NC-303 Receiver Project

Discussion in 'Technical' started by W5HRO, Apr 8, 2014.

  1. W5HRO

    W5HRO Administrator

    Been putting the final touches on the NC-303. WD-40'd all of the tube sockets, pots and wafer switches, cleaned and aligned everything, replaced the HV supply cap and installed a new audio transformer for 8-ohms out. Also replaced the SO-239 connector and replaced the power cord with a nice long grey colored 3-prong grounded cord. Just have a couple of more things to go and it's finished. Not too bad for a 1958 thru 1962 vintage ham bands only receiver.

  2. W5HRO

    W5HRO Administrator

    I'm still working on this receiver and replacing a few more out-of-tolerance components along with adding a zener or two to regulate the screens on a couple of tubes.

    I've noticed that with the Heathkit IG-42 generator connected the alignment on 20-meters is really flakey and there is some interaction gong on with it. It somehow de-tunes the output of the generator and changes its output frequency. Has anyone ever run into this before? I have a hunch its a harmonic related issue with either the receiver or the generator.

    Anyway, It's a good thing I left this receiver out of the cabinet leaving it turned on and have been slowly replacing a few remaining components because I've had two of the number 47 lamps burn out behind the dial along with the one for the meter in the past week and the AGC clamp diode just crapped out, the old 1N1692 CR1 diode. I just replaced it with a 1N4002. Hopefully in the next week or two everything will be 100% because I want it reliable and working perfect for next Fall.

    The pic in my first post has the center #47 lamp behind the dial burned out. If anyone likes working on 50+ year old receivers like these then just leave them on until all of the weak links crap out. Otherwise you will be pulling them out of the cabinet 15 times before its over. Replace the main necessary components first then re-burn them in again like the factory did back in the day.

    One thing I did do was attenuate the level being injected into the 6BE6 product detector, pin 7. The level is way too high stock and when listening to an SSB station it's really distorted. What seems to work best is changing R55 from it's stock value of 33pF and making it 50pF then adding a 150pF capacitor from pin 7 to ground. The difference is like night and day now :icon_thumbup: Attached below is the stock circuit showing C55 and the 6BE6 (V8) circuit in general.

    Another thing I'm playing with is adding a cap on the AVC bus to ground at the accessory socket. A 1uF electrolytic cap from there to ground makes SSB reception better and louder, but on AM it holds the voltage too far negative after a strong signal comes on. What would be really good is to figure out a way to add the cap for SSB, but then disconnect it when on AM.

    One final thing I also noticed was having the one #47 lamp burned out in the dial changed the meter calibration. I replaced all of the #47 lamps in the receiver first and the meter reacted completely different. What that tells me is any variation with the 6.3V filament voltage and current changes the 12AT7 S-meter amp tube calibration and is the reason why it won't stay zeroed and constantly moves around. I'm debating now on whether to just rectify the 13.5 volt winding of the filament transformer and mounting a couple of 6V TO-3 package regulators on the side of the chassis and just DC regulate all of the tube filaments in the whole receiver. I really don't like altering receivers like that, but it would greatly improve it's performance and solve the problem.

  3. KM1H

    KM1H Guest

    I use a 6BY6 in place of both 6BE6's in my NC-300, better signal handling. The design is 1954-55 era and the 303 has changed little except for sideband switching.

    Meter bulbs run at about 5.5V with a series resistor.

    Id also check your AC line voltage; I use bucking transformers here to run gear at 110-114V or so. Mucho cool and even Hallicrafters drift is very low.

  4. W5HRO

    W5HRO Administrator

    I might try that before I slide it back into the cabinet, thanks. I'm sure I have more than one 6BY6 laying around.

