...Working on this Magnavox hi-fi for a friend. It sure has a lot going on inside! I replaced all of the paper caps as a start. The AM crackled and broke in and out if I rocked the detector IF. The FM was weak and stations tuned very broadly. The eye tube didn't deflect at all for either band. I started by tearing down the AM detector IF, and replaced the insides with ceramic caps on the outside. Thought that would cure it. Still problems. Rocking the tube sockets changed things. You name it. Replaced a multi-section 'lytic that tested perfectly, both for value and leakage, but had visible physical leakage. That improved things slightly (spurrious oscillation on each station). After disecting an FM IF and finding the silvered mica capacitor to measure perfect capacitance and leakage, and after examining other FM IF transformers to find no defects, I was ready to pull my hair out. However, after close examination of the circuit board it turns out that a great majority of the solder connections were fractured for whatever reason. Each circular joint had a circular fracture, isolating whatever was soldered in there from the circuit board. I resoldered all of the connections, and now the AM works perfectly, and the eye tube works a bit, too! ...But the volume is too low. It distorts with barely any power. ...Went over the pre-amplifier and cathode-follower circuits without finding any trouble. Replaced an electrolytic that didn't need replacing. Finally I moved on to the main amplifier (in a separate chassis). I went over all of the resistors, and they measured perfectly except for one 2.2 Meg that fed the screen grid of the 1st audio amp/phase inverter. Now the thing is super powerful! ....But I have to put back together the FM discriminator IF. YAY! How much could go wrong!? Thankfully a great majority of the caps in the RF chassis are ceramic or mica.

The AVC circuit has so many shunts. It doesn't need to be very powerful because it controls the gain of multiple tubes, and not just one, but it would be nice to have an eye tube that deflected a lot on strong stations. I finally removed the 1.5 Meg resistor that shunts the AVC .1 MFD filter to chassis. There's a shunt in the FM discriminator circuit and one in the AM detector circuit, and possibly some others, so I don't see the need for this extra shunt. All of the circuits connect together through various high value resistors, and remain connected together regardless of what band is selected, so all shunts are always taxing the AVC circuit. The AVC filter cap has plenty of paths to discharge through already. Now the eye tube deflects a bit more, though not like it would on a typical 30s set. I'll see if I have to reconnect the resistor once I get the FM section going. I might instead change the 470K resistor in the eye socket to a 1Meg, which will allow the eye to deflect more. I cannot find any flaws in the AVC circuit, so its low output must simply be due to the rather excessive amount of resistors shunting to the chassis.

T.

Thomas,
Those old pc boards were probably wave soldered, and unpredictable compared to today's reflow process. I believe it was caused by the expansion/contraction of the copper/solder in comparison to the paper/phenolic substrate. I found the same thing on a neighbor's really old garage door opener, and it was weather related. Cool dry mornings were a no go, but damp cool/hot weather didn't cause a problem. After close inspection under a magnifier, reflowed approx 50% of the ass'y. Has not failed in over 3 years.

marv

:...Working on this Magnavox hi-fi for a friend. It sure has a lot going on inside! I replaced all of the paper caps as a start. The AM crackled and broke in and out if I rocked the detector IF. The FM was weak and stations tuned very broadly. The eye tube didn't deflect at all for either band. I started by tearing down the AM detector IF, and replaced the insides with ceramic caps on the outside. Thought that would cure it. Still problems. Rocking the tube sockets changed things. You name it. Replaced a multi-section 'lytic that tested perfectly, both for value and leakage, but had visible physical leakage. That improved things slightly (spurrious oscillation on each station). After disecting an FM IF and finding the silvered mica capacitor to measure perfect capacitance and leakage, and after examining other FM IF transformers to find no defects, I was ready to pull my hair out. However, after close examination of the circuit board it turns out that a great majority of the solder connections were fractured for whatever reason. Each circular joint had a circular fracture, isolating whatever was soldered in there from the circuit board. I resoldered all of the connections, and now the AM works perfectly, and the eye tube works a bit, too! ...But the volume is too low. It distorts with barely any power. ...Went over the pre-amplifier and cathode-follower circuits without finding any trouble. Replaced an electrolytic that didn't need replacing. Finally I moved on to the main amplifier (in a separate chassis). I went over all of the resistors, and they measured perfectly except for one 2.2 Meg that fed the screen grid of the 1st audio amp/phase inverter. Now the thing is super powerful! ....But I have to put back together the FM discriminator IF. YAY! How much could go wrong!? Thankfully a great majority of the caps in the RF chassis are ceramic or mica.
:
:The AVC circuit has so many shunts. It doesn't need to be very powerful because it controls the gain of multiple tubes, and not just one, but it would be nice to have an eye tube that deflected a lot on strong stations. I finally removed the 1.5 Meg resistor that shunts the AVC .1 MFD filter to chassis. There's a shunt in the FM discriminator circuit and one in the AM detector circuit, and possibly some others, so I don't see the need for this extra shunt. All of the circuits connect together through various high value resistors, and remain connected together regardless of what band is selected, so all shunts are always taxing the AVC circuit. The AVC filter cap has plenty of paths to discharge through already. Now the eye tube deflects a bit more, though not like it would on a typical 30s set. I'll see if I have to reconnect the resistor once I get the FM section going. I might instead change the 470K resistor in the eye socket to a 1Meg, which will allow the eye to deflect more. I cannot find any flaws in the AVC circuit, so its low output must simply be due to the rather excessive amount of resistors shunting to the chassis.
:
:T.
:

Wish I could have found this before I tore apart the AM detector IF. I suspected that it was faulty because I rocked it and reception changed significantly, which is often a sign of oxidized silvered mica caps. For all I know it could have been just fine. It is fairly easy to check for shorts across and between coils due to silver migration, but it is rather difficult to check for inaccurate capacitance due to oxidation and poor contact, without disassembling the transformer and disconnecting all of the coils.

