The forum software allows only 20 images per post. This one will get split up.
The Palomar 300A gets older every year. Hasn't been made since 1978 or '79. But the customer says he wants it fixed, and he'll come up with the tubes. That's the magic combination for us, so he got his wish.
But not without a bump or two along the way. First rule of the Palomar 300A:
Assume nothing.
This one has the plug-in circuit board with three discontinued relays on it. Wasn't that big a problem, since I located a source for the R10-E3604 RF relay. The pins are in a pattern not like any other cradle relay of this size. It has a grounded metal shield between the contact array and the coil. That's a good thing for a RF circuit. But not so much for the poor schmuck who needs to buy them. Far as I know, it hasn't been made in decades. Lucked out and found a few. Always a good idea to clean the contacts on old-stock relays. Just a strip of copy paper and 99% isopropyl. This is a preventative measure, but worth the small effort before soldering them into the pc board. We replaced only the two outside relays on the board. The center one is the receiver preamp. Left that out of this repair.
If you're going to skip the preamp it needs to be bypassed. All it takes is a solder bridge across two adjacent foil traces. If you find the two sets of three pads on the missing center relay, follow the trace leading from the center pad of the three. This will lead you to the receive side of the antenna relay on the right of the pic. That's where we bridge across them.
Next issue is filter caps. The big HV filter caps had already been replaced, so we didn't need to. But there's another factory quirk that requires attention, the bleeder resistors. In this case, you need one on each filter cap, mainly to insure that they divide the voltage equally between them. This is easier than it sounds. They will mount just fine on the lead wires of the caps as you see here. No need to remove this pc board.
Not yet, anyway.
So here is a question I get from time to time. What are the resistance values of the parts on this board, providing negative grid bias to the tubes. I scribbled the value in Sharpie (tm) on the unmarked ones. They run hot enough to burn away the markings after the first couple of decades. The smaller 2-Watt used to be 27 ohms, but the 33 ohm you see was as close as I could come. Seems to work just fine. This board is NOT found in any of the 300A schematics circulated on line. The later version with two filter caps is what you'll see posted for download, not this one.
Here is a quirky detail you may decide to skip. The 16 (12) Volt DC power that runs the relays is fed into the screen grids of the two driver tubes. Pins 3 and 11 of each driver socket were originally soldered to ground. Here you see that they have been isolated, a bypass cap added, and a jumper wire joining them between the two driver sockets. The gray wire feeds 16 Volts DC from the wire that powers the relay circuit board. First thing we do is add a 10k 2 Watt resistor to ground from pin 11 of the front driver tube. Really doesn't matter which of the four pins it's connected to. This one is convenient.
The gray wire comes from the edge connector socket. We unhook it and insert a 1N4007 rectifier diode. This serves to protect the low-voltage circuits from any surges that might occur inside a driver tube. Trust me, several hundred Volts is not kind to the keying circuit and relays.
The next item of interest is to remove the ham-band coils in the plate circuit of the driver and final tubes. That, and to add parasitic chokes to the plate connection of each of the four final tubes.
If your 300A has only a bare wire between the plate cap of each tube and the plate choke in the center you'll need to add these. The factory slid a couple of black ferrite beads over each plate-cap wire at the factory. But the four final tubes run too hard for those beads. They get hot, crack, fall off and nobody ever knows they were ever there. Remove the top and the bare wires look like they belong there. Not exactly.
You don't need pics to remove the driver coil and all the wires that lead away from it. This coil is wound on a 3/4" form. As you see it, the driver plate tune resonated at around 1/3 mesh.
Pretty close, but the target is to make the tune capacitor peak as close to one-half mesh as possible. Nothing magic about having the control peak at its midpoint, but there's a strategy here. If the driver tubes get changed, this peak position will most likely change. If we send it home peaking in the middle, this provides leeway in both directions. Might prevent the need to molest the coil later on. Might.
