Performance of tube recievers on higher bands.

[ke7trp]

I have been fighting a loosing battle trying to get a Tube receiver to perform on the higher bands.  I just installed the 5 band mosely beam and spend my time on 15, 17 and 20 meters AM. 

I tried to use my NC183D on 15 meters, The performance was piss poor.

Back on the bench,  I did an RF alignment and gained some performance but not much.  I found weak tubes. I put new tubes in and did an alignment again and some gain was found.

I then started checking voltages and found the B+ was 252 which was supposed to be 245.  That was fine.  The screen volts said 105 in the book but I found 250!!!   

I started checking resistors and  I found 5 bad resistors. Weird. I checked all these about 6 months ago when I recapped this thing.  I replaced them and voltages came down.  I then replaced the Screen resistors 47K with new ones and that did the trick.  Now I had 103volts with no signal and RF gain up.   

At this point, I performed another RF and IF alignment.  On the 2 low bands, I can hear .1 microvolt from the 26B sig gen.  This is as low as the Generator goes. When I go to 50 microvolts, I am right at S9 on the meter.  When I go to 40 DMB, There is 40 on the meter. Great.. Perfect... This radio is strong and a hot performer on 80 meters.

Now on the two higher bands this radio seems weak.  I have to go to .3 microvolts to hear the signal.   With 50 microvolts into the radio and I get about S5 to S6.

My question is about the performance of these recievers up here... Is this normal performance?  I have no experience with this to understand if the radio is working as it should or not.

I swapped each tube one by one with good tubes. No change.  I checked every resistor in the RF amp and IF amp stage and all caps are replaced.  The alignment is spot on and the image is in the right spot on the OSC.  Each RF amp trimmer has a nice peak.  Each IF can has a sharp peak.

I feel there is nothing else to check here and I feel that I have the receiver running as best I can with all voltages correct. I am looking for confirmation the radio is indeed runining at peak performance up on these high bands. Is this what I can expect?

C

[kb3ouk]

From what I've always heard, yes that's as good as it gets for most tube receivers at that high of a frequency. most average receivers do poorly once you pass 15 or 20 meters. The only tube radio I know of that does good clear up to 10 meters is the R-390A. so that's that. but I can tell you one thing, your S meter is right on if it is giving an S9 reading at 50 microvolts.

[Steve - K4HX]

Some performance numbers on a properly working 183D here, but only for 20 meters. There is a drop off compared to 40 meters.

http://www.w1vd.com/NationalNC-183D.html

[ke7trp]

Thanks guys.  I just did not realize that the drop off in performance is so drastic.  I am going to test some other recievers here to get some baseline data.

C

[W1VD]

Expect MDS to fall off 3 - 6 dB as compared to 20 meters for most tube receivers ... low to mid -130 dBm numbers being typical. This level of sensitivity may be adequate depending on your antenna / 'background' noise. Listen to (or measure) the receiver noise output (tuned to a vacant frequency) when switching between a 50 ohm resistor and the antenna. If the noise comes up an easily noticeable amount (~6 dB or more) when switching to the antenna no further receiver sensitivity is required - the system is already limited by the 'background' noise.         

[VE3LYX]

There may well be a drop off in performance at higher frequencies but I would not say it is because it is a tube radio. If you want a real piece crap try a Heathkit Mohican recver. I have given it the award as THE WORLDS WORST SHORTWAVE RADIO. It is solid state but couldnt hear an arcwelder in the next room. Tubes have and do regularily perform way up the spectrum . Some tubes do well up there , some dont. The prefered ones are indicated in tube manuals and info. I often see a 6SN7 where a 6sl7 would be a better choice. Both fit and work in the same socket. Same goes for the popluar dual triodes. 12ax7  12ay7  etc etc All will work but some are better suited to higher frequencies. Often they get subbed (because they will at least work.)
I would be very surprised if transistors in a similar circuit and design worked any different. I wouldnt blame tubes. I sometimes think it would take a hundred transisitors to equal one tube. Like comparing a german shepherd and a poodle. One is a DOG. The other? Well you know!
Don VE3LYX
(If you can carry your dog it is not big enuf)

[Tom WA3KLR]

Clark,

Several years ago we had the following thread on the NC183D performance.  I had overhauled my father's NC183D.  The emphasis was on the image rejection but I did show some subjective sensitivity numbers.  

http://amfone.net/Amforum/index.php?topic=19922.0

Note: The NC183D's input impedance is not 50 Ohms, it is around 300 Ohms.
A proper L network (or 4:1 balun) at the receiver input for matching to 300 Ohms could improve the apparent sensitivity by quite a few dB.

