Forums :: NEW! Web Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives
Support This Site: Contributors :: Advertise


It is currently Dec Wed 08, 2021 3:31 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 24 posts ]  Go to page 1, 2  Next
Author Message
 Post subject: Choke hum on LM386 Transmitter
PostPosted: Aug Fri 13, 2021 2:51 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 5021
Location: Texas
I've been fiddling with mating my frequency synthesizer to an LM386 transmitter. I didn't have any chokes on hand so I wound enough #36 wire on a short ferrite rod to get 1.2 mH. This is the choke between the LM386 output and the transistor. And it's a source of hum.

Listening to the transmitter on a nearby radio, there's steady 60 Hz hum unless I curl my fingers around the choke. It's about 1.5 inches long and 3/8" diameter, so there's plenty of surface area to pick up stray AC fields. I expect that the problem will go away when I buy a proper choke, maybe even a toroidal, but I want to understand what's happening here.

The hum goes away only when I shield the choke with my body? Why?

I can wrap the choke in foil and ground the foil, but no joy. I'm testing the transmitter using battery power. When I ground the foil to my scope probe (scope plugged into mains), the hum is much louder than not connected to mains.

1. What is it about the human body that more effectively shields than typical shielding techniques?

2. Is there a way using typical parts to emulate the effect of the human body?


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Fri 13, 2021 3:06 am 
Member
User avatar

Joined: Jan Fri 06, 2012 8:47 pm
Posts: 7616
Macrohenry wrote:
2. Is there a way using typical parts to emulate the effect of the human body?
At Motorola in one of the circuit labs we were having a problem. An RF circuit would only behave properly when the technician touched his finger to a certain part of the circuit. There was talk of severing a finger and closing the lid on the offending circuit (finger-in-a-box). We also tried to figure out how to make the electrical equivalent of his finger. We eventually found a solution, and the technician got to take all of his body parts with him that day.

Sometimes it is a combination of body capacitance to ground, resistive losses and De-Q'ing a tuned circuit that combine to make a body part do what nothing else seems to accomplish.

It's one of the mysteries of life.

Curtis Eickerman

_________________
http://curtiseickerman.weebly.com


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Fri 13, 2021 4:23 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 5021
Location: Texas
Eickerman wrote:
Macrohenry wrote:
2. Is there a way using typical parts to emulate the effect of the human body?
At Motorola in one of the circuit labs we were having a problem. An RF circuit would only behave properly when the technician touched his finger to a certain part of the circuit. There was talk of severing a finger and closing the lid on the offending circuit (finger-in-a-box). We also tried to figure out how to make the electrical equivalent of his finger. We eventually found a solution, and the technician got to take all of his body parts with him that day.

Sometimes it is a combination of body capacitance to ground, resistive losses and De-Q'ing a tuned circuit that combine to make a body part do what nothing else seems to accomplish.

It's one of the mysteries of life.

Curtis Eickerman


Ok, thanks, that actually is helpful. One thing I didn't try was a resistor with a capacitor, both parallel and series. I'll bet my chokes arrive first.


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Fri 13, 2021 1:20 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1628
Mysteries of life eh?

Just a guess here, but most likely what is going on is that when you touch the choke, you are not providing any shielding (explaining why shielding doesn't work). Exactly the opposite is happening and you are acting as an RF signal radiator, or improving the radiation from your existing antenna.

The hum you are receiving likely represents re-radiated signal. The signal appearing at the receiver from the transmitter's antenna directly, is not as effective as the re-radiated component of it from mains wiring in your dwelling and that is where & why the hum originates. Your body is acting like an antenna and radiating a better direct signal to your receiving antenna. It suggests the orientation, efficiency,tuning to resonance of your transmitter antenna needs improving, often to solve the problem a resonant loop works well. Or a better grounding system.

The other explanation that could give the same result would be that by touching the choke, you are increasing the radiation from your antenna. The transmitter output is accelerating charges in your antenna wire to produce the EM waves and by providing an artificial ground plane (your body) the antenna's RF current is increasing, resulting in better direct reception at your receiving antenna.


