Using an SB-220 (HL-2200) With One Tube

[WB2RON]

One of the 3-500z's in my HL-2200 developed a plate to grid short (a know manufacturing defect at Eimac) and the only damage other than the tube that I could find was a blown 0.8 ohm grid current meter shunt. Because of covid, we are not receiving any packages in the mail so for now I would like to just operate the amplifier using one tube expecting 100 watts carrier instead of 200 watts that I usually operate. I have a very limited supply of parts but I was able to put together a bunch of resistors in parallel to get to 1.6 ohms to replace the 0.8 ohm shunt resistor. I know that my grid current readings will now be double the actual current but I'll just take that into account when operating.  My question is whether I should I leave the defective tube in the circuit (filament continuity checks good) with the plate cap (HV) removed. My concern is that if I remove the tube the single tube filament voltage with rise and that the current limiting that the filament transformer provides will not be enough for only one filament, both of which will shorten the life of the exixting tube. I probably shouldn't worry too much about this since I will eventually replace both tubes. Also, are there any concerns as far as loading of the input signal with or without the second tube. The grids are directly grounded for both tubes so if I leave the defective tube in the circuit should I remove the grid connection to ground on the defective tube even though there is no high voltage present to that tube? This all assumes that the tank circuit has enough tuning range for the higher source impedance of operating with only one tube. Also, any other comments or suggestions will be appreciated.

Ron
WB2RON

[WQ9E]

Ron,

Most 3-500Z shorts are grid to cathode shorts because the clearance is exceedingly tiny, grid to plate shorts aren't common.

You can safely run one tube and I would leave the second tube in to load the filament transformer but do disconnect the grid connections to the defective tube.

I am just about to start building a SB-220 that I bought last year still packed in its original box.  It will be a fun COVID/winter project.  The first Heathkit I built was a SB-220 in 1976.  This time I won't install the bias filter cap backwards

Rodger

[WB2RON]

Thanks Rodger for the advice. The plate assembly is leaning over and shorting against the grid as a result of the weld on the plate assembly coming loose. This was a known defect with a particular batch of tubes from Eimac. The other tube probably has the same defect but hasn't let go yet.
I think your advice to disconnect the grid from ground is good advice since there is now some pressure from the plate assembly against the grid that could increase the chances of it shorting against the cathode which wouln't be good. By the way, is there a simple fix such as adding a fuse that would protect the filament transformer if a grid to cathode short occurred in the good tube?
Good luck with the SB-220 kit. It sounds like a great project. Will you have to replace any of the capacitors since they are old although they have never been used? Do you think the 3-500z's will still be good?

Ron
WB2RON

[Detroit47]

Why aren't you receiving packages? Did they quit delivering?

N8QPC mo

[WQ9E]

Ron,

You could separately put an inline fuse in the feed to the filament/bias transformer to protect it BUT if you have the stock Heathkit setup, the grid chokes should let go before the filament transformer primary is hurt.  What you would likely lose is the bias secondary which could be fused with a small fuse.

The plate meter is the most likely thing to go with SB-220 issues.  Some of the aftermarket boards like the W7RY board I am using have protective devices for it.

Rodger WQ9E

[KD6VXI]

Ron,

You could separately put an inline fuse in the feed to the filament/bias transformer to protect it BUT if you have the stock Heathkit setup, the grid chokes should let go before the filament transformer primary is hurt.  What you would likely lose is the bias secondary which could be fused with a small fuse.

The plate meter is the most likely thing to go with SB-220 issues.  Some of the aftermarket boards like the W7RY board I am using have protective devices for it.

Rodger WQ9E

Meter protection isn't rocket science, but it will effect metering accuracy.

Typically a set of back to back diodes (with enough in series to allow for the metering voltage) is applied across the metering terminals.

Problem is this.  If you put JUST enough to protect the meter at full scale, it will statt to be Jon linear towards full scale as the diodes start to conduct.  Add any heating effects, the problem gets worse.

If you put enough diodes in series for this not to be an effect then you'll still pin the meter during a short condition and negate the use of the protection mechanism.

It's a conundrum that can be fixed by using a variable voltage supply and putting a proper meter scale on the face of the meter to take into a count any non linearity.  The changing effect as the diodes heat up, not so easy to work around.

Myself, I out the diodes in.  I rarely operate any meters at full scale so it's not a problem.