[WQ9E]

Try a number of 6BE6 tubes to select the best one in the first converter position.

In most general coverage receivers, the L/C ratios are suboptimal on the higher ranges and you will probably be happier with the performance of a ham band only unit on 15 and 10.

[K5UJ]

Rodger beat me to it--I was wondering how a 183D would compare to a hot NC300--weren't they roughly from the same time period?  I have heard the NC300 is pretty sensitive up through 10 m. 

Regardless, Clark it's great to be getting into a rx and doing that troubleshooting and alignment.  I need to learn how to do things like that.

I agree with Don about tubes--the gain with a tube is incredible.  And, if you have a dog and it's not big you're wrong 

[ke7trp]

THanks alot for posting.  I never doubted Tube technology but did doubt the performance of common tube receivers from the 40s and 50s. I ASSumed that they would work just like they do on 80 meters where I use them.

W1vd, Thanks for the info.  Min discernable signal. How does this test compare to me using the calibrated microvolt scale? I am trying to understand how these figures compare to my .1mv

I did find that the HRO60 is very very hot up there on 15 in comparison to the 183.  My HRO60 has alot of gain and this one will hear .1 microvolt on 15 meters and with 50micro in, its showing 30 DB over.  I found my winner in the shack. It layed waste to all other recievers here on the high bands.  I will install it with the king 500 today and get on those high bands and enjoy myself. I need to find a 15 meter band spread coil now!

I think we are in agreement that the 183D is working the best it can. The SX28A also went through the ringer this week in trying to make it work up there with good performance. It is way way down up there on 10 and 15 meters. 

I am going to button up the 183D and put it back in use with the T368C on the low bands

C




 

[WQ9E]

Before you write off the NC-183D completely do try a couple of different 6BE6 tubes because gain will fall off rapidly with frequency if the oscillator section of the tube isn't performing at peak.  I have found 6BE6 tubes to be a source of problems in many receivers (low gain and prone to microphonic issues), tube testers are not sufficient-you need to try a few to see if that helps.

I have found 7036 tubes to work very well as a 6BE6 substitute, they were designed as "logic switchers" in early tube type computers and are more rugged and reliable in my experience.   They are a direct plug-in EXCEPT the envelope is taller so you will need a different shield.

[ke7trp]

I swapped in 6 brand new tubes last night trying to get some gain. I got two db so far

[WQ9E]

Are you getting a clean and distinct peak with the antenna trimmer control? 

Pay attention to Tom's comment about the input impedance.

If you have a favored vintage receiver with which you are pleased with the IF bandwidth and audio but the sensitivity (and perhaps bandspread and stability) are lacking on the higher bands then think about building a simple crystal controlled converter to bring 10 and 15 meter coverage down to the 40 meter range.  You can find construction circuits (both tube and transistor) in the older ARRL handbooks.  As I recall for awhile they were running an article on a 80 meter DC (direct conversion) receiver with switchable converters for the other bands, these converters would provide a simple design also.

[W7TFO]

Clark, what Tom KLR said about the input impedance is really important.  A lot of users think a radio is deaf at certain frequencies but the culprit is matching the antenna impedance to the receiver properly for proper power transfer of that tiny amount of power.

Many accomplish this via a outboard preselector, and that is perfectly acceptable.  At times all that would be needed is a passive LC network, such as that marketed by Meissner and others  in the past.

Some (read most) RX of the era had balanced antenna input design, with an adjacent ground terminal usually pressed into service to make it an unbalanced input more suited to low-impedance coaxial feed.  As the frequency goes up, the losses doing this can be several dB.