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Fri 13, 2021 9:15 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 5021
Location: Texas
Thanks for the replies. Trying things further, I notice that the hum diminishes 90% when I unplug my turned-off oscilloscope, which is within 1 foot of the transmitter.

I've been using the scope to monitor the output waveforms. So as not to load the transmitter, I hook the scope to my 19" square passive loop tuned to the transmitter frequency. I notice that even with the scope turned off but still plugged in, when I disconnect the probe from the loop capacitor, the hum increases tremendously. I would also like to understand the cause of this. The loop is within 1 foot of the 6 foot transmitter antenna.


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Fri 13, 2021 10:14 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1628
Macrohenry wrote:
Thanks for the replies. Trying things further, I notice that the hum diminishes 90% when I unplug my turned-off oscilloscope, which is within 1 foot of the transmitter.

I've been using the scope to monitor the output waveforms. So as not to load the transmitter, I hook the scope to my 19" square passive loop tuned to the transmitter frequency. I notice that even with the scope turned off but still plugged in, when I disconnect the probe from the loop capacitor, the hum increases tremendously. I would also like to understand the cause of this. The loop is within 1 foot of the 6 foot transmitter antenna.


With regard to this type of problem, I mentioned on another thread once, that there is a way to demonstrate it in a controlled manner. The way I have done it is to have a transmitting loop and at the radio, a receiving loop or ferrite rod antenna. The advantage for the experiment is that with the receiver sitting somewhere with its rod on some physical axis, the transmitter loop can be rotated for a null because the transmitter loop is very directional. Then the receiver will receive more re-radiated signal from the house wiring. Also then, the receiver can be rotated a little to optimize that signal. And then the transmitter antenna re-rotated for a null. In the end an angle is found for both antennas where the hum modulation in the received signal is severe. Also with this experiment I was able to prove that there was no actual hum modulation in the transmitted carrier originating directly from the transmitting loop. These are some of the interesting effects seen in the near field of a transmitter. Further away (many wavelengths) the signal's electric and magnetic field ratio evens out and this is why you do not commonly get the effect from far off MW radio stations, but under some circumstances, it still can happen.


In essence exactly the same thing is happening with whip antennas vs the radio receiver wire antenna. So any modification of the antenna arrangement at the transmitter, its position, adding a ground plane or a conductor leading to it can rotate the electric and magnetic axis of the signal around, and result in more or less favorable signal at the radio receiver, with the potential for more or less sensitivity to re-radiated signal with hum modulation on it. So connecting your body, test instruments, test loops at the transmitter, with a whip antenna, can alter the effect. This is one reason I moved to resonant loop transmitting and loop receiving antennas for pantry transmitters as they can be rotated for max direct signal pickup and the hum issue is not usually a problem. If you are using a transmitting loop and still getting the problem, make sure it is tuned to the transmission frequency and rotated for a signal maxima at the receiver.

(If you mainly have vintage transistor radios, a transmitting loop is better anyway. Tor tube radios, all you need to do is put a ferrite rod in a plastic box with a tuning capacitor and a few turns for a coupling coil to feed into to Tube radio's antenna inputs, a twisted wire pair works fine.Then you can tune that and rotate it for a signal maxima and your hum problems will go away and you will have less noise pickup as well)


Last edited by ACORNVALVE on Aug Fri 13, 2021 10:28 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Fri 13, 2021 10:20 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 17044
Location: S. Dartmouth MA 02748-1225 USA
Coil/ferrite rod combination are an AC field pickup device. Wind another choke in pies if need be on a wood, ceramic, phenolic or air core... Good catch discovering the choke as the source of the AC... May still catch some AC field. There are ferrite cups as a possible or a toroid...

_________________
List' & I will Enchant Thine Ear


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Fri 13, 2021 10:32 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1628
Chas wrote:
Coil/ferrite rod combination are an AC field pickup device. Wind another choke in pies if need be on a wood, ceramic, phenolic or air core... Good catch discovering the choke as the source of the AC... May still catch some AC field. There are ferrite cups as a possible or a toroid...


