The AM Forum

Just got in a very nice unmolested NC-183D receiver. While prepping the receiver for performance measurements, one step is to verify voltages at all tubes per the manual. Measurements starting right with the 1st rf amplifier and continuing through the i-f were significantly off...even though the receiver did function. The reoccurring theme going stage by stage was high plate voltage and very low screen voltage. In each stage the 47k ohm 1/2 watt screen dropping resistors were at fault. The resistors measured anywhere from 120k to 2.9M! Replacing the resistors with new brought each stage back close to factory spec voltages. Virtually all other resistors in the receiver were close enough to not require changing. Could the factory have gotten a bad batch of 47k ohm resistors?

Later I ran across a post by Rodger WQ9E in a different thread commenting on bad 47k ohm screen dropping resistors in his receiver. Coincidence? Using factory voltages the dissipation for each resistor was found to be as follows:

NC-183D

1st rf	.42 w
2nd rf	.42 w
1st i-f	.31 w
2nd i-f	.31 w
3rd i-f	.33 w
phase inverter (plate)	.28 w
phase inverter (cath)	.28 w

Common practice (at least nowadays) is to limit dissipation to 1/2 the rated value. One would guess that the resistors started life as 47 k ohm and over time got 'cooked' driving the values unusually high. Pix of the resistors below.   

This got me wondering about the HRO-60. Again, using factory voltages the dissipation for each resistor was found to be as follows: 

HRO-60

1st rf	.54 w
2nd rf	.26 w (100k ohm)
1st i-f	.10 w
2nd i-f	.31 w
3rd i-f	.31 w
phase inverter (plate)	.10 w
phase inverter (cath)	.10 w

Think I'll go back and take another look especially at the 1st rf stage in the HRO-60!

The 183D did work as received...however the changes made a significant difference in performance. A few other minor repairs were required. Next up is an alignment (although it's probably pretty close as is) followed by a trip to the 'torture rack' for receiver performance testing.

A bunch of 47k ohm 1 watters are on order...

   



 

             

What you found is typical of both receivers.  I have had two of each model and found that other resistors have changed significantly like the screen resistors.  The 183D that I have now didn't have that many bad screen resisors.  Another problem is that if you try to unwind the wire around the tube socket pins, they are easy to break. 

Watch out on any National product.

Yea, what he said! ! !

I have run into the same problem on jst about EVERY National receiver from that era that I have worked on. Either the screen dropping resistors, and/or the fuzz inverter plate and cathode resistors are trash. I have also found similar problems on NC300s as well. they look fine (not burned or cracked) but they go way high in resistance.

                                                           The Slab Bacon

I assume "they" means the socket pin.

That's why I always just poke the wire through the hole in the pin and maybe form a hook and slightly crimp it before soldering.  None of that "firm mechanical connection before applying solder" B.S.

I have seen some factory construction where they wrapped each wire 5-6 turns.  I don't know what they were trying to accomplish.  Maybe assembly line connections tended to fall apart before they reached the person who did the soldering, but one or two wraps would have easily been enough.  I usually  don't bother to unwrap the wire.  Just clip off the old one and solder to the pin, wrapped wire and all, if there is enough room.

The screen dropping resistor probably handles more power than some of the others, so it might be more prone to change value over time.

Brian I don't know why you have sensitivity problems with either the 183D or HRO 60.  I need about -105 db to hear a discernable signal on my 183D and with the RF gain fully on, I see -90 to get S9 on 10 meters.

Maybe you have an IF section misalignment?

The ones that I have seen that can really get you in trouble are the feed resistors to the oscillators. Lack of HFO injection and a weak BFO are two common complaints. These resistors typically handle more juice and they drift up even more.

I found an ARC-5 with 20 volts on the BFO tube. It actually still oscillated! but the BFO was weak to say the least. 

Mike Wu2D

A properly working NC-183D definitely has plenty of sensitivity for even a fairly quiet location.  Page 15 of the manual provides stage gains for the converter and IF stages so with an output meter and/or scope and signal generator you should be able to narrow down the problem area pretty quickly. 

After I took care of the screen resistors on mine the only remaining issue was low sensitivity on the top range which was due to insufficient oscillator injection.  That took a while to find but it turned out to be a kinked wire on the oscillator coil for that band which apparently reduced the Q a bit.

Several 50;s national receivers I've had (183, 183D, HRO) have had 22K resistors actually all read 220K.  I suspect National might have had an issue with its supplier.

I have never been a fan of IRC resistors.  If memory serves, that is what National used.  Back when I used the new ones, I found them to be very loose in the tolerance department compared to Allen Bradley.

Closer view of the faulty NC-183D 47k ohm resistor next to a new AB. The 183D 47k resistors have a rougher molded body surface and the color bands are not as bright and as well defined. Some of the resistors in this 183D are AB looking types though. Wonder which type is in your receiver Brian? The 2.9M resistor was on the 2nd rf amplifier.





 

I have never run across one, but some metallic film resistors are said to be moulded into a  case that looks identical to carbon composition resistors.  I have found film resistors to be very fragile and susceptible to self-destruction from only moderate overloads - perfect companions for solid state devices.  Metallic film fuses might be a better description.

Maybe someone replaced the original carbon comps with those films, which later blew.

I'm going to look hard at the resistors in my RAO-7 when I get to it - one thing I did notice is that they are very very large for 1/2 watt size. closer to normal 1 watt sized. (I've never seen a schiz of it, so I dont know anything about what might be going on.)

Some years back the body color of a resistor denoted what kind it was, blue, green, beige, etc.  Has that been abandonded or does it still stand?

Okay...now we're making some progress...

First, as I noted earlier this is an unmolested receiver except for a couple black beauties. All of the resistors are (were) original with original solder joints intact. There is a mixture of the 47k looking resistors and AB looking resistors.

Rechecked the resistors and one of them indeed measures 2.9M, Brian. Looks like your receiver got the 'good' resistors.

I cracked open a standard AB resistor and it is shown on the left in the pix below. The core is a solid carbon slug as one would expect. Cracking open the defective 47k, shown on the right, presented a problem as they tend to disintegrate. The core appears to be thin and hollow. Between the core and the outer coating is a chincy filler material. These resistors are considerably less robust than the AB types...which probably explains why they went so high in value with the power they were dissipating. Other resistors of this type that are not dissipating much power have retained their value pretty well.       

 

My dad's NC-183D I worked on had 2 distinct vendors of the 47k resistors.  The one vendor's resistors were all o.k.  The other vendor's 47k's were all bad.

When I overhauled my old Heathkit HW-101 it had many resistors high in value.  Many of them turned out to be under-rated in wattage when I measured the voltage drop with the new resistor and did the math.  So  I had to replace many of them again with higher wattage.

In my experience it wasnt only National. Ive had bad 22K, 24K, 47K especially across the board. Even in a 75A2 there were a few.

In the last HRO-60 I did about 3-4 months ago I replaced 18 resistors if I remember correctly. After that a SX-42 was about as bad as was an earlier HQ-129X. Toss in several Zeniths, Philcos, Silvertones, etc and all have similar problems. Between my own and customers I do typically 1-2 radios a week.

Ive found many radios are quite sensitive to tube quality. Using only NOS tubes I could see a big difference on the output GR meter. I selected the best performers and easily exceeded stage gain specs on the HRO-60 for example and using 10M for the RF and conversion stages. I went thru maybe 4-5 6BE6's in the first mixer. a couple of 6BA6's, and even a few 6SG7's in the IF.

Also note that socket voltages are specified at a 115VAC line.

As far as breaking socket pins there are plenty of spares to choose from ;D  I snip the component close to the socket or tie strip and carefully unwrap the lead. Never cared for the TV repair shop look of splicing. Thats fine if its a S-38 but not in a real radio.

Carl
KM1H

I have seen real carbon composition resistors modified to make higher values by filing notches in the side of the resistor, through the case material, to remove some of the carbon material from the slug.

Ive done the filing a notch thing many times Don when maintaining original appearance. Notch to desired value, a bit of Bondo and repaint which is real easy on the 30's dogbones.

I always save the pulls from radios that get carbon or metal films.

Carl
KM1H

Why?

Carl
KM1H

A bit much for 1/2 watt resistors considering they're in a receiver that runs 'continuous'...no?

Calculations are fine but did you actually measure at the required 115VAC?

The reason I ask is that after rebuilding the last HRO-60 and wondering similar about the resistors I did the old touch test on both RF stages with the AC at 111V thru a bucking transformer. The 1/2W replacements were barely above room temperature and IMO operating well within ratings.

As I mentioned earlier I replaced 18 out of tolerance resistors and many had no appreciable current thru them.

Of the five 47K RF and IF screen droppers I just measured them at 59K, 69K, 114K, 325K and 1.3M. Should I assume that the 2 highest came from the RF amps? Was the increase due to DC or RF heating or both?

Before Id run out and blame engineering remember that they had no control over the vendor. National was as cost conscious (cheap) as Hallicrafters and the Purchasing Dept scraped the bottom. I saw the result of that thinking in other products while at National.

During the years I worked in the Service Dept it included hundreds of postwar receivers in the NC-173 to HRO-60 era which all used similar vendor components. Except for resistors cooked by leaky caps it was a non issue at the time when some of the sets were over 20 years old. That certainly disproves any reliability guessing.

What happened to components during the next 40 years in no way can be blamed on engineering.

Carl
KM1H
National Radio 1963-69

ABs are the pick of the litter. Almost all carbon comps go high in value with age. Those particular resistors with the rough finish are especially prone to going high or being found open.  Most of us that restore antique receivers from the 30s and 40s routinely check every resistor for value.  Wax caps are either replaced or rebuilt with new components hidden inside. Engineers never anticipated that a 1930s communication or home appliance would still be used 60 or 70 years later. 

Pete

No, actually they will, and do, go up in value sitting a drawer, New Old Stock. They aren't stable devices.

A few isolated examples that have held value doesn't prove much.  High voltage may affect them, but
how much use have many of these radios over the past 60 or so years?? Most overheated (discolored resistors
tend to cook downward in value, in my experience.

Allen Bradley manufactured resistors are generally more stable than those of other manufacturers.
This subject has been beaten to death on other Forums. Here is one recent discussion:

http://antiqueradios.com/forums/viewtopic.php?t=115577

and a good website for further reference:

http://www.cliftonlaboratories.com/carbon_composition_resistors.htm

I've pretty much abandoned carbon comps and now prefer to use carbon film and metal oxides.

Peter k1zjh

After readind all this. I just happen to have a Collins 75A-4 on the bench and changing out the 7 deadly caps along with other bad caps. I snipped one side of a few screen resistors and guess what, two to four times there normal value. Looks like a good project for tomorrow.
Regards,
Gary

Repeated photos in the forum body just waste bandwidth for those on dialup. If its of any use to the thread please make it as an attachment.

Carl
KM1H

The pictures you posted in this thread are the ones that take up excessive bandwidth, they take forever to download and then really show nothing that a few words wont accomplish. Others post the small jpg's as attachments that can be clicked on to expand. Those download fast.  Learn how to resize your photos for attachments.

Im stuck on dialup until my T-1 microwave radio comes back from repair. DSL isnt even an option in this area due to Verizon putting the fiber upgrade on hold and I dont have cable. That limits me to what a US Robotics 56K external modem can eke out of a poor phone line that runs around 33K tops and sometimes folds back to 19.2K.

Its part of the benefits of rural living on top of a hill and at a dead end road.

I'll appreciate your cooperation.