A modern broadband balun here if coax is used may be called for, with the added bonus of not coupling in common-mode noise.

73DG

[ke7trp]

Sorry. I should have posted that i am using a 75 ohm to 300 ohm balanced converter. It did help somewhat. 

[KM1H]

The antenna mismatch is important as a 50 Ohm antenna or generator will kill the front end performance. A 4:1 balun is an easy way to go and if you wind it check its performance with a MFJ 269B or similar.

For my sensitivity measurements of an AM radio I always measure the AM SNR in the widest bandwidth with a 30% modulated signal and always include 10M. MDS is rather useless unless youre chasing DX on CW with maximum selectivity cranked in.

Then be sure those resistors on both switch sections are checked.

The 183D uses antique 6BA6 tubes in the front end, the first of the remote cutoff miniature pentodes from WW2.

What it needs are the same tube swaps I mentioned to you a few years ago for a SP-600. A 6GM6 for V-1 and 6BY6's for both 6BE6's. Dont even waste time trying 6BE6's unless they are Telefunken.

Reduce the 6GM6 voltages per http://tubedata.tigahost.com/tubedata/sheets/135/6/6GM6.pdf
by changing R-6 and R-7. Also unground the cathode and install a 56 Ohm 1/2W with a .05 disc bypass.

It may oscillate and if so put a 10-30 Ohm carbon resistor right at the grid pin. And a 100 Ohm at the screen pin with a .01 or .02 disc at both ends to ground with very short leads.

You can experiment with leaving the suppressor grounded or connected to the cathode.

Once aligned you will likely find too much gain so add some resistance in the cathode of V-2 along with another .05 disc.

A 6BZ6 will also improve V-1 but realize you will have to reverse pins 2 and 7 functions as well as follow tube manual values.

National did some weird things to the 183D front end which I dont care for at all.

Coincidentally, Jay and I were chatting on the phone yesterday on another subject and then switched to BA's when I asked why does he stop his testing at 20M. The 390A was mentioned by me as being an overhyped performer on 10M. Just looking at the tube lineup will tell you its far from ideal and strictly early 50's tech.
Another thing Ive found over many decades is that most home testing and even some civilian labs pay little to no attention to generator and cable/connector signal leakage and attenuator accuracy. All will give you readings much better than reality.

Yes, a NC-300/303 is a very hot package on 10M. With selected tubes a typical 10M 10dB SNR on AM is .4 to .45uV and with a new 6GM6 it gets down to .35uV or so. Going to a 6BY6 for the 2nd conversion doesnt improve the SNR but does help with overload. Dont use a 6BY6 where the original 6BE6 has AGC applied.
Contrary to a few misinformed folks the 6BA7 is a very acceptable mixer when used properly.
While Im on the NC-300 subject, changing to a crystal controlled 2nd conversion makes a drastic stability improvement. After a short warmup I can regularly set to a 10M SSB qso while working on something else and dont have to retune for 30 minutes+. Cost has gone up, it cost me $29 each plus postage last month from Interntional Crystal and I had to buy 3 for the ones in the shop. Accuracy was well  under 100Hz for these high tolerance babies.

There are a few cures for the SX-28 but thats a different thread.

Carl
 

[WA5VGO]

I'm not sure I'm completely understanding the thread. The NC-183D is not the quietest receiver ever built, but it's certainly not short of gain. If it is working properly, you should find that if anything, it has too much gain. This is equally true of the HRO-60. With both of them, I run the rf gain control cut back - even on 10 meters. I've used a number of each of them and always found this to be true. On the other hand, if you're refering to actual sensitivity - the ability to hear a signal above the internal noise of the receiver - it probably about par with most of the receivers from that generation.

Darrell

[W3RSW]

Concerning the perceived shortcomings of tubes, a lot of Motorolas and Collins used 6AK5's.
Plenty hot up through 6 meters.

I agree with those pointing out the compromised coil/cap 'Q' front end shortcomings in some general coverage receivers.   Same old phenolic coil forms, way too much capacitance, long leads to band switches, circuits and tubes lifted from the all American five, etc.