    My line voltage does move around where I live in San Jose and that's exactly what it is. It doesn't just effect only the meter zero either, I'll be listening to a perfectly tuned-in SSB station and then they will start moving up or down a little. You can actually see the dial lamps get a little brighter or dimmer when it happens too. That's one of the prices you pay for living in California, the power grid stinks. With just over 1 million people living in San Jose in the sparsely populated layout it can get pretty bad at times and I live down south where its worse.

    What I might do is just get the two TO-3 package regulators and mount them on the side of the chassis where the transformer and 4H4 is. I could use one diode from the 13.5V winding to get a half-wave out then use the 4H4 socket to mount a big 47000uF or 2200uF cap that pugs in. I could do something like epoxy it down into an old octal tube base so the cap plugs in just like the 4H4. I need to go through the tube manual first and see what each of the tubes draw for filament current then total it up. If it’s less than 1A (which I doubt) then I would only need one regulator.

    I did the same thing to my BC-348 and used a 24V regulator for all of the filaments. My HRO's are ok because I only use them for AM, but the 348 will have the Q multiplier which will make the slightest frequency drift an issue just like SSB reception with this 303.
  5. W5HRO

    W5HRO Administrator

    Ok, all of the tubes minus the lamps and xtal calibrator tube draw 3.9A total. So it would take about five 6V 1A regulators to handle it :confused:

    So, using two regulators and just regulating the 6AH6 oscillator, 6BA7 mixer, 6BE6 2nd converter and the 12AT7 meter amp tube is about the only practical way if using regulators. Those 4 tubes would draw just a little over 1A. Could still throw in the 6BE6 product detector/cwo tube as well. 5 tubes total.
  6. KM1H

    KM1H Guest

    Put those 5 tubes on a DC regulated line, minimal components. It could all plug into the 4H4 socket

  7. W5HRO

    W5HRO Administrator

    I don't think two TO-3 regulators with heat sinks will fit there. Maybe two TO-220 package ones on a single heat sink back to back, but if they get too hot and short, 13.5V to all of the filaments :icon_wtf: The 5Y3 would probably help heat them both up some more too since it's right next to it.
  8. W5HRO

    W5HRO Administrator

    Interesting, I did a search around the web and found some dual cathode +6.2V 5W zeners, but I'm not sure if they are 5 watts total on each side or 5 watts on both sides combined. 5 watts would be too small, but if the cathodes are connected together and used in parallel they would work if that totaled 10 watts. Those could be mounted on a octal tube plug or base easy.
  9. KM1H

    KM1H Guest

    A TO-220 LM-317 is good for 2.2A typical. Put it all under the chassis, sink the regulator, use a mini trim pot, set at 6.0V and forget.

  10. W5HRO

    W5HRO Administrator

    I had looked at that a long time ago when I was considering installing that small regulator board, but there was not enough room under there for it. Maybe if I find a small enough heat sink and solder all the parts under there separately I can make it fit.

    The extra parts in the circuit attached below would give it some protection. If a heat sink wont fit then the only other option would be to mount it on a tube plug with a heat sink then have most of the other parts underneath.

  11. WD5JKO

    WD5JKO Member

    When I try to use a series pass regulator and run out of input voltage (ripple min too low at full load) then I need to recess the options. Bigger input cap, more AC from transformer, higher VA rating of transformer, switch to Schottky diodes in bridge (these have about 1/2 the forward drop of silicon diodes), go to FW CT rectifier to eliminate two diode drops with transformer at 2X the voltage with CT., lower the load current, or PUNT, and switch to a shunt regulator.

    An off the wall idea for a shunt regulator is to stay with AC filaments, use a FWB and a 50W 5v zener. The FWB connects as follows: AC sides to each AC filament rail, + to zener cathode, and - to zener anode. This will make a 6.3v flat topped wave where the peak is constant when the AC mains varies a bit. The zener could be a power transistor coupled to a TL-431 adjustable zener diode.

    Lots of ways to do this.

  12. W5HRO

    W5HRO Administrator

    Full wave center tap ground is out of the question because it will split the 13.5VAC in half. A bridge is the only option. See the attachment below.

