SB-221 Amplifier Modifications and Repair

Purchased from the Heathkit Electronic Center
5050 FM 1960W #126
Houston, TX 77069

This amplifier was purchased 10/27/1982 and assembled by myself, N4BOS shortly thereafter.  Before assembly, the parts were purchased from Heath to add 10 meters back into the amp (the stock SB-221 amp deleted 10 meters because of the CB band) and it was assembled using the SB-220 assembly manual.  A delay circuit for "soft start" was added as well as an electronic bias circuit and an external hour meter.  

Due to a move to a no antenna area, the amp was packed away from about 2000 to 2015 when it was put back in service.

2/2015: 130.8 Hrs
When I built the amp, I installed an electronic bias switching circuit from the May, '74 issue of QST.  That circuit had issues and resulted in distortion (for unknown reasons) so I had redesigned the circuit back in the '80's.  That circuit had failed so after reading numerous articles about how the 3-500 anodes (plates) need to get red for the getters to work, I removed the circuit from the amp and put it back to stock configuration. 

Installed new high voltage capacitor module from "Paul Kraemer" (W0UYA) <elespe@lisco.com> at Electronic Specialities in Iowa.  This module contains 8, 330μF, 450V capacitors instead of the original 8, 200μF capacitors  Since the power supply is arranged in a voltage doubler arrangement, the effective capacitance is now 82μF as compared to 50μF originally on each side of the doubler.   As a result, the power supply is supposed to deliver slightly higher voltage under transmit which should equate to slightly higher output power. 

Replaced the original diode board with one from Harbach Electronics.  "The RM-220 is a replacement rectifier/metering board for the Heathkit SB-220 & SB-221. It features modern 1000PIV 3A diodes, new meter shunt resistors, a precision high voltage meter divider string and a high-current zener diode replacement string. The board also features built-in meter protection diodes."   The board as supplied WOULD NOT FIT so I had to build an offset mount so the board would clear everything.  The new board was over .100 inches wider from the screw hole to the edge of the board and it ran into the cage around the high voltage section.  Disappointing that the design was so far off.  I moved it "north" about .250 inches.

Installed the Harbach SK-220 "Soft Keying" interface.  "The SK-220 safely acts as an amplifier interface to your modern transceiver to the Heathkit SB-220 or SB-221. This unit is mounted inside the amplifier and transforms the 120VDC key closure to approximately 0.7VDC at 1.5mA.  Without it you run the risk of severely damaging the keying circuit in your transceiver."   This board also did not fit as designed because the SB-221 has a connector for the high voltage transformer in the way so the board had to be turned 180 degrees.  No big deal but the instructions didn't mention that.

Incidentally, I talked to Jeff Weinberg of Harbach and he was not aware of either problem.  As a result, he redesigned the diode board to be the same width as the original Heath board (4 inches) and he said he is going to include my picture of the soft key board installed in my 221 so people know how to mount it.

Resistive Bias Mod: I also did the "WY6K - An Essential Mod to the SB-220 and SB-221".  There are two versions of this modification,.  I did the second that positively biases the tubes off with 120 volts through a 100K resistor.  This circuit biases the tubes into cutoff in RX mode by pulling the cathode up toward the 120 volt bias supply voltage through the 100k resistor.  There will be some voltage drop across the resistor, but the cathode will still be close to 100 volts and firmly in cutoff.  If the cathode shorts to ground through the grid, the current is limited to about 1 ma and the transformer will not melt down.  Voile.

This mod is completed at the TR relay.  Here is the before picture and here is the after.

When the T/R relay is switched into TX mode, an 8 diode string on the new Harbach diode board (replacing the original 5V zener) is connected to the center tap of the filament transformer, taking the cathode to + 6.8 volts and biasing the amp as intended for class AB2 operation.  This is called "resistive bias", whereas the original circuit used "voltage bias".

In addition, I removed all the bypass capacitors and chokes that connected the 3-500Z grids to ground and firmly grounded them with .016 X 1/4 inch copper straps.  Here's a picture of that.  This was recommended in most of the articles about using the 3-500Z tubes in grounded grid configuration.  While you're there, it's a good idea to check all of the pin connectors on the 3-500z sockets.  On most of the sockets Heath used, the clip that tensions the pins can be removed and the contacts can be aligned to better fit the tube pins.  Also the clip can be bent to more evenly apply pressure to the pin of the tube.  I found several that didn't fit well and fixed them.  This will not only help with the grid connection but the 3-500's have been known to get so hot as to melt the solder in the filament pins (they're handling ~ 15 amps!) so a good connection there is essential.

Filament/Bias transformer failed:  The amp was powered up and the keying relay would only buzz when keyed.  It was found that the 22μF capacitor on the 120 volt bias supply was HOT.  The bias winding in the filament transformer was shorted so the transformer was removed and sent to Gary Brown (1-207-942-5745) at Transformer Rewinding Service in Maine to be rewound.  I have no idea what caused the failure.  Since the capacitor was so hot, one would suspect either the diode was in wrong (it wasn't) or the capacitor was bad but it tests good. 

