NC-300 alignment question - HF oscillator tracking

[W3BH]

Gentlemen:

The HF oscillator (V7) in my 300 is tracking high, about 1.9 kcs. The filter appears to be still very close to spec at 2215 kcs. Everything else in the alignment sequence has gone pretty close to what the manual suggests apart from L2B where it is hard to see a true dip at 160 kcs off 2215. The receiver appears to perform as well as the last one I had 40 years ago (perhaps I don't remember it that well. Hi).

The question is, how can I best tweak up the the HF oscillator frequency to spec and would it actually be worth it. The attached image of the oscillator circuit shows what looks like a Hartley with C47 and C48 being candidates for checking. I don't want to unsolder and pull anything unless it's going to be worth the trouble. If a 1.9 kc (after proper warm up) difference between the mixed injected frequency (out of V2) and the center frequency of the xtal phasing filter matters a great deal, I could rebuild the oscillator and add a trimmer for future tune ups?

The circuit can be quickly viewed here:
Which components in this circuit need checking first?

https://picasaweb.google.com/lh/photo/RCC48NruNjgI_tqEzs5vTg?feat=directlink

Any comments would be greatly appreciated.

Thanks,

Jim
W3BH

[Jim, W5JO]

Is the tracking problem on all bands or just the upper ones.  The alignment of L6 is critical to tracking and linearity for that thing on the upper bands.  If you still have an adjustable coil in there, I would convert it to a crystal oscillator.  That was a factory mod because of problems with it's frequency moving around depending on moisture and heat.  Check the BAMA site for that mod.  You can get a small crystal that will allow you to put the entire circuit inside that can on the board after you remove the coil.

[W3BH]

Thanks to Jim for pointing me at this reference: 

In these instances we recommend crystal control of the second conversion oscillator as per the attached schematic. it should be remembered that there are now two crystals in the circuit which must be properly coordinated if the receiver is to work at maximum efficiency Normal production tolerances in the grinding of the crystals may allow a situation to exist where the difference in frequency of the crystal in the crystal filter (2215 kc) and the crystal in the second conversion oscillator (2295 kc) is not exactly 80 kc (second IF).

If this is so the second IF strip must be adjusted accordingly either slightly higher or slightly lower to compensate.

This is something I might consider after I've worked out the HF oscillator problem - if it is a problem.

If my understanding is correct, for example, the incoming signal, say 2.000 MHz is combined with the output of V7 (6AH6) which I can measure as 4.217 which gives a difference of 2217 which is produced in the first mixer which is right before the crystal phasing unit which features a 2215 xtal with three xtal positions and an off. If my mixed signal of 2217 goes to the second converter and the second conversion oscillator (V3) and T6 tunes the second converter oscillator to 2297 (2297-2217=80) I'm looking good in the xtal phasing off position. If I switch in the xtal phasing at that point, I'm hitting the xtal at 2215 with a 2217 signal. The signal is not centered on the filter thus presumably degrading its effectiveness, at least in the two narrowest positions? Would National have seen this as 'out of spec'? What about replacing the xtal in the phasing unit with a 2217 xtal then adjusting the second converter and the 80 kc IFs accordingly? Head starting to pound...

So as I see it, to bring it all back into spec, I need the RF oscillator to produce a signal for the mixer which is very near or on 2215 kcs. Right now it appears to be 2 kcs out.

I may be splitting hairs here and very well might not thoroughly understand the circuits - what you get when you let liberal arts majors into ham radio.. Sending the R-390A off to Chuck Rippel or Miltronix when I have the money,  that's for sure.

Thanks,

Jim
W3BH








 

[WQ9E]

Jim,

What you are describing is a calibration error and it won't impact the performance other than your dial calibration would be slightly off.  Using the example in your previous post, when your receiver dial indicates that you are tuned to 2.000 you are actually tuned to 2.002 and that could be corrected with the calibration control.

Tracking error is more generally used to refer to the case where the HFO, mixer, and RF stages are not staying properly aligned across the band.  For example perhaps everything is perfect with the dial set at 2.000 with the HFO operating at 4.215 and the mixer and RF coils simultaneously peaked at that frequency.  But maybe when the dial is set at 1.800 the HFO is at exactly 4.015 but the mixer stage is peaked at 1780.  This would be a classic tracking error and would somewhat reduce sensitivity.

I would set the calibration as accurately as possible from end to end and then adjust the mixer and RF stages for best tracking.

Required disclaimer, My Ph.D. is in business and not engineering so caution advised

[W3BH]

Excellent Roger -

I believe that could account for what I've been experiencing. I just had not thought of it in those terms. Checking calibration and tracking of the HFO, mixer and RF stages as a system should keep me busy for a while and provide further information.