    All of the tubes in the receiver only draw 3.9A. If I plug in the xtal calibrator and/or accessories then just a little more, but it's still well under 7.5A.

  13. WD5JKO

    WD5JKO Member

    My Knight V44 project with DC regulated filaments suffers one problem. The power transformer runs hotter then it did when the VFO had AC on the filaments. This is because making 6V regulated at 300ma (1.8 Watts) took 12V at 300ma into the regulator (3.6 watts). The half wave voltage doubler circuit has high peak currents every other half cycle, much higher then when the filaments ran on AC. So the transformer core + copper losses increase. This is offset somewhat by the pause the next half cycle. The bottom line is that that filament winding had to provide > 2X the VA it did before. The transformer is noticeably hotter.

    If there was a way to regulate the filaments with the same transformer, and without all the extra heat. Well there is, and it could be a simple buck type switch mode power supply. I include a picture of one that can provide up to 5 amperes, and run either constant voltage, or constant current. There are a number of these on Ebay, and this one costs only $5.09, Buy It Now. Look for variations to this text string, "5A Constant Current Voltage Buck Boost DC Converter Module".

    With a device as mentioned, the losses go way down on that NC-303 filament winding / transformer since the DC current going into the module will be about 50% what it would be with a linear regulator.

    There is one downside though of a SMPS, and that is noise. The PWM rate is likely > 100 Khz, so the spurrs will be farther apart. That module could be enclosed in a small aluminum sealed box with feed through capacitors for the input and output. Taming the noise issue might just be doable.


  14. W5HRO

    W5HRO Administrator

    Attached below is some additional reading when it comes to some simpler SCR regulation circuits. You can greatly reduce the power dissipation with linear regulators using this approach.

    Attached Files:

  15. WD5JKO

    WD5JKO Member

    Those SCR circuits are interesting. The much younger me used to dabble in that stuff. I developed an affection for unijunction transistors too. The 2N2646 comes to mind, as well as the Ne-2 used to fire the scr(s). I once came up with a dual 555 to trigger the SCR where one part was the usual phase shifter, and the other was a multi spike scr fire arrangement like what Autotronics came up with in the 1970's with their MSD ignition system. I had an MSD-2.

    About 25 years ago now, I was working for a defense contractor. The project was a towed UAV that trailed behind a large aircraft such as a C130, or even Air Force One. There was a 1000' tow cable on a reel that was only 1/4" diameter. Inside the woven sheath were four twisted pairs of #30 wire, and a single fiber optic cable. I got to use 2 pairs for power, and the other two pairs were for RS-232. The fiber was for real time ECM (electronic counter measures) jamming information. This UAV had an onboard computer, a guidance system, and position control ailerons on each of four fins. All this took about 250 watts of power....from 1000 feet of 2 pair #30 wire gauge!

    The power supply needed two stages where the first was a pre-regulator. The 4 ailerons were servo controlled by an LM-12 power op-amp from National Semi (now Ti). Those LM12's were great parts...a legacy part today. The computer was a dos based 286 which booted on a floppy. This thing with the gyro based guidance computer could hold any position behind the aircraft. The power supply was all linear stuff, and it was chosen to use 480V AC 400 Hz down the tow rope. A heavy transformer in the nose of the UAV stepped down the volts and provided needed ballast weight. There was plenty of air cooling available to keep the heat sinks cool. With the pre regulator scheme, high line AC with low load DC was OK, as was low line AC with full load DC. Losses were reasonable this way.

    As to the statement that switching regulator suck, that is often true. That said, the usual SCR circuit used with phase control to chop up the AC incoming wave then "blows". Given the choice of a spur every 100Khz or so with a switcher, or a spur every odd multiple of 60hz up to at least 5 to 10 Mhz, it go with the switcher. Switchers got a bad rap from their early crappy days when the switching rate was around 10-20 Khz. A TV horizontal output at 15.something Khz was an everyday example of switchers putting spurs on 80m, and one always landed on 3880. The newer stuff with PFC correction are much better. Unfortunately the FCC looked the other way when the Asians poured out crap that was supposed to be part 15 compliant. So out went the filters, since they only drove up the cost.

