The magic eye is interesting. What will you use it for?

 

Connect it to the AGC like they were in old radios

They have been used for VU meters in old tape recorders and in modern times,
and also in the past they were part of some test equipment.
A resistor/capacitor tolerance tester for example, you'd connect your component and
select a range, and then turn a dial that would open the magic eye when the dial was
indicating the % tolerance of the component under test... assuming the components in
the actual test instrument were the best tolerance that could be produced I suppose.

I was thinking about the AGC connection, it could be handled with a new micro since
it doesn't need to interact with the first. So long as the scanning thing can be turned
on or off with a "scan' button, the other part of the circuit can be used to store channels
when the first part pauses at a station.
Also, I was informed that every AM broadcast channel in my country are spaced 9kHz apart.
So the freq counter could just search 9kHz intervals and test the AGC at every 9kHz interval.

 

A bit of progress... A ferrite loop antenna which means I was able to ditch one of the cans
that I relocated to under the chassis when the stepper motor was installed,
and the spring which tensions the tuning gang shaft to provide some resistance for the
stepper motor, so not as much clanging around like heavy machinery when it turns.

The extra tuning gang is usually a small trimmer between one end of the coil and earth.

 

This is what real radios look like.

 

Connect it to the AGC like they were in old radios

They have been used for VU meters in old tape recorders and in modern times,
and also in the past they were part of some test equipment.
A resistor/capacitor tolerance tester for example, you'd connect your component and
select a range, and then turn a dial that would open the magic eye when the dial was
indicating the % tolerance of the component under test... assuming the components in
the actual test instrument were the best tolerance that could be produced I suppose.

I was thinking about the AGC connection, it could be handled with a new micro since
it doesn't need to interact with the first. So long as the scanning thing can be turned
on or off with a "scan' button, the other part of the circuit can be used to store channels
when the first part pauses at a station.
Also, I was informed that every AM broadcast channel in my country are spaced 9kHz apart.
So the freq counter could just search 9kHz intervals and test the AGC at every 9kHz interval.

This all sounds like a good plan to me. I have a radio in my kitchen that uses digital tuning and scans every 10kHz between stations. I guess ours in America are that far apart and their individual bandwidths are less than that but greater than 5kHz, I think.

In case of your scanning at discreet intervals like this your not really using feedback from the AGC signal to determine where to stop your motor and begin a short routine of tuning back and forth to get the maximum AGC voltage. Your device is analog and should use this method I think. Going the other way you'd be counting motor steps to determine channel location would you not? Now I think I see it: you'll go the analog route to find the stations and the step count to memorize them and vector directly to them when you press a memorized location button.

 

That is a cool machine. I liked the way it sounded. Ever think of giving it wheels?

 

Thanks for the compliments

With the frequency counter as feedback from the radio
the 9kHz intervals should be exactly where the frequency counter says.
I don't have to count steps, I'm only stepping in the right direction relative to the
current frequency to move in the direction of the desired frequency.

When I do look at the AGC, and find that it's strongest,
I expect that to be where the frequency counter is displaying a number divisible by the 9kHz,
Then it should be just a matter of checking the AGC each 9 kHz increment
to see if there's either a station at that interval or not.

Made a new video
https://www.youtube.com/watch?v=XnErlKFDMmo

Got two motor speeds now too.. the slower one kicking in when within 20kHz of the desired frequency).

 

Thanks for the compliments

With the frequency counter as feedback from the radio
the 9kHz intervals should be exactly where the frequency counter says.
I don't have to count steps, I'm only stepping in the right direction relative to the
current frequency to move in the direction of the desired frequency.

When I do look at the AGC, and find that it's strongest,
I expect that to be where the frequency counter is displaying a number divisible by the 9kHz,
Then it should be just a matter of checking the AGC each 9 kHz increment
to see if there's either a station at that interval or not.

Made a new video
https://www.youtube.com/watch?v=XnErlKFDMmo

Got two motor speeds now too.. the slower one kicking in when within 20kHz of the desired frequency).

You appear to know what you're doing and I admire your work, but here's the way I would do it if I had your setup. The information about a valid station's presence would be a matter of its position with respect to the stepper motor shaft. I don't know enough about stepper motors offhand but I'm pretty sure you could make the motor go to a particular position by counting the number of steps it takes to get somewhere. This would mean the program driving the motor would always know the tuner position. So if someone wanted to go to the station at 1480 and it was presently sitting at station 1260 the number of steps to get there could be found by simple arithmetic. But for display purposes determine an algorithm that determines a frequency for a corresponding shaft position and display that. Conversely, when you enter a frequency in kHz that algorithm will translate it into a motor shaft position. In addition, when you find a station you want the radio to remember using the AGC just push a memory button and the shaft position will be recorded in EEPROM.

 

If you just assumed the frequency was always on the same spot on the dial,
that doesn't account for drift unless you also keep checking the ACG once you get there.

The program could easily know the shaft position, etc. but that is actually less accurate
than looking at the tuned frequency.
The numbers on the display aren't made up.. it is the actual radio's local oscillator frequency
minus the 455 kHz intermediate frequency that is being displayed.

It knows the tuned frequency that is on the display to a decimal place more than is being displayed.
That's much more precise than one step of the stepper motor.
In fact, the decimal is not displayed because it is always changing.

 

If you just assumed the frequency was always on the same spot on the dial,
that doesn't account for drift unless you also keep checking the ACG once you get there.

