Heathkit SB-230
Last updated   Home


Heathkit SB-230


Panel

Go to the SB-230++ page

Contact: 


The Heathkit SB-230 is a linear amplifier for amateur radio use, covering the 80, 40, 20, 15 and 10 m bands. Its maximum input power is 1200 W PEP in SSB mode. It requires less than 100 W drive to produce full output. The SB-230 was sold in kit form by Heathkit from 1974 to 1978. Occasionally unbuilt kits can still be found and they sell at record prices in auctions (more than $ 800 on Ebay).

The SB-230 uses one single 8873 EIMAC triode in grounded-grid configuration. The 8873 was designed for conduction cooling. A finned radiator is mounted on the amplifier rear panel. A square block of BeO (Beryllium Oxide) is used as thermal link to transfer heath from the external tube anode, which is at RF and HV (2300 V) potential, to the radiator (at ground potential). BeO is an excellent electrical insulator and also has very good thermal conductivity. A thermal silicone compound, as used for power transistors, improves the heath contact of the two faces of the BeO block with the anode and the heath sink respectively.

8873 tube. The flat anode
ensures good contact with the BeO block BeO block

Unfortunately BeO is a highly toxic and carcinogenic material, so the assemby manual recommends extreme care in handling the block. With today safety standard is doubtful if a similar kit could be sold to the public. The SB-230 was the only commercial ham linear amplifier, together with the Henry 2K Ultra, that used this tube.

The tube anode temperature should never exceed 250C. A thermal switch on the heath sink turns off the tube in case the temperature rises beyond that value.

The 8873 tube is electrically equivalent to the 8875 and 8874 (3CX400A7) and differs only for the cooling method. The 8875 uses transverse-flow forced-air cooling whereas 8874 uses axial-flow forced-air cooling. The 8873 has the lowest anode dissipation: 200 W. The 8875 has 300 W and the 8874 400 W. Of those EIMAC tubes only the 8874/3CX400A7 is still in production and can be purchased new, albeit at a price around 400 USD. The 8873 is rare and can occasionally be found on the Internet auctions and at flea markets. NOS tubes are sold at high proces.

EIMAC LOGO 8873 8874 8875 Data Sheets


SB-230 Specifications

1974 list price: $ 339.95

The SB-230 is capable of 600 W PEP output at 50% duty cycle in CW mode. This corresponds to 1000 W DC input quoted above, with an amplifier efficiency of 60%. In SSB mode the PEP power is around 700 W. At 100% duty cycle (RTTY or in key-down condition) the amplifier must not exceed 400 W input, corresponding to the 200 W anode dissipation with some safety margin.

See here for the schematics.
See here for the Heathkit assembly manual.
(in case of broken links and if you have difficulties in locating them I can email the files to you)
Substituting the 8873 tube

The SB-230 has many estimators because of the elegant design, the simple circuit and the relatively roomy chassis that makes mods simple. However the scarcity of the tube poses problems. Many projects exist to replace the 8873 with something more readily available. They fall into two broad classes: Useful links to SB-230 retrofit projects can be found below.
There have been speculations on the possibility of modifying the 8874 for conduction cooling. This would require to replace the finned radiator in the 8874 with a solid metal block (possibly a salvaged 8873 anode). Although possible in principle, the operation requires industrial facilities not available to the normal amateur and the risk of damaging the (expensive) 8874 tube is high. Thanks to Tony I0JX for this information.


My SB-230

I purchased my SB-230 on Ebay (USA) and the shipping to Europe cost almost as the beast alone. The seller had no idea if the unit was working (except for lighting up) so the risk of a DOA (Dead On Arrival) was big.

Luckily the BeO block was intact and the tube lighted up. Then I could proceed to more serious tests. The HV supply worked, but the meter read about 22% low. The reason: the three 1 Megohm resistor had increased value because of age. Replaced them with "new" ones (but try to find 2 Watt Allen Bradley resistors today!)

I am grateful to PA3GOS and K4POZ who helped me with advice and support during the restoration and the various tests.




catalog page Initially I could not get any grid-current reading. I found the grid feed-thru capacitors were leaking to ground. I replaced them with ceramic discs and tightened up properly the grid contact ring (which was loose). In the process I learned that you must be very careful in putting the thermal paste between the anode and the BeO block: too much will squeeze out on the BeO block sides and can cause an arc between the anode and the radiator flat face. I use Dow Corning 340 Thermal paste as suggested by EIMAC.
catalog page Following tests indicated that the SB-230, once excited, developed terrible parasitic oscillations (tens of watts). I failed to destroy my FT-450 exciter. The 33-ohm grid resistor disintegrated. Following K4POZ's suggestion I changed the wiring of the 8873 plate circuit. The assembly manual illustrates the two possible schematics but it does not explain which to select. In my case the change resulted in stable conditions so I strongly recommend it.
Click for larger image The default anode circuit wiring according to the Heathkit assembly manual. The tuning capacitor is connected directly to the 8873 anode.
Click for larger image Conduction cooled tubes gained some popularity in the seventies and eighties because of their noiseless, solid-state-like operation. The other linear amplifier using the 8873 was the Henry Radio 2K Ultra (see picture) introduced in 1972. The 1980 ARRL Radio Amateur's Handbook featured a 2 kW linear amplifier project also using a pair of 8873. The 8560A tetrode (similar to the 4CX250B) was also conduction cooled and a project using it was presented in the same Handbook. The 8560A still appears in the Svetlana catalog.

The SB-230++ Project

Finally after several tests I concluded that the 8873 tube was exhausted. I could not get more than 200-300 W out of it. After some fruitless attempts to purchase a replacement I decided to substitute a russian GI-7BT tube. The refurbished linear became the SB-230++.

Click here to go to the SB-230++ page (under construction).


HOME
© 1994-2011 Televideo.ws