I made the following receiver to pick up the signal from my part 15 AM transmitter to monitor the audio.
If I increase the RF gain any using the RF gain control or I move the receiver closer to the transmitter (sits about 4' from it now) I get the following demodulated audio with the transmitter being fed a 400Hz sinewave.
With no modulation I can increase the RF gain or move the receiver closer to the transmitter and the RF waveform still looks good.
If I use my scope on the RF output with no modulation I get 8.1Vpp at 1MHz with the RF gain set to minimum which based on calculations is 1.77.
It seems like to me like it may be a problem with the detector circuit which is the 1N5711 given the RF waveform based on the +/- supply voltage can go up to 26Vpp minus how close the output can come to the rail voltages of the LM118 and going much above the 8.1Vpp at the output of the LM118 causes the issue.
One thing I cannot remember is what the 27k resistor is for.
If I increase the RF gain any using the RF gain control or I move the receiver closer to the transmitter (sits about 4' from it now) I get the following demodulated audio with the transmitter being fed a 400Hz sinewave.
With no modulation I can increase the RF gain or move the receiver closer to the transmitter and the RF waveform still looks good.
If I use my scope on the RF output with no modulation I get 8.1Vpp at 1MHz with the RF gain set to minimum which based on calculations is 1.77.
It seems like to me like it may be a problem with the detector circuit which is the 1N5711 given the RF waveform based on the +/- supply voltage can go up to 26Vpp minus how close the output can come to the rail voltages of the LM118 and going much above the 8.1Vpp at the output of the LM118 causes the issue.
One thing I cannot remember is what the 27k resistor is for.
The 27k resistor grounds the detector at DC so its output voltage level is defined.
I wonder if the TL081 is going into phase-reversal on the peaks, or such such effect. Try monitoring the -ve supply rail as well to see if its being disturbed.
I wonder if the TL081 is going into phase-reversal on the peaks, or such such effect. Try monitoring the -ve supply rail as well to see if its being disturbed.
The LM118 or either of the TL082 stages?
Also I have a frequency converter connected to allow me to use a 6-9.1MHz receiver as well and the noise also appears in the receiver's audio.
Now if I don't modulate the audio I can turn the RF gain up quite a bit and still get a proper 1MHz sinewave.
With the receiver where it is the max RF voltage I can get to the detector with max RF gain is 19.82Vpp.
Now could the 27k resistor be too small or large in value?
I cannot remember exactly how I determined 27k was what was needed there.
If the. 01uF cap feeding the detector was a problem I'd see it with the unmodulated RF waveform.
One thing I can do is temporarily disconnect the detector diode and see how the modulated waveform does when the RF gain is turned up.
Also I have a frequency converter connected to allow me to use a 6-9.1MHz receiver as well and the noise also appears in the receiver's audio.
Now if I don't modulate the audio I can turn the RF gain up quite a bit and still get a proper 1MHz sinewave.
With the receiver where it is the max RF voltage I can get to the detector with max RF gain is 19.82Vpp.
Now could the 27k resistor be too small or large in value?
I cannot remember exactly how I determined 27k was what was needed there.
If the. 01uF cap feeding the detector was a problem I'd see it with the unmodulated RF waveform.
One thing I can do is temporarily disconnect the detector diode and see how the modulated waveform does when the RF gain is turned up.
Last edited:
It is quite difficult to assess a schematic if I don't have the actual circuit on my bench and I can poke around with my scope probe. So here some thoughts.
The resistor of 27k is quite non-critical. It only clamps the detected voltage to a defined level. When it is too small it only loads the LM118 too much.
The output voltage is 10Vp. With a 13.3V supply voltage this is almost the limit. You should check the TL082 for the VU output. It gets 10Vp at its input and is set to amplify 5 times. At no load. That seriously drives the opamp into non-linear operation, with all possible effects on the power lines. I would suggest to disconnect the TL082 for the VU part and see what is happening.
Why do you want a 20Vpp output on the audio and set the VU amplifier to 5x? Just asking.
FWIW you are using a period correct oscilloscope 🙂
The resistor of 27k is quite non-critical. It only clamps the detected voltage to a defined level. When it is too small it only loads the LM118 too much.
The output voltage is 10Vp. With a 13.3V supply voltage this is almost the limit. You should check the TL082 for the VU output. It gets 10Vp at its input and is set to amplify 5 times. At no load. That seriously drives the opamp into non-linear operation, with all possible effects on the power lines. I would suggest to disconnect the TL082 for the VU part and see what is happening.
Why do you want a 20Vpp output on the audio and set the VU amplifier to 5x? Just asking.
FWIW you are using a period correct oscilloscope 🙂
Agreed.
The voltage is +13V and -13V as it's a dual supply so I can have a max of whatever the OP-AMP can come to the rails.
The 100k audio level control sets the level of the audio.
I don't need 20Vpp out of the RF amp stage. Just was stating that's about the max it can do before distortion.
Now what if I replaced the 27k resistor with three 100k trimpots with one for the converter output, one for the counter output and one for the detector output?
