Unwanted Tone - DC Power Woes

[MicMaven]

My friend is a blind ham radio operator. He wants me to construct a DC voltage buck converter or other device that will power an old Radio Shack Model 32-2040 amplified speaker from his ham station power supply. The power supply is an Astron 35M linear type which produces 13.8 v (nominal 12v.) The amplified speaker requires 4 C-cell batteries (6v DC) or 9 v DC from a wall block supply, with its coaxial connector center pin Negative, and ring Positive. He mixes audio from his transceiver with audio from his computer before sending the mix to the amplified speaker (or alternatively, headphones.) The whole point is to have his computer use text-to-speech technology to produce audible confirmation of what he types in his computer log, whilst he listens to stations on the radio. I devised this simple remedy, and he says it works very well, and other blind operators have modeled their stations similarly.

To further his desire to eliminate the 9v wall block supply, I assembled a simple power cable with a DC buck converter purchased on eBay.com that reduces 13.8v to 9v, and it works as expected ... EXCEPT ... it appears to introduce a constant ~800 Hz tone in the speaker. This does not occur if he powers the speaker on batteries or with its original 9v wall block power supply. Nor does this occur if he uses the buck converter cable with a separate power 12v supply. The tone only occurs if he uses the buck converter cable with his main station power supply. Moreover, the tone does not occur if he powers the amplified speaker directly connected to the main station supply with a different cable (i.e., bypassing the buck converter.) This last connection is, however, contraindicated as it stresses and causes distortion in the speaker. Nevertheless, I mention it as I believe it is a valuable clue.

Consequently, the deleterious tone occurs only if the buck converter cable is used between the 13.8v station power supply and the amplified speaker. I believe the buck converter is causing the 800 Hz tone. I suspect various odd theories, such as a ground loop or feedback loop, of sorts, or some problem with the polarity of the different DC audio devices in the chain. The buck converter has its own voltmeter which might, could, maybe emit EMI which interferes with the amplified speaker. Or, it is something entirely beyond my grasp at this point!

I am tasked with finding a way to power his amplified speaker from his station 13.8 power supply which eliminates the unwanted ~800 Hz tone.

Possibilities?

I suppose I could try a different buck converter, but I fear it, too, will prove problematic (assuming the neon meter is not the cause.)

Perhaps an alternative is a simple L-pad or T-pad voltage divider, although lack sufficient expertise to select appropriate component values therefor.

Perhaps there is an even better option???


The operator is a good friend, and he serves the community by organizing and providing emergency communication services in the community, so I do not want to let him down. Besides, it is more fun to solve the problem than give up and use the dumb old wall wart supply!

Can anyone suggest a simple, low cost solution I could build to save the day? Please let me if I left any necessary information out, and I will endeavor to supply the same.

Thanks loads in advance.

James

Note: I attach a photo of the DC buck converter and a crude system chart indicating how it sits in the mix.

 

[Khron]

Is there something else in the system working in the ~150kHz area, that's about 800Hz off and resulting in that as a beat-frequency?

 

[Brian Roth]

If the speaker's current draw is less than an ampere, a 7809 linear regulator is a possible alternative. It will require a heat sink and a couple of bypass caps.

Many examples are available. Here is one:

https://www.mouser.com/ProductDetail/onsemi/MC7809CTG?qs=%2B9/cbd0IE0RZOsET7SgWUA==

Bri

 

[MicMaven]

Is there something else in the system working in the ~150kHz area, that's about 800Hz off and resulting in that as a beat-frequency?

I do not think so.

It starts and stops immediately upon connecting and disconnecting the buck converter. It is curious it is a single frequency, sine wave type tone - slightly higher than the 700 Hz tone most operators use to monitor continuous wave (CW) Morse Code transmissions. In contrast, most EMI/RFI in the shack manifests a more random frequency static-type noise.

What has me treed is, why it induces a tone when it is connected to the station power supply, but not when it is connected to a separate power supply. This kinda, sorta, almost suggests a ground loop or phase reversal sort of issue.

Most station equipment has center pin Positive and ring Negative, but the speaker thing has it reversed. And yet, I am at a loss to explain how or why that should matter. But then, I am clutching at straws at this point.

I ordered a different converter without an on-board voltage meter to determine whether the voltage display is the problem.

Thanks for playing along. / James

 

[MicMaven]

If the speaker's current draw is less than an ampere, a 7809 linear regulator is a possible alternative. It will require a heat sink and a couple of bypass caps.

