482 Afterburner Plus-Too Much Drive Power Damage

[Dmans]

Update/delay to Grid Bias & Input tune mod on this little unit.
(Moving this to a new thread/Title may generate some much needed help)

A little background on this unit first. I picked this up for next to nothing a couple years ago (advertised as working with normal output-YEAH RIGHT!) I went through the unit and replaced the filter capacitors, the rectifier diodes, bypass capacitors in the power supply, RF chokes, SO239 connectors (Lightning damaged I think) all internal coax jumpers, cleaned the switches and replaced the pilot lamp and Tubes. It has worked flawlessly(150w peak on SSB) since then.

It seems I inadvertently (Senior Moment!) left this unit inline while using my Kenwood on sideband (about 60 watts peak) and caused some issues. The amp was turned off at the time but RF was passing through.
(Note to self-GET A COAX SWITCH!)

After a few minutes I smelled that familiar electrical burning smell. Opening the unit, I found resistor R5-330 ohm burned. Replacement yielded no relay engagement of the amp when driven with 3 watts dead key. Modulating the 3 watts, the relay engaged but buzzed and did not stay engaged. Sideband was the same result. When engaged, amp output was normal. Further attempts at repair included testing Q1 (Tested BAD!) and replacement of Q1-Ditto on Q2. (Replacements were NTE 159 for Q1 and NTE 123A for Q2) Replacement of C20, R6 and R7. C21,C22 and C26 were also replaced with same results. New relays were tried with same results.
Currently the amp will power up, the Pre-Amp works and K1 relay energizes with SSB input but does not hold in.(Tried larger value(1000uf@35v versus 500uf@15v stock) C22 with same result). K1 relay will engage with AM input but only when modulated. It seems that input wattage (3w) is not enough to make Q1 and/or Q2 conduct (double checked value of C20-all is per original design-.0023uf) but holding K1 engaged is a separate issue I think? I am measuring 17vdc at the + side of C21 at rest and about 10Vdc when relay is energized. Just to be sure current was not falling, I replaced D3 with a 1N5054 I had on hand, results were the same.

Is it possible RFC 3 is damaged from high RF current? (I don't currently have an extra for replacement)
I have not yet replaced C23 but it occurs to me that high RF current could have taken this out as well??

I am running out of ideas with this one. Any and all help is appreciated.
73's
David

 

[wavrider]

measure value of R5.
or jump across R5 and see if it keys as it should.

 

[Dmans]

wavrider,
I measured 327 ohms across R5 in circuit with 2 different meters. I jumpered across R5 and……..wait for it……….PROPER OPERATION!

Now several questions come to mind-
1. Does there need to be a some resistance before C20 as a measure of protection?
2. Does the value of the resistor determine at what RF input the relay energizes?
3. Will leaving this resistor out (and installing a jumper in its place) have a detrimental effect on the amp in the long run?

I have studied several schematics of "Classic" CB amps and the relay keying schemes are similar on most with some variances (maybe depending on what transistors are used?). The value of C20 (as in this Hygain) is different in many. In some the resistor is before C20 and in others it is after. I really do not use AM mode so I am not concerned whether 1 watt will energize the relay or it takes 4 watts to energize. My plan is to add grid bias and input tune to this little gem but need to start that journey on a good platform with everything working "as designed" first.

I am inclined to solder in a 100 ohm 1/2w resistor in place of R5. What are your thoughts?
Many thanks for your input! I was pulling what is left of my grey hair out.
73's
David

 

[nomadradio]

If R5 checked okay, I'd guess that shorting across it is only compensating for a problem that still remains downstream from it. It's there to limit RF current feeding into Q1 to a safe value.

I would double check to make sure Q1 is turned around the right way. Looks as if the designer had sideband in mind with this design. You want the keying circuit to activate with only a fraction of a Watt drive, to prevent it from clipping off the first part of a word after you stopped to take a breath. Using two transistors is the routine way to do this. Q1 is effectively a preamp driving Q2, the one carrying the relay's coil current We use a PNP darlington transistor, just to reduce parts count. The circuit in the Hy-Gain is is a darlington circuit, but with one transistor of each polarity.

Pretty sure this circuit would work with Q1 installed in reverse, but with limited sensitivity.

Just a shot in the dark. The RF choke RFC3 should not affect the keying circuit. It only serves to block the radio's drive power from leaking into the relay coil. The 6800pf capacitor C23 is nearly a dead short for 27 MHz RF, so there should not be much RF voltage on the emitter of Q1 twith C23 shorting the RF voltage (only) to ground. Replacing RFC3 with a jumper wire should not affect the keying circuit's performance at all.

(edit)

Just noticed that on the grid-bias thread you mention the DC power-supply voltage measured at the relay was low.

This thing should show around 16 Volts DC on standby and drop to 13 or 14 when keyed. Low power-supply voltage will make the keying circuit a bit "lazy" like this. Probably the more-important detail to fix first.

73

 

[Dmans]

Nomad,
I did install a 33 ohm 1/2w in place of the 330 ohm at R5. Results seemed normal. (i.e. it keyed up in AM with 1 watt drive. Output on SSB about 120 watts peak. Still shows about a 2.1 reflect going straight through the amp and also when keyed.)
I did previously measure resistance of R9 in circuit but can't at the moment remember the results. I would say it was very close to the spec'd 1 ohm or I would have had reason to remove it. Down side is I don't have another 1 ohm on hand just now.

I was quite suspicious of the NTE transistors I installed as replacements and removed the sub for the 6516 twice and checked it with a diode tester and a cheap Chinese transistor tester to insure it was good and to make sure I installed it correctly. I did also remove the NTE replacement for the 2N696 as well but only once after the initial install.

I suppose it is possible that the NTE replacements are not "exact" replacements to the originals but close enough to function in most circuits??

I did not measure closed relay voltage after this latest effort but in retrospect I probably should have. It occurs to me that "close enough" transistor replacements could also reflect in less than ideal closed relay voltage??

Spent the afternoon studying the chassis and schematic trying to figure where to stuff 10lbs of flour in a 2lb sack with reference to the tuned input and grid bias additions looming on the horizon but it seems a closer inspection is in order with this one first.

Would a different value cap at C7 help the input reflect until I can figure out how to stuff more components (i.e. Tuned input) into an already crowded place?

I appreciate your input, knowledge and time.

73's
David

 

[Dmans]

A quick update to this saga.
I peeled back the curtain to check voltages on the keying relay and to find out why my pass through (not keyed) SWR was above 2.1. I found that my replacement of R5 resulted in a "cooked" 33 ohm resistor (blackened but still measuring 32.7 ohms??). This after less than a half dozen key ups. I replaced R5 with the specified 330 ohm resistor and noticed a jumper from the circuit board trace (at the input of R5) to pins 11 & 12 of the K2 relay was not soldered well. I cleaned this up and re-soldered and checked pass through (not keyed) SWR at 1.4.
Further checks revealed a poor connection from the K1 coil (junction of C21 neg and C22 pos) to the collector of Q2 on the back side of the circuit board. I cleaned this up and re-soldered the connection and measure 14.5 vdc when keyed and 15.5vdc un-keyed on the K1 relay. The problem of not staying engaged on transmit is still there. I suspect I may have performed "Capital Punishment" on Q1and/or Q2 from this poor connection (and not enough resistance on my R5 substitute) as I measure less than .5vdc on the base of Q2 when keyed. I suspect it should be 1 volt or more??
I did remember to check resistance of R9-it measures 1.2 ohms.

(Side note/Personal Observation-These white fiberglass looking circuit boards with gum wrapper traces suck!)

That was all I had energy for today. Perhaps 2 more weeks of recuperation will allow more time and clearer thoughts.
73's
David

 

[nomadradio]

Maybe now it will work with a 330-ohm part at R5.

One thing that R5 does is to prevent the input of the keying circuit from pulling too much RF current from the radio's transmit power on standby. The standby-side SWR should be a lot lower with 330 ohms at R5.

73

 

[Dmans]

Nomad,
The standby SWR is 1.4 now with a 330 ohm in R5 slot and repairing the bad jumper connection to the K2 relay base. I think my next swipe at this (when energy and attention span permits) will be to replace Q1 and Q2 while paying particular attention to lead connections and orientation.

If/when that goes well, I think I have worked out the physical installation of the tuned input components and the control grid bias circuit and will proceed in that direction.

Thanks for the tips and support.
73's
David

 

[Dmans]

The Afterburner Plus Lives!
I replaced Q1 & Q2 with new transistors. I used a NTE 159M (T018 Package) in place of the previously installed (and lethally executed!) NTE 159 (T092 Package) for the Q1 location and installed a NTE 123 (T039 Package) in place of the previously installed NTE 123A (T018 Package) for the Q2 location.

The new NTE159M is a round metal transistor as opposed to the small plastic NTE159 first tried. Although the original Motorola MPS6516 was a like size to the NTE159, the larger metal cased NTE159M seems to me will be better, stronger, faster……..(70's TV show tagline-extra points if you name the show!). Same thing with the new T039 for Q2.

To prevent any further Senior moments, I also added a new piece to the radio bench. (seen below)

Next step, Control Grid Bias and Tuned Input.
Thanks to All who contributed help. Stay tuned!

73's
David

 

[wavrider]

Six Million Dollar man, Steve Austin, actor Lee Majors
Glad you got the amp working for ya

 

[Dmans]

Wavrider,
100 Bonus points for you! Spend them wisely!
Thanks for your help and direction!
73's
David

 

[wavrider]

I remember that series, I had to go outside to turn the tv antenna so dad could get a clear picture on the old Curtis mathis console tv whenever the show came on