Many articles published on preamps/preselectors for those bottom line receivers years ago, even when they were new.

[Steve - K4HX]

Conversion chart here. Remember this is only for 50 Ohms.

http://www.moseleysb.com/mb/mv2dbm.html

Or you can calculate dbuV [ 20log (your measured sensitivity in uV / 1 uv)] and subtract 107 to get dBm.

Min discernable signal. How does this test compare to me using the calibrated microvolt scale? I am trying to understand how these figures compare to my .1mv

[KM1H]

Concerning the perceived shortcomings of tubes, a lot of Motorolas and Collins used 6AK5's.

Its a great tube for radar IF strips and FM as Moto knew well. Its also a sharp cutoff device which destroys AGC action, something Collins knew little about in the beginning. It was an early cure all for VHF hams until they figured out that when more than 2 stations were on the band all hell broke loose. It certainly didnt last long in a TV tuner.

Its like the old Hammy Hambone swap of putting a 717A in the front end of a set that needs help on 10-15M but is already hot on the lower bands and expecting it will work well. Its a 6AK5 on its side underneath the mushroom. I had it both ways with a HQ-129X. A 717A for 15 and a 6SG7 for 40/80 as a Novice and later added 10 plus 20 which was a transition band and used both.

My first HB 6M converter was a Handbook 6AK5/6J6 which was rapidly changed to a 6BQ7/6X8 during those wild Cycle 19 days of the mid to late 50's. Now I didnt have spurs from one end of the band to the other.

[ND9B]

Now on the two higher bands this radio seems weak.  I have to go to .3 microvolts to hear the signal.   With 50 microvolts into the radio and I get about S5 to S6.

.3 uV sounds pretty sensitive to me for any HF band. (The lowest scale on my HP 606B in 1 uV.) When you connect the antenna to the receiver, does the noise come up? As long as you can hear the noise floor of your antenna, you won't miss any signals.

BTW, I have a Kenwood R-5990 solid state receiver that is totally silent on 10M with the antenna connected (in the absence of a signal of course). This is good example of poor sensitivity.

Bobby Dipole ND9B

[Pete, WA2CWA]

BTW, I have a Kenwood R-5990 solid state receiver that is totally silent on 10M with the antenna connected (in the absence of a signal of course). This is good example of poor sensitivity.

Bobby Dipole ND9B

What's a R5990?

My Squires Sanders is very quiet, even with the antenna connected, until I tune across a signal.

[WA1GFZ]

A few things you need to consider when operating 15 and 10 meters.
1. the background static level is usually very low
2. A beam antenna being directive will reduce background noise even more
3. The feed line loss can be an issue
4. You can't compare signals riding above noise like 80 and 40. When the band is open signals will have very low background noise.

MDS of .3uv is not bad on 10 meters but if you have 3 or more db of coax loss that could be your problem.
You want to use good coax between the generator and RX when playing at these levels. You really need double shielded cable like RG223, RG142, RG400.
I can copy strong stations on 75 by the leakage of single shielded cable between my generator and test RX. 

[Jim, W5JO]

Something taught to me years ago was check your test equipment and interconnect cables before you use them.  I never rely on cables being good until I check them, it is particularly important in commercial applications.  I have seen generators that the attenuator is inaccurate or the connector on the cable has been used and twisted so much the center conductor bearly touches the center pin.

We seem to take the test equpment and cables for granted but it bears checking regularly.

[ke7trp]

Wishfull thinking.  My new Belden Cable has about .5 DB loss for 100FT that I am using on 15 meters and about .6 on 29 mhz.  To verify the cable after install I hooked the dummy load up out on the yard and used the 259B.  It read 1.0, 52 ohms ect.

Alot of the ham band 50s tube recievers are just down on the higher bands.  Are they usable?  Sure.

One exeption to the rule is the HRO 60 which seems to match or exceed the Icom 756 Pro I use as a bench mark.  No noise at all with 20 to 30 DB over AM signals pouring into the shack on 15 meters. The way it should be.

C