Gary suggested rewinding the primary for 125/250 volts.  He said Heath designed the transformer for 117 volt primary and that at 125 volts, the core is nearly saturated.  He said the transformer would be much better if the primary was rewound for 125 so that's what he did.

New Bias Power Supply Designed and Installed:  Because the bias winding in the original Heath filament/bias transformer had failed and because it is relatively fragile, I decided to install a separate bias transformer.  In the event of a bias problem again, it will be MUCH less expensive to repair.  I selected a Triad FP88-65 PC mount transformer.  This transformer has two 44 volt secondaries and a 110/220 volt primary and will deliver 65ma at 88 volts.  It costs just under $11.00 from Mouser Electronics.  The bias supply needs to deliver roughly 30 ma to power the keying relay (~26ma), the bias for the 3-500's (~1 ma), the current for the soft key board (~1ma) and ALC current (~2ma).  The new power supply is mounted under the amplifier near the 120 volt terminal strip and is wired for 220 VAC only. 

Each secondary winding is spec'd at 115-145 ohms (230-290 total DCR) so the transformer will only deliver about 120ma when short circuited.  Open circuit, the transformer delivers about 185 VDC through the half wave rectifier.

                                                         

I realize the wiring of the transformer appears a bit weird and you might think the windings are in buck configuration but I drew it that way because of the numbering of the pins on the actual transformer.  When using the transformer on 240 volts, pins 2&3 are wired together and the input is applied to 1 and 4.  Pins 2 and 3 are NOT side by side (as you can see HERE) on the transformer so I drew it to represent the way the pins appear on the transformer.  Likewise on the secondary; pins 6 and 7 are not side by side so to series the windings, pins 6&7 are tied together and 5 and 8 are the output.

The 120 volt, 5W zener (1N5380) has a temperature coefficient such that at operating temperature, the voltage rises to about 130VDC.  At that voltage, it is only conducting about 15ma.  Since the bias supply also runs the keying relay in the SB-221 and the ALC supply.  When the relay is keyed on, it pulls about 23ma and the voltage drops to about 110VDC.  The zener is then non conducting so the transformer load is never more than about 30ma for the relay current, the ALC supply, the soft key board and the bias to the 3-500Z's.

Amp Wired for 220 Volts Only
In addition, per Heath Service Bulletin SB-220-27, (MSword document) when the rewound filament/bias transformer from Gary was reinstalled, I rewired it for 250 VAC only.  The new bias winding output (red wires) were covered with red heat shrink tubing and not used.  That bulletin is very interesting and takes some thinking to understand but if there is a malfunction in the power switch or the high voltage transformer, the filament transformer can be damaged.  While it seems unlikely, apparently it happened often enough that Heath recommended the change to all amplifiers that are run on 240 volts.

4/13/2015 Amplifier put back on line:
The rebuilt amplifier now works great.  Unkeyed high voltage is now 2400 in CW and 3100 in SSB with 246 volts input.   The key down, no signal plate current runs ~100ma in CW and 150ma in SSB.  I measured the filament voltage from the rewound transformer before and after installing the tubes and found the open circuit voltage to be 5.287 volts AC and with the tubes installed, it is running 4.86 volts.  Tuned to 14.260 and driving a 50 ohm oil filled dummy load, it does 1100 watts key down in CW and over 1600 peak in SSB when driven with my TS-520S.  Since the maximum allowed power on the amateur bands is 1500 watts peak, this amplifier produces the full legal limit.  Not bad!

Now, here's a "freebee" for reading all this stuff.  There's a device out there called a "3898 pecker"....  Look it up on Google.  It's real.  The best link I have found is at http://www.iol.ie/~bravo/Woodpecker.htm

You can read all about it there but almost as a lark, I built one, exactly as shown in the article.  I got all the parts (except for the 74LS13) from Mouser for about $15.00.  The 74LS13 is now obsolete but I found a guy that had 25 for a good price so I have a few spares.  The pecker puts out a very strange waveform that is fed into the audio in of your transmitter and then you tune the finals (or your amp) in SSB.  The result is EXACT tuning with almost no stress on the finals, even if you're way off resonance.  You have to use a peak reading wattmeter to tune because the duty cycle of the pecker is so short. The article says 10% but looking at it on a scope it may be less than that.  Using my Autek WM 1 peak reading meter, my TS-520S shows 145 watts in SSB and if I switch the meter over to average, it shows LESS THAN 1... yes ONE watt!  I don't believe the absolute watt numbers are right because of the strange waveform but I do believe the ratios.

I hope this helps others make this great amplifier even better.  With these updates and mods, the SB-220/221 amplifiers should be around for a long time to come.

73,

Steve - N4BOS