The False Superadio - a Construction Project

This page is an installment in my series - what makes a Superadio? The basis of this construction project is a radio given to my by my late father. It was a premium given out by Time Magazine for subscribing. And, I might add, a very cheap one - almost useless. It has sat on my shelf for years unused:

Sensitivity on AM and FM - poor. Selectivity on both bands - poor. Sound - tinny. No stereo separation on FM. All in all - not a radio I would use, because there are only 3 or 4 AM stations I bother with, and three of them are extreme DX. FM is not a usable band in Houston anyway - nothing on the air I care to listen to.

I wondered - what is inside this pathetic little radio - anything of use? Could it somehow be improved to the point of usability?

I cracked open the case by removing five screws, and took a look inside:

Well there is the answer why stereo on FM - the speakers are connected together in parallel! This is a mono radio, plain and simple.

Next, I took the board out to have a look:

Tiny AM ferrite bar = poor AM sensitivity. There is one obvious place to improve the radio. The dial "string" is not string at all, it is a piece of flexible plastic. Everything about this radio screams "CHEAP". Worse yet, the tuning knob is connected directly to the shaft of the tuning pot. It is going to be really hard to tune this radio, even if improvement can be made.

I decided at this point that the original case, tuning hardware, and speakers were unsalvagable. I want to put in a decent ferrite bar that won't fit in the old case anyway. But - is the circuit board capable of making a better radio?

Let's have a closer look at the RF portion of the board:

Single IC, one cheap ceramic filter on AM, a cheap ceramic filter for FM, two gangs for each band. But - I knew from past experience with a Mary-Kate and Ashley radio, these single chips can perform really well, when provided with an adequate antenna and good ceramic filters. A quick feasibility study confirmed, it can make a really good radio. Just how good? Good enough to proceed.

Of real interest is the three capacitor locations marked for longwave - a future project for sure!

I have a small supply of surplus SR-3 cases and parts, so these will form the chassis and case of the new radio. But before I can even think about that - I have to do some preparation work. All components on the bottom side of the board will be in the way and have to be removed. And there are other modifications as well.

The first task is to get rid of those lousy cheap IF ceramic filters. A 150 kHz filter for FM, and a three stage filter for AM should do the trick:

I can definitely do better than the small rectangular loopstick antenna. I removed the bar. And I also had to remove the holder, which involves taking a lot of wax off the board and rotating the holder to remove it from the board. The wax is useful for positioning the coil on the new ferrite bar, so I saved it.

I don't need the LED - but perhaps some day I might want to use these connections for a dial light:

The frequency response is tailored for small speakers. It has a correspondingly small output coupling cap:

There is plenty of room for a larger cap:

The existing volume slide potentiometer has to be removed - a spare SR-3 volume control will be used instead. The slide potentiometer, though, is 10k, and the SR-3 volume is 100k. But it works ---

The new headphone jack for the radio will be a spare headphone jack from an SR-3. This way, it will fit nicely in the holder already on the chassis.

Both jacks were removed - the DC one for use in a different project.

There are switches for the on / off / alarm and band. Both of these need to be removed:

The on / off / alarm switch was problematic to remove - I lost a could of pads and broke off some switch pins. It is no consequence - I won't be going back to the old case ever again. The power switch was wired in first, it will of course be the switch for an SR-3. The yellow wire is used to connect the ground, which previously was made through the metal body of the switch. This is a common practice - when single side PC boards have problem making connections. They will use every trick to make a connection.

The new band switch will be a band switch for an SR-3:

A hole needed to be drilled to get the antenna wires from behind the dial to the chasis mounting area:

I later discovered I could get better performance with the brown wire (far end of the winding) separated from the rest, so I drilled a second hole just for it.

The new ferrite bar is similar in diameter to the SR-3 ferrite bar. But the mounting clips for the SR-3 bar are too large. The original SR-3 bar used a material that reminded me of medical tape to enlarge the diameter. So I used medical tape to expand the diameter:

And the ferrite bar was mounted. Two guides are meant to hold the ferrite bar to exactly 200 mm, but the optimum position for the coil is slightly off the end of the rod, so one guide had to be cut off so the coil could hang over the end:

The most critical - and difficult matter in mounting the board is the tuning system. I intend on using the existing SR-3 pulley, dial string, and pointer, so that means that the tuning capacitor shaft needs to line up with the hole in the chassis that accomodated the tuning pot. But there is another problem. The SR-3 pulley is made for a tuning pot, not a tuning capacitor. Shaving down a broken tuning pulley that was used for a capacitor yileds a hub that I can use to put torque on the capacitor, when screwed tightly to the SR-3 pulley:

The hole in the existing SR-3 chassis is too small and had to be enlarged to accomodate the hub:

Once the position is established, three holes can be drilled to place 4:40 mounting screws and the board mounted to the chassis. The antenna wires were re-attached to the board, the board mounted, and the dial strung:

I had to make a small PC board to hold the band switch:

The volume control was mounted with hot melt in the position it would be on an SR-3 board:

I had hoped that using the original SR-3 pulley would make the dial scale line up. It does not. The problem is that the tuning potentiometer adjusts over a larger angular range than a tuning capacitor. The result is a compressed scale. I taped some paper to the dial and marked a new scale:

There was a problem - the speaker interfered with the board so the radio couldn't be closed. A smaller speaker was mounted on a piece of cardboard, and that solved the inteference issue - mostly. I still need a long term solution, but at least I can screw the case together:

The original radio only used AA batteries, so I put two battery holders in the battery compartment for AA's.

The finished radio is nothing short of amazing! Sensitivity is almost Superadio quality on AM - but FM is not as good. Selectivity on AM is also very good, much better than an SR-3 but not as good as an SR-2 or original SR.

So - why is the performance so good? This radio does not have a tuned RF stage, and yet is extremely sensitive! The only conclusion I can come to is that the ferrite bar is doing a lot of the work of packing signal into the SR front end. But although this is a serious DX radio now, it is not good as an SR. The overall noise floor appears to be higher - and that may be due to the lack of the tuned RF stage getting rid of some extraneous noise from locals spread across the dial. With only one pole low and one pole high, the antenna stage alone only has 20 dB / decade roll off in the stop band instead of 40 dB / decade if the radio has a tuned RF stage.