HME HM 150 Bi Linear Amplifier Question

[nomadradio]

Direct quote from Helge Granberg/Norman Dye's "Radio Frequency Transistors", page 194 first paragraph:

"In addition to gain reduction, negative feedback lowers the effective input impedance of the device(s)."

Not just my opinion/observation. He doesn't say "lowers" by how much, but it's usually enough to see on the input-side SWR in my experience.

73

 

[JB383RI]

You are experiencing what's called self oscillation of the amplifier. Where it will begin to operate as an oscillator and produce a second output on another frequency that has a higher reflected power.

When the oscillation is strong enough, it will sustain enough RF to keep the keying relay latched, after you unkey the radio. The issue causing this problem, is a lack of negative feedback.

This is easy to add by placing a .1uf cap in series with a 100 ohm 2 watt resistor and placing these across the base - collector junctions of both output transistors.

Thanks Shockwave. Just to be clear I did fix the problem where it was staying keyed up on its own it was a bad keying transistor. Are you saying the NFB Mod is for that problem or for the other problem where the reflect shoots up.

Also that problem seems to happen after I have been talking and the amp gets warm.

Thanks Nomad as well I appreciate all the info I can get.

 

[Shockwave]

Thanks Shockwave. Just to be clear I did fix the problem where it was staying keyed up on its own it was a bad keying transistor. Are you saying the NFB Mod is for that problem or for the other problem where the reflect shoots up.

Also that problem seems to happen after I have been talking and the amp gets warm.

Thanks Nomad as well I appreciate all the info I can get.

I was under the impression the keying relay only stayed latched, after you were transmitting and unkeyed the microphone. If that is the case, this is very likely related to a lack of negative feedback. Changing the transistor probably only made a difference because it has a lower gain and is not being held keyed when the amplifier produces an oscillation.

The increase in reflected power, can also be corrected with negative feedback. I see both problems as related, even though you are no longer seeing the first. If the keying transistor were bad, the most common result is the relay will not key, or it stays latched on power up, without ever keying the microphone.

While it is possible for a transistor to leak and develop this symptom, it is very unusual for a small signal transistor like this, to do that. They typically short, or short and then burn right open.

 

[JB383RI]

I was under the impression the keying relay only stayed latched, after you were transmitting and unkeyed the microphone. If that is the case, this is very likely related to a lack of negative feedback. Changing the transistor probably only made a difference because it has a lower gain and is not being held keyed when the amplifier produces an oscillation.

The increase in reflected power, can also be corrected with negative feedback. I see both problems as related, even though you are no longer seeing the first. If the keying transistor were bad, the most common result is the relay will not key, or it stays latched on power up, without ever keying the microphone.

While it is possible for a transistor to leak and develop this symptom, it is very unusual for a small signal transistor like this, to do that. They typically short, or short and then burn right open.

So I threw the transistor on a semiconductor tester and it tested as a diode. Once it was replaced it no longer would stay latched. If you think the NFB Mod will help with reflect I can still try it anyway. The weird part is I used this amplifier a lot and never had a problem until the relay stayed latched. Then the high reflect started. Replaced the transistor and the latch problem cleared up but still the reflect will shoot up after the amp has been on and in use for a bit. I would think if it was a NFB problem it would have had this problem from the beginning. I will try it anyway and see what happens.

 

[Shockwave]

Direct quote from Helge Granberg/Norman Dye's "Radio Frequency Transistors", page 194 first paragraph:

"In addition to gain reduction, negative feedback lowers the effective input impedance of the device(s)."

Not just my opinion/observation. He doesn't say "lowers" by how much, but it's usually enough to see on the input-side SWR in my experience.

73

I wasn't suggesting that there will not be a measurable reduction in input impedance, with the addition of negative feedback. I was simply saying that I have never seen anything close to a reduction from 50 ohms, to 16.5 ohms, which would indicate the 3:1 change, that lowers impedance.

My main point was this amplifier already has a compression variable capacitor across its input and that it should have enough range to bring the impedance back into alignment. Does your experience suggest otherwise?

 

[nomadradio]

Hey, it's all in the result. If you turn the trimmer screw all the way to one extreme without obtaining a null in the reflected power, this suggests that fixed components have not been chosen for an optimal result. Likewise if you see a pronounced peak in output at one setting with a high reflected, and see a dip in reflected input at a different setting but with reduced peak power this suggests a stability issue in the making.

What you will or won't need to tweak further is something you won't know until you twist the adjusting screw and observe the result.

73

 

[JB383RI]

Hey, it's all in the result. If you turn the trimmer screw all the way to one extreme without obtaining a null in the reflected power, this suggests that fixed components have not been chosen for an optimal result. Likewise if you see a pronounced peak in output at one setting with a high reflected, and see a dip in reflected input at a different setting but with reduced peak power this suggests a stability issue in the making.

What you will or won't need to tweak further is something you won't know until you twist the adjusting screw and observe the result.

73

So funny you should say that. When I see the lowest amount of reflected power along with the highest amount of output is when the trimmer screw is turned all the way clockwise.

 

[Shockwave]

So funny you should say that. When I see the lowest amount of reflected power along with the highest amount of output is when the trimmer screw is turned all the way clockwise.