Carl
KM1H

Having lived on top of a mountain, in the middle of nowhere, using nothing more than wifi with a yagi and parabolic reflector, I can sympathize with Carl.

But, I find it funny that people who are stuck in the backwardness of non-high speed internet always want everyone else to post low bandwith.

Carl, we'd appreciate it if you'd move.  Somewhere with high speed internet.

Thanks for your cooperation :)

(tongue firmly in cheek).

--Shane

What, move and give up a kick butt location Shane? Its worth 10-20 dB over the bottom dwellers.

The only real bad thing is that no 3 phase service is available to build amps with tubes containing five zeroes as you do 8) :o

It will probably be several weeks before Im back on microwave so bear with me.

Carl
KM1H

I have issues with the use of 47K dropping resistors for the screens. I had the use of a NC 183-D as in douche! when I was out in Sorryexcuse NY. I found that the receiver overloaded on very strong signals. When backing down on the RF gain pot the signal levels would drop off then drop off very rapidly like falling over a cliff.The S meter did not take much signal to get a reading but would pin easily.Take a meter and measure the screen voltage with no signal. Then measure the screen voltage with a strapping signal. The screen voltage was much higher . What is happening is the"Shoveling 10-1000 Against the Tide syndrome. The AVC voltage comes screaming down the line telling all of the gain controlled stages to come down in the gain department. The screen current falls off causing the voltage drop across the 47K resistors to decrease there fore causing the screen voltage to increase and therefore turning up the gain. The fix was simple : I borrowed the 105 volt  source from the 0B2 regulator. I fed that voltage through 4.7K resistors (value not critical) to the two RF stages and the IF stages. The 183-D behaved like it should after that. I could monitor my own signal without overload. The RF gain control worked smoothly. My own signal would pin the S meter. Other signals indicated were more believable . The meter behave more linearly.I have done these screen voltage mods to my Hallicrafters SX-62 and the Collins 75A4. It has resulted in better performance in dealing with a wide range of signal strengths. This technology is applicable to any high B+ receiver with improvement. Try it, you'll like it
Tim WA1HnyLR

In my homebrew receivers, I run the B+ at 150 volts, but have the screen dropping resistor at the mixer at 470k, the 1st IF at 47k, and the 2nd IF at  100K.

Maybe I should try the screens at the 105 volts (regulated) that the LO runs at, and smaller values of screen dropping resistance.

The unamplified agc seems to do fine on everything other than when I transmit, but I am always open to improvements...

Brett

Years ago I worked on the TPX47 IFF RADAR. It used a high value screen dropping resistor with a cap to ground at the screens. When a radar pulse came through the screen current would discharge the cap to a low value. This killed the IF gain until the cap charged up again just in time for the next closest possible pulse.
This trick was used in the six IF stages and was called antijam.
I wonder if communication receiver designers were that smart dealing with static.

Since the receiver is still on the bench it was easy to have a look...

On the 1st and 2nd RF, 1st, 2nd and 3rd IF, with no signal into the receiver, the screen voltages are close to the book values in the 105 - 115 VDC range. With an input of -50 dBm (1 mV) the screen voltages rise to the range of 135 - 158 VDC. (Note there is less power dissipated in the screen resistors under this condition).

I think Tim is referring to on frequency BIG signal overload of the i-f and audio stages because of the rising screen voltage vs. increased negative grid bias. This may well be a valid concern. However, this is a different situation than being tuned to a weak signal up or down the band with little or no AVC action. If the AVC isn't activated the tubes are operating, presumably, in their most linear region and one has to accept the associated level of IMD, cross modulation, image rejection etc.

Frank...interesting tidbit. There's quite a difference on how well these various receivers handle static. Could be an interesting study unto itself...             

