Waters 334 Dummy Load Wattmeter Question.

[w5kcm]

I am in the middle of repairing an old Waters 334 dummy load wattmeter
for a local ham. He had picked this up at a hamfest and the meter
didnt work. I have removed the oil tank and cut an opening in the back
to access the internal components. The oil has been drained into jars
and I will put the oil back after I make the repairs.
Ok, here is the question.
The failure is the diode. it looks to have the original diode. Black
glass marked SG5409. I do not see a modern cross or the exact type
still available. Is there a better or more durable RF diode that will
work in this circuit? Temporarily, I soldered in a 1N4148 and this
seems to work well at the 100 watt level. My concern is at a KW will
this hold up? What about the possibility of stringing several 1N4148's
in series to increase the voltage capability? I want to put in a very reliable diode before I seal the tank up again.  I know that most of the
RF is going into the 50 ohm resistor and the diode is just seeing a
small sample. The diode is in series with four 2.7 ohm resistors and
then has a 15K fixed and a 10K pot before it gets to the meter. This
is on the 1KW position. Any suggestions will be appreciated. 
73, Randy, w5kcm

[N3DRB The Derb]

I'd worry more about if all the fixed resistors are still on the money. most modern diodes are pretty stout devices. You'll nned to cal the thing against a known standard like a bird setup with the new diode. unless I'm missing something, I think almost any diode would work FB.

[KD6VXI]

Most wattmeters use a small signal (germanium, schottkey, etc) diode for the rectifier, so that they work well at the QRP and QRPp levels. 

The coupling network, be it a bridge, xformer, etc. is usually what limits the signal to the diode... IE, the range switch is between the RF sampler and the rectifier.

I have an Altronics, good to 50 kw, but it doesn't include a wattmeter.  I also have a legal limit oil cooled wattmeter, a Barker n Williamson that I use for every day usage, and it is as I stated above...  The detector is a  small signal diode, and it has 10, 100, 300 and 1kw, all AVG scales....  It goes from a resistor network that selects the proper range, then to the detector... The detector is thus ALWAYS in it's linear range (in theory, at least).


--Shane

[w5kcm]

Thanks Shane and Derb for the responses. I think I will try a geranium diode in there and see what happens. 

[w5kcm]

Ok this is what I ended up doing to the Waters dummy load wattmeter. Possibly in the future someone may want to fix one.
On the back of the oil tank where the thermal switch and the so239 connector is located, at the top of the back is located a small hole plug that is soldered sealing the oil. I used my 100 watt soldering gun to get the solder soft to pry this metal cap/plug. The oil (Toxic Transformer Oil) can then be drained from this hole. Using a very small funnel some alcohol or other light solvent could be put inside to swish around and clean out the remaining oil. Now, I cut about a 2" square hole in the back using a small stone type cutting blade in the drimmel tool. You will have to work inside at the front and back ends of the tank, so cutting a larger hole or cutting holes in both ends may help. I removed all the internal (tank) meter components. The RF is sampled off the large dummy load resistor near the grounded end so the RF level is pretty low. I used some small teflon 50 ohm coax and soldered the center conductor to the connection on the 50 ohm dummy load resistor. Soldered the shield to the metal ground connection. Now cut the coax and trimmed it back to solder the other end to the small feed through capacitor at the back. Solder the shield to ground on both ends.
Now, cut a square of flat copper sheet to fit over the hole in the back and soldered all around this until you have a solid complete seal. Put all the oil back into the tank and resolder the hole plug. Checked for leaks and if all looks well you are ready to build the meter circuit. The meter circuit is built on a small 5 or 6 point solder terminal strip. I found a couple Germanium diodes, ECG/NTE 109. These have a 100 volt rating. Not knowing exactly how much voltage I would have at the 1 KW level, I put the two diodes in series for a 200 volt capability. Used a couple .01 caps, one at the 2 diodes junction and one at the output to ground. Connect the coax from the feedthrough at the back of the tank to the anode of your germanium diode. Ground the shields at both ends. Mount the terminal strip on one of the meter mounting bolts and connect the output (diode cathode) to the small wire that used to be connected to the feedthrough on the back of the tank. You can now apply RF and calibrate the meter on each range. If you use 100 watts it should be fairly accurate up to 1KW. Doing this mod should be no problem on the HF bands, using this on VHF or UHF may effect the meter accuracy. 73, Randy

[K1ZJH]

You should be able to calculate the peak RF voltage by simply measuring the resistance at the tap point on the non-inductive resistor and going from there based on a 1-kW input rating.

Pete

[WA1GFZ]

germanium diodes are junk. throw in some 1N5711s if you want low level accuracy.
I use them in my HB watt meter. In 25 years only lost the reflected diode due to a lightning hit with the antenna connected. 1n4148s are very good diodes but have a .7 volt junction and may not give accurate readings at low power. 5711s are hot carrier with about .4. My Watt meter design is a copy of the Collins but scaled for higher power.

[w5kcm]

With the meter circuit now outside of the oil tank. It would be easy to measure the peak voltage on the diode and try different type of diodes if needed. I gave the Waters back to the owner and he pumped 1 KW into it and reported that all is well. If it should fail later on. I will try other diodes. Thanks for all the comments. 73, Randy