The program could easily know the shaft position, etc. but that is actually less accurate
than looking at the tuned frequency.
The numbers on the display aren't made up.. it is the actual radio's local oscillator frequency
minus the 455 kHz intermediate frequency that is being displayed.

It knows the tuned frequency that is on the display to a decimal place more than is being displayed.
That's much more precise than one step of the stepper motor.
In fact, the decimal is not displayed because it is always changing.

As you just said, you will use the memory of the shaft position only as a rough guestimate, then you use that AGC thing to fine tune via strongest signal. In programming terms, this is a subroutine that drives the stepper one way and the other until the maximum voltage is determined.

Speaking of the AGC, you can use that, but if you're wanting to develop your own dc signal as an input to your comparator you can put a diode detector at the output of the last IF stage, but you can also put it at the audio output, i.e., the speaker output. This saves you from having to incorporate a signal not of your own derivation so that you are in control of the whole process.

 

That's an interesting idea using the audio instead of the AGC.
It might not look so good if the audio were displayed with some kind of
indicating device for tuning tho.. the LED or eye would become a VU meter.
I have already considered sending audio from the pic into the radio's audio amp
for indicator beeps when the dial is about to turn, etc.

Speaking of AGC... where might I find it?

This is missing the power supply to make the rest of the hosted image clearer.

I have devised a temporary way to select one of five stations,
and a demo to cycle through tuning the same five local stations.
This is the first time it has tuned any other than the one station

 

The line at the very bottom (left) is the AGC. It is derived from the diode in the 6SQ7 valve. It should be DC but it will not be a pure DC like a battery and this may be a problem. Look at it on an oscilloscope to see if it is useable. You'll need to dc couple the scope so be sure you put the probe's ground on the radio's ground and don't assume the chassis is at dc ground, check it first. Once you're set up turn on the radio and tune to a strong signal and note the voltage value and see if it wavers and by how much. Then tune to a weak station and note the voltage value; it should be less. The voltage should vary with signal strength. If you don't have an oscilloscope just use a VOM or DMM on the dc setting.

 

Ok thanks, it's easy to get to from the top, as it's one of the terminals sticking out the top
of the double tuning gang, and I've already looked at it on a scope, and it didn't appear useable,
but the scope was AC coupled, that might make a difference then, thanks again

 

Ok thanks, it's easy to get to from the top, as it's one of the terminals sticking out the top
of the double tuning gang, and I've already looked at it on a scope, and it didn't appear useable,
but the scope was AC coupled, that might make a difference then, thanks again

Your welcome. You might just put a VOM at that point now and see what happens with no signal, weak signal and strong signal. If there's no variation you've got nothing to work with.

 

Enjoyed reading this thread.

Nice creation there . . . I too like the "steampunk" vibe it gives off.

 

Enjoyed reading this thread.

Nice creation there . . . I too like the "steampunk" vibe it gives off.

Thanks for the comment It wasn't supposed to be this way,
I'm not the arty type, but once the brass gears went on I loved it
Apparently I should have cut slots into the chassis, and used a black leather
dog collar to fix the stepper motor, but that information arrived too late,
and the best I could do is cover the silicone bead with some old transformer tape
and then beeswax

Paul, work is getting in the way as it does, I'm hoping there's something to see there.

 

I tried the scope with DC coupling and the voltage seems to go below zero
when a station is tuned.. a smooth enough signal, but only 2 or so Volts difference.

I don't know if there's a way to use an ADC pin to detect below zero voltage directly.

I've tried this on a breadboard with LM324:
http://electronics.stackexchange.com/questions/3105/how-do-i-measure-a-negative-voltage-with-a-adc
I used pots for both resistors and can't get it to light an LED from the output
which I think it should even with + 2 Volts.
It appears my chassis is ground since that's what I was using with the scope.
Might have a look at the circuit's output on a scope.

Edit... that's -2 Volts sorry , not 20!

 

It might be easiest just to make your own detector. All it is is a diode with a fixed capacitor and a resistor. See those circuits associated with the crystal radio thread. Your radio needs negative voltage for its AGC control. To do that they put the diode pointing toward the source of its IF. All you need to do run a wire to the same location (just to the right of transformer 24) but point your germanium or silicon diode away from this source to get positive voltage. A change of 2 volts should be plenty of information for your comparator. For your resistor use a 47k which is high enough so as not to load that portion of the circuit. The capacitor should be about .002 microfarad. I prefer germanium diodes for this purpose since it has a smaller voltage drop across it than silicon (.2 verses .7 volts).

 

Thanks, I'll give it a shot. I had enough to keep going for the weekend.
I got two scan buttons working, but at the moment they just increment the
desired frequency in steps of 9kHz.

It's cool that you don't control the stepper directly, but the target frequency.
It's an interesting feeling to get ahead of the motor and wait for it to catch up.

 

Well Paul, you were right about the AGC, and it does matter which side of the coil
you take it from the radio for the magic eye. Take it from the other side of the coil,
and it works, but introduces noise, and attenuates the radio. On the correct side it goes a charm

This particular magic eye I got from hamfest on the weekend (6E5), was from an old
capacitor checker that must have been left on for long periods.
The part of the eye that normally glows at zero volts is very dull compared to the
rest that opens out to actually indicate something, but it's ok, they are still available from eBay.

I've still not read the AGC level into a micro yet... but the eye is cool

https://www.youtube.com/watch?v=fmHnpZtEHIw