[R-390] R-392, IF stage gain

Sun Jan 9 04:25:01 EST 2022

Hi Dave,

Thank you for having given a so detailed reply, you are very kind.

I shortened still long my message to be not too much tedious, but I lost really a lot of time on this point, especially to fix the AGC voltage drop problem, that, at the end of the day (=many weeks 😀), resulted to be not a problem. According my tests, the AGC voltage drops because the IF amplification drops with the temperature, but that has no effects on the sensitivity, i.e. on the S/N ratio.

I also went thru the analysis of the capacitors inside the IF transformers. I don’t know the ZM-30 but have a ZM-11/B, that, I suppose, is the father of ZM-30. It allows to perform the test you explained me, applying a voltage up to 500V and going down to measure the effective resistance of the device under test. ALL my IF capacitors were rather bad, with an insulation resistance from 300 to 700 k ohm. I GOT IT! I thought and replaced all of them (almost 80€…). But not, I could not notice any change in the practical behavior. So I concentrated on a single IF stage (V506) and changed every passive component. To be surer that the temperature had no effect on the resistors, I carried them out of the deck. So, the only component that remained original was the T601 coil (T601 has no secondary winding, the signal is taken out by mean of a capacitor. The one that seems to be the secondary, is only a RF impedance). And fully removed the rectifier V603. I also checked the tubes directly in circuit and I also separated them in classes.

The problem is absolutely the same: when I apply the heat to T601, the V506 gain goes down…. (It can be seen as well on the oscilloscope) But, I as I could check, this phenomenon does not seem to affect significantly the receiver sensitivity, as you can easily check by adjusting the gain of V505 by mean of the potentiometer R532, that allows a wide range of amplification change but that is not very effective on the sensitivity of the receiver. So I archived this problem and (almost) forgot it.

Coming back to V503, my opinion is that the manual is not so perfect as I could imagine and the stage gain is only indicative. In my case, V503, as you said, seems to work correctly. The voltages are OK and nothing suggests that something is not operating properly, or at least that is the conclusion of a tube-beginner like me… Furthermore I noticed that also at Collins they could have had some doubts, because in the later models they added a second potentiometer on the cathode of V504, that, according my opinion, was added to compensate for the low gain of V503.

However, we mad guys don’t like to give up and I have furthermore the temptation to buy some new 26A6 and continue to investigate…😜

Thanks again, yours

Gianni

> Il giorno 8 gen 2022, alle ore 21:20, David Olean <k1whs at metrocast.net> ha scritto:
> 
> Hello Larry and Mr. Becattini,
> 
> I do enjoy playing with the R-392 receivers and have brought four of them back from the dead.  That being said, I always have taken the alignment instructions with a grain of salt.  I usually skip trying to couple in with a small coupling capacitor if I can avoid it. What I can do is to try to duplicate the coupling method used by Mr. Becattini and record my results. The IF tubes running on 28 volts can show some variability and bear no resemblance to what a tube tester might say. The best way to test them is in a unit and monitor relative stage gain as you change tubes.  I sorted all my tubes and color coded them with dabs of paint.  Now all I have to do is not lose my color code!!
> 
> The important thing to do is have the IF gain peak right at 455 kHz. An accurate 455 kHz signal is important.  Do not rely on a dial setting. I do the same thing with the PTO. I take it out of the receiver and install it in a test fixture with an accurate dial so I can check the end points and the linearity using a good frequency counter. After that, I have never worried much about stage gain and concentrated on whether each stage seemed to be working properly.  The IF stages all have silver mica capacitors resonating the IF windings and I find that these can cause trouble in the following way. At low signal levels, all looks good, but, when you increase the RF signal level, strange things start to happen. You will hear a crackling noise and your AGC voltage will not increase properly as the levels are increased.  You might also find that your AGC voltage will drop over time with a constant rf signal applied, as components get warm. I spent a number of hours removing Vitamin Q capacitors in the AGC circuits  and testing them to find that they are perfect after all the many years since they were manufactured.   I did find that a few silver mica postage stamp capacitors inside the IF cans would start to leak or actually arc over at very low RF voltages. I have two capacitor checkers, a Military ZM30 bridge (which is very cool) along with a General Radio 1657 RLC bridge. They both give similar and very accurate readings, but the crackling does not show up when testing the capacitors as there is no significant RF voltage across the capacitors during the tests!!.
> 
> My solution is to use a regulated adjustable HV supply and connect the positive lead to one side of the capacitor, and then read the DC voltage with a very high impedance voltmeter on the other lead of the capacitor. I use a Fluke DVM and reference it all to the negative lead on the power supply. Raise the voltage slowly and you can watch for leakage on the voltmeter. I found that some of the capacitors were fine up to about 8 to 12 volts! Above that, and they started arcing and making the crackling sound. A new capacitor fixes everything.  This method of supplying a high DC voltage across the cap is an accurate and effective way to test for leakage. It will expose problems not seen with RLC bridges even if they can measure dissipation.  There are SM capacitors in the RF coils as well, and, if you pay attention to which bands seem to have the noise, you can localize it to a particular set of coils used for those bands. A high frequency scope will allow you to find the right stage. It can be difficult especially in that chain of IF stages in the R-392, to determine where the noise is coming from.  A bad cap in the plate circuit will also show a little noise back into the grid stage of the same tube, so always keep that in mind.
> 
> Once I had all my IF and RF stages working properly, I never bothered with checking the stage gains. I figure that, if the receiver will not meet its' performance levels, then I would check into it further.  These radios have no problem meeting their spec, and you can run into a point of diminishing returns very fast.  If you note that a stage does not tune up properly, than that is cause for worry.
> 
> I will play around with a nicely working R-392 and get back to you all with what I find as compared to paragraph 104.
> 
> Dave K1WHS
> 
> On 1/7/2022 10:50 PM, Larry H wrote:
>> Hi  Dave,  It looks like from your posts that you have some R-392 experience.  Would you mind looking at the attached and see if you can help  him out.  Nobody has responded to his request.  Thanks.
>> 
>> Regards, Larry
>> 
>> ---------- Forwarded message ---------
>> From: *GIOVANNI BECATTINI via R-390* <r-390 at mailman.qth.net>
>> Date: Wed, Jan 5, 2022 at 6:15 AM
>> Subject: [R-390] R-392, IF stage gain
>> To: <R-390 at mailman.qth.net>
>> 
>> 
>> Hi,
>> 
>> I am sure a little bit pedantic and tried to check if my R-392 (1951, Collins) fulfills the IF stage gain chart (par. 104 of TM 11-858). I started from this point a long sequel of tests and component substitutions, with some obscure behavior.
>> 
>> First of all: how do you do the tests? The manual is not precise about that. They say to apply the siggen output thru a 50nF capacitor. But in this way the generator output impedance is not matched, so, when you think you apply n-microvolts, you are really applying the twice and the value of the table are widely exceeded. So I simply multiplied by two the nominal output values of the signal generator.
>> 
>> In this way, I am “rather" aligned with the requirements (4KC BANDWIDTH) for the last 3 tubes:
>> 
>> V506: 46 effettive mV (already doubled), it should be 50..70mV
>> V505: 23 effettive mV (already doubled), it should be 22..32mV
>> V504: 32 effettive mV (already doubled), it should be 25..40mV
>> 
>> But to achieve that, I had to reduce to minimum the gain of V505, acting on R532 (the potentiometer in series with V505 cathode).
>> 
>> Also the first two tubes V501 and V502 have a right gain. The problem is V503. I am rather sure that every component is OK but, if I want to keep on the table, I had to add a 3K3 resistor in parallel with the 6K8 to have the about 1 correct gain factor.
>> 
>> To be short: V503 stage is gaining too little, V505 is gaining too much and I am not able to explain why. Most the passive components are new. The tubes are 110% in the tube tester and also swapping them the results are not so much different.
>> 
>> There is furthermore a mistake, I believe, in the manual: in par. 115, letter (i), they say to adjust R532 so that you have -2.5V with 300uV on J510. That is not compatible with the stage gain chart, that requires 130…180uV instead of 300.
>> 
>> By the way, I can state that the overall sensitivity is not very much affected by the IF gain: getting more gain does not imply a better S/N and so no better sensitivity. That is true particularly for R532.
>> 
>> So, you could ask: if anyway the IF chain has the right gain and the gain it is not very important, why are you losing your time on that? That’s the more difficult question…. :)
>> ______________________________________________________________
>> 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