Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives :: Books
Support This Site: Contributors :: Advertise


It is currently Dec Tue 11, 2018 10:56 am


All times are UTC [ DST ]





Post New Topic Post Reply  [ 13 posts ] 
Author Message
 Post subject: Yet Another AM Transmitter: LDR modulated proof of concept
PostPosted: Feb Mon 12, 2018 6:45 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3652
Attracted to the simplicity of an op-amp based transmitter, I was curious how well it would work if a crystal-controlled signal was modulated by light dependent resistor (LDR) in the feedback loop. It works, per the circuit shown.

The LDR is driven by a diode, which ordinarily would pass only the positive excursions of the audio waveform. This produces severe distortion that results in unrecognizable audio. To get the diode to pass both positive and negative, it requires a virtual ground that is offset from audio input ground. A 3.4 volt bias accomplishes this, which is just the right amount for the 5V max swing from the LM386 source.

DISADVANTAGES:

Requires dual supply for decent signal strength
Requires adjustable second power supply for LED section of LDR
Output fidelity is limited by the response speed of the LDR
Adjustment pots are needed in the feedback loop to accommodate different samples of the same LDR (LDR specs are not all that uniform)
Input volume from source is critical; the sweet spot is narrow

ADVANTAGES:
Construction simplicity
Niftiness and curiosity factor that unconventional circuits attract


Attachments:
TransmitterLDR.jpg
TransmitterLDR.jpg [ 30.03 KiB | Viewed 1119 times ]
Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Mon 12, 2018 10:57 pm 
Member
User avatar

Joined: Aug Sun 02, 2015 7:34 pm
Posts: 884
Location: Yukon, OK USA
Interesting. I will have to give that a try. Don't have a 318 on hand but should not be a problem. Have a spare LDR. I have some cheap adjustable regulators so the additional voltage won't be an issue.

Edit; when running this configuration I assume you have the oscillator running continuously with a +5v supply. What supply voltage are you using on the 386?

_________________
Rick
AB5ON

Forest Gumping through life


Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Sat 17, 2018 8:00 pm 
Member
User avatar

Joined: Aug Sun 02, 2015 7:34 pm
Posts: 884
Location: Yukon, OK USA
Got the opamp in and breadboarded the circuit pretty easy. Had some trouble getting my LDR functioning so I put an LED in that could be observed to see that the audio was modulating the light. I ended up biasing the LDR to slightly on between the supply and ground then applied the audio through a cap and that got it working. Not sure what I did wrong in the schematic version. It had a narrow window to get everything set but end result was a good sound that covered my house and one more each direction. I used a 7805 off the +12 side to get the +5 for the oscillator.

Later I was changing the bias on the LDR and POOF, it quit working. :-( Must have gone too far. Kept playing with the circuit and tried the audio into pin 2 of the opamp. Just left a 10k pot in the feedback loop. After finding a sweet spot it also sounded pretty good and had decent range.

Interesting use of an opamp for this. One benefit to this approach is that your compressor also uses a bipolar supply so the two could pair well.

It was enjoyable to try out a new approach.

_________________
Rick
AB5ON

Forest Gumping through life


Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Sat 17, 2018 8:47 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3652
I'll bet the sweet spot varies noticeably from LDR to LDR.

I also got it to work taking the LDR out of the feedback loop and relocating its resistor as a shunt to ground from the non-inverting input. This arrangement keeps intact the traditional op-amp fixed relationship of input and feedback resistors. For that I had to add a lot more bias voltage, so much that it worked out that I was able to do away with a separate power supply and connect the negative LDR lead directly to the -12V source.

Looking at LDR specs, some of them have a very slow 35ms response. That would kill the highs. Now I'm wondering about a FET instead of LDR. So many projects, so little time!


Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Sat 17, 2018 9:12 pm 
Member
User avatar

Joined: Aug Sun 02, 2015 7:34 pm
Posts: 884
Location: Yukon, OK USA
Macrohenry wrote:
Looking at LDR specs, some of them have a very slow 35ms response. That would kill the highs. Now I'm wondering about a FET instead of LDR. So many projects, so little time!


Post if you do it with a FET. I have some on hand and the circuit is already on the breadboard.

_________________
Rick
AB5ON

Forest Gumping through life


Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Sun 18, 2018 1:28 am 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 566
Location: Equinunk PA 18417
Have a look at page 2 of this pdf.
https://m.eet.com/media/1140508/120601di.pdf
Seems like it would work much better than that ldr.

edit: "work much better" instead I mean "have wider bandwidth"

_________________
-Richard


Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Sun 18, 2018 4:08 am 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3652
Attachment:
Image2.jpg
Image2.jpg [ 19.62 KiB | Viewed 964 times ]
Excellent building block, Richard. The RF oscillator goes into R1 and is modulated by the voltage at the FET gate. Per the chart, it could accommodate a variety of input voltages, all controlled by a 100K pot for RA. RA would be set at 27K for audio coming from the LM386 that outputs 5 V p-p. Per the green line on the graph, that would make about 10dB difference in output. That's more than enough for 100% modulation, so the 386 would need to be dialed back a little.

Best to use an op-amp with greater bandwidth than the TL071 shown.


Attachments:
Image1.jpg
Image1.jpg [ 18.4 KiB | Viewed 964 times ]
Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Sun 18, 2018 4:13 pm 
Member
User avatar

Joined: Jan Fri 02, 2009 11:32 pm
Posts: 566
Location: Equinunk PA 18417
I have wanted to experiment with this for rf modulation for many months but never seemed to get around to it, and it was on my mind when you first posted your idea to use an ldr. Seems like this fet approach could work with a fast-enough opamp (not TL071). Physical layout will be of utmost importance. I built some stuff with OPA658 a few years ago and, coming from an audio background, quickly learned that you don't even get out of the starting gate without ground plane construction, very short leads, surface mount components. SO EASY for it to be unstable with parasitics. I still have several of them but OPA658 is a horror to experiment with, which is probably why I haven't even attempted to try it here. You'll want some gain, a useable bandwidth of 20 to 50mHz seems like a compromise that won't invite more trouble than its worth. I don't have much experience with this type of thing but I'm sure I don't have anything in my parts bins that fits the bill.

Audio distortion will go up when RA and RB are not approximately equal, they are specifically there to leverage the fet back towards ac linearity. You could make one of them variable in order to "fine tune" the fet for lowest audible distortion on the modulated output.

_________________
-Richard


Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Mon 26, 2018 3:16 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19323
Location: Warner Robins, GA
Very interesting.

So many different solid state transmitter designs here lately.

If I had the time and money I'd build them all and see which works the best.

That would then require me to build a VU meter box so I have some way of visually knowing when I'm at 100% modulation so that I don't need a VU meter for each transmitetr I tested.


Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Tue 27, 2018 7:23 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3652
I having trouble sourcing a J271 FET. It seems to have parameters quite different from more common FETs. I can get a SMD, but who likes soldering those? I'll try biasing a PNP bipolar instead.

I've thought about building a simple overmodulation indicator that uses a comparator to compare the unmodulated carrier with the modulated carrier. An LED would flash when the modulated peaks exceed 2X the value of the unmodulated carrier.

Not shown in the circuit below is an attenuator to reduce the carrier to a workable voltage range. Then as shown, rectifiers convert to DC both the unmodulated and modulated carriers. The rectified DC of the unmodulated carrier is fed into one input of a window comparator. The other input is fed with the output of an adjustable voltage regulator whose output is manually adjusted to match the rectified carrier. The window comparator LED indicates when it's matched.

Now with the voltage regulator output equal to the rectified carrier voltage, the regulator's voltage is multiplied 2X and fed into one input of a second comparator. The other input of the comparator would be the DC rectified modulated waveform. Thus, when modulation peaks exceed 2x the unmodulated carrier, the comparator's output LED flashes. Adjust it such that it just stops flashing.

If necessary, it would use a fast op-amp as a precision rectifier. I need to think about whether this would be necessary, as it would more than double the parts count.

It might be cooler to have an LM3914 or another bargraph of its ilk. It would be about the same number of parts and I've not decided whether it would be worth it. The VU-like movement would be neat to look at, though.


Attachments:
Overmodulation Indicator Macrohenry.jpg
Overmodulation Indicator Macrohenry.jpg [ 48.88 KiB | Viewed 841 times ]


Last edited by Macrohenry on Feb Wed 28, 2018 4:25 am, edited 2 times in total.
Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Tue 27, 2018 9:14 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19323
Location: Warner Robins, GA
It's simple.

1. One LM 386 circuit.
2. Trimpot to adjust the signal to the 386
3. VU meter.
4. Resistor in series with VU meter.

Can be powered from the B+ supply to the transmitter.


Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Tue 27, 2018 10:26 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 3652
Tube Radio wrote:
It's simple.

1. One LM 386 circuit.
2. Trimpot to adjust the signal to the 386
3. VU meter.
4. Resistor in series with VU meter.

Can be powered from the B+ supply to the transmitter.


I can see how that displays audio volume, but how does it display overmodulation? Do you sync the output first with scoped modulation so that when the VU meter is in the peaks, that's when the scope shows overmodulation?


Top
 Profile  
 
 Post subject: Re: Yet Another AM Transmitter: LDR modulated proof of conce
PostPosted: Feb Wed 28, 2018 1:12 pm 
Member
User avatar

Joined: Jan Thu 01, 1970 1:00 am
Posts: 19323
Location: Warner Robins, GA
Yes you set the level control to the VU meter circuit such that overmodulation as seen on a scope is shown by the meter needle moving into the red.

If you try to set it using a single audio frequency the meter will read too low in normal operation so it must be set using music as the source.

Concerning the meter, the correct VU meter is one that will properly deflect the needle with an AC input voltage.

Also you may have to experiment with the resistor value in series with the meter as just directly driving it from the 386 makes the whole thing way too sensitive and is nearly impossible to get it set right.

Here's the meter circuit. One of those LM-386 boards from Ebay will work for this.

Attachment:
VU meter.png
VU meter.png [ 68.72 KiB | Viewed 809 times ]


Top
 Profile  
 
Post New Topic Post Reply  [ 13 posts ] 

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