The AM Forum

THE AM BULLETIN BOARD => Technical Forum => Topic started by: KA3EKH on February 17, 2020, 01:19:37 PM



Title: Weird Osc problem
Post by: KA3EKH on February 17, 2020, 01:19:37 PM
Have been playing around with this oscillator circuit in my MAK transceiver for a while now and have been having an issue or two. The problem is that the MAK is intended for use in the field powered by a car battery. I do not have the original power supply that provided +175 to the receiver and the TX oscillator and built up a solid state supply for both the B+ and the HV for the PA and modulator.
I am running around 150 volts for the B+ and with 13.5 volts on the input of the power supply the radio works without issue but below 11.5 or so the B+ drops down to 125 volts and the transmitter oscillator will not start every time. At 10.5 the oscillator fails to start at all.
This radio has a crystal in the receiver also and the receiver works without issue down to 10 volts although it’s not as loud it still works. The issue is with the transmitter.
The question is am I dealing with a B+ issue or something else? At first I thought it was B+ related but not one hundred percent of that. The original transmitter oscillator used huge primitive crystals and now I am using modern rocks and am wondering if this is not a crystal capacitance issue that causing the oscillator not to start under less than ideal conditions?
I am attaching a copy of the schematic of the oscillator. It’s a 12A6 tube with a grounded cathode and a tuned resonance circuit in the plate that directly feeds the PA tube grid. There is no directly visible feedback path at least that I can see but I have seen this same approach used before in prewar and ww2 design. Did designers back then count on the added capacity of the huge primitive rocks and the modern crystals just don’t provide that? Would a small amount of capacity across the crystal help it start osculation? Or is this just an issue due to the reduced B+ voltage?
Or maybe it’s something completely different like reduced filament voltage cuts the gain of the tube below the point of self-exciting?
An additional item of interest is that the PA plate tuning affects oscillation. If I tune dead center on the plate tank dip the oscillator will not fire up the next time I key the microphone but by tuning a little on either side of the dip the oscillator works just fine, except the plate voltage issue.
Maybe when you consider the PA and its tank that is what is kicking this thing into oscillation in kind of some weird form of Armstrong oscillator? Although there is tons of shielding between the PA tank and the oscillator tank.





Title: Re: Weird Osc problem
Post by: K8DI on February 17, 2020, 02:17:56 PM
I am attaching a copy of the schematic of the oscillator. It’s a 12A6 tube with a grounded cathode and a tuned resonance circuit in the plate that directly feeds the PA tube grid. There is no directly visible feedback path at least that I can see but I have seen this same approach used before in prewar and ww2 design.

The feedback path is through the “ground” connection. The screen is effectively the plate of the oscillator, and the bypass cap creates an rf path from the screen (‘plate’) through the crystal to the grid.

You might try putting the tube in a tester and deliberately setting the filament voltage control low to see how the tube holds up at ten volts filament. It could just be weak.

Ed


Title: Re: Weird Osc problem
Post by: DMOD on February 17, 2020, 08:59:57 PM
Have been playing around with this oscillator circuit in my MAK transceiver for a while now and have been having an issue or two. The problem is that the MAK is intended for use in the field powered by a car battery. I do not have the original power supply that provided +175 to the receiver and the TX oscillator and built up a solid state supply for both the B+ and the HV for the PA and modulator.
I am running around 150 volts for the B+ and with 13.5 volts on the input of the power supply the radio works without issue but below 11.5 or so the B+ drops down to 125 volts and the transmitter oscillator will not start every time. At 10.5 the oscillator fails to start at all.
This radio has a crystal in the receiver also and the receiver works without issue down to 10 volts although it’s not as loud it still works. The issue is with the transmitter.
The question is am I dealing with a B+ issue or something else? At first I thought it was B+ related but not one hundred percent of that. The original transmitter oscillator used huge primitive crystals and now I am using modern rocks and am wondering if this is not a crystal capacitance issue that causing the oscillator not to start under less than ideal conditions?
I am attaching a copy of the schematic of the oscillator. It’s a 12A6 tube with a grounded cathode and a tuned resonance circuit in the plate that directly feeds the PA tube grid. There is no directly visible feedback path at least that I can see but I have seen this same approach used before in prewar and ww2 design. Did designers back then count on the added capacity of the huge primitive rocks and the modern crystals just don’t provide that? Would a small amount of capacity across the crystal help it start osculation? Or is this just an issue due to the reduced B+ voltage?
Or maybe it’s something completely different like reduced filament voltage cuts the gain of the tube below the point of self-exciting?
An additional item of interest is that the PA plate tuning affects oscillation. If I tune dead center on the plate tank dip the oscillator will not fire up the next time I key the microphone but by tuning a little on either side of the dip the oscillator works just fine, except the plate voltage issue.
Maybe when you consider the PA and its tank that is what is kicking this thing into oscillation in kind of some weird form of Armstrong oscillator? Although there is tons of shielding between the PA tank and the oscillator tank.


You can try these three things 0NE at a time.

1) Reduce the capacitance from the screen grid to B+. It is usually 100 pF or so but may need reducing.

2) Place a 33 pF capacitance across the 50k grid resistor.

3) Increase the grid resistance to 68k or higher.


Phil - AC0OB


Title: Re: Weird Osc problem
Post by: WU2D on February 17, 2020, 09:19:48 PM
I have an idea but it is a gimmick. No it is a gimmick. Solder a piece of # 22 solid insulated wire to the grid about 2 inches long. Wrap it around the wire going to the plate with about 6 turns.
AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands