Valve receiver Kenwood Trio KW-55 repair

[SpecialK]

I bought an old tube receiver that needs some love.  A couple of the pictures in the ad show what looks like red plating of one of the EL84 output valves.  So I figured I would go over the unit before powering it up.

As expected a lot of paper caps need replacing.  If you look at the schematic C315, is across the AC line and this is just a paper cap 0.02uF.  So should this replaced with a regular film cap or an X2 safety cap?

Also, the R308 330 ohm resistor isn't present.  R309 is present but has a 2200uF 15 electrolyic across it.  This seems a bit suspicious to me, as it would probably affect the apparent cathode bias.  Could this explain the replating alone?  It does look like the grid capacitors have been replaced with non-polar electrolytics many years ago, one is an orange drop. 
 
 

[Circlotron]

An electrolytic cap is far too leaky to be used as a grid coupling cap!
Schematic please, so we can see what you are talking about.

[SpecialK]

Sorry, I don't know why it wasn't attached in the first post.

But yeah, I tested the three non-polar electrolytics anf the orange drop with 60V applied I was getting around 50uA of leakage each.

[TimFox]

Are you sure you have NP electrolytics at C120, C121, C220, and/or C221?  Ordinarily they would be paper or film capacitors at 50 nF.  Note that "orange drop" does not specify the dielectric.  The CDE/Vishay/SBE/Sprague 715P Orange Drop capacitors are polypropylene film-foil and should work in those locations.  I'm not aware of NP electrolytics as small as 50 nF.  Perhaps a previous owner tried to improve things by increasing the capacitance?  50 nF and 470k give a corner frequency of 6.8 Hz, so it might make sense to increase the capacitance to, say, 220 nF, but that might affect the stability of the feedback loop and should be tried carefully.  Red-plating of the EL84s could indicate excess leakage of the C120, etc. grid capacitors, or a shorted cathode bypass C134 (shown on the diagram as 40 uF (common to both pairs).  You shouldn't get 50 uA of leakage on a film capacitor at below its rated voltage.  The 715P datasheet shows > 2 x 10¹⁰  at 85 C, or < 5 nA at 100 V (20 times less leakage at 25 C).
Note the old trick of feeding the cathode current through the heaters of the preamp tubes V14 and V15 for cheap DC.  A problem there could also cause red-plating by decreasing the negative bias (grid to cathode).

[BrokenYugo]

If you want to be good and proper, yes, always use the appropriate safety caps in across the line and line to chassis positions.

I think some of the early "Orange Drops" (among other dipped radial lead film caps, like the red/brown ElMenco caps) were actually "difilm" (mixed dielectric, paper and mylar), I've never personally tested one at voltage, but they supposedly don't leak/short. Were these tested with at least one lead desoldered?

[Circlotron]

Unusual that V16 should be run off a 5V winding on the tranny, not 6.3V like most of the others. Or does V16 mean something different?

[TimFox]

I also don't see any explicit wiring to V16 heaters (and almost no pin numbers for the other heaters explicitly drawn).  I first thought that maybe the transformer already existed for a 5AR4 rectifier circuit, and was kept for the silicon rectifier version, but the high-voltage winding uses a voltage doubler, but it is only rated at 0.3 A (as required for one 12AX7 with parallel heaters).  It was probably done to reduce hum in V16, the first tube in the power-amp section, with the hum-balance pot, just as the DC feed to the preamp tube heaters (most sensitive to hum).  The only good reason I can think of to use 5 V on a 12AX7 heater would be to reduce grid current at only 0.9 V grid bias. 

[SpecialK]

Yeah.  I will take a couple pictures of those capacitors after dinner.  I think the 50uA leakage was actually at 30V.


I didn't fully understand how the 12AX7 heaters were being fed.  Pin 9 was vacant, which I think means there must be 12.6 being fed to pins 4 & 5.  I don't see a 12.6V winding so maybe there's a double there?

[SpecialK]

Red plate from listing.

[SpecialK]

Those blue aluminum cans are connected to the EL84 grids.  Not my pic, I will take a better one.

[TimFox]

Pins 4 and 5 of a 12AX7 are the extreme ends of the heater, and would normally require 12.6 V.  The center-tap is pin 9, and 6.3 V operation is between pin 9 and pins 4 and 5 tied together.  (Pin 9 is left open for 12.6 V operation.)  Perhaps, the 5 V indication on the wiring diagram is a typo?  A 12AX7 might operate with 5 V on the 6.3 V connection, but almost certainly not on the 12.6 V connection.
What are the markings on the blue grid capacitors?

[SpecialK]

[SpecialK]

This was across the cathode resistor R309

[Circlotron]

This was across the cathode resistor R309

Shouldn't be a problem, but I'd rather see it in place of C314.
Sudden thought - check whether the 470K grid leak resistor of the hot valve hasn't gone open cct. That would prevent that valve getting any bias voltage, and also wouldn't pull the grid down to the correct voltage after the preceding anode voltage rises at power-on.

[SpecialK]

Well I had it on the bench for a while tonight, but I've put it away now.