Chokes with large numbers of turns and iron cores can act as hum pickup devices for local low frequency magnetic fields, or other signals, this is the basis of tele-coils in hearing aids as one example.

In this case though the hum was going away with no alteration of the low frequency magnetic environment the choke is in, but when the choke was touched with the finger.


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Sat 14, 2021 4:58 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 5021
Location: Texas
Looking further into my extensive junkbox, I found a couple of inductors much smaller than the one I wound. I installed a 5 mH inductor between the LM386 output and the transistor. It's about the size of a pencil eraser. A smaller 100 uH inductor goes to the antenna that's tuned to 1680 kHz.

Although wrapping my fingers around the new inductor does reduce hum a little, it's nowhere near the drama I experienced with the home wound inductor. BUT what does have the same previous effect is touching the positive terminal of the 12 volt battery. And guess what, even touching the ground quiets the hum.

I attached a clip lead to the ground of the circuit board and connected a test probe lead to that and let it trail from the workbench to the floor. That also quiets the hum, maybe 90% as much as fully touching the battery terminal. And then clipping same to the drain of the FET, again letting it trail, also quiets.

There's another effect. Unless the trailing lead is attached, with audio input levels where receiver volume is comparable to local stations, the audio is distorted, choppy, sounding overdriven. I have to turn it down some to get clean audio. But with the trailing lead attached, the audio clears up, even FM-like. I can even turn up the source louder until it over-modulates.

So I found a way to fix this. I have a 9" clip lead with one end attached to pin 1 of the LM386, where the positive terminal of the 10 uF cap goes. I run this lead in parallel with the antenna and it effectively removes the distortion. I didn't yet know if this is radiation where shielding would help or capacitance, or what. So I bypassed LM386 pin 1 to ground with the first cap I saw lying on the bench. 0.05 uF. Oh my, what a difference. The signal quieted and the volume almost doubled. My scope is off, so I check tomorrow what's actually happening.

All this fits on a 3.8" x 2.5" board, which I plan to duplicate for anyone who wants a 386 transmitter that's frequency synthesized. Here are the traces so far. It does benefit from the increased gain by using the 10 uF cap with the LM386.


Attachments:
Image4.jpg
Image4.jpg [ 553.26 KiB | Viewed 1232 times ]
Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Sat 14, 2021 10:07 am 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1628
Macrohenry wrote:

Although wrapping my fingers around the new inductor does reduce hum a little, it's nowhere near the drama I experienced with the home wound inductor. BUT what does have the same previous effect is touching the positive terminal of the 12 volt battery. And guess what, even touching the ground quiets the hum.

I attached a clip lead to the ground of the circuit board and connected a test probe lead to that and let it trail from the workbench to the floor. That also quiets the hum, maybe 90% as much as fully touching the battery terminal. And then clipping same to the drain of the FET, again letting it trail, also quiets.



"Guess what" ?

All of this is no surprise at all and perfectly consistent with the information I provided you with on the previous posts.


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Sat 14, 2021 6:48 pm 
Member

Joined: May Tue 30, 2006 4:46 pm
Posts: 11031
Location: Santa Rosa, CA
Quote:
The way I have done it is to have a transmitting loop and at the radio,


I never considered this! Would a tuned loop be legal according to strict P15 rules? It has a lot more than 10 feet of wire, but its physical dimension is less than 10 ft. I would use a 23" embroidery hoop.

It would probably be very directional, but that would be fine for transmitting to a particular radio at home.

Rich


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Sun 15, 2021 2:01 am 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1628
Rich, W3HWJ wrote:
Quote:
The way I have done it is to have a transmitting loop and at the radio,


I never considered this! Would a tuned loop be legal according to strict P15 rules? It has a lot more than 10 feet of wire, but its physical dimension is less than 10 ft. I would use a 23" embroidery hoop.

It would probably be very directional, but that would be fine for transmitting to a particular radio at home.

