X-Force 600HD Smoked

[Crusher]

I can’t tell what cap you have across input transformers, but it should be 470pf or 500pf. Not 330pf. Not sure if those are 2879 or 2879C? Should take around 230-240pf to tune input once caps across input are changed. If 2879HG, then needs about 170pf to tune output. If HG2879C needs about 130-135pf to tune output.

Your oscillation problem is due to biasing. HG do not like biasing. What voltage are you seeing at the base? While the Toshibas liked about 0.62-0.65V, the HG need something like 0.50-0.52v. To put them right on edge of turning on. After some experimenting, HG do not like bias, easily run away. I will not bias HG transistors. There is a substitute diode that will provide a lower voltage to base, almost turns them on but not quite. They seem to do ok this way. I will see if I can find diode to substitute. But with diode used now, they are turned on hard.

I’d also recommend more bypassing of the B+ strip. May also parallel another 100ohm 2W resistor across the resistors for feedback. Lower gain some but help stabilize better too.

 

[Crusher]

Is the bias on all the time, Shade, or switched by relay or transistor?

I'm very surprised at the RF when not keyed also.

I believe the bias is switched on when keyed. Uses caps and chokes to block DC and AC. Depending where at in circuit when keyed. I don’t like that design. Rather use a 3rd pole on relay.

 

[2NC995]

It looks like the bias is switched with the relay so it would not be on in standby. Maybe it was when the trombone was in too far and the transistors were getting warm like they were conducting? If they were conducting, it wouldn't have been DC to burn my finger so it must have been some kind of RF. This is a Benjamin Franklin moment for me. haha

I’m with you. Over my head for sure, but I appreciate you sharing all the details as you delve further into this.

Just speculation, but with no input you would think it must be self-oscillating in order to feel RF on the output. I’ve had the pleasure of being burned a couple times over the years. Feels as described - hot. Remove the RF and the parts aren’t hot.

Another random thought...I agree that the output will effect the input SWR. But, why not try two capacitors and a coil on the input side of things?

I believe Shockwave recommenced 2x 464 trimmers, separated by a 1/2” 5 turn coil...and go from there. I haven’t messed with my old amp, but that’s what I plan to try when I get around to it.

I’m curious as to any responses about the SWR varying with drive level. All my amps do this to some degree. I just make sure it’s good at the level I plan to run it. As far as SWR on sideband...I just don’t bother looking. Haha.

 

[Crusher]

On the diode for biasing. I think they used 1N5408. Try replacing with 1N5420. Will forward bias 0.475v roughly. Much better than the 0.65V roughly. They are turned on so hard, acting like a switch.

 

[Crusher]

By the way, the 6-8” long wire from brass cap to output is a bad idea. Lead should be much shorter.

 

[Shadetree Mechanic]

I can’t tell what cap you have across input transformers, but it should be 470pf or 500pf. Not 330pf. Not sure if those are 2879 or 2879C? Should take around 230-240pf to tune input once caps across input are changed. If 2879HG, then needs about 170pf to tune output. If HG2879C needs about 130-135pf to tune output.

Thanks for the info Crusher, The transistors are the RFParts brand HG 2SC2879 but they are not the "C" version. The capacitors on the input capacitors are 330pf. I am going to start putting another order together for RFParts.

 

[Shadetree Mechanic]

Your oscillation problem is due to biasing. HG do not like biasing. What voltage are you seeing at the base? While the Toshibas liked about 0.62-0.65V, the HG need something like 0.50-0.52v. To put them right on edge of turning on. After some experimenting, HG do not like bias, easily run away. I will not bias HG transistors. There is a substitute diode that will provide a lower voltage to base, almost turns them on but not quite. They seem to do ok this way. I will see if I can find diode to substitute. But with diode used now, they are turned on hard.

On the diode for biasing. I think they used 1N5408. Try replacing with 1N5420. Will forward bias 0.475v roughly. Much better than the 0.65V roughly. They are turned on so hard, acting like a switch.

I did a quick voltage check with my Fluke 16 meter and it showed 0.388v on the peak hold setting. The transistors that I removed were DEI so maybe they are closer in operation to the HG?

