Maco 300 gets 47-year tuneup.

[Shockwave]

Gonna need a better way to screen customers' tubes before tearing into an amplifier. A tube tester just doesn't work them hard enough.

The older the tubes get the riskier they seem to get.

73

The most reliable sweep tube tester that I had was an old two tube, cathode driven 6LF6 amplifier. Pulled one tube socket to install a 9 pin 6LQ6 socket. Aligned the tank circuit so the plate tune could hit resonance on a single tube. Switching to a variac as the filament transformer, allowed testing of everything from 6LQ6, to 40KD6. With a little effort you could even add a third socket to test the 6KV6 pinout. Adding a decent glitch resistor prevented tube failures from damaging the "tube tester". You can test for breakdown voltage and emissions at RF frequencies, rather than DC.

Also, letting old tubes sit with just the filament voltage running for several hours, can help raise the breakdown voltage, and prevent internal tube arcs.

 

[nomadradio]

letting old tubes sit with just the filament voltage running for several hours,

We learned to do that to every old russky tube we bought. You can monitor the gas content by putting a 1 megohm resistor in line with a 100 Volt DC supply. The positive side goes to the cathode, the negative side to the grid. A voltmeter across the 1 meg resistor will read one Volt for every microamp of gas leakage current. With the soviet-era metal tubes you could see that reading drop by the hour. When it stopped dropping the tube was ready to try at full operating voltage.

73

 

[Shockwave]

We learned to do that to every old russky tube we bought. You can monitor the gas content by putting a 1 megohm resistor in line with a 100 Volt DC supply. The positive side goes to the cathode, the negative side to the grid. A voltmeter across the 1 meg resistor will read one Volt for every microamp of gas leakage current. With the soviet-era metal tubes you could see that reading drop by the hour. When it stopped dropping the tube was ready to try at full operating voltage.

73

The only Russian Surplus military tube amplifier that I have that still runs reliably, is one that I built for a pair of GI7bs, that required a GS35b to be installed in order to make it reliable. I understand that others have had a better experience with them. Mine were all extremely gassy. I even had one with so little vacuum left, the filament wouldn't light. It just sat there and drew a lot of current.

 

[Amperex]

When the ultra expensive 8950 tubes die, use cheap 6LB6 with two tubes each filaments wired in series. 6LB6 is same tube as 8950 with different filament voltage.

 

[nomadradio]

all extremely gassy

I guess we just lucked out. YMMV. And yeah, I did have one GI7B go kapow. The others worked okay.

73

 

[Shockwave]

It's in the way, blocks proper air flow, doesn't contribute enough additional power to justify blocking the way for a proper plate choke with one parasitic choke for each tube. The power lost by using four final tubes instead of five won't move the S-meter at the other end by more than the thickness of the meter needle.

Pretty much.

73

Forgive me for having a difference of opinion here however, in my mind it's not the extra power the 5th tube makes. It's removing 20% of the load off of the other four finals, that are already being pushed and essentially still receiving the same amount of drive power that was feeding five tubes.

I also understand that the layout in this amplifier sucks in every aspect from cooling to self oscillation. In other words, you have to work with what was available to start with. It's essentially a Palomar 300A now, with improved stability. You put a lot of time into this amplifier and it looks infinitely better than it did to start with.

 

[Shockwave]

When the ultra expensive 8950 tubes die, use cheap 6LB6 with two tubes each filaments wired in series. 6LB6 is same tube as 8950 with different filament voltage.

The 6LB6 is close to an 8950 however, it has 300 milliamps less peak cathode current and slightly less output capability as a result.

 

[Cable Guy]

The 3300uf filter is more than enough to smooth out the negative DC bias voltage. No more hum with the carrier turned down.

The rectifier diode in the center of the three has its cathode (banded end) connected to the 12 Volts AC heater supply The anode now feeds negative DC to the 3300uf filter cap. The top of the three diodes feeds bias to the grids of the driver tubes. Any increased "grid-leak" voltage appearing on this pin of each driver tube will stay here, and won't charge up the filter cap. It's called a "blocking" diode. The lower one does the same thing for the four final tubes.

Nomad, could tell me what part diode you chose for the "cathode to relay center pole" diodes and around the grid bias filter cap? I can't make them out from the pics, thanks!

 

[nomadradio]

The one on the final cathodes is a 6 Amp 1kV. Some brands call it a "6A100", others call it a "6A10". The one against the back panel carries the driver tubes' cathode current. A 1N5408 is fine for the two driver tubes. Might also be okay for the finals, but when overkill is cheap I tend to go for it.

73

 

[Monk]

So here is a rare treat. A Maco 300 with low miles. Upshot of this is that the owner has six good tubes. It's the 7-tube version, but no problem. I won't send a Maco 300 home with seven tubes in it. The fifth final tube comes out. More on that later.

Naturally each of the four final tubes will get its own, separate parasitic choke.


At least we know its birthday.

The two driver tubes each have a proper parasitic choke.

Fortunately only one wire has to get spliced when the center tube sockets gets removed.

You'll need a tiny plate to cover the empty tube socket hole and support the plate choke.

The new plate choke occupies the center of the four driver tubes. Provides a stable anchor to attach four parasitic chokes. This is the same choke used in the Palomar 300A and the Heathkit SB220. Honest.

So here is my pet peeve with Maco. They didn't ground the shield on this coax. It leads from the Load control to the output side of the relay.

The shield is just left loose at the front end, too.

The receiver preamp just goes away. The entire circuit board with the white relay on it.

Abracadabra! It's gone.

Just one problem. This board also contains the keying circuit. But not a problem for long. Got keying circuits. On hand.

This pic shows three issues fixed. The output coax is now teflon and has the shield grounded to boot. There is a tiny keying-circuit board visible under the radio jack and a rectifier diode has been inserted between each set of tubes and the relay's center pole.

Didn't shoot any "intermediate" pics before all three fixes were in place.

Yes, the output coax braid is grounded at the Load control, too.

There is a visible rectifier diode leading from the final tubes' cathode circuit to the center pole of the relay.

The cathode connection on the two driver tubes has a 3-Amp diode between it and the relay's center pole. The four finals have a 6-Amp diode between them and the relay. This serves to isolate the two separate cathode circuits. Maco simply tied them together at the relay. Sounds simple, but we had a recurring problem that a driver tube would "FLASH!" when you unkey. Not healthy for the tubes. Stumbled across this fix. Someone suggested it to me, but 35 years later I have no idea who to give credit. Long story short, this stopped the flash problem.

Yeah, the original hookup was causing some kind of oscillation, but I don't know what. A fix that works is where I move on to the next chore.

Oh, and speaking of oscillation. That's what leaving the braid on the output coax "floating" with no connection will cause. This turns the coax braid into an antenna, beaming unwanted feedback into the amplifier's input circuits. This is our preferred fix. Yes, this coax is skinny, but it's coax. The losses are low enough it will handle whatever four sweep tubes can deliver.

There are more chores remaining on this one, both upstairs and downstairs, but it was time for dinner.

More when it gets finished.

73

 

[Monk]

Nomadradio,

An edit is in order. Sorry not nit picking but.........
The new plate choke occupies the center of the four driver tubes.
You did mean "Driven tubes" not driver tubes, right?

Monkradio

 

[Monk]

The most reliable sweep tube tester that I had was an old two tube, cathode driven 6LF6 amplifier. Pulled one tube socket to install a 9 pin 6LQ6 socket. Aligned the tank circuit so the plate tune could hit resonance on a single tube. Switching to a variac as the filament transformer, allowed testing of everything from 6LQ6, to 40KD6. With a little effort you could even add a third socket to test the 6KV6 pinout. Adding a decent glitch resistor prevented tube failures from damaging the "tube tester". You can test for breakdown voltage and emissions at RF frequencies, rather than DC.

Also, letting old tubes sit with just the filament voltage running for several hours, can help raise the breakdown voltage, and prevent internal tube arcs.

shockwave, I did the same by building single tube amps for 811-572 tubes and also for 3-500z,400z, 4-400, 4-250, 4-125
Variac on the plate transformers to "Bring them up slowly".

Monkradio

 

[Cable Guy]

The one on the final cathodes is a 6 Amp 1kV. Some brands call it a "6A100", others call it a "6A10". The one against the back panel carries the driver tubes' cathode current. A 1N5408 is fine for the two driver tubes. Might also be okay for the finals, but when overkill is cheap I tend to go for it.

73

I have some 10a10's but they're huge so I ordered some of the 6a10's and 1n5408's. What are the 3 diodes around the grid bias filter cap, between the 12v filament supply, the 3300uf cap and the driver/final grids? Are those 1n5408? Thank you Nomad.

 

[nomadradio]

You did mean "Driven tubes" not driver tubes, right?

I meant 'final' tubes. Just a typo. Sometimes they sneak through.

The negative grid-bias diodes are just 1N5408. Overkill, but they're cheap when you buy enough of them. One serves to rectify the 12.6 VAC heater supply. The other two isolate each grid circuit from the fat filter capacitor. And from each other.

73