Rich


With the 10 foot of wire limit, it also raises the question if you would have to count the length of the turns in series with a loading coil it might have. And with the loop, the the total length of the wire in it, or just the outside perimeter geometry of the loop ? And, since any earth conductors, which also act as radiators, should the length of those leading to the earth be included in the antenna measurement ?

Sometimes regulations might not cover every contingency, but they might do, so it pays to check. One thing is for certain, that is due to the fact that most pantry transmitters have an antenna structure that is tiny compared to the wavelength of the drive signal (MW band), the radiation resistance is very very low. So, the primary thing limiting the transmission range is the 100mW limit on the drive power. (The radiation resistance is a fictitious resistance that converts the antenna drive power into EM waves).

All other things equal, if one compares the radiation resistance of a linear wire as a radiator, the important parameter is its length, or effective height he. On the other hand, in the case of a loop, it is the product of the number of turns and its cross sectional area that determines it. So ideally the regulations would have specified a standard loop with the same radiation resistance as the 10 feet of wire to have created a comparable antenna design and a "level playing field" that included both loops and whips in the regs.

I have attached the equation 12, which shows the radiation resistance of loops vs whips. Solving the equation to make the two values equal and considering a one turn air cored loop, then the loop's area A would be (10 foot x wavelength) / 2pi . This creates an interesting complication, that the comparison between the whip and the loop, for radiation resistance, is wavelength or frequency dependent.

At a frequency of 1MHz, the wavelength is 300m or about 985 feet. This would make the equivalent cross sectional area of the one turn loop 1560 square feet and its diameter about 44 foot to have the same radiation resistance as the 10 foot whip. If the loop had 10 turns, it would be about 14 ft diameter, much bigger than anyone is making. I don't think you have to worry about any practical sized loops you make exceeding any regs with respect to the 10 foot antenna wire. (Perhaps someone else could check these calculations in case I have fouled them up).

Even though these figures make the loop look inferior for the size of it, compared to the whip, the advantage is in the directional nature of the near field magnetic component of the radiation and the receiver's relative insensitivity to electrical noise & interference, if it also has a loop or ferrite rod and it is orientated correctly. Also, a higher RF current in the loop's wiring, compared to the whip, is achieved at resonance because of its parallel tuning capacitor across its terminals.

Due to the fact the actual radiation resistance is so low for these small whips and loops, due to the long wavelength, the transmission range remains fairly limited with the 100mW regulation. The main thing you can learn from the attached equations, being stuck with the low power limit and you want to improve the transmission range, is to note that the radiation resistance is proportional to the 4th power of the frequency, for a loop and the square of the frequency for a whip. A very steep relationship, more so for the loop. So for arguments sake, if you shift the transmission frequency from 800kHz to 1600kHz for a loop, then your antenna radiation resistance increases by a factor of 16 (or a factor of 4 for a whip). So it pays to transmit as high up the medium wave band as possible due to this factor.

Generally I build my transmitter circuitry with attenuators, so regardless of the antenna configuration, I can turn down the level below 100mW so it does not transmit well to more than say 20 to 30 foot further than where my receiver is in the next room and falls into noise at that point, this completely avoids any issues of interference to others, this is well under 100mW in the 50mW range I have found. The purpose and intention of a regulation should always be complied with, even if there are any technical ambiguities specifying it.

When you add a conductor between your circuit to ground, that conductor becomes part of your transmitting antenna as like your "antenna" proper, it has RF currents in it and it radiates signal too. Some people have found their pantry transmitters for this reason appear to work better with an earth and hardly any of what you would call an antenna. This also completely confused someone testing out Nikola Tesla's idea of ground current transmission, Their "earth conductor" leading from their transmitter to ground was simply a radiating antenna, though they called it an "earth" and they detected remote waves in another conductor connected to ground. The only way to do the experiment properly was to have buried in the ground both the transmitter and receiver for the experiment.

One important point about building pcb's which are for RF application, they always should have a ground plane. Obviously this is more important at HF,VHF and UHF frequencies. Otherwise odd effects occur even trying to make scope measurements or test on the pcb as different results occur depending on where the scope earth clip is placed.