Likewise, if the operator wants to use it above or below the 40 channels, this provides leeway in both directions, peaking at the center on channel 20.
The final plate coil gets replaced with fatter wire. Just one problem. The Load control doesn't quite look right.
The sharp-eyed reader will see that the plates are spaced exactly the same on the final Load control as the final Plate Tune control. Definitely not right.
On a hunch, I paralleled two disc caps on the Load control, to add about 140 pf to it. Was almost enough.
Here's what the factory final Load cap in another 300A looks like. Lots more plates, much closer together. And a lot more capacitance.
Took off the Load knob to change it out, and this explains it all. Missing screws from the switcheroo. Oops.
A word about how the 300A is constructed. They used machine screws in two sizes. Short and extremely short. Here is a screw from the final Load control.
The screws that attach the control to the front panel need to be short so they won't interfere with the capacitor's rotating parts. Other very-short screws are sunk into very short threaded spacers, holding down circuit boards. A screw that's too long bottoms out before the shoulder of the screw head can touch the board it's meant to to fasten down.
I'm incredibly fond of Philips head screws. I won't put the original slot-head screws back into the amplifier.
Here is a tool that has paid for itself so many thousands of times over it's hard to fathom. Shortens a screw down to 1/8 of an inch, or more if you need it. Don't know what we would do without it.
For reference, here's the max capacitance of the bogus final Load control.
Here's the new Load control. Only has two sections, but it has four times the capacitance of the control we found in it.
The sharp-eyed reader will spot the so-called "safety" choke. The small black RF choke from the Load control's hot side to ground. This serves to keep any stray DC voltage across the control's plates at zero. This way, the gap between the plates only has the RF voltage across it. Without this choke some of the power supply's 800 Volts DC gets divided across the plates of the Load control. No joke. Tends to make the air gap break down and arc, with the two voltages DC plus RF across the air gap. I'll guess that the factory load control went bad just this way from arcing over, and someone thought a control with a wider air gap would "fix" the problem. Nice try, no cigar.
To be continued.
The Palomar 300A gets older every year. Hasn't been made since 1978 or '79. But the customer says he wants it fixed, and he'll come up with the tubes. That's the magic combination for us, so he got his wish.
But not without a bump or two along the way. First rule of the Palomar 300A:
Assume nothing.
This one has the plug-in circuit board with three discontinued relays on it. Wasn't that big a problem, since I located a source for the R10-E3604 RF relay. The pins are in a pattern not like any other cradle relay of this size. It has a grounded metal shield between the contact array and the coil. That's a good thing for a RF circuit. But not so much for the poor schmuck who needs to buy them. Far as I know, it hasn't been made in decades. Lucked out and found a few. Always a good idea to clean the contacts on old-stock relays. Just a strip of copy paper and 99% isopropyl. This is a preventative measure, but worth the small effort before soldering them into the pc board. We replaced only the two outside relays on the board. The center one is the receiver preamp. Left that out of this repair.
If you're going to skip the preamp it needs to be bypassed. All it takes is a solder bridge across two adjacent foil traces. If you find the two sets of three pads on the missing center relay, follow the trace leading from the center pad of the three. This will lead you to the receive side of the antenna relay on the right of the pic. That's where we bridge across them.
Next issue is filter caps. The big HV filter caps had already been replaced, so we didn't need to. But there's another factory quirk that requires attention, the bleeder resistors. In this case, you need one on each filter cap, mainly to insure that they divide the voltage equally between them. This is easier than it sounds. They will mount just fine on the lead wires of the caps as you see here. No need to remove this pc board.
Not yet, anyway.
So here is a question I get from time to time. What are the resistance values of the parts on this board, providing negative grid bias to the tubes. I scribbled the value in Sharpie (tm) on the unmarked ones. They run hot enough to burn away the markings after the first couple of decades. The smaller 2-Watt used to be 27 ohms, but the 33 ohm you see was as close as I could come. Seems to work just fine. This board is NOT found in any of the 300A schematics circulated on line. The later version with two filter caps is what you'll see posted for download, not this one.
