Yes ... the inevitable ... a new ferrite coil inspired the creation of a 4th Ferrite Fox Radio. I'll get to the radio itself in a bit, but first I wanted to revisit some of what was posted in Dan McGillis' earlier post on making a Well Behaved MK-484
My new Ferrite Fox had the same overloading problem others have noted. I decided this time I was really going to get in there and understand what was going on. So I fired up the scope and meter and went to it.
Below is the schematic of the circuit for the TRF radio as it is depicted in the spec sheet...
Figure 1 shows the normal function of the AGC in this circuit. The bias current for the input to the TRF IC is derived from the voltage at point A
in the schematic - which varies according to the volume at the output of the IC. When there is little or no signal, the voltage at point A
is around .9 volts, and the drive bias is high, making the gain high. With a strong signal the voltage at A
falls to around .6 volts, and the gain of the IC is driven down.
It works out that using a 1K to 1.5K resistor for Ragc provides a very nice AGC range for most radios. The AGC range can cover 30db, which is about a 30:1 gain adjustment. This worked very well on one of my radios ( Ferrite Fox #3 - with the contra wound coil ) - but couldn't handle the strong stations on this latest radio.
So, let's look at the circuit for the Well Behaved
Figure 2 shows the changes made to try and tame the wild TRF. Basically, the supply voltage and bias voltage are made adjustable. I tried this approach to see what could be achieved. Indeed, you can tweak the radio to work nicely on more powerful stations ... but you give up a lot of the 30:1 dynamic range in many cases ... and may need to retweak a lot if you go from listening to DX stations, to armchairing a local Clear-Channel station.
I didn't like the way the supply voltage could wander about. You couldn't really preset it and forget it ... especially when you consider that battery voltage will change over time. The full gain voltage at A
was always best around .9 volts, but it really wanders around with the two adjustments. After some tweaking and experimenting, I decided I needed to lock the supply voltage in place with a regulated supply ... so I modified the circuit as shown in figure 3.
Now, with a fixed supply voltage, you only needed to set the bias with Radj.
Reducing the Ragc from 1K to 470 ohms is needed to keep the voltage at A
high enough to allow the bias adjustment to work. If a strong signal was able to drive the voltage at A
all the way down to .6 V then there would be no latitude for the bias adjustment to be made. But the 470 ohm resistor also halved the dynamic range of the AGC. Adjusting the bias to further limit ( tame ) the drive for a strong station limits the dynamic range of the AGC action even more. Indeed, the adjustment becomes more of an RF limiter setting than an AGC adjustment. But, I'm not complaining, it does allow the radio to handle strong signals very elegantly !!
So, what can be done to get the best of both worlds ???
How about this approach...
This is the new trick that I used in Ferrite Fox #4. A switched pot is used for the RF gain adjustment. When the RF gain is set so the wiper is set fully to point A
of the circuit, the switch opens and a full 1.47K Ragc is in place to allow the radio to work with the full 30:1 dynamic range of the AGC circuit -- as the designers had originally planned. When a strong signal requires RF limiting, the RF Gain adjustment is tweaked, closing the switch and reducing Ragc to 470 ohms ... and the setting is adjusted for the needed bias.
This really works very well, and it became the final approach I used for the Ferrite Fox #4 radio.
But there are other approaches that could also be tried ... for instance ...
What the heck is going on with Figure 5 ?? you ask
... Well, I had considered what might happen if you made the bias adjustable in a positive direction. It might be of some use where you have no locals, and want to tweak the gain even higher than the AGC would normally set it ... but that never seems to be the case. I just wanted to let you know someone has actually tried this approach ... and you can forgedaboudit
Here is the one approach that might yet be the best solution to a no-fiddle TRF radio ...
If you use a ferrite loop with a coupling winding, it might be possible to adust the number of turns to reduce the level at the TRF chip so that the 30:1 dynamic range of the AGC would cover the full range of potential signals. This might require a lot of experimenting, but could result in a big payoff. I'll leave this one for further experimenting ...
So ... here is how the Ferrite Fox #4 was finally wired ...
I have been going through my 1N4148 diodes with a meter to hand select diodes with lower forward voltages in order to get the supply voltage to set at 1.2 volts. This has been one key to my radios working well.
The adjustable RF Gain switch circuit required my removing 3 resistors from the circuit board and working them into the wiring of the switch/pot on the back panel of the radio as seen below ...
You can also see the ferrite rod that precipitated the new radio ... a 7" long heavy ferrite from a vintage Meissner antenna coil, that I rewound with 176/46 litz to make a really sensitive antenna coil ( maybe too sensitive
Buttoned up, the radio sounds great.
I hope this post wasn't too tedious ... I know some of this has been discussed a bit in the past ...