I believe it draws substantially less than 1A.

I will search for an appropriate schematic as a model and confer with my local electronics Guru who formerly designed and built radio equipment for NASA, the Air Force, Lear Jets, and etc., after he returns from hunting deer this week.

Can you suggest a string of words to facilitate my search? I propose "DC" "buck" "converter" "voltage regulator" "circuit" "schematic" -- any thoughts on the matter?

Thanks for the suggestion. / James

 

[ruffrecords]

It is possible the 13.8V supply does not like the heavy current pulses the SMPS likes to draw. A local energy store might help. Try adding 1000uF across the input to the SMPS.

Cheers

Ian

 

[Brian Roth]

I believe it draws substantially less than 1A.

I will search for an appropriate schematic as a model and confer with my local electronics Guru who formerly designed and built radio equipment for NASA, the Air Force, Lear Jets, and etc., after he returns from hunting deer this week.

Can you suggest a string of words to facilitate my search? I propose "DC" "buck" "converter" "voltage regulator" "circuit" "schematic" -- any thoughts on the matter?

Thanks for the suggestion. / James

The 78xx series of three terminal regulators has been around for eons and is widely used even until this day. Here is a data sheet link to one manufacturer's offerings (it covers the variants, with the last two digits in the part number being required voltage....ie, 7809 for your application).

https://www.mouser.com/datasheet/2/308/1/MC7800_D-2315963.pdf

On the first page you see the schematic. A 7809 and two capacitors. You want the TO-220 package and heat sinks for that are very common.

If you want more info, you can search for something like "3 terminal linear voltage regulator".

Bri

 

[MicMaven]

Is there something else in the system working in the ~150kHz area,

Mr. @Khron :

Um ... er ... after further review, I suspect you are on to something here!

The buck converter employs a LM2596s power converting IC that operates at a switching frequency of 150 kHz.​

Coincidence? I think not!

Perhaps, this suggests a replacement board without voltage meter shall prove equally problematic because it also employs a LM2590s converter chip. Perhaps I can suppress the tone with one or more EMI/RFI squelching techniques typically used in the shack. Considering this is DC-DC conversion, I doubt an isolation transformer will work. And, because it seems to enter the speaker circuit through its power cable, or it otherwise affects the amp circuit, it is part and parcel of the final audio output reaching the speaker element. Um ... er ... or is it ? (scratching head whilst typing . . .)

I ALSO wonder whether OTHER, newer and more efficient converter ICs are cleaner, or do they simply impose a different frequency tone because they use a different switching frequency?

Of course, it all depends on whether the 150 Hz switching thing is causing the audible tone. But i suspect you are on to something here.

Shoot! - Perhaps the whole project is doomed!

James

 

[MicMaven]

The 78xx series of three terminal regulators
Here is a data sheet link to one manufacturer's offerings

On the first page you see the schematic. A 7809 and two capacitors. You want the TO-220 package and heat sinks for that are very common.

THANK YOU for the Data Sheet and references. A cursory, incomplete, reading of the Data Sheet suggests this is not a switching type converter, so it may not produce the sort of noise the LM2590s appears to produce. The simple schema on P.1. seems easy enough to implement.

I will continue to research in this direction. / James

 

[Brian Roth]

THANK YOU for the Data Sheet and references. A cursory, incomplete, reading of the Data Sheet suggests this is not a switching type converter, so it may not produce the sort of noise the LM2590s appears to produce. The simple schema on P.1. seems easy enough to implement.

I will continue to research in this direction. / James

There is no switching....the linear regulators are purely "analog". Downside (seems like not a big deal in this application) is the regulators produce some waste heat requiring a heat sink. Millions (billions?) of these regulators have been used in commercial and DIY gear for decades.

Bri

 

[Khron]

Um ... er ... after further review, I suspect you are on to something here!

The buck converter employs a LM2596s power converting IC that operates at a switching frequency of 150 kHz.
Coincidence? I think not!

That's why i put that particular number out there, i was wondering about anything else there switching around 150kHz, BECAUSE i looked through the LM2596 datasheet and saw its nominal switching frequency is that.

 

[MicMaven]

It is possible the 13.8V supply does not like the heavy current pulses the SMPS likes to draw. A local energy store might help. Try adding 1000uF across the input to the SMPS.