Jay,
      You said a mouthful there! ! Now that would make one hell of a study / topic of discussion!  8)  8)

As an experiment, I tried various things with the screen supply to the mixer and IF amp tubes on the homebrew last nite.
The results were nil, I tried 4.7k ohms off the regulated 105 volts, and since the voltage was higher on the screens, the gain went up into the saturation point where strong broadcast signals on 40 meters maxed out the S meter and the additional signal strength agc voltage did not show on the S meter.
(signals went up to 50 over and stayed there).
Turning the IF gain down (cathode pot in the 1st IF) would lower the S meter and signal strength variations were back.....maybe I have to increase the cathode resistors...

Normal screen voltage at the tubes was around 50 volts, with 150 on the plates.
With the 4.7k resistors off the 105 volts, it was around 100 volts.

I even ran the mixer screen off the TR relay 15 volts and the gain was down but the thing worked fine!.

I then got the adj power supply out and put 40 to 50 volts on the screens, with the 4.7k resistors, but did not notice any real difference between that and stock settings.

I need to do more work, hook the scope up to various IF points and see what stage is saturating with real strong signals, if one is, I dont hear any distortion, but the S meter stops moving up past some point in signal strength.

I also put the BFO on, set it to hetrodyne at about 30 Hz, and tried adding a .47uf cap to the agc line to slow it down, I heard no difference in the audio. I then added another .47 and heard no change, the S meter was very sluggish, but no audible change in the 30 Hz or broadcast audio quality.

Not sure what is up there....the normal AGC time constant seems about like medium on the R390a, looking at the meter.

Some day I will pick up a distortion analyzer and run amuck with it...

Brett

 

The RF stages and the second and third IF stages of this receiver are grounded cathode.  This makes for a lot of gain.  Where I have seen most of the saturation has been in the third IF.  A strong signal is hot where it gets there and the additional gain really pumps it up.  This is also where the AGC is sampled so it is seeing a very hot signal

On the early model radios the S meter was driven off the AGC line and in the later models they modified the AGC a bit and used 6SN7 as phase inverter and as a meter driver.  In the first model it was very easy to pin the meter as has been stated, but in the later model it did not pin as easily.

One could insert cathode resistors in these stages to help reduce the gain throughout the receiver as in the HRO 60 then apply other fixes as desired.  There was a 2 or 3 part article about this topic in ER some years back.

Since the 183D is designed to work up thru 6M that explains the gain availability.

I guess a few here arent aware of that or the fact that the radio has a RF gain control that works rather well.

An external step attenuator is another option.

I can also see where slightly leaky tubes in the RF and IF stages could cause an overload condition. Nothing like little or no grid bias to make things hot

Carl
KM1H

All I'm saying is lower voltage drop across the resistor means less power dissipated by the 47k 'suspect' resistors.


You're going to have to help me with this one. In the HRO-60, for example, there are cathode resistors on each of the RF and IF stages. However, they are bypassed. Stage gain should remain the same in that case.

Carl...I'm with you on the proper use of the rf/af gain controls on the old receivers.

And the NC-183D also has a deck of the band switch that has resistors to control the gain band by band.  When I got mine, a previous owner had rewired that section so that it ran at full gain on all bands-once I corrected the out-of-spec screen resistors I also put it back as National intended.

Rodger WQ9E

There is nothing wrong with adding cathode resistors to the 183D as in the 11/04 ER article. The author is quoted as saying it is an alternative to using the RF gain control which he didnt care to use. However that has a long term effect as tubes age. The HRO-60 uses a selected resistor in the 3rd IF; I used a trimpot in mine and set the full IF string gain as per the manual.

The 6BA6 is a pretty poor tube above around 20M and a 1st RF stage change can make a big difference. Going to a 6GM6 with suitable component changes and reducing the 2nd stage gain to compensate will really lower the noise and wake up the higher frequencies. Making that change in a HRO-60 brought the 10M AM sensitivity to under 0.5uv for 10 db S+N/N and that is mighty good.

Most hams in the 50's had pretty simple antennas and overload wasnt so much an issue as today. I use Beverages on 160-40 and overload is not an issue. I also use Beverages on the modern gear for 160-80 and even 40 at times.

Carl
KM1H