FUN!

T.

:Wish I could have found this before I tore apart the AM detector IF. I suspected that it was faulty because I rocked it and reception changed significantly, which is often a sign of oxidized silvered mica caps. For all I know it could have been just fine. It is fairly easy to check for shorts across and between coils due to silver migration, but it is rather difficult to check for inaccurate capacitance due to oxidation and poor contact, without disassembling the transformer and disconnecting all of the coils.
:
:FUN!
:
:T.
:
Your friend will have a nice unit when you finish. I hope to pick up a nice old Magnavox from that time period. I have a box style Motorola from 1958, record changer only with amp.

For my information what's the B+ measurement to the eye tube and what's the amp number? Thanks.
Bill

The amp number is 170-M, and the eye tube voltage is, for FM, AM, and Phono, as follows: 140, 180, and 230. The eye tube takes its B+ after a resistor that feeds the RF circuitry, and this varies depending on switch position. I have considered reconnecting the eye tube before the resistor, which would keep it closer to 230, but at lower voltages it will last longer, and might deflect more (??). It goes to 230 because in the Phono setting, all of the RF circuitry is disconnected.

T.

:The amp number is 170-M, and the eye tube voltage is, for FM, AM, and Phono, as follows: 140, 180, and 230. The eye tube takes its B+ after a resistor that feeds the RF circuitry, and this varies depending on switch position. I have considered reconnecting the eye tube before the resistor, which would keep it closer to 230, but at lower voltages it will last longer, and might deflect more (??). It goes to 230 because in the Phono setting, all of the RF circuitry is disconnected.
:
:T.
:
:
:T.
:
On my RCA K81 I used a 6E5 in place of the 6U5. I couldn't get the 6U5 to close right. The 6E5 is ok, sometimes overlaps on strong stations. The B+ is 275 and I guess I should knock it back sometime.

Doesn't the schematic for your tuner indicate 170V and taken with switch in phono? Maybe changing the .1 cap to .05 would give more deflection. I'm now thinking about using my 6U5 and adding about 1 meg to shunt my .05 cap to ground. Or experiment with different values. Magnavox design has 20 more years on mine. But come to think about it the shunting resistor has been around too.

Bill

Increasing or decreasing the capacitor size won't change deflection. It'll only speed up or slow down the rate at which deflection occurs (and also speed up or slow down the reaction time of the AVC circuit).

Also, the shunting resistor exists in all radios--usually it's the volume control. However, in this particular radio, there are many shunting resistors, all of which serve a different purpose. I assume that the low value (I believe that it's 250K, but I can't remember for sure) resistor in the AM detector circuit is to load down the AM so that it is closer in volume to the FM. It also serves to complete the detector circuit. If you look at the schematic you'll see that any other shunts are a way off from this resistance wise, and so there wouldn't be enough of a load on the detector. The extra shunts you find in the FM circuit are, of course, part of the ratio detector circuit, and are necessary. That these circuits are all connected to the eye tube at all times makes for quite a load on the circuit. A shunt of some sort is necessary so that the AVC capacitor will discharge when there's no signal, but all of these shunts reduce the signal to the eye tube significantly, and the extra 1.5Meg resistor across the AVC filter cap I find to be excessive.

T.

I made a new rivet for the discriminator transformer. Then I put a small dab of dielectric grease on each side of the silvered mica, and put the transformer base back together using the rivet I made. After checking capacitance for accuracy and stability, I soldered the coil wires back in place and put the transformer back in the radio. The FM performs well enough. Good selectivity for the most part. Could be more sensitive, but I haven't gone over it with a signal generator yet. Unfortunately the sweep generator I was given I have yet to repair. One station actually does deflect the magic eye. The rest don't. I will try a larger antenna. Also, at the high end one station dominates from about 104 to the end. Selectivity goes away there. Could be an alignment issue. I'll have to check it out.

It is too bad that manufacturers didn't originally coat the silvered mica capacitors with silicon grease. Could have eliminated a lot of headaches later, though with this set I strongly believe that my headaches were all due to a LOT of faulty solder connections, and not at all due to oxidized silver mica capacitors.

I am curious about the 1st FM RF coil. It consists of about 3 turns of stiff wire, and has one end soldered to the stator of that section of the tuning condenser. The other end snaps into a hole of the condenser frame by pressure. Unless it is pre-tensioned, it doesn't touch the metal frame well, if at all. What's up with that? Contact definitely affects reception.

T.