  16. W5HRO

    W5HRO Administrator

    I looked around and I only have 3A Schottky diodes here in the standard DO-41 axial lead package. They make them with higher current values just not in the DO-41 package.
  17. WD5JKO

    WD5JKO Member

    In my last post in this thread I showed a switcher supply that was on Ebay for a little over $5 each. On a whim I bought two of them. So I hastily hooked one up Monday while preparing for the Monday night 20M AM group on 14,330 (about 6:30pm Central time).

    I used the 13.8V supply feeding my Flex 3000 rig. For a load, I wired in an 812 tube (6.3V @ 4 Amps) using a 4 pin ceramic socket. When I turned it on, the Constant Current (CC) led went on. So I turned up the current pot until the CC led went off, and the Constant Voltage (CV) LED went on. Then I adjusted the CV pot for 6.3V. The tube was lite up normal for an 812.

    Looking at the Flex pan adapter, I fully expected the noise floor to have risen, or to see spurs across the 20M band. Since the switcher was using the same power source as the Flex, I expected to hear a buzz or hash in the audio. I saw or heard none of these undesirable traits. I need to explore the other bands, but on 20M, there was no ill effects on the receiver caused from the switcher.

    After several minutes the switcher was getting quite warm. It was putting out 25 watts with no heatsink. It appears to need to be bolted to a small heatsink, and isolated with a thin RTV rectangular pad. I shut it down until I can add a small heatsink. When hot switching it to the 812 tube filament load, the switcher goes CC as the filament warms, and in about 3 seconds the CV led comes on as the CC led goes off.

    I am not saying that this approach is viable for the NC-303 project, but since I mentioned it, and we've debated the pros & cons of switchers versus linear regulators, and linears with an SCR pre-regulator, I thought I'd add my switcher story to the mix.

    There are several of these on Ebay right now. Search with EIN # 6913281574266

  18. W5HRO

    W5HRO Administrator

    How much current are those things good for? How much will they output?
  19. WD5JKO

    WD5JKO Member

    Here is what they say on the auction package:
    Size: 5 x 2.6 x 1.2 cm
    Input: 5V-30V
    Output: 0.8V-28V
    Output current: 0.1A-5A
    Maximum power: 30W
    Note: This product can only lower voltage and can Not raise voltage.

    I was running it pretty hard. For smaller DC filament needs, like 6.3V at 1 amp, I doubt it needs any heatsink unless the ambient gets pretty hot. I was brutal to this thing, feeding it from an unfused 13.8V 30 amp supply, and using an 812 filament for a load. I hooked up the input HOT since my Flex was on. There was a bright orange flash initially, and I thought I'd done it now. Turns out that was an LED...There are three of them, "Power", "CC", "CV". Pretty neat, and it survived!

    On your NC-303, those LT regulators will do a good job. The problem is the AC to DC conversion with FWB and capacitor input filter. The diode charge current lasts but a fraction of each half cycle, and the peak current can be 10X what the transformer winding was designed for. So the winding resistance is a bigger factor as is the transformer leakage inductance. Then them diodes are always having two in series for double diode drops. For these reasons, the VA rating of the transformer winding must be de-rated when making DC this way. A lot depends on individual circumstances, components used, etc. Still a 50% derate is not uncommon to get the transformer temperature to what it was before with a resistive AC load.


    Attached Files:

  20. W5HRO

    W5HRO Administrator

    Finally got around to working on this project again. Mounted a 50A bridge rectifier and started wiring everything up, but as soon as I ground the negative terminal of the bridge the output dies.