The good news is, this VSWR is actually likely to drop lower with the addition of negative feedback. When your trimmer capacitor is fully clockwise, it is adding the maximum capacitive loading to the input and reducing the input impedance. That is the direction that is producing a lower input VSWR for you, and it is also the same direction that adding negative feedback will produce.

Another thing to keep in mind, this VSWR will shift depending on which position the clarity switch is in. In order to see the actual input to the amplifier, the clarity switch would have to be in full power. These switches tend to screw up the input impedance because they are not true 50 ohm attenuators.

They are simply extra resistance added in line with the input. This extra resistance causes the impedance to rise above 50 ohms and is likely why this amplifier is tuning with that capacitor fully compressed. That's assuming we are not in the full power position.

If it still does not have enough range on the variable capacitor after negative feedback is added, all you have to do is add another 50 or 100 picofarads in parallel with that trimmer, and it should pull into alignment.

 

[JB383RI]

The good news is, this VSWR is actually likely to drop lower with the addition of negative feedback. When your trimmer capacitor is fully clockwise, it is adding the maximum capacitive loading to the input and reducing the input impedance. That is the direction that is producing a lower input VSWR for you, and it is also the same direction that adding negative feedback will produce.

Another thing to keep in mind, this VSWR will shift depending on which position the clarity switch is in. In order to see the actual input to the amplifier, the clarity switch would have to be in full power. These switches tend to screw up the input impedance because they are not true 50 ohm attenuators.

They are simply extra resistance added in line with the input. This extra resistance causes the impedance to rise above 50 ohms and is likely why this amplifier is tuning with that capacitor fully compressed. That's assuming we are not in the full power position.

If it still does not have enough range on the variable capacitor after negative feedback is added, all you have to do is add another 50 or 100 picofarads in parallel with that trimmer, and it should pull into alignment.

Thanks Shockwave I will try it and see what happens.

As far as the type of capacitor for NFB as well as if I need to put one in parallel with the trimmer should they be Silver Mica, Ceramic Disc or Electrolytic?

Also as far as the NFB goes does it matter in which the order of the cap and resistor are connected in series to the base and collector? For example the cap connected to the base and resistor to the collector or vice versa.

 

[Shockwave]

Thanks Shockwave I will try it and see what happens.

As far as the type of capacitor for NFB as well as if I need to put one in parallel with the trimmer should they be Silver Mica, Ceramic Disc or Electrolytic?

Also as far as the NFB goes does it matter in which the order of the cap and resistor are connected in series to the base and collector? For example the cap connected to the base and resistor to the collector or vice versa.

The negative feedback cap can be a simple .1 uf ceramic disk. Generally the cap is connected to the collector and the 100 ohm 2 watt is connected to the base. Capacitors that must hold a precise value in tuned RF cicuits like in parallel with the input, should be dipped silver mica.

If you can read a part number on the side of the variable compression cap, I can give you a value for an appropriate fixed value to go in parallel.

 

[JB383RI]

The negative feedback cap can be a simple .1 uf ceramic disk. Generally the cap is connected to the collector and the 100 ohm 2 watt is connected to the base. Capacitors that must hold a precise value in tuned RF cicuits like in parallel with the input, should be dipped silver mica.

If you can read a part number on the side of the variable compression cap, I can give you a value for an appropriate fixed value to go in parallel.

Ok thanks. Its a 466 trimmer, from what I could find it looks like a 105pf 175V mica trimmer.

 

[Shockwave]

Ok thanks. Its a 466 trimmer, from what I could find it looks like a 105pf 175V mica trimmer.

The 105 is its lowest limit, set fully counterclockwise. It is 480 picofarads fully clockwise. That is more padding than I expected to be in this location. Therefore, you will need a larger fixed value in parallel with it.

Something between 200 and 300 picofarads should find your best match. You don't want to go too large otherwise you will reverse the problem and be stuck at minimum capacitance and still not hit the optimum value.

 

[Shockwave]

The more I think about this, the more I think we should get another opinion. In my mind, requiring more than 480 pf across the primary of this input transformer, is a strong indication of a mismatch in the turns ratio or, capacitive loading across that transformers secondary. How many wraps are on this transformer and what is the value of the large dipped silver mica cap, soldered to the top of it?

I'll admit, I'm not all all familiar with this specific RF transistor however, something seems wrong with the value of required primary loading and is why I would like input from others on this one.

 

[linearone]

People make a big deal about turns ratio and capacitors on these transformers. Yeah they all make a difference but remember you're just making a handshake. If it tunes out it tunes out. You'll notice differences but many combinations will work

 

[nomadradio]

I'll admit, I'm not all all familiar with this specific RF transistor

Solid State Systems, the "SSS" on the part was bought by maybe TRW? Then that operation was spun off to somebody else. Can't say I ever remember seeing a data book for their RF transistors. Pretty sure this dates it to around 1980 or so.

If they're like the other brands' transistors of the day, they'll take scandalously little drive power before they saturate.

The knob marked "clarity" is labeled backwards, if it's actually a "Low/Med/High" control. Highest "clarity" number would be at the lowest drive level, not the highest.

And if it's actually wired backwards, highest power on the left, it's labeled correctly. Overdriving a linear that's over 40 years old will be easier than it looks. Mobile radios are a lot meaner now than they were 45 years ago.

73