The bias network has what appears to be a couple of inductors and a capacitor to ground on the sand bar input. I assume to shunt any RF to ground?

 

[Shadetree Mechanic]

The biasing diodes themselves, it seems that they took the numbers off. Haha. And then the same inductors and capacitor along with another bigger capacitor.

 

[2NC995]

Shade, Just thinking out loud here, but I have to wonder if that long input lead is contributing to tuning issues, and maybe even the oscillation? I usually see the coax terminated at the capacitor, which is connected directly to the input combiner?

 

[Shadetree Mechanic]

I’d also recommend more bypassing of the B+ strip. May also parallel another 100ohm 2W resistor across the resistors for feedback. Lower gain some but help stabilize better too.

Ok so for bypassing on the B+ you mean the 104 disc caps on the 12v rail? Would you normally have like 6 or 8 instead of the 4 that are there?

And the feedback resistors would be these? Just parallel and on the same disc cap, or a separate cap?

.

 

[Shadetree Mechanic]

By the way, the 6-8” long wire from brass cap to output is a bad idea. Lead should be much shorter.

Shade, Just thinking out loud here, but I have to wonder if that long input lead is contributing to tuning issues, and maybe even the oscillation? I usually see the coax terminated at the capacitor, which is connected directly to the input combiner?

The final question is about the long wires on my input and output tuning capacitors. Should they be coax like 2NC955 suggests? I suppose that if I could use the correct fixed caps then it would not be an issue. And I probably wouldn't need an output coil either.

Sorry if I am asking too many questions, just trying to get this thing ironed out. Thanks Guys!

 

[Shadetree Mechanic]

I’m with you. Over my head for sure, but I appreciate you sharing all the details as you delve further into this.

Just speculation, but with no input you would think it must be self-oscillating in order to feel RF on the output. I’ve had the pleasure of being burned a couple times over the years. Feels as described - hot. Remove the RF and the parts aren’t hot.

Another random thought...I agree that the output will effect the input SWR. But, why not try two capacitors and a coil on the input side of things?

I believe Shockwave recommenced 2x 464 trimmers, separated by a 1/2” 5 turn coil...and go from there. I haven’t messed with my old amp, but that’s what I plan to try when I get around to it.

I’m curious as to any responses about the SWR varying with drive level. All my amps do this to some degree. I just make sure it’s good at the level I plan to run it. As far as SWR on sideband...I just don’t bother looking. Haha.

I am still thinking about that input network, I will have to see if I have any 464 trimmers. Thanks!

 

[2NC995]

I am still thinking about that input network, I will have to see if I have any 464 trimmers. Thanks!

Maybe we can stir up @Shockwave to confirm I’ve articulated his suggestion correctly!

As I recall, two 464 caps, and 5 turns of wire around a Sharpie was the staring point.

 

[Crusher]

Before going any further. The DEI that were in it, did all 4 blow? If not, was it a transistor on each 2 pill section or were both transistors bad on one side? Trying to understand if maybe there was a problem before you changed devices.

Also, when you checked voltage, think you said it was 0.388V. Was this at the base of each transistor. ( what I’m getting at, does 1 or 2 show different voltage?)

 

[543_Dallas]

Why is that brass tuning cap in the amp to begin with? Someone may have thought it looked heavy duty, but it's in the way and unnecessary. All of that stray inductance from the long lead is counterproductive.

Rip out the trombone and save it for a 6 pill or larger amp that needs it. Run the wire straight from the output to the relay (where you put that coil) and solder in a good size cap to ground like every Dave copy box has been for the last 15-20 years. Long lead lengths are bad so keep everything short.

Do the voltage checks Crusher mentioned. If some pills are pulling more current than others the thing will be out of balance and keep failing until all are matched.

You have the right idea on the feedback resistors. Just take another 100 ohm 2 watt and parallel it with the one that's already there. More negative feedback will make it more stable.

Stick with the nominal cap values and figure out why it isn't stable before you jump too far down the rabbit hole.