[Boatanchors] AR 88 Discussions

[Richard Knoppow]

----- Original Message ----- 
From: "Mark Johnson" <mvjohn at sympatico.ca>
To: "Boatanchor List" <Boatanchors at mailman.qth.net>
Sent: Sunday, February 02, 2014 2:57 PM
Subject: [Boatanchors] AR 88 Discussions


> Team,

     I may be able to give you some hints. I finished 
rebuilding an AR-88F recently but am still detailing it. I 
avoided removing the gear box but think will have to in 
order to get it as smooth as I think it should be.
     The F version has some differences from the A and D 
versions. It was designed to be part of the DR-89 diversity 
receiver. This consists of three AR-88F receivers along with 
the combiner, tone keyer and monitoring units.  The 
receivers are somewhat modified for use in this system. I 
think the most important modification is to the AVC system. 
The first and third IF amplifiers are connected so that 
their cathodes return through a 5K pot located on the front 
panel and used to trim the gain so that none of the 
receivers dominates the diversity combiner. In addition the 
fixed bias on the second IF is changed by going to the gain 
pot through a 15 ohm resistor and bypass capacitor. In 
standard receivers this cathode is returned to ground 
through a 150 ohm resistor which is unbypassed. The reason 
is to provide some feedback to compensate for any Miller 
effect detuning of the IF transformer.  Another change is 
the application of a small amount of positive bias to the 
AVC buss to reduce the AVC delay.  On normal receivers the 
minimum voltage on the bias line is about 1.8 volts. This is 
reduced to around 1.0 volt with the modification. I am not 
sure of the reason for this change. My guesses are that it 
makes the diversity switching more sensitive. It also lowers 
the noise level of the receiver very slightly.  I have 
removed these modifications on my receiver because my 
measurements show they make overloading more likely. 
However, they can very easily be reinstalled. I also shorted 
the trimmer pot because I kept bumping it. It has no 
function on a stand alone receiver. The line to the trimmer 
is made from the line that was originally intended to go the 
a signal strength meter.  I was able to obtain a meter and 
installed it.
    The IF can be aligned without a sweeper.  There are 
fairly complete instructions in the British versions of the 
AR-88 manuals as well as in the CR-88 manual.  The key is to 
do each of the overcoupled stages separately. The four 
transformers are first peaked at the crystal frequency while 
in the No.2 position. Then put the selectivity in No.1 and 
investigate the peaks using a counter on the generator. One 
will be quite a bit larger than the other. Peak this, it 
will not come up to the first peaks amplitude. Now, measure 
the exact separation of the two peaks, it should be exactly 
12 khz. If not go back the larger one, set the generator 
exactly 6 khz from the established center frequency (the 
crystal filter frequency) and peak it. Then go exactly 12 
khz to the other side and peak that. You will find that it 
takes several alternations to get the two peaks to be 
exactly the same amplitude and exactly 12 khz apart. After 
this check in position 2, you should get two peaks of the 
same amplitude by separated by about 7 khz.
    The remaining IF transformers are not overcoupled and 
can be just peaked on the center frequency.  Once all are 
peaked properly you should be able to see a relatively flat 
topped response in position No.1 which is about 12 khz wide 
at the corners and about 16 khz wide at the 6db down points.
    I used a Hewlett-Packard 606-A coupled using scope probe 
as the source and a Hewlett-Packard 400H VTVM with a scope 
probe as the detector. Of course I am reading RF and not DC 
but it makes no difference. On one stage I was getting a 
rapid variation, not sure what was causing this.
    Now, more.
    The dial calibration of my receiver also is off in the 
mid range. The spec is 0.5% and it does easily meet the spec 
but I am suspicious that the tuning capacitors are not quite 
centered even though they _look_ centered. For one thing 
they are stiffer than I think they should be.
    The method of checking centering, assuming no plate 
bending has been done, is to set the cap for maximum 
capacitance and adjust the bearing for _minimum_ 
capacitance. Because the capacitance is proportional to the 
inverse of the spacing it is not a linear function. If you 
think about it as the plates of the rotor are moved toward 
the plates of the stator the closer side will increase 
capacitance faster than the further plates lose it. So, the 
minimum is when they are exactly centered. I have not done 
this yet on this receiver but have done it to others.  The 
AR-88 has two sections of variable capacitor; one is for the 
oscillator and mixer stage, the other for the two RF stages. 
The oscillator should be easy to do by listening to a 
frequency at the low end of band three or four. The RF 
section will take some other method. I am going to see if I 
get a sufficient indication using a grid dip meter to do the 
adjustment.
    If my suspicions are correct the result of the centering 
error will be that when the ends are set correctly the 
center will read low which is what I get. The cap on my 
receiver has inspection or lock paint on the bearings of 
both sections and on the screws securing the frame so its 
possible its just as good as it can get. The stiffness 
bothers me.
    The AR-88 is an exceptional receiver. It has quite low 
noise level, very good sensitivity right up to 30 mhz due, I 
think to very low loss coils and good RF tubes, it is 
astonishingly stable, all the tuned circuits have 
temperature compensating ceramic caps.
    The receiver was designed without any electrolytic caps 
and except for the headphone jack, meter on sets which have 
them, and the gain trim pot on the F nothing is mounted 
directly on the front panel.  The bare chassis weighs about 
80 lbs and a receiver in a desk cabinet weighs about 100 
lbs. The bypass caps are oil filled paper in tubs. Most of 
these have degraded seals and have leaked oil. Be careful, 
the oil may be a pcb. I have had no success in restuffing 
caps so replaced mine with film caps mounted on terminal 
strips using the hardware from the old caps. The RF 
compartment may have Micamold or Solar paper caps. These 
look like mica caps but are not and will be bad. Replace 
them with film or ceramic caps. Some early AR-88s have 
lozenge shaped caps of an odd pink-violet color. AFAIK, 
these are RCA made mica caps (not silver mica but stacked 
mica) and are probably still good. I think mica was in short 
supply during WW-2 and so the paper caps were used.
        The audio amp is single ended by has enough feedback 
around it to keep the distortion low.

  There is a great deal of information on the AR-88 and its 
relatives at the Western Historic Radio Museum site.  Some 
of the handbooks are at BAMA but look under the military 
category as well as RCA.  Again these are outstanding 
receivers and worth a lot of work to make right.
     I am typing off the top of my head and have probably 
forgotten something.


--
Richard Knoppow
Los Angeles
WB6KBL
dickburk at ix.netcom.com