality. China,are you listening????
I also noted any movement of the antenna,say like in a windstorm,also makes the PLL move around.
WEEBLE WARBLE WOAH!!! No kidding? Well that does indicate very bad coupling and isolation design if moving the antenna around by wind causes the thing to go warbling all over the place!
Your right..its no wonder they gave them away for 99 bucks...perhaps it should have been .99 cents! 😀
Another good example of poor design is found in the LPB unit's exciters. These things had bulky stand up trimmer pots, and low and behold they put one for the drive control! You could put your hand 4 inches away from the thing and watch the power level shoot through the roof, resulting in blowing out the finals or if your lucky, just the fuse. I solved this problem by replacing that bulky stand up pot with a surface mount trimmer pot and totally eliminated that issue, then I replaced the entire exciters with Chris Cuff's C-QUAM boards.
RFB
"except replace any ceramic capacitors in the PLL circuit with the silver mica types or tantalum types as well as replace the crystal with a more precise cut and better tolerance crystal." RFBurns
I replaced some caps in the audio section with higher quality polypropylene (and increased/decreased some values). The sound improved times 10. It went from barely enjoyable (musically) to very reasonable and almost on par with the best oldies (1970's or older) in the area. Lucky to have one oldies station, but I like more modern music.
Crystals are not that expensive. I had an old piece of radio test gear (Sencore SG165) with a 10.7 Mhz crystal that drifted. I could not find a used 10.7 Mhz in the right package to save my life. Also you have to get the right cut, capacitance, parallel or series use. There are manufactures that will make you a new one, to your spec for about $24. I am assuming the PLL crystal is a standard Freq, so it might be an off the shelf item and cheaper.
This is true..I should have been more (excuse the pun) precise. Maybe my brain PLL crystal is a few PPM off and I need a new one! 😀
Yes custom crystals are more expensive versus the bulk variety. Bulks can be good in tolerance as well, I just have had a wide range of varying experience using bulk manufactured crystals and is why I said that. I always custom order my crystals even though I can jolt on down to the local electronics supply shop and pick up a bulk manufactured crystal for 30 or less. However I have always ran into issues with those so I custom order mine and although far more expensive, it is well worth the cost as every single crystal I have had made, even ones I have that are 15 years old, are still within their 10PPM spec. I always specify 10PPM specification no matter what. It just saves me the headache later on. (corrected typo)
Well good luck with those TH units. They sound like a lot of fun to experiment with as well as that auto antenna tuner unit. I have read about those and they sound pretty neat to work with.
RFB
Patent # 6973294 shows the complete schematics for the TH. The quality of the scan is marginal, but it gives a lot of hints without giving away exact component values.
Figure 4A tells the PLL story in sufficient detail to explain the PLL warble problem. The patent text refers to transistor 80 and transformer 82 as being part of the PLL circuit. Looking at figure 4A, the PLL circuit is powered by the same source (+12V?) as the final RF output stage. The PLL circuit is a discrete implementation, using a varactor diode as the frequency determining component in the voltage controlled oscillator. Such a discrete implementation may or may not be good, depending on design information that isn’t visible in the patent, but powering the PLL from the same source as the RF output amplifier is definitely bad form.
Every time I go to this patent to look at various details, I marvel at how over designed it is. The automatic tuning feature consumes a huge portion of the design. That would be expected, but then when you look at the complexity of the basic Modulator/RF portion, there are a huge number of components unnecessarily involved. Some how, it all doesn’t add up. Reported performance is not at all commensurate with the complexity.
Dumping old inventory at $99 makes sense. That seems to be what they are worth. Charging $500 or more for new units doesn’t make business sense. In even the best case, the sales volume will only be a tiny fraction of the volume required to make such a product profitable. I’m talking hundreds of thousands of units minimum. The perpetuating $99 sale price is an indicator that the current seller inherited a very large inventory of unsold units. There’s no way he’s going to make any profit unless he got them for free (a possibility), and as long as the fire sale is ongoing, he is unlikely to sell many of his higher price units. I predict this will all end when inventory of the old $99 transmitters is depleted.
By the way, the FCC has limitations on the modifications a manufacturer can make and retain certification, so don’t expect any great stuff from any new TH units.
PhilB
Great digging on the patent. I have heard that supplying power to the RF and PLL from separate supplies could be a good mod.
As far as Talking House, I've called them a few times to chat about different things. From my memory this is what I recall. The idea of using a low powered radio to sell homes was the brainchild of some estate folks in the Chicago area. The design was farmed out to a British company and manufacture in China. Skip ahead, bankruptcy and Radio Systems Inc of New Jersey bought out the existing supply, about about 5000 units are available, pennies on the dollar. They say they sell quite a few.
There are many on the market and people don't have the patients to mess with getting them working. You see them on iPay all the time. They usually go for about $50 on average. I bought four for about $120 to the door. I ended up selling three of them for what I had into the four, so I got one for free. They were still in the box with the plastic cover on the display.
Radio Systems has been around for over 20 yrs and has a line of radio studio equipment, cable and connectors, microwave studio links and a 10 watt LPAM transmitter to name a few things. They were started to support college and community radio. They are part of a bigger company. They felt this low powered part 15 radio would supplement their product line. They changed the name for i.A.M. radio and made some modifications to the sound and target them for schools, church, business and hobby as well real estate. I got the signs and how to market and sell your house DVD, books and brochures with my TH units.