Golly, I apologize. I do not understand your suggestion. The station supply is a heavy duty 35 Amp linear (transformer) type, I believe it can take a lot of abuse in stride. The only switching-mode power supply is the original 9v wall block supply, which bypasses the linear supply. I do not believe the buck converter is, itself a power supply, although it appears to have a 150-kHz fixed-frequency internal oscillator on board. I think.

What would the 1000 uF capacitor do? (Revealing my ignorance! - Although, every day is another opportunity to learn something new. In contrast my wife already knows EVERYTHING - just ask her! . . .)

Happy trails. Thanks for pitching in !! James
_________________

(PS, I am only kidding, as my wife is truly a paradigm of humility and patience.) / JR

 

[ruffrecords]

The buck converter is an SMPS - it is a dc to dc SMPS. It works by charging up an inductor in part of the 150KHz clock cycle and dumping it via a diode to its output capacitor in the remainder of the cycle. So the load it presents to the 13.8V power supply is a chopped 150KHz square wave mitigated by whatever capacitor the buck converter has at its input. The 13.8V supply may be able to supply lots of current but there will be inductance in the leads from it to the buck converter which could be significant at 150KHz. Adding large local charge storage device at the buck converter input (a 1000uF capacitor) will significantly reduce the 150Khz component.

Cheers

Ian

 

[MicMaven]

The buck converter is an SMPS - it is a dc to dc SMPS. It works by charging up an inductor in part of the 150KHz clock cycle and dumping it via a diode to its output capacitor in the remainder of the cycle.

OK - Now I get it! I understand the charging cycle bit after reading about buck converters this afternoon, but none of the literature described the converter chip as an SMPS, itself, so that aspect escaped my grasp.

(Imagine me slapping my forehead and muttering, "Gee, I could have had a V-8!"

(a 1000uF capacitor) will significantly reduce the 150Khz component.

OK, now I get it. And I will try it, even if I build something like Mr. @Brian Roth suggests, and see which works best. Parenthetically, I originally acquired the buck converter for my own desktop power supply project, and pressed it into service when my friend, Tom, asked how he could eliminate the 9v wall wart supply. Great - this gives me multiple options to try at, apparently, little to no expense. How cool is that ? ! ? !

THANK YOU for the clear and direct uptake. I appreciate how much information is shared by so many in this forum, and how so many egos have been checked at the door, so one is free to learn without stigma or fear of being told he is as dumb as he sometimes feels!

Happy trails. James

 

[MicMaven]

... why i put that particular number out there ... BECAUSE I looked through the LM2596 datasheet and saw its nominal switching frequency is that.

Oh ... I seeeeee. Ahem ... And here I thought you guys are all clairvoyant with super-genius intuition and just know these things!

But, seriously folks, I think I am beginning to understand the problem and possible solutions. It seemed so easy when I took the job to just stick the buck converter between the supply and the speaker and take a bow. It is frustrating when the cure turns out to be worse than the disease. The good side is, my friend is extremely patient and understands we are both just slogging our way through the muck of ignorance, so there is no pressure beyond my pride at stake.

Thanks, again, for pitching in and playing along. James

 

[ccaudle]

It is curious it is a single frequency, sine wave type tone -

No o'scope available to just check the power in and out of the switching regulator to verify the noise frequency is on the power lead?

I believe it draws substantially less than 1A.

Just stick a multimeter in line with the power feed and measure the current when it is playing something louder than you would ever intend to use.
There should also be a label on the back of the indicating max power consumption. You can also estimate if you know how long it can run on batteries. A quick online check indicated that alkaline C-cell batteries are about 8Ah, but there is probably some fine print around the voltage level they consider the cell to be depleted. So it would be a rough estimate, but divide 8 by the number of hours it can run on alkaline batteries and will give a ballpark current consumption.

 

[MicMaven]

No o'scope available ... ?

Nope.

Just stick a multimeter in line with the power feed and measure the current when it is playing something louder than you would ever intend to use.

I will try that when I can get over to his house. In the mean time, can I test it with another load I may have available at my house?

There should also be a label on the back of the indicating max power consumption.

Nope. Darn it.

You can also estimate if you know how long it can run on batteries.

I will try to test that . The procedure described is straightforward - thanks for the suggestion.

Bottom line - at this point I can only guess it consumes less than 1 A. My research suggests it has a TCA830SM / ABB43 amplifier chip ... so I am researching that. As far as I can tell, a related chip specs out 1.5A "peak output power." But, shoot, I don't know!