I ran the unit up slowly on the variac minding the current.  I got it up to full line voltage, and with only three working bulbs and without tubes it was consuming about 28W.  Seems a tad high.

I measured the voltage on the EL84 side of those caps and found one had 20V!  That one corresponds to the red plated tube in the pictures.

[SpecialK]

So my findings are that JCP caps (look like electrolytics?) should be replaced with the supply filter electrolytics.  I should get a bunch of resistors too, the carbon comp ones are generally reading high.  The dogbones are great!

The switchgear seems good, not stiff nor spongy.  Looks good on inspection.   Generally it's pretty clean, no corrosion.  It has 18 of the 22 tubes.  The unloaded B+ was ~ 380VAC, and the EL34 heater supply was a little under 7VAC at the socket.  The output transformers pass signal too.   So it seems to have good bones.

[floobydust]

It looks like an audiophool was doing mods for better sound? Like putting in boutique Jensen coupling capacitors and adding parts. I laugh at the 2,200uF Nichicon bung ready to explode, mounted on top of a hot power resistor 

Check the 0.05uF caps "pure copper foil" paper in oil? for leakage or shorts, the four to the O/P tubes are critical and one might be why one tube is cooking.

What I would do is confirm each output tube has proper bias and the group (cathode) has the -21.8VDC. Just don't run it for long.
Note the O/P tubes cathode currents generate DC filament voltage for the phono/tape preamp tube. It's a delicate balance and if you don't have proper currents there the O/P tubes bias will be wrong, as a group.

[SpecialK]

It looks like an audiophool was doing mods for better sound? Like putting in boutique Jensen coupling capacitors and adding parts. I laugh at the 2,200uF Nichicon bung ready to explode, mounted on top of a hot power resistor 

Check the 0.05uF caps "pure copper foil" paper in oil? for leakage or shorts, the four to the O/P tubes are critical and one might be why one tube is cooking.

What I would do is confirm each output tube has proper bias and the group (cathode) has the -21.8VDC. Just don't run it for long.
Note the O/P tubes cathode currents generate DC filament voltage for the phono/tape preamp tube. It's a delicate balance and if you don't have proper currents there the O/P tubes bias will be wrong, as a group.

Yeah, I did notice that electrolytic was on the hot resistor.  My guess is that 16V isn't the right value either.  I mean if you have the cathode bias currents of four tubes running through a 270 ohm resistor, the voltage must run more than 16V.  100ma x 270 ohms = 27V on the low side.  Might be twice that much current?  I don't know if I'm doing that right.  So yeah it's coming out and I will add the 330 ohm back as per schematic.  I think.

Any way, it's fun to tinker I suppose.  My project EL34 amp sounds pretty good playing music while I type this.




 

[TimFox]

Yeah, those blue capacitors are fancy oil-filled units with fancier electrodes.  If your 50 uA measurement at 30 V is correct, that would be a huge bias shift into the 470 k grid-to-ground resistors (theoretically, 25 V), while a proper 715P at much less than 5 nA would be only <<2.5 mV bias shift.  Those fancy capacitors, however, should not have anywhere near 50 uA leakage, so they may be objectively faulty.  Note that the heater voltages of the preamp tubes are part of the cathode bias circuit for the output tubes, and must be included in the calculation.  This was a common practice back then:  I first encountered it with a Fisher amplifier, ca. 1967.

[SpecialK]

Alright.  I'm looking at components.  I think I like carbon film resistors as a mostly aesthetic purpose.  I don't think baby blue resitors look right and are hard to read.


I'm not sure if I like the yellow Mallory style axials or the orange drop types.  The Mallory probably fits better, but some of the brownish "drop" capacitors would probably be a little more tastefully subdued.  The yellow seems to scream attention and look "modded" to me.

 

[mvno_subscriber]

So happy to see others annoyed at the neon signal color of many new components! 🤠

If you have the patience, printing the capacitor value on brown paper (or better yet, stamping it using a stamp with interchangeable letters), then wrapping it around the original part works well. Or use paper tubing, or better yet - melt the wax of the old ones, rip out the guts and stuff the new one inside. However, that last one might take too much time and patience.. good luck!

[SpecialK]

Note that the heater voltages of the preamp tubes are part of the cathode bias circuit for the output tubes, and must be included in the calculation.  This was a common practice back then:  I first encountered it with a Fisher amplifier, ca. 1967.

Tim, what is the purpose of this?  I'm assuming that giving the phono preamp tube heaters a DC supply minimizes the hum.  Wouldn't there be a bunch of tubes in the radio section that should be even more sensitive? 

If I don't plan on running the phono preamp tubes, do you suppose I should jumper 82 ohm 5W resistors across pins 4 & 5 of each socket?   

[TimFox]

The tubes in the radio section do not have high gain at 60 or 120 Hz.  However, in an RIAA preamplifier, the gain at low frequencies is very high, falling off above 50 Hz.
If you do not install the preamp tubes, you will need to insert appropriate resistors.  The 12AX7s each pull 150 mA at 12.6 V or 300 mA at 6.3 V, depending on their wiring.  82 ohms/5 W for each tube would work.