From: Garey Barrell 
Subject: Re: [drakelist] R4C xtal bank bad


Garey Barrell  made an utterence to the drakelist gang
----------------------------------------------------------------------


Eric Webner wrote:

> Garey,
>  
> Several weeks ago when I discovered the play in one of the bandswitch 
> wafers, I set the wafer where it should be and used adhesive  on the 
> inner wafer and shaft. It appears to work very well, and with no 
> binding. I wouldn't recommend that for all the wafers, but just one 
> seemed ok.
>  
> Regarding my low output, I may be ok. First of all, I was not loading 
> it to full power, since one of the errata sheets that Drake put out 
> said to load it for only 340 mA in order to avoid any spurious output. 
> After reading your latest message, I loaded 80 and 40 meters to full 
> power, obtaining 135-140 watts measured on my Bird wattmeter. Loading 
> it to full power ran the plate current to 500 mA!

A plate current reading of 500 mA tells you that your 3.3 ohm meter 
shunt has gone up in value.  Probably to about 5.6 ohms!  :-)

>  
> I also took your suggestion and checked a few components. R44, the 680 
> ohm resistor, is good. The cathode resistors are good and very well 
> balanced. The screen resistors checked out as well. However, the 
> resistor across the meter measured 5.6 ohms. I disconnected one end 
> from the circuit board and the resistor snapped in half. I couldn't 
> come up with another one, so I hope to try a 2.2 ohm unit (measures 
> around 2.6) so I can get on the air this weekend.

That will "work", but you really need to replace it with a 3.3 ohm 
unit.  For "emergency" purposes, you can measure the 2.6 as accurately 
as you can. then adjust the bias voltage for 100 mA of current through 
that resistor by measuring the voltage drop across said resistor as 
accurately as possible.  Then note the reading that gives you on the 
plate meter.  If it reads 50 mA (or whatever) that becomes your "bias 
adjust" setting for the finals.  Then just Tune and Load for maximum 
output, using the plate meter to adjust for minimum plate current.

Then, when you get a "real" 3.3 ohm resistor in there, again accurately 
measure 100 mA through that resistor and pick a "select at test" 
resistor to make your meter read 100 mA.  I wouldn't worry too much 
about the "340 mA limit", since the meter isn't that accurate anyway.  
The only way to determine the "cleanest" operating point is with a 
spectrum analyzer, and it just isn't that critical.  If you're driving a 
5 KW amplifier then you need to worry about it.

>  
> Do you recommend that I vary the other resistor in order to achieve 
> accuracy on the meter, or can I play around with values on the shunt 
> resistor (the 3.3 ohm resistor)?
>  

See above.

> One other observation: my low power output could be attributed to this 
> meter shunt resistor being off, as that would cause me to back off on 
> the loading quite a bit. This is also why I didn't want to load it to 
> full power (500 ma???) for fear of melting a couple sweep tubes. It 
> also makes me wonder about my bias. Is there a voltage one should 
> shoot for, or is getting the bias (idling) current still the 
> preferable approach? Do I need to realign everything if I need to 
> reset the bias?
>  

This is exactly the case.  When you thought you were loading to 340 mA, 
you were probably only drawing 150 mA, resulting in low  output.   THE 
MOST IMPORTANT CURRENT READING IS THE BIAS SETTING.  This is why you 
"calibrate" the metering circuit at 100 mA.  These meters only cost 
about $1.25 and are NOT highly accurate, linear meters.  So we make a 
"reference" point for the  _important_  meter reading and use it as an 
"indicator" only for the far less important absolute plate current 
reading at full power.

73, Garey - K4OAH
Atlanta

----------------------------------------------------------------------
On Behalf of Garey Barrell 

---