Observation;
What has me treed is, it works fine business on a separate power supply - either the original 9v wall block supply or a separate regulated 12v supply - but NOT on the larger Astron RS-35M power supply we want to use. And I still harbor an unfounded, ignorant suspicion (i.e., WAG) it may have something to do with the fact the Astron station supply puts Positive on the center pin but the 9v wall block puts Positive on the ring of the speaker's coaxial DC connector. But I lack sufficient training and experience to understand how that works.

THANK YOU for weighing in on my little conundrum! James

A quick online check indicated that alkaline C-cell batteries are about 8Ah, but there is probably some fine print around the voltage level they consider the cell to be depleted. So it would be a rough estimate, but divide 8 by the number of hours it can run on alkaline batteries and will give a ballpark current consumption.

 

[ccaudle]

Is there something else in the system working in the ~150kHz area

Possibly a harmonic of the 150kHz feeding into the receiver power feed? I would expect a radio receiver to have really good power filtering, though.

I suspect various odd theories, such as a ground loop

You could partially eliminate other devices from that problem by powering the speaking from the buck converter, and checking to see if the noise is present with nothing else connected to the speaker, just the power.
Then plug headphones into the receiver and see if the noise is present on the receiver output with nothing else connected, just the buck converter powering the speaking sharing the same main power supply.

If the noise is present in the speaker with nothing else connected, then the noise is from the power supply into the small amplifier, typically you would address that with additional bypass capacitance.

If the noise is not present in the speaker, but is present in the headphone output of the receiver, then the receiver is demodulating the power supply somehow, either leakage through the power lead, or leakage into the antenna input. Antenna input doesn't seem very likely unless the receiver is tuned to a frequency which is a harmonic of, or close to a harmonic of 150kHz.

If the noise is not present in the speaker alone, nor in the headphone output of the receiver, you can start making combinations to see when the noise shows up:

1. receiver headphone output directly to speaker only
2. mixer output to speaker only
3. receiver headphone output to mixer, no connection to speaker, check with headphones at mixer
4. receiver headphone output to mixer, mixer output to speaker
5. laptop output to mixer, no connection to speaker, check with headphones at mixer
6. laptop output to mixer, mixer output to speaker
7. receiver headphone output to mixer, laptop output to mixer, no connection to speaker, check with headphones at mixer
8. back to full setup, receiver headphone output to mixer, laptop output to mixer, mixer output to speaker

Very tedious I know, but if you want to find exactly where the noise is being injected you have to be systematic and look at all the combinations.

Likely the most straightforward approach would be to just waste the power of a linear regulator with a large enough heatsink to handle dropping the 5-ish volts from the main supply down to 9V. Brute force and wasteful of power, but if the receiver is already run from a linear power supply it will probably not be the dominant waste of power in the whole setup.

 

[moamps]

Bottom line - at this point I can only guess it consumes less than 1 A. My research suggests it has a TCA830SM / ABB43 amplifier chip ... so I am researching that. As far as I can tell, a related chip specs out 1.5A "peak output power." But, shoot, I don't know!

Observation;
What has me treed is, it works fine business on a separate power supply - either the original 9v wall block supply or a separate regulated 12v supply - but NOT on the larger Astron RS-35M power supply we want to use. And I still harbor an unfounded, ignorant suspicion (i.e., WAG) it may have something to do with the fact the Astron station supply puts Positive on the center pin but the 9v wall block puts Positive on the ring of the speaker's coaxial DC connector. But I lack sufficient training and experience to understand how that works.

Take a DMM and check the continuity between the sleeve pin on the input signal jack and the + or - pin on the power supply connector from the active speaker. If the active speaker works well with an external 12VDC power supply, then you probably have a problem that with the active speaker the +pin of the power supply is connected to the common/ground of the signal. If so, then the simplest way is to insert a small audio transformer to the audio input of the active speaker to separate the common/grounds of the power supply and the active speaker. And then you can power the active speaker directly from a large power supply.

 

[MicMaven]

Take a DMM and check the continuity between the sleeve pin on the input signal jack and the + or - pin on the power supply ,.... insert a small audio transformer to the audio input of the active speaker to separate the common/grounds

Busted. You caught me!

(Now, I am embarrassed because I hoped nobody would mention doing that! I THOUGHT of doing that, but forgot to take one of my little isolation transformer projects when I visited him to test the converter cable. It may be several days days before I can return to test it.)

Duh ... Thanks for the uptake. / James