Here is a quirky detail you may decide to skip. The 16 (12) Volt DC power that runs the relays is fed into the screen grids of the two driver tubes. Pins 3 and 11 of each driver socket were originally soldered to ground. Here you see that they have been isolated, a bypass cap added, and a jumper wire joining them between the two driver sockets. The gray wire feeds 16 Volts DC from the wire that powers the relay circuit board. First thing we do is add a 10k 2 Watt resistor to ground from pin 11 of the front driver tube. Really doesn't matter which of the four pins it's connected to. This one is convenient.
The gray wire comes from the edge connector socket. We unhook it and insert a 1N4007 rectifier diode. This serves to protect the low-voltage circuits from any surges that might occur inside a driver tube. Trust me, several hundred Volts is not kind to the keying circuit and relays.
The next item of interest is to remove the ham-band coils in the plate circuit of the driver and final tubes. That, and to add parasitic chokes to the plate connection of each of the four final tubes.
If your 300A has only a bare wire between the plate cap of each tube and the plate choke in the center you'll need to add these. The factory slid a couple of black ferrite beads over each plate-cap wire at the factory. But the four final tubes run too hard for those beads. They get hot, crack, fall off and nobody ever knows they were ever there. Remove the top and the bare wires look like they belong there. Not exactly.
You don't need pics to remove the driver coil and all the wires that lead away from it. This coil is wound on a 3/4" form. As you see it, the driver plate tune resonated at around 1/3 mesh.
Pretty close, but the target is to make the tune capacitor peak as close to one-half mesh as possible. Nothing magic about having the control peak at its midpoint, but there's a strategy here. If the driver tubes get changed, this peak position will most likely change. If we send it home peaking in the middle, this provides leeway in both directions. Might prevent the need to molest the coil later on. Might.
Likewise, if the operator wants to use it above or below the 40 channels, this provides leeway in both directions, peaking at the center on channel 20.
The final plate coil gets replaced with fatter wire. Just one problem. The Load control doesn't quite look right.
The sharp-eyed reader will see that the plates are spaced exactly the same on the final Load control as the final Plate Tune control. Definitely not right.
On a hunch, I paralleled two disc caps on the Load control, to add about 140 pf to it. Was almost enough.
Here's what the factory final Load cap in another 300A looks like. Lots more plates, much closer together. And a lot more capacitance.
Took off the Load knob to change it out, and this explains it all. Missing screws from the switcheroo. Oops.
A word about how the 300A is constructed. They used machine screws in two sizes. Short and extremely short. Here is a screw from the final Load control.
The screws that attach the control to the front panel need to be short so they won't interfere with the capacitor's rotating parts. Other very-short screws are sunk into very short threaded spacers, holding down circuit boards. A screw that's too long bottoms out before the shoulder of the screw head can touch the board it's meant to to fasten down.
I'm incredibly fond of Philips head screws. I won't put the original slot-head screws back into the amplifier.
Here is a tool that has paid for itself so many thousands of times over it's hard to fathom. Shortens a screw down to 1/8 of an inch, or more if you need it. Don't know what we would do without it.
For reference, here's the max capacitance of the bogus final Load control.
Here's the new Load control. Only has two sections, but it has four times the capacitance of the control we found in it.
The sharp-eyed reader will spot the so-called "safety" choke. The small black RF choke from the Load control's hot side to ground. This serves to keep any stray DC voltage across the control's plates at zero. This way, the gap between the plates only has the RF voltage across it. Without this choke some of the power supply's 800 Volts DC gets divided across the plates of the Load control. No joke. Tends to make the air gap break down and arc, with the two voltages DC plus RF across the air gap. I'll guess that the factory load control went bad just this way from arcing over, and someone thought a control with a wider air gap would "fix" the problem. Nice try, no cigar.
To be continued.
Last edited:
