TL-922 no plate or grid current at idle, where should I start checking?

[NN7B]

I purchase a damaged (cosmetically, and broke both original tubes) Kenwood TL-922. I took everything apart, cleaned, and inspected it. The amplifier looked new in all respects internally, but I knew it had been sitting for a long time. I purchased a matched set of new RF parts 3-500ZG tubes, and put her on the variac to warm up over 24 hours. Everything seemed good, and checked out, so hooked her up to my trusty TS-820. Initially, specs were within normal ranges, and I tuned up in a dummy load on CW. Plate current and grid current were within spec when the linear was placed in operate mode, and at idle. I moved over to SSB, and re-checked the tune, and it seemed fine. I hooked the amp up to my antenna, and started tuning again on CW, then SSB when all of a sudden my wife smelled something burning ( I didn't smell it at first, but then a faint wift of burning electronics, but not severe). I at some point realized I was not getting RF out, and the RF meter needle was no longer swinging full scale like it had earlier. I was reading the wattage put out by my 820. The tubes still glow, and HV reads correct, but no plate or grid current at idle. If I still transmit as if to tune up in operate mode, I can still dip the plate, and max RF out, but only puts out what my exciter is putting out. I tried adjusting the AGC control, but no change. I'm now ready to start looking around again inside for something "smoked". I was hoping that one of you experienced fellows with the TL amplifiers could help guide me in a closer direction as to what I might have fried. Thanks for the help!!

73'
Damon
NN7B

[W1QJ]

When you say "I can still dip the plate" do you mean on the exciter? If you have no plate current or grid current how can you "still dip the plate"? Unless you mean the exciter? If you get no idle current when you key the amp, does the HV meter drop a tiny bit when you key? If not, check to see if you have an open plate choke. That could be the smell you smelled. The plate choke could have burned open.

[NN7B]

Actually, when the exciter is fully tuned up, then I go to tune up the amplifier, when I key down, I do have a small amount of plate current showing on the meter, which will dip with the amplifier plate control. And, I can then watch the RF meter, and adjust the load control, and tune capacitor to max RF out, which when peaked, equals the amount of RF out my TS-820 is putting out barefoot. The manual states that once the exciter is tuned, and you flip the linear switch over to operate, and NO key down (at idle), you should have around 100mA plate current when in CW mode, and 200 MA in SSB mode. This was true when I first tuned up the amp, however, now, there is no plate current or grid current at idle when switched over from standby. The only thing I could see that was a little charred in the amp was the covered strap that runs from the band switch to the tank coil, the longest strap, probably the 3.5 mHz, which was the band I was trying to tune up. The strap was a bit close to a sharp bracket edge, and I'm wondering if I might of had some arcing through the insulated cover (white woven fabric cover) into the chassis bracket. I've simply moved it further away from everything metal around it. It's almost as if the operate/standby switch is inoperative. I need to go back to the schematic and see if there is a relay in the "operate" mode that may be smoked too. I'll check continuity in the plate choke, but can't see anything visible at first glance being open yet. Thanks for the help!

[AF6LJ]

is this on more than one band?

If so.....
I would suspect something in the amplifier grid circuit has gone up in smoke.

[WB2WIK]

Actually, when the exciter is fully tuned up, then I go to tune up the amplifier, when I key down, I do have a small amount of plate current showing on the meter, which will dip with the amplifier plate control.

How much is a "small amount?" If it's not about 100mA (CW) or 200mA (SSB) when idling, something's wrong. Idling (amp keyed but no RF drive) there shouldn't be any grid current at all.

What is the indicated high voltage (meter reading)?

Do you hear the keying relay "click?"

If the B+ is normal and you don't have normal idling current, sounds like something with high resistance is in series with the path, either in the anode or cathode line. Doubt both tubes are bad, especially since they're new.

[AG6K]

I purchase a damaged (cosmetically, and broke both original tubes) Kenwood TL-922. I took everything apart, cleaned, and inspected it. The amplifier looked new in all respects internally, but I knew it had been sitting for a long time. I purchased a matched set of new RF parts 3-500ZG tubes, and put her on the variac to warm up over 24 hours. Everything seemed good, and checked out, so hooked her up to my trusty TS-820. Initially, specs were within normal ranges, and I tuned up in a dummy load on CW. Plate current and grid current were within spec when the linear was placed in operate mode, and at idle. I moved over to SSB, and re-checked the tune, and it seemed fine. I hooked the amp up to my antenna, and started tuning again on CW, then SSB when all of a sudden my wife smelled something burning ( I didn't smell it at first, but then a faint wift of burning electronics, but not severe). I at some point realized I was not getting RF out, and the RF meter needle was no longer swinging full scale like it had earlier. I was reading the wattage put out by my 820. The tubes still glow, and HV reads correct, but no plate or grid current at idle. If I still transmit as if to tune up in operate mode, I can still dip the plate, and max RF out, but only puts out what my exciter is putting out. I tried adjusting the AGC control, but no change. I'm now ready to start looking around again inside for something "smoked". I was hoping that one of you experienced fellows with the TL amplifiers could help guide me in a closer direction as to what I might have fried. Thanks for the help!!

73'
Damon
NN7B

 Damon --- New 3-500Zs or ZGs do not have reduced HV leakage after running the filament for 24-hrs before use. . .
• TL922s have 500wv filter caps that have only 300v on them on the CW tap, so running a long-idle 922 on the CW tap for a few minutes is plenty of cap reforming time before going up to the SSB HV tap.
WARNING: those persons who believe that amplifier engineers are semi-infallible demi-gods should definitely stop reading at this point.
• Before running a factory-stock 922, I would.
1. Change the 922 from +110V-cutoff bias to resistor-cutoff bias so that T2 will not burn up in the event that a tube has a filament to grid short. This is a simple mod. that is discussed in:
http://www.somis.org/QSK922.html
which could save you from having to have T2 rewound after it burns up from a filament-to-grid short.
2. Put a paralleled pair of cross-polarized (arrows pointed in opposite directions) 3A avg, 200a-pk Si diodes across each meter movement to protect it from glitches.
3. Replace the carbon-comp plate/anode current meter shunt resistor, and the carbon-comp grid current meter shunt resistor with modern, more stable, MOF resistors of the same value and W-rating. If you have a DMM, checking the accuracy of the grid and plate/anode I-meters is a good idea.
4. Replace the small choke that's in series with the HV+ (L2, 12µH) with a 15Ω, 120Joule, 10Watt WW surge-rated R. This "glitch R" will prevent a filament-to-grid short during a glitch by limiting the peak discharge I from the filter cap bank to <200A.
- note - If you need one of these surge-rated resistors, we sell Ohmite's-version for $1.77 each.
5. Optional - A not-well-understood problem with 3-500Zs is that a pair of them have 0.3pF of feedback-C. This sounds like no biggie until you use a dipmeter and discover that the parasitic-resonance in the anode circuit (measured at C-dc-blocker) of the 922 dips c. 120MHz - f. depending somewhat on where the Tune-C is set. . . At 120MHz, 0.3pF has c. 4000Ωs of XC. This means that there's a not-on-the-schematic, 4000Ω signal path between the output back to the input of tubes that have an amplification-factor of 130! This is obviously not good news unless you are building a VHF power oscillator. The fix for this potential problem is to build a better R/L device to reduce VHF amplification even more.
- note - The way that a R/L VHF-"suppressor" device reduces VHF-amplification is by reducing the parallel-equivalent VHF load-R presented to the anodes..
http://www.somis.org/YtoZ.GIF
The 922 has a R/L device/"suppressor" but unfortunately the factory's device does not go far enough in reducing VHF-amplification.
. To make a R/L VHF suppressor reduce VHF-amplification more, lower it's VHF-Q - which is a pretty simple thing to do: increase L and increase R. However, decreasing VHF (30-300MHz) -Q - and VHF-amplification - is going to result in an additional power-loss at 28MHz. Decreasing the 100MHz-Q of the 922's original Q=5.5 R/L suppressors to a Q of 1.5 will do the job of reducing VHF-amplification enough to tame a 922 - however, the tradeoff is c. a 2% power drop at 28MHz - or about 0.1db or 1/60 of one S-unit.
• JPG of a pair of 3-500ZGs in a 922 with VHF-Q=1.5 suppressors and a 15Ω, 120J glitch-R:
http://www.somis.org/
• A tradeoff with lower-VHF-Q suppressors is increased heating during operation at 28MHz - so one must come up with a design that can safely handle the extra heat on 10m. One way around this is to use overload-resistant MOF resistors instead of carbon-comp resistors and to dissipate some of the heat in the L part by using resistance-wire for the L instead of silver-plated strap.

. THE CIGAR TEST:
- note - hold cigar in a plastic clothes pin, stand on a plastic matt, and do not contact the cigar when it is touching a suppressor-R --- unless of course you want to meet the Virgin Mary sooner rather than later.
. During A0 or RTTY operation on 10m, a properly-designed VHF suppressor should be able to ignite a cigar within 30-seconds. If the cigar does not ignite, you rather obviously need more heat so shut down, unplug the amp from the mains, open a bottle of Beck's or Saint Pauli Girl, kick back. wait until the voltmeter indicates <100V, add another turn to L-supp, and try again.
- note - If the suppressor resistors get hot enough to change color on 10m, the Cigar Test is probably not needed.

6. Bending the first plate of the Tune-C rotor so that the breakdown V is reduced to about 2/3 of original will cause the Tune-C to arc during a parasite rather than arc across an open contact on the bandswitch - possibly saving you from having to replace the band$witch.