I could then adjust each to where it's optimum for the detector, the counter and the converter. I would then not need to worry about the detector receiving too much RF voltage which is likely what is causing the problem.
As far as I can see the 27k doesn't do a thing for gain or attenuation. You can replace it with 10k and 100k and see what happens. Should not be anything.
However, you measure on the output of the lower TL082. You measure 10V pos and 6V neg, distorted. The gain of that stage is 1x. So at the + input you have probably 20Vpp. That is really high in an opamp circuit with +/- 13V. I get those readings from you scope screen, Y is 5V/div.
Then the same 20Vpp is connected to the + input of the upper TL082. This stage is configured for 5x gain. You are driving is high into saturation. I don't know what that might do to the power rails.
Can you try to adjust the 100k pot (audio level) such that the output of the lower TL082 is just 2Vpp. How does that look?
However, you measure on the output of the lower TL082. You measure 10V pos and 6V neg, distorted. The gain of that stage is 1x. So at the + input you have probably 20Vpp. That is really high in an opamp circuit with +/- 13V. I get those readings from you scope screen, Y is 5V/div.
Then the same 20Vpp is connected to the + input of the upper TL082. This stage is configured for 5x gain. You are driving is high into saturation. I don't know what that might do to the power rails.
Can you try to adjust the 100k pot (audio level) such that the output of the lower TL082 is just 2Vpp. How does that look?
The scope I like because it has slower phosphor which makes it nice for viewing waveforms under 60Hz, however it does not have a variable V/DIV pot, a variable TIME/DIV pot and a variable trigger level pot so it's fairly limited in some respects.
That was measured on the upper TL082 . The lower TL082 is the same only lower in amplitude.
Ok I see what's going on now.
I measured the output of the RF amp and according to the scope screen with a 400Hz modulation signal I am getting 20Vpp out. So as the RF signal level increases due to an RF gain increase or moving the receiver closer to the antenna, I start hitting how close the LM-118 can come to the rails.
So what I can do is replace the 3k resistor in series with the RF gain control with a 10k resistor. That will give me a gain of 1.5-2.
Thinking about it maybe I should have added a buffer for the frequency converter. The buffer would provide a low impedance for the nearly 12' cable to minimize issues caused by the cable capacitance and I could add a 100k trimpot in place of the 47k resistor. If I use a TL082 I can use two 100k trimpots and have a variable output for the counter as well.
The converter has a 68k input impedance and the output that feeds the converter has a 6.8k series resistor. So that will drop the output by about 1 volt plus whatever the cable capacitance drops if any.
For the 6-9.1MHz receiver I use the antenna trimmer control to adjust the received signal strength. With the adjustable output buffer circuit I will not need to detune the receiver with the antenna trimmer control.
That was measured on the upper TL082 . The lower TL082 is the same only lower in amplitude.
Ok I see what's going on now.
I measured the output of the RF amp and according to the scope screen with a 400Hz modulation signal I am getting 20Vpp out. So as the RF signal level increases due to an RF gain increase or moving the receiver closer to the antenna, I start hitting how close the LM-118 can come to the rails.
So what I can do is replace the 3k resistor in series with the RF gain control with a 10k resistor. That will give me a gain of 1.5-2.
Thinking about it maybe I should have added a buffer for the frequency converter. The buffer would provide a low impedance for the nearly 12' cable to minimize issues caused by the cable capacitance and I could add a 100k trimpot in place of the 47k resistor. If I use a TL082 I can use two 100k trimpots and have a variable output for the counter as well.
The converter has a 68k input impedance and the output that feeds the converter has a 6.8k series resistor. So that will drop the output by about 1 volt plus whatever the cable capacitance drops if any.
For the 6-9.1MHz receiver I use the antenna trimmer control to adjust the received signal strength. With the adjustable output buffer circuit I will not need to detune the receiver with the antenna trimmer control.
Last edited:
You wrote you measured the audio. That is on the lower TL082. So the whole discussion is moot because you did not tell you were measuring on the upper TL082. So you should have indicated that as the VU output at the opamp output. Or whatever.
Sorry, but you are making random assumptions all the time without any logic, and then propose random changes without any explanation as why that should work. Instead you should approach the problem in a structured and linear way.
So please make the measurements as pointed out in the figure below.
For each measurement indicate clearly what V/div you use on the Y scale.
Signal level should be such that AC demodulated signal is approx 1Vpp. That signal is on point D.
You might move further away from the Tx or detune the input circuit to achieve that.
That should give approx 7Vpp on point E
And 1 Vpp on point F
Use DC coupling for all measurements except F and G
For G and H use AC coupling and 0.5V/div.
The audio signal is also on the upper TL082.
I already know an RF gain above about 1.7 will cause the modulated RF output to hit the rail voltages on the LM118 output so reducing the available gain will solve that problem.
Adding the pots and buffers is so that I can tweak the RF level to the counter and external frequency converter.
I can either detune the circuit to achieve that or turn down the 100k audio level pot.
Which would you prefer I do?
1. Leave the audio level pot where it is and detune the circuit.
2. Adjust the audio level pot and leave the circuit tuned.
I can take the measurements tomorrow afternoon.
