Mods to Phillys

[Joe Miller]

Were Philly's modified in later production runs without updating the schematic?

I found the IF AGC tube's cathode grounded rather than connected to ground through a 600 ohm resistor with a bypass cap as shown on the Rider's schematic.

Was this a factory change or the result of some one before me making changes?

Thanks,

Joe

[Bill Liers]

Joe,

The chokes are captive to the can and the caps are soldered to the choke leads so, the leads all have to be fed through when removing the can. Happy capping!

[Scott Seickel]

There is a recent thread called "Bass Choke Renewal" that has a couple of pics and may be helpful.  I would post the link, but it looks like a copy and paste does not work on this forum software.

[Bill Liers]

Hope all had a relaxing holiday. I have the Philharmonic I'm presently restoring up, aligned and running well. I measured the sensitivity of the radio at 1 MHz. and she is down under 1uV, so I'm sure we're ready for some noise testing.

I have some recently installed lighting in my kitchen which is very noisy in the middle of the broadcast band. Monitoring the junction of R23 &24 you can observe the combined audio and detected DC from the 4th IF. On the Sens. diode, the sensitivity side is jumpered so there is no difference in signal level with the tube in or out.

The first trace shows inter-station noise plus large spiking at 120 Hz. with 1 volt per box indicated and 6H6 removed. Audio at this point is harsh. The second trace shows tube installed and warmed up. Noise slowly diminishes as tube warms. As you can see, the spiking is cut almost in half.

The third trace shows audio plus noise on a low power station. Noise peaks vary at this point but large ones get through and one can be seen about mid screen. Audio is noisy but can be understood with the tube removed. With the tube back in, the spiking is limited to approximately the peak audio value and although still noisy, is much more comfortable to listen to.

Result: Certainly better than nothing for some types of listening and good to know that Scott didn't stoop to simple marketing tactics. I fail to see however why this would be called a "Dickert" filter.

[Joe Miller]

Bill,
Thanks for the info. I still don't see how the sens diode can ever conduct?

BTW what alignment procedure did you use.? I know Norm has documented his approach (I'll look for it). Did you use his?

Joe

[Bill Liers]

Joe,

Apparently, some strong impulse (very short duration) signal can get through can get through as the detector is just beginning to conduct on weak signals. Imagine the tube detector diode as a S.S. diode with a small value of capacitor across it. During this time. R23, 24 divider polarity is favorable to the sens diode turning on and thus shunting the audio line to ground for an instant. I'm surprised that it works as well as it does.

For the alignment, I start with the procedure in the Philharmonic technical document, which effectively peaks the IFs at the minimum bandwidth setting. I then run a gain and sensitivity check on the IF chain. If I feel that a tube is contributing too much noise, I swap it out and start again. I then run my signal generator back and forth in small increments to see that the gain falls off evenly on both sides of center frequency. This is a critical step for best result.

From there, you can start down the long road of checking for symmetry at each step as you increase bandwidth. If for instance, the skirt of the bandpass drops off slowly on one side, you are allowing more noise through than necessary which usually goes along with a poor roll-off in frequency response etc. There are no specs for this of course, but I find that if symmetry is maintained, the system works well and sounds good.

I also found an RF spectrum analyzer a while back for a song on ebay which needed only a power transistor. With this luxury item, you can watch exactly what you're doing. It sure makes alignments go faster.

regards

[Bill Liers]

While we're on the subject of testing, I decided to check the limits of overload and distortion at the detector. There had been an article published a while back which stated that Philharmonics had "an inherent tendency to overload". I've felt that any distortion cause by high signal levels must be due to a lack of control in the RF and IF AVC amplifiers. Since the Philharmonic I'm working on had a few noisy tubes and a few weak ones, NOS tubes were selected for every position from first RF to detector plus the AVC amplifiers. This gave the opportunity to test a set with maximum gain for the ability to control such as well as possible.

Once the signal was increased to about 20 uV, the detected audio was sufficiently noise free so as to show good waveform. From there, the signal was increased progressively until distortion was evident. The photo attached shows the detected signal at an input level of 100,000 uV. Distortion became evident above 500,000 uV., nearly half a volt at the antenna terminal. Now that, my friends, is a radio.

[Norman S Braithwaite]

Bill:

Thanks for posting the test results.  How would the change in voltage divider resistors employed by EH Scott Radio Laboratories for the later production affect impulse noise suppression?  Some time ago I learned of the "Dickert" noise suppressor option in Scott literature.  I will have to re-read the literature to find the reference.

Joe:

The information that I posted for alignment reflects the Scott procedure but in a manner that is easily understandable.  Working from the original Scott instructions becomes a research project each time you need to align a Philharmonic receiver.  As with the instructions for the Full Range High Fidelity Receiver (AW23), the Philharmonic instructions are "sufficient" for field service but they are definitely not the procedure used before the receivers were shipped.  Balancing the aprons of the IF bandpass characteristic is very important for good quality AM reception.  Unbalanced aprons yield high audio distortion as well as noise and adjacent station interference.  The IF alignment method described by the EH Scott Radio Laboratories does not address balancing the aprons of the IF bandpass.  The best method today is by sweep alignment, initially peak aligning as described in Scott literature then backing off some adjustments slightly to balance the aprons.  If sweep aligning, you can see how minor adjustment of each trimmer affects the bandpass characteristic.  Three adjustments change the center frequency, three affect one apron, and three affect the other apron.  I find and adjust the trimmer or trimmers that have the greatest influence for balancing the aprons.  Adjustment of this trimmer or trimmers will affect a minor reduction in the peak (receiver sensitivity) but a satisfactory balance can be found with minimal reduction in the peak.

Norman

[Bill Liers]

Norman,

I would say that the change was less to do with the suppression and more likely a change to increase audio level slightly. Is it possible that the sets with the newer values also had a tapped volume control?

[Norman S Braithwaite]

Bill:

Going from recollection (I am away from my resources) I believe the revised voltage divider resistors were in sets with tapped volume controls.

Norman

[Bill Liers]

Norman,

Of course, the tapped control will reduce the perceived level through the first third of the controls operation with the drop in mids and highs. I noted this when changing the control in this radio to the tapped type. The divider change would have "normalized" this. That would be my guess.

If you could confirm the correlation, it would cement the theory and be good reference material.

[Norman S Braithwaite]

Bill:

I will check and report back when I return from visiting relatives.

Norman

[Norman S Braithwaite]

Yes, the revised resistor values correspond with use of the tapped volume control.  One more circuit mystery solved!

Norman