[R-390] Filter testing

Sat Aug 9 00:06:04 EDT 2014

Bob,  As long as you're sure c553 is good and the resistance checks out as suggested, there's not much other testing you can do.  So, when installed, the output level of all four filters should be some what the same.  If not, there is a problem.  I had one that was low and it was a bad cap in the filter can.

I copied the following in part from the 'pearls' of wisdom at: 
http://www.r-390a.net/Pearls/IF_deck_filters.pdf

Date: Thu, 5 Apr 2001 15:31:02 -0400
From: "Barry L. Ornitz" <ornitz at tricon.net>
Subject: [R-390] Testing Mechanical Filters
Thomas Warren, W4PG, asked about testing mechanical filters, specifically if there is a
way to do this without rebuilding the stages before and after the filter.
Short answer - yes and no.
Several people suggested using a signal generator on the filter input and a scope on the
filter output. This will tell you if the filter is completely dead, but little else. Kevin,
VA3SU, has the right idea about filter impedance matching. This is an absolute MUST
DO. Without this, you have little chance of getting meaningful results. And even when
doing this, you have do do the tests with great care.
First you must know the filter input and output impedances. The new Collins
mechanical filters of today are generally designed for 2000 ohms input and output
impedance. But the filters of earlier days had widely varied impedance characteristics.
The R-390A filters were designed to be driven from a plate load into the grid of the next
stage. Their input and output impedances are thusly different from most other filters
found in later receivers. The input impedance of the R-390A filters is somewhere in the
vicinity of 5000 to 10,000 ohms (I do not have the exact value handy). The output
impedance is high, typically 50k TO 100k. Both the input and output must be resonated
with external capacitors too, which is different from most modern filters which have the
capacitors built in. The typical (quality) signal generator has a 50 ohm output impedance,
while an oscilloscope generally has a 1 Megohm input impedance. A line level meter will
have its own input impedance too. All of these will need to be matched to what the filter
expects.
For a 50 ohm signal generator, place a resistance in series with its output, the value being
the desired filter input impedance minus the generator internal impedance, or one half this
impedance (see below). The signal generator output meter may be calibrated as the
output voltage into a 50 ohm load, or the output voltage with no load - you need to know
which for your particular generator. If it is into a 50 ohm load, then you will need to
terminate the generator with a second resistor such that the impedance in parallel with the
series resistance of the other resistor and the filter input provides 50 ohms. In this case
the first resistor is the desired filter input impedance minus 25 ohms. If the signal
generator output circuit is calibrated to read the open circuit voltage, the second resistor is
not needed and the first resistor is the filter impedance minus 50 ohms.
Whew! This sounds complicated - lets do an example. Let's assume the filter input
impedance is 500 ohms and the generator output meter reads the voltage into 50 ohms.
The resistor between the generator and the filter is 500 - 50/2 = 475 ohms. The sum is
500 + 475 = 975 ohms. To get a 50 ohm parallel combination, a resistor of 52.7 ohms
will be needed from the generator output terminal to ground. The 475 ohm resistor will
connect to the generator output terminal and to the filter input terminal, the other filter
input connection being grounded. The actual voltage input to the filter will be 500/975 =
0.5128 times the generator indicated voltage, or 5.8 dB lower than the generator output.
Now if the generator output meter reads the open-circuit voltage, the series resistor is 500
- 50 = 450 ohms. The filter input voltage will be one half the indicated generator voltage
or 6 dB lower.
Similar calculations can be done to match the output impedance of the filter.
In the case of an R-390A filter, the 5 to 10K input impedance is so much higher than 50
ohms that you can use a 50 ohm shunt resistor and a series resistor of the same value as
the input impedance with little error in the first case and just a series resistor equal in
value to the input impedance in the second. The filter input voltage will be about 6 dB
below the generator voltage or approximately one half.
Likewise on the filter output with an R-390A filter, you can load the output with a 50 to
100K resistor and the scope probe and not use a series resistor.
To test, tune the generator to 455 kHz and set the output voltage to a convenient value.
[In modern filters designed for low impedance circuits, you should remember to not over
drive the filter or you will get nonlinear responses.] Adjust the filter matching capacitors
to give the maximum indicated voltage on the scope. You can then move the frequency
slightly to measure the filter response.
You might think that a sweep generator on the input and an RF probe on the output
would display the filter response curve. It will, but not without a considerable amount of
care.
First the sweep rate must be extremely low. Then the RF probe output must be
exceptionally linear down to small signal levels.
The ideal way to measure the filter response is with a network analyzer, and I have done
this. I used an HP-3577 network analyzer on a number of filters I have purchased at
hamfests over the years. To get the published response from a narrow CW filter (250 Hz ......

Regards, Larry

On Friday, August 8, 2014 6:55 PM, Norman Ryan via R-390 <r-390 at mailman.qth.net> wrote:

Hmmm, test with no power applied?  Okkkkay...

In that case the quickest way, IMO, would be a simple continuity test with an ohmmeter.  If you read roughly 50 ohms on each coil and close to infinity from the terminals to ground, chances are good that the filter under test is OK.

Bear in mind you have to switch around the bandwidth control to get true readings.

As far as I can tell, the next test is to put the IF deck in the receiver, power it up and check alignment per the manual.  Alignment often isn't needed unless you suspect the slugs and trimmer caps have been messed with.

Replace C553 with a high quality film foil cap such as has been suggested recently by Charles.

Good luck,
Norman

On Friday, August 8, 2014 5:31 PM, rbethman <rbethman at comcast.net> wrote:

>
>
>All,
>
>With an IF deck with no power applied, what would be a test set up, 
>filters in place and wired, that would indicate that they can/will pass 
>a signal?
>
>I have multiple signal generators, and a TEK 5440 scope.
>
>Would setting to 455Kc on the sig gen, first looked at and synced with 
>the scope, then applied to each filter input, and look at the output do 
>this?
>
>I'd also ask if it really matters which direction the signal is applied?
>
>Bob -  N0DGN
>______________________________________________________________
>R-390 mailing list
>Home: http://mailman.qth.net/mailman/listinfo/r-390
>Help: http://mailman.qth.net/mmfaq.htm
>Post: mailto:R-390 at mailman.qth.net
>
>This list hosted by: http://www.qsl.net
>Please help support this email list: http://www.qsl.net/donate.html

>
>
>
______________________________________________________________
R-390 mailing list
Home: http://mailman.qth.net/mailman/listinfo/r-390
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:R-390 at mailman.qth.net

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html