Concerning the measurements would you prefer I just post the approximate voltages or post photos of the waveforms?
I'll likely use my Telequipment D54 oscilloscope as it's the only one I for sure know is within reasonable cal given I calibrated it maybe 3 years ago using calibrated equipment at work after adding voltage regulators to the main supply voltages so that the calibration would not change with varying line voltage.
Concerning the 100k pots I may use two 100k resistors in parallel for 50k to see how well the circuit operates with a 50k resistance in place of the 27k resistor.
1. Leave the audio level pot where it is and detune the circuit.
2. Adjust the audio level pot and leave the circuit tuned.
I can take the measurements tomorrow afternoon.
Concerning the measurements would you prefer I just post the approximate voltages or post photos of the waveforms?
I'll likely use my Telequipment D54 oscilloscope as it's the only one I for sure know is within reasonable cal given I calibrated it maybe 3 years ago using calibrated equipment at work after adding voltage regulators to the main supply voltages so that the calibration would not change with varying line voltage.
Concerning the 100k pots I may use two 100k resistors in parallel for 50k to see how well the circuit operates with a 50k resistance in place of the 27k resistor.
Last edited:
Please detune the circuit. I want a low input signal, not attenuation halfway the circuit. Leave the 100k pot (AUDIO LEVEL) at max.
Please post the wave forms. I am interested in the wave forms. Make a note of the Y V/div
Use 400 Hz or 1000 Hz as LF modulation. Accuracy of this frequency is not important. Modulation level 80-90% is fine.
I don't care about amplitude accuracy. If the voltage level at point C is approximately 1 Vpp it is fine. Everything between 1 Vpp and 2 Vpp at point C will do.
You can leave the 27k resistor as as. When there is a reason we can change that. For now it looks fine.
Why is the resistor of 2.4M chosen for such high a value? Any specific reason? I think you get away with it because the TL082 has a FET input. A bipolar opamp might choke on it.
Please post the wave forms. I am interested in the wave forms. Make a note of the Y V/div
Use 400 Hz or 1000 Hz as LF modulation. Accuracy of this frequency is not important. Modulation level 80-90% is fine.
I don't care about amplitude accuracy. If the voltage level at point C is approximately 1 Vpp it is fine. Everything between 1 Vpp and 2 Vpp at point C will do.
You can leave the 27k resistor as as. When there is a reason we can change that. For now it looks fine.
Why is the resistor of 2.4M chosen for such high a value? Any specific reason? I think you get away with it because the TL082 has a FET input. A bipolar opamp might choke on it.
The 2.4 meg resistor is there so that when the audio on/off switch is set to off the top TL082 has a DC path to ground so that the VU meter still operates properly.
I chose that value so that I wouldn't load the detector down too much.
Originally I did not have the audio on/off switch, but added it later and then discovered with the switch off the VU meter wouldn't work right due to no DC path to ground.
The idea behind changing the 27k resistor is it would be a perfect place to add two 100k pots to feed buffers for the RF counter out and frequency converter out.
I think part of the reason I chose 27k is because with the .01uF cap it forms a high pass
I can only figure it was done so that there's no chance of the coil to pick up the 60Hz line voltage.
I chose that value so that I wouldn't load the detector down too much.
Originally I did not have the audio on/off switch, but added it later and then discovered with the switch off the VU meter wouldn't work right due to no DC path to ground.
The idea behind changing the 27k resistor is it would be a perfect place to add two 100k pots to feed buffers for the RF counter out and frequency converter out.
I think part of the reason I chose 27k is because with the .01uF cap it forms a high pass
I can only figure it was done so that there's no chance of the coil to pick up the 60Hz line voltage.
Here's the measurements. 10X probe used.
I wound up not having to detune the circuit.
I don't like how my cellphone camera makes the trace a little thicker than it actually is.
A .5V/DIV
B .5V/DIV
C .5V/DIV
D 20mV/DIV
E .5V/DIV
F 20mV/DIV
G 10mV/DIV
H 10mV/DIV
Looks like I do need larger filter caps for better filtering. I suspect the only way I'm getting away with the 1,000uF caps is because the OP-AMPs are dual supply and most if not all the ripple voltage will be cancelled on the output of the OP-AMPs.
It's likely I'll leave the 1,000uF caps in place (not easy to desolder from that type of perfboard without pulling a trace) and wire in larger caps and secure them to the piece the board is on.
Here's the receiver.
I wound up not having to detune the circuit.
I don't like how my cellphone camera makes the trace a little thicker than it actually is.
A .5V/DIV
B .5V/DIV
C .5V/DIV
D 20mV/DIV
E .5V/DIV
F 20mV/DIV
G 10mV/DIV
H 10mV/DIV
Looks like I do need larger filter caps for better filtering. I suspect the only way I'm getting away with the 1,000uF caps is because the OP-AMPs are dual supply and most if not all the ripple voltage will be cancelled on the output of the OP-AMPs.
It's likely I'll leave the 1,000uF caps in place (not easy to desolder from that type of perfboard without pulling a trace) and wire in larger caps and secure them to the piece the board is on.
Here's the receiver.