Attachments:
RAD1.jpg
RAD1.jpg [ 330.1 KiB | Viewed 1195 times ]
Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Sun 15, 2021 2:57 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 5021
Location: Texas
This is an enlightening discussion of RF. Thank you for taking the time to explain all of this.

Today I took the transmitter outside away from all of the radiated noise in my house. Hum and hash are really bad in my house because we have computers and other equipment spread all throughout the house.

With the transmitter n the backyard there's absolutely no hum in the signal. With fewer electrical distractions I get a much better idea of its capabilities. 50 feet away if so has a strong signal with FM like quality. It close to pegs the receiver bargraph only if I turn up the receiver gain all the way. I'll determine the output wattage last thing. It may be more than 100 milliwatts as it is. I saw a similar circuit that had a 500 ohm resistor between the LM386 and transistor. That was with a 2N7000, and I'm trying different transistors, FETs and bipolar.

I'm also observing what happens with different output inductors. Maximum signal so far occurs when the inductor is the highest value possible so that the capacitor is at minimum value. This is consistent with an observation Bob Weaver made. The variable capacitor is now set at 4 pF, and the inductor measures 157 uH at 1680 kHz. Small inductors with fine wire lower the output noticeably.


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Sun 15, 2021 2:52 pm 
Member
User avatar

Joined: Jun Sun 22, 2008 2:03 am
Posts: 1068
Location: Burke, VA 22015
I'd like to make a couple of suggestions.

In your circuit, as seen from the layout, the MOSFET's grid is floating. It should be tied to ground with a resistor (e.g. 100KOhm).

In any type of switching circuit involving high RF current, you should pay attention to where all the AC signals circulate. When laying out the circuit board, extra care should be made to separate the "loops" to mitigate cross-talk. When the antenna circuit is tuned, it can carry very high circulating current, limited only by the circuit's Q factor. You have to make sure that it doesn't share with other sensitive input loops. Here is one suggested layout. A ferrite, paralleled with a resistor and bypassed by a 0.1uF ceramic capacitor, was added to prevent the RF signal from getting back to the LM386 output.
Attachment:
AM_1.png
AM_1.png [ 23.36 KiB | Viewed 1144 times ]

_________________
Binh


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Sun 15, 2021 5:15 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 5021
Location: Texas
bb.odin wrote:
I'd like to make a couple of suggestions.

In your circuit seen from the layout, the MOSFET's grid is floating. It should be tied to ground with a resistor (e.g. 100KOhm).

In any type of switching circuit involving high current, you should pay attention to where all the signals circulate. When laying out the circuit board, extra care should be made to separate the "loops" to mitigate cross-talk. When the antenna circuit is tuned, it carries very high circulating current, limited only by its Q factor. You have to make sure that it doesn't share with other sensitive loops. Here is one suggested layout. A choke paralleled with a resistor bypassed by a 0.1uF ceramic capacitor were added to prevent the RF signal from getting back to the LM386 output.
Attachment:
The attachment AM_1.png is no longer available


Thanks, Binh. I carry high regard for your knowledge and contributions and your additions make sense. A couple of comments followed by questions:

First, I'm trying not to have conflicting goals within the continuum that has optimum theoretical design on one end and simplest practical design on the other end. My goals lean toward the latter. I want this to be the least expensive best performing circuit with the fewest parts. If that means omitting one resistor, so be it. Then again, the goal is also to have a 100% duplicatable reliable circuit. If that means adding a couple of resistors, no big deal, although every inductor added increases expense and variability, as inductors are far less standard than resistors. If a certain inductor is specified, there's far less guarantee it will be available next year than with resistors. That's probably nitpicking, but the point is, if I can do without an inductor with negligible effect on the performance and applicability of the circuit, I'll lean toward omitting it.

Second, why is the 2 mH choke without the additional decoupling you suggest not sufficient to keep RF from getting back into the LM386? (Right now I'm using 5 mH and it previously worked with 1.5 mH.)