The ones Radio System modifies and sells go for $500. I don't know how much better they are. I inquired about what they do but they were mum, but changing a few caps in the audio signal path does wonders. It is not a bad system. However if I am going to mount it outside with an antenna, the talking house is not the way to go. SStran or Ramsey may be. Are there any other good AM radio kits? Rangemaster and ChezRadio is too much money.
The infamous C-Cuff C-QUAM exciter/transmitter. For the price, you cant find a better deal.
RFB
powering the PLL from the same source as the RF output amplifier is definitely bad form.
Well the unit is also turning and burning a standard DC motor as well which obviously is going to sag that supply rail somewhat. Then there is the way the thing is sensing the signal level for the auto-tuner and the large amount of stray inductance in the entire circuit. I would say that the auto-tuner circuit itself is contributing to the warble especially when you have folks saying the wind moving the antenna causes warble.....naturally since the thing is improperly coupled through the final stages and sensing circuitry.
So basically this thing is not only fighting to maintain a maximum signal on the wire due to varying inductance from motion on that wire from wind, but also fighting itself at the same time because of the stray inductance inside the thing as well as the sagging of the power rail to drive that tuning motor.
My suggestion would be to separate the supply rails and improve the coupling of the sensing circuits for better isolation. Get rid of the DC mag motor and replace with a servo and driver circuit on its own supply rail.
I would bet if you take an RF probe you could find RF bleeding all over the place in that thing. Because I do not see adequate RF bypassing at all in there. Heck adding some effective bypassing may cure a lot of the problems, but obviously the design could stand to be revised heavily.
Would make a great weekend "Talking Home Improvement" project for sure!
RFB
Well, I'm scratchin' my head.
The mod I did to NET my TH output frequency works great. I check the freq periodically, maybe once a week. I've only had to tweak it once. It's a little touchy to ZERO it in so I'm happy being within 2 Hz of center.
As for all this banter about instability, I've not noticed any problems. The warble is gone and isn't affected by windage.
As for the audio bandpass, I boost the low end drop the mids a little and boost the high end just a smidge at the end of the audio chain to compensate for the transmitter audio response. This flattens out the transmitter audio response and then the settings on the console look normal.
For $9.99 from Ebay (plus shipping), I'd say I'm doin' OK. I use the TH ATU so the transmitter can LEGALLY be in the house as the certification includes using the remote antenna. The INFO SPOT system sold by ISS is based on the TH with ATU and they advertise using up to 300 feet of coax.
Well, I'm a lucky duck as my TH is doing a great job for me.
I think I am detecting two different interpretations of the word “warble” in this thread.
The warble described in most posts is caused by a shift in carrier frequency (FM) coinciding with the carrier amplitude modulation. This has been correctly identified as a PLL instability problem due to poor regulation of the PLL power supply or RF coupling from the TX output stage back into the PLL circuit.
However, some posts seem to using the word “warble” to describe the heterodyning of the transmitter carrier with another station on the same frequency due to the carrier frequencies being somewhat different. This will happen even with perfectly frequency-stable transmitters. If the frequencies differ by say 5 Hertz, you won’t hear the sub-audible heterodyne as a tone, but you will hear it as a flutter with a frequency of 5 times a second. A trimmer cap added to the crystal oscillator will allow you to bend the frequency difference to near zero. I have done this to the point where the difference is less than 1 Hertz, but it is really difficult to eliminate the effect entirely. Even a difference less than 1 Hertz will be heard as a slow change in the received signal, rising and falling with a period of several seconds.
Using an SSB receiver is the only practical way for the average hobbyist to diagnose PLL instability. Start with an unmodulated carrier. Tune the SSB receiver so the carrier produces a beat tone of about 1 or 2 kHz. With no modulation, the tone should sound fairly stable and pure. Then, turn up the modulation. If the carrier tone goes erratic and changes pitch in synchronism with the modulation, then you have PLL instability. If the beat tone doesn’t sound stable with no modulation, then you have a crappy PLL or a really crappy SSB receiver! Some minor carrier tone changes with modulation are likely in any relatively inexpensive transmitter, but when the FM level is very small, they won’t degrade the signal quality noticeably.
Bending the crystal frequency with a trimmer cap can zero out heterodyne effects with another station, but it will NOT improve the stability of the crystal oscillator or the PLL, and will not improve distortion caused by severe frequency instability.
Phil
I think that for the price of those TH units and how the circuits are laid out on the PC board and component types used, as well as the intended purpose for them, a little weeble warble is not surprising.
It makes for a wonderful platform to experiment with and find improvements for these transmitters..the very core of what we are all doing...and that is our hobby in low power radio.
As the saying goes..."Happy Tinkering"! 🙂
RFB
"At least I can feed the TH ATU with coaxial and keep the TH unit indoors, legally. What is the rule on that? Seems that TH got an exception that does not apply the hobbyist folks. I'm not worried."
Not really. I suppose it could be argued, but perhaps the ATU is considered a tuner, just as if it were in the cabinet, so the cable isn't a transmission line, it's just a long shielded wire (standard 75Ω TV coaxial cable) to the tuner. The ground wire is the shield on the cable, so in theory, while it does ground, it creates a canceling effect to the unmatched signal. I tested it awhile back, and the cable length didn't make any significant difference between 6 ft. and 25 ft.,i.e., the signal strength was about the same. In fact, IIRC, it was a hair less at the longer length.
Now I'm wondering if maybe it's less than 100mw due to cable loss.
Test question: In this case, which is considered the final stage?