[W1QJ]

 Damon --- New 3-500Zs or ZGs do not have reduced HV leakage after running the filament for 24-hrs before use. . .
• TL922s have 500wv filter caps that have only 300v on them on the CW tap, so running a long-idle 922 on the CW tap for a few minutes is plenty of cap reforming time before going up to the SSB HV tap.
WARNING: those persons who believe that amplifier engineers are semi-infallible demi-gods should definitely stop reading at this point.
• Before running a factory-stock 922, I would.
1. Change the 922 from +110V-cutoff bias to resistor-cutoff bias so that T2 will not burn up in the event that a tube has a filament to grid short. This is a simple mod. that is discussed in:
http://www.somis.org/QSK922.html
which could save you from having to have T2 rewound after it burns up from a filament-to-grid short.
2. Put a paralleled pair of cross-polarized (arrows pointed in opposite directions) 3A avg, 200a-pk Si diodes across each meter movement to protect it from glitches.
3. Replace the carbon-comp plate/anode current meter shunt resistor, and the carbon-comp grid current meter shunt resistor with modern, more stable, MOF resistors of the same value and W-rating. If you have a DMM, checking the accuracy of the grid and plate/anode I-meters is a good idea.
4. Replace the small choke that's in series with the HV+ (L2, 12µH) with a 15Ω, 120Joule, 10Watt WW surge-rated R. This "glitch R" will prevent a filament-to-grid short during a glitch by limiting the peak discharge I from the filter cap bank to <200A.
- note - If you need one of these surge-rated resistors, we sell Ohmite's-version for $1.77 each.
5. Optional - A not-well-understood problem with 3-500Zs is that a pair of them have 0.3pF of feedback-C. This sounds like no biggie until you use a dipmeter and discover that the parasitic-resonance in the anode circuit (measured at C-dc-blocker) of the 922 dips c. 120MHz - f. depending somewhat on where the Tune-C is set. . . At 120MHz, 0.3pF has c. 4000Ωs of XC. This means that there's a not-on-the-schematic, 4000Ω signal path between the output back to the input of tubes that have an amplification-factor of 130! This is obviously not good news unless you are building a VHF power oscillator. The fix for this potential problem is to build a better R/L device to reduce VHF amplification even more.
- note - The way that a R/L VHF-"suppressor" device reduces VHF-amplification is by reducing the parallel-equivalent VHF load-R presented to the anodes..
http://www.somis.org/YtoZ.GIF
The 922 has a R/L device/"suppressor" but unfortunately the factory's device does not go far enough in reducing VHF-amplification.
. To make a R/L VHF suppressor reduce VHF-amplification more, lower it's VHF-Q - which is a pretty simple thing to do: increase L and increase R. However, decreasing VHF (30-300MHz) -Q - and VHF-amplification - is going to result in an additional power-loss at 28MHz. Decreasing the 100MHz-Q of the 922's original Q=5.5 R/L suppressors to a Q of 1.5 will do the job of reducing VHF-amplification enough to tame a 922 - however, the tradeoff is c. a 2% power drop at 28MHz - or about 0.1db or 1/60 of one S-unit.
• JPG of a pair of 3-500ZGs in a 922 with VHF-Q=1.5 suppressors and a 15Ω, 120J glitch-R:
http://www.somis.org/
• A tradeoff with lower-VHF-Q suppressors is increased heating during operation at 28MHz - so one must come up with a design that can safely handle the extra heat on 10m. One way around this is to use overload-resistant MOF resistors instead of carbon-comp resistors and to dissipate some of the heat in the L part by using resistance-wire for the L instead of silver-plated strap.

. THE CIGAR TEST:
- note - hold cigar in a plastic clothes pin, stand on a plastic matt, and do not contact the cigar when it is touching a suppressor-R --- unless of course you want to meet the Virgin Mary sooner rather than later.
. During A0 or RTTY operation on 10m, a properly-designed VHF suppressor should be able to ignite a cigar within 30-seconds. If the cigar does not ignite, you rather obviously need more heat so shut down, unplug the amp from the mains, open a bottle of Beck's or Saint Pauli Girl, kick back. wait until the voltmeter indicates <100V, add another turn to L-supp, and try again.
- note - If the suppressor resistors get hot enough to change color on 10m, the Cigar Test is probably not needed.

6. Bending the first plate of the Tune-C rotor so that the breakdown V is reduced to about 2/3 of original will cause the Tune-C to arc during a parasite rather than arc across an open contact on the bandswitch - possibly saving you from having to replace the band$witch.

All of which has nothing to do with your present problem.

[AG6K]

All of which has nothing to do with your present problem.

 IME, there are several failues that occur in 922s that can result from not having addressed the above issues.

[K0BG]

The TL922, and TL922A are fraught with design problems, most of which have nothing to do with VHF issues. Most of the bypass caps are incorrect types for the use, and some of the signal paths are way too long, as is the fact the grids are floating (not hard to ground). Most of the problems can be easily fixed, Rich's palaver to the contrary. The best information on these amps, and how to make them play correctly is on Tom Rauch's (W8JI) web site. He is, after all, one of the best amplifier design engineers out there.

[NN7B]

O.K., I understand, and have read those suggestions before, but just wanted to see if the Amp would fire up before I started making modifications. I would at least like to get this amp to send out a signal before I start making mods. I was hoping someone here could perhaps suggest a starting point to trouble shooting this problem. Something has definitely gone south, because it worked for a short bit before this happened. I have downloaded all of the mods I could find for this amp, including pictures, etc. Keying relay does click, and I'll have to re-check my notes on the other issues. Thanks for the input so far, but I haven't got a direction yet.