--Shane
KD6VXI

[Opcom]

If a higher-voltage (2x or 3x) stack of protection diodes is used so the meter is accurate full scale, which is worse on a meter, the mechanical abuse of pinning the meter or the electrical abuse to the meter's coil?
The question presumes that the fault time is shortened by a HV fuse blowing.

[KD6VXI]

If a higher-voltage (2x or 3x) stack of protection diodes is used so the meter is accurate full scale, which is worse on a meter, the mechanical abuse of pinning the meter or the electrical abuse to the meter's coil?
The question presumes that the fault time is shortened by a HV fuse blowing.

Pat,

That is the million dollar question.

As I put in my thread, I try to use meters that typical usage will be half to two thirds scale.  Then I don't have to worry about inaccuracy.  Jim Tonne's Meter software is pretty affordable and ensures I can have Green, Yellow and Red scales added to my faces, any scale I want, etc.

I think this is the best method.  Then you can ensure the meter will never be pinned.

--Shane
KD6VXI

[WB2RON]

It turns out that there are other issues with the amplifier than just an open shunt resistor. Before testing the amp I removed the tubes and I also installed a 1 amp fuse in series with the replacement shunt resistor as a precaution. Also, before applying power I measured the resistance between the plate connectors and the chassis and it measured approximately 200k in accordance with the resistance check noted in the manual for the HL-2200. I then applied ac power and a loud buzzing sound was coming from the amp along with a full scale reading on the plate current meter. I immediately shut down the amp and it turns out that the 1 amp fuse that I added in series with the replacement shunt resistor was now open. Now with the open fuse I again applied ac power and there was no plate current or noise and the high voltage was reading normally on both cw and ssb positions of the mode switch. The fan was also working normally. This is the same state that I noticed when the tube originally shorted taking out the shunt. My only test equipment for troubleshooting is a digital and an analog volt/ohm meter. I had already checked all of the diodes and resistors on the meter board and all appeared normal. Suggestions please on what to look for going forward.
To summarize:
Heard a big bang
Saw normal high voltage before turning off amp
Found short between plate and grid on one tube
Found open 0.8 ohm shunt resistor
Replaced shunt resistor with 1.6 ohm and series 1 amp fuse
Removed both tubes and turned on amp
Heard buzzing and saw full scale plate current
Found open 1 amp fuse in series with replacement shunt
Turned on amp with open fuse and high voltage now normal and no buzzing

Ron
WB2RON

[Detroit47]

I would disconnect the HV from the base of plate choke and try it. You may have a shorted plate block that only shows up under HV or the bypass on the bottom of plate choke.

Johnathan N8QPC

[KD6VXI]

I would disconnect the HV from the base of plate choke and try it. You may have a shorted plate block that only shows up under HV or the bypass on the bottom of plate choke.

Johnathan N8QPC

I had this exact problem with the doorknob at the base of the plate choke.  The SB220 I had would tick from the day I bought it in 1999.  It lasted 17 years until it shorted.

--Shane
KD6VXI

[Detroit47]

Have you worked on the amplifier? Or is it top secret?

[WB2RON]

Yes, I worked on the amplifier. My goal was to find the short causing the high plate current. I began by disconnecting all of the external hv components external to the meter/rectifier board. One by one I added the components back and after all the of the components were connected the hv was still functioning. At this time I'm not sure what was causing the short initially but I put this project on hold until I get diodes to put across both meters for protection expected from a friend on or about February 8th at which time I will continue with this repair.

Ron

[WB2RON]

I have completed the repair of the HL-2200 and am operating with just one tube. It appears that the only problem that I had with the amplifier was the one shorted tube and a blown 0.8 ohm grid meter shunt resistor. The short that I was seeing after replacing the resistor I believe came from one of the plate caps touching the cabinet cover as this short has completely gone away. To operate with one tube, I left the defective tube in it's socket to load the filament transformer but did disconnect the plate cap and ungrounded the grid as suggested by Rodger, WQ9E. My output power using just one tube is 750 watts in CW mode and 925 watts in SSB mode as read on my LP-100a power meter operating into a dummy load on 80 meters. When operating on AM, I reduce the drive to 100 watts carrier and am getting 500 watts on modulation peaks as read on the peak power meter which closely agrees with my 122% positive modulation as read on my REA Mod Monitor.

Ron
WB2RON