Third, to your point about the floating gate: I had previously included a 33K resistor on the gate of the FET, located in the ellipse below. It made no difference in the performance, so I omitted it. Seeing my layout with new eyes, I note that it was placed before the capacitor to the gate, not after, so the gate would have been left floating anyway, right? Assuming 33K is as practical as the 100K you suggest, how much difference does it make having it before or after the input cap to the gate? And right now, I'm using a bipolar MPSA18 instead of a FET. Same thing apply to the floating base?

In the context that the fidelity and range are already as good as anyone might expect with a 386 transmitter, I think I know the answer, but I'd like your comments, thanks.

EDIT: Yikes, I just noticed that the 33K would be connected to +12VDC instead of ground. I'll have to redraw that. This brings up the question of what effect there is when defloating to the positive rail instead of to ground.


Attachments:
Image1.jpg
Image1.jpg [ 190.93 KiB | Viewed 1138 times ]
Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Sun 15, 2021 9:59 pm 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1628
I agree with Binh's remarks and good catch on the missing gate resistor !

As I mentioned, the better way to make the pcb is to have a ground plane. Generally this would be the entire pcb top. The power supply rails and other tracks are on the bottom. At every point there is an earth, it is soldered on the top side, or via a plated through hole if there is one, this includes any components, IC pins, bypass capacitors etc. It reduces the chances of circulating currents inducing voltages and causing feedback pathways that result in instability. Also, of note, for anything RF application , with IC's, it pays to NOT use IC sockets and solder the IC's in directly.

A mosfet's gate, being an insulated structure always requires a resistor to discharge it from the electrostatic perspective. Also, many mosfets have a relatively low g-S breakdown voltage and it requires a Zener diode added, typically around 12 to 18V depending on the mosfet's rating to prevent the gate-source terminals being over-voltaged. Some, but not all mosfets have this diode inside them. These are the kind of mosfets that you can leave kicking around in your junk box with their BJT neighbors, without too much fear the gate has been damaged by electrostatic charges.


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Sun 15, 2021 11:15 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 5021
Location: Texas
ACORNVALVE wrote:
Also, of note, for anything RF application , with IC's, it pays to NOT use IC sockets and solder the IC's in directly.

I'm looking into having boards made professionally where ground plane is a cinch using their software. Excellent benefits at no additional cost.

Given the wide variability in users soldering and construction skills, I do think there's something positive to be said about including IC sockets. If the parts are being supplied in a kit where the user gets just one chance at soldering an IC without destroying it, as long as the circuit works just as well with sockets, I'd provide sockets. Those who need them can use them, and the pros can leave them out.

ACORNVALVE wrote:
A mosfet's gate, being an insulated structure always requires a resistor to discharge it from the electrostatic perspective. Also, many mosfets have a relatively low g-S breakdown voltage and it requires a Zener diode added, typically around 12 to 18V depending on the mosfet's rating to prevent the gate-source terminals being over-voltaged. Some, but not all mosfets have this diode inside them. These are the kind of mosfets that you can leave kicking around in your junk box with their BJT neighbors, without too much fear the gate has been damaged by electrostatic charges.


How does this apply to a BJT? What benefits come from keeping the base unfloated?

As always, thanks for the education!


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Mon 16, 2021 1:29 am 
Member

Joined: Jan Thu 16, 2020 12:29 am
Posts: 1628
Macrohenry wrote:

How does this apply to a BJT? What benefits come from keeping the base unfloated?

As always, thanks for the education!


It doesn't really apply to a BJT, although if the drive signal for any reason reverse biases the the B-E junction to over about 5 to 7V for many transistors, the junction will zener, and in the long run this slowly degrades the transistor's hfe. To prevent this a plain diode can be put across the junction (B-E) terminals if required, limiting the reverse voltage to 0.7v.

The BJT, if you are capacitively coupling into it, also requires a base to emitter resistor, in many designs this could be part of the bias network, but at least the one resistor is still required. The collector to base leakage in Silicon transistors is low, but still, if there is no resistor the transistor tends to want to turn itself on and is not thermally stable. But also, given that the B-E junction acts like a diode, all that happens without a base resistor to discharge the capacitor, and with capacitive coupling, is that the capacitor charges up to the peak voltage and no longer passes the AC drive signal, except just a trace of it on peaks.

There are some circuits that exist where a transistor's base is floating but its not a common scenario in electronics, sometimes transistors were wired as noise sources or optical sensors like this. One other example is a transistor with a diode across the B-E junction, no resistor, and a long wire connected to the base. In this case there is a very tiny amount reverse leakage from the diode and the base floats up until the transistor is nearly on the verge of conducting. This is an electrostatic discharge detector. But all of these things are the oddball exceptions to the rule. On power output stages with bjt's often the base resistances are kept below 1k to help improve thermal stability.

Thermal stability of power stages is an interesting topic. It was more problematic in the days of germanium transistors because of the higher collector to base leakage currents. A lot of the time, its not a problem in switching stages as the device is in saturation or has a very low voltage drop when turned on and the dissipation low. The main area where thermal stability of an amplifier stage is problematic is in the class A power output stage and it requires careful analysis to get it right. An interesting fact is, that for a class A stage, if the collector voltage is less than 1/2 the power supply voltage it will be unconditionally thermally stable. This is not often the case though as the load is a transformer or choke and the collector DC voltage is closer to the supply rail. Then the stage may or not be thermally stable depending on how much emitter degeneration there is, the source resistance of the base bias network and the size of the heatsink. Often, many designs you will see, the thermal considerations get neglected and then people are left wondering why something overheated and failed.


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Mon 16, 2021 2:07 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 5021
Location: Texas
Thank you! I'll see what a base to emitter resistor does. I've monitored the plastic MPSA18, and sometimes it does get warm. But never hot. When I first started trying different transistors an IRF520 works about as well as an MPSA18, and even it would get hot sometimes but mostly it didn't. Which contributes to the notion of thermal instability.

I've seen some designs like this that were referred to as class E transmitters, however they had a perfect square wave driving the gate of a 2N7000 FET. Someone pointed out on this forum that this configurationis not true class E if the transistor drive signal is not a clean square wave. My understanding is that if the transistor drive signal is a clean square wave, the transistor would be switching either on or off and not subject to the inefficiencies of Class A or B functions.

The output of the frequency synthesizer is a curved triangle wave. I thought about setting the frequency to double and then adding a JK flip flop to provide a perfect square wave to the gate or base. But the circuit is performing so lovely that I just didn't want to complicate it unnecessarily.


Top
 Profile  
 
 Post subject: Re: Choke hum on LM386 Transmitter
PostPosted: Aug Mon 16, 2021 7:20 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 5021
Location: Texas
Using the bipolar MPSA18 the emitter resistor has no effect on performance I can perceive Even with the design goal of as few parts as possible, I may include it anyway if only for proper design and whatever protection it provides. Plus if someone wants to use a FET it will be there.

Here's a video of the current configuration. I've decided to go with a variable inductor as the primary tuning component using a five foot antenna wire. The 157 mH inductor is ten feet of #36 wire on a 3/8" form made adjustable with a 1/4" ferrite rod. Of all the inductors I tested, including a Litz Miller 6300, this home wound inductor produces the strongest output. It's not easy to source a commercial variable inductor that's this inexpensive.

In the YouTube video below, the music source is an FM radio. The first half of the video shows that the volume is constant across the back yard and the lack of hum and noise doesn't change that distance. The second half has various input volumes going from under modulated to over modulated.

https://youtu.be/uXdms3ITJyM
https://youtu.be/7CVb-cwCFZY

Fun piece of music sounds good far from transmitter
https://youtu.be/_M1BUnnKl0Y


Top
 Profile  
 
Post New Topic Post Reply  [ 24 posts ]  Go to page 1, 2  Next

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 2 guests



Search for:
Jump to:  




































Privacy Policy :: Powered by phpBB