This unit is in excellent physical shape and was the 11th from the last NC-300 made. My feeling from the start was that it could be pretty close to factory spec. There is no sign that it has even been touched internally since delivery and there are no apparent modifications or repairs. But we both know what 50 years can do to components, so I will take your suggestion and see if it can be calibrated and aligned to near spec or not. If not, I'll start checking components, or just leave it as is. It plays ok and looks great.

Many thanks,

Jim
 

[WQ9E]

Jim,

Sounds good and I wouldn't worry too much about a ~2 Khz. calibration error as that is well within the range of the calibration reset control.  Even the single range VFO/PTO units of the time could have close to this much variance from end-point to end-point and still remain in spec.

As I recall, didn't some of the last of the NC-300 units already have the second conversion oscillator crystal controlled?  I thought that was a running production change towards the end but Carl would know for sure.  My NC-300 has the crystal controlled version so if not done during production it was later modified very neatly.

I use my NC-300 paired with a DX-100 and they make a nice pair for both AM and CW.  I am sure you will enjoy using yours.

[Jim, W5JO]

What I read is that the tracking was the same across the bands, ~ 2 Kc.  If that is the case L6 can be the cause.  It is extremely difficult to get on frequency and you may have a lot of trouble keeping it there.  That coil seems to have very high Q therefore any jolt and it move the core ever so slightly, causing it to jump frequency. 

I had one that would be on calibration until the next time I turned it on when it would be significantly off (2-5 Kc).  By changing that oscillator it stabilized enough that calibration was the same each time it was used and did not move if I accidentaly bumped the radio.

[W3BH]

Thanks Roger,

The serial number on this unit is 481-1726. I think that puts it right at the very end of production for the NC-300. The second converter is not crystal controlled. I've now managed to obtain good calibration and tracking across 160. The unit came with an xtal calibrator and I checked that against my HP 5382A and against an RF-590 receiver with xtal oven. The calibrator is a great help in aligning the 300. The 300 I had in the 60s did not have one. I can see now that I am expecting perhaps too much of this technology and 2kcs one way or the other with the cal set will work fine. But what I've also learned here is that properly aligning a receiver of this vintage is a lot of painstaking hard work - no glory. I've got poor performance on every band above 160 now and will have to work my way through and hope that it won't be at the end a matter of trying to juggle all 6 bands to obtain reasonable performance out of each. Thanks for the comments and advice. I hope I will soon have the 300 mated to the VII and get back on AM...

Sounds good and I wouldn't worry too much about a ~2 Khz. calibration error as that is well within the range of the calibration reset control.  Even the single range VFO/PTO units of the time could have close to this much variance from end-point to end-point and still remain in spec.

As I recall, didn't some of the last of the NC-300 units already have the second conversion oscillator crystal controlled?  I thought that was a running production change towards the end but Carl would know for sure.  My NC-300 has the crystal controlled version so if not done during production it was later modified very neatly.

I use my NC-300 paired with a DX-100 and they make a nice pair for both AM and CW.  I am sure you will enjoy using yours.

[W3BH]

Thanks Jim,

It was quite early on in the  game when my attention began to focus on L6. As you say, it's an accident waiting to happen. For a start, it does not turn smoothly and obtaining a peak with it is neigh impossible. I'm not taking it apart to apply some graphite or other, so I guess I'm stuck with that. I did finally manage to find the second peak the manual talks about, which is the wrong one and a fraction of a turn into the coil from the first peak. Perhaps jerk by jerk I can accidentally peak L6 at the right place and hope it stays put.

So I guess if I'm going to do any hacking I can start with the second converter and crystal control it. What I still am slightly unsure of is where the xtal in the filter is actually peaking so I will know what frequency to have the L6 xtal cut at. I suppose ballpark will probably be fine in this context. I could just replace both xtals and I'd have 2215 - 2295 - to 80. That mod appears in the BAMA doc.

I ran the IF into my RF-590 the other night and it really sounded great, better than the 300 itself. So perhaps some new tubes and audio tweaks are in order as well. I love the dial, the sound on cw and the rather individualist design. When it came out it looked rather space age and like no other receiver. It still does.

Thanks

What I read is that the tracking was the same across the bands, ~ 2 Kc.  If that is the case L6 can be the cause.  It is extremely difficult to get on frequency and you may have a lot of trouble keeping it there.  That coil seems to have very high Q therefore any jolt and it move the core ever so slightly, causing it to jump frequency.

I had one that would be on calibration until the next time I turned it on when it would be significantly off (2-5 Kc).  By changing that oscillator it stabilized enough that calibration was the same each time it was used and did not move if I accidentaly bumped the radio.

[WQ9E]

Jim,

My favorite setup for quickly calibrating a vintage receiver is to use a modern digital shortwave portable equipped with memories.  First, figure out the correct HFO frequency for each end of each band. Typically it is equal to the first IF plus the indicated dial frequency ("high side tracking") but a few receivers set the HFO for dial frequency minus first IF ("low side tracking").  The manual will usually indicate and tracking the HFO on the wrong side will result in major calibration error across the range.  Then set up a memory in the receiver for each calculated frequency.  The HFO will be clearly audible in the nearby receiver and you can quickly switch from memory to memory as you go between the low and high end of the band.  This is faster than also tuning a signal generator, does away with potential loading effects of a frequency counter, and avoids accidentally tracking the oscillator on the wrong side.

This same receiver is useful for determining which oscillator in the receiver is responsible for drift, whether hum or noise on the signal is resulting from a problem in one of the oscillators, and can also be used to set the other oscillators on frequency (other conversion, BFO, etc.).

I use a Grundig YB-400 but any of the decent small digital units with at least 1 Khz tuning and variable BFO will work fine.  It is one of the more useful test instruments on my bench.

[W3BH]

Thanks Jim,

You've just made my day. I have a Harris RF-590 that resolves down to a cycle and has 100 memories. I find it more useful on the bench than in the shack and this proves that once again. For me it is always the last word on frequency accuracy and when you run it down to audio frequencies it produces audio output for each frequency right on pitch. It's a far better bench tool than ham receiver, that's for sure. I'll use it to implement your suggestion.

I was tracking the 160 meter HFO signal with it, but never thought of doing what you have suggested. That should give me a very clear picture of how each band's HFO is playing in terms of stability and linearity. And as you said, the loading of the frequency meter on my rather ancient signal generator is significant. I very much need a new phase lock loop digital readout signal generator of the modern era rather than the tube based sweep generator I'm using now. Hi.

I very much appreciate this suggestion - never too old to learn new tricks. I'll let you know how it goes.

Jim

[KM1H]

I couldnt even begin to count the number of NC-300's that I converted to xtal plus others did the same.

My time there at National was after the 300/303 had come and gone but every 300 that came in for service was given the xtal at no charge; it was a wire lead version and soldered in place.

I also converted many NC-183D and HRO-60s while there as the free running oscillator is a good source of drift that many blame on the front end and go crazy trying to fix it.

Carl
KM1H
National Radio 1963-69.

[WQ9E]

Carl,

There is a good chance my NC-300 has your finger prints in it   It does have the crystal modification.

[The Slab Bacon]

but the real question izz is a 1.6KC drift / change really enough to worry about??
come on guys I cant even see those little division lines on mine unless I put my cheaters on!!    Hell, without em I'm lucky if I can the numbers   

Now, to get serious,
Karl, (or anyone else out there ) Have any of youse guys ever seen a PUBLISHED mod for a 300 / 303 that involves cutting the HFO coils to move / change either the bandspread or band coverage of the 300 / 303?? I have one nice good working 300 and another one (in real nice cosmetic condx) that has been relegated to "parts junker" status. Some A/H has cut / trimmed the oscilator coils and changed the some of the fixed mica padders to do what ,I'm not sure. The freq coverage is so far off of the original that you cannot even get much sensitivity out of the front end. (Not enough range in the trimmers). I bought it cheap a few years ago, and when I saw what was done to it, I just tossed it on the parts pile. I have also ran into someone else who ended up with one that was similarly butchered up.
(I bought it cheap enough in an auction that the crapstal calibrator was worth what I paid for the whole radio)

Just wonderin if any of youse guys have run into this??

What was this A/H thinking     

[KM1H]

The only hack mod I can remember was modifying the converter band Frank. Is yours hacked on all bands? Mabe the AH forgot to ses what side the oscillator was on, grabbed his whiskey bottle for nerve lotion and went to show them stoopid injuneers a thing or 3.

[The Slab Bacon]

The only hack mod I can remember was modifying the converter band Frank. Is yours hacked on all bands? Mabe the AH forgot to ses what side the oscillator was on, grabbed his whiskey bottle for nerve lotion and went to show them stoopid injuneers a thing or 3.

Yea, Carl, the A/H cut the oscillator coils on all of the bands. IIRC he also removed a plate or 2 from the tuning cap as well. I kinda think he was trying to spread out part of the band for a slower tuning rate.

I tried to perform the "Vulcan Mind Tap" on it, but couldn't become one with it so I just tossed it on the junker pile.......................

[WQ9E]

The previous owner must have been dedicated to a single mode and tried to spread out that segment of the band.   Maybe the same guy who decided to add a noise limiter to one of my NC-100X receivers by drilling a hole for the control through the center of the National diamond logo.

Look upon it as your opportunity to build the ultimate AM ghetto receiver.  You could make it a monobander with 3875-3890 spread across the entire scale.  On second thought tossing it on the parts pile was the best idea.

[KM1H]

I could use a real nice lid

[The Slab Bacon]

I could use a real nice lid

there are plenty of them all over 75M     

[WQ9E]

You obviously overlooked the key word "nice" in Carl's want