Quad 837 Grounded-Grid Linear Amplifier

(probably built by Roger A. Jacobsen, K9KLN)

front_view

(top metal screen, side skin & tubes removed for better view of components)

Rear panel showing pl-259 RF Input, 115VAC 4-pin Jones Plug & HV B+ Input jack	Side view with skin removed showing tube socket and meter wiring
Top View with metal screen removed	Detail of tube shields and plate choke
Detail of switched-capacitor output tank	Detail of pi-network with sliding-tap coil

Schematic diagram: created using 'xcircuit ver. 1.7' -- thanks to Tim Edwards for a wonderful program
(fan omitted in drawing)

This amplifier is similar to other published designs according to references in Usenet posts:

  1. E. L. Hoover, W9SAR, and R. L. Peck, W9MOW, "A 200-Watt Grounded
     -Grid Linear Amplifier," QST, June 1955, pages 21-22, 128;

  2. An 807/1625 version described in the 1957 ARRL Handbook

  3. "Sweep Tube Linear-Amplifier Design" by Doug DeMaw, QST, July 1968,
    reprinted in the last ARRL SSB manual mentions the design.

I would appreciate hearing from anyone who recognises the chassis layout or
other aspects of this amplifier's design; if you have access to the 1957
Handbook I'd appreciate knowing if our's looks like the one described therein.

Have you seen our design elsewhere?

We received this nice response to our requests; it is from a U.S. ham (w0 call) (name withheld pending permissions):

I built this one as a kid just after acquiring my General Class Amateur Radio License. It's done on a hand made steel chassis, housed inside a Heathkit HX-20 transmitter cabinet. In those days people generally used 811As or UE572s for this sort of thing but I liked the 837's indirectly heated cathode (thus no filament choke is required) and its modest filament power requirement As vacuum tubes go, the 837s made for a pretty compact amplifier.

My design is quite similar to the one shown on your web site with the following exceptions:

1) For safety, rather than metering plate current, mine uses cathode circuit metering.

2) My output PI network uses switched taps on the inductor, rather than a roller mechanism.

3) The output capacitor in my PI network is a variable unit augmented by several switched
capacitors.

4) In addition to covering 80/75, 40, 20, and 15 meters, mine also works on 10 meters at
reduced efficiency (the output capacity of the parallel 837's is pretty high for 10 meters).

My power supply is homebrew as well and provides a switchable 1200, 1500, and 2000 VDC plate voltage. All 837 tube designers are no doubt now spinning in their graves at the thought of a 2 KV plate voltage but I never had a moment's trouble with it in low duty cycle, SSB service. Key down, it drew about 500 MA with 2 KV on the plates. I suspect that this power level
would have been unacceptable in CW service and other, higher duty cycle modes but I was always very careful with this amp and ran it mostly with 1.5 KV on the plates. Cooling was provided by a small fan that I stole from the freezing compartment of an old refrigerator.

Being a destitute electrical engineering student, I never was able to measure output intermodulation distortion or, for that matter, even output power. The 837, of course, was never designed for zero grid bias, linear operation so this again dictated care in operation. I drove it lightly with less than 50 watts from the single 6146 in my HX-20 transmitter. In any case, on the air reports were uniformly good with regard to audio quality and power gain in on/off testing. I suspect that the amp provided around 600 watts of output on the lower bands, with 2 KV on the plates. It
probably was not far behind the 811A amplifiers of the day.

The power supply filter capacitors will need replacing but maybe I'll put it back on the air someday. In the meantime, I'll just look at it occasionally and smile. It cost me maybe 50 bucks to build and I still have four good, spare tubes.

Please e-mail 'msg _AT_ cybertheque _DOT_ org if you have
any other information on the design, chassis layout or experience
operating this type of amplifier.