    After looking at the factory schematic again then ohming out the winding and to ground I know why. The one end of the winding is shorted to the iron inside of the transformer. I saw that in the beginning, but I assumed it was just the way the schematic was drawn up and wasn't real. It was just some unnecessary thing an engineer must have done at National. Maybe they thought it would remove heat, provide more current, or reduce noise somehow. The only way to use the winding would be to unhook all of the transformer's wires, remove it, then take the housing apart and that's not happening. I'm sure it just needs to be cut loose unless it was done in the middle of the core where you can't get to it, but its not worth finding out.

    What I'm going to do is mount a separate filament transformer inside. I hate doing it, but the 13.5VAC winding it not usable with bridge rectifier because of that internal ground and if I go halfwave there wont be enough voltage when loaded to keep the regulators stable ::)

  21. W5HRO

    W5HRO Administrator

    Update: I ordered and installed a Hammond 166Q14 which is a 14V @ 6A transformer. Should be more than enough :icon_thumbup: That will remove the big load from the power transformer anyway which is probably a good thing.

    I'm building up this NC-303 to be a workhorse and I want it stable and reliable. Without regulating the filaments it wont be.

  22. W5HRO

    W5HRO Administrator

    It's been a while, but I decided to use and already received twenty L7806CV's from Mouser and will start installing them. There is also a metal plate behind the S-meter where I could mount the final one for the dial and meter lamps as well. Also, after thinking about it I don't think its a good idea to connect the HF oscillator and 2nd converter to the same regulator. The HFO needs to be left on it's own supply line and fed with coax or be well bypassed.

    I'll post an update when its finished, but I'm now considering the below instead. The mixer really isn't necessary and I would like to also separate V8 and V9 as well, but that would require 5 regulators total. I may still do that before its over though.

    Regulator 1
    V3 = 6BE6 = 0.30A = 2nd Converter

    Regulator 2
    V7 = 6AH6 = 0.45A = HF Oscillator

    Regulator 3
    V8 = 6BE6 = 0.30A = CWO/Product Detector
    V9 = 12AT7 = 0.30A = S-Meter Amp

    Regulator 4
    I1 = #47 Meter Lamp = 0.15A
    I2 = #47 Dial Lamp = 0.15A
    I3 = #47 Dial Lamp = 0.15A
    I4 = #47 Dial Lamp = 0.15A


    Attached Files:

  23. W5HRO

    W5HRO Administrator

    Here is an update. I have the main supply installation, bridge and the first two regulators done and working well. Both the HFO and 2 converter tubes are on there own individual regulator and the voltage at both tubes is exactly 6.15Vdc. The tube of L7806CV's I bought are from the same lot so they should all output 6.15V. The inputs are sitting at a little over +19V which is a bit hot, but still ok. I will see if it drops any after I install the remaining regulators this weekend. Then if it's still a bit hot I will add a small value resistor between the two terminals seen in the attached Filter Caps pic below to drop it down some. Those are the two 4700uF caps at the output of the 50A bridge and the resistor would be between the two caps (outside terminals). The 3rd 4700uF cap is mounted close to the first two regulators. I also installed push-on terminal connections to the bridge in case I decide to use even a bit heftier bridge before its over or if the current one was to go bad.

    Anyway, I will mount the regulator for the product detector (PD) tube where I have it in the PD Reg Location pic. Both the PD and the S-Meter amp tubes are right there to the left of it. As for the S-meter lamps and dial lamps I will mount the regulator on the plate right there behind the meter as seen in the Lamps Reg Location pic. It's the perfect location because the wires and supply feed are already right there. I might use two regulators in parallel there so they run cooler. The Hammond transformer is 6 amps so its more than enough to drive several regulators.

  24. W5HRO

    W5HRO Administrator

    Yesterday I also finally changed out C130 which connects to the mode selector switch. I had run out of 1uF caps a while back so I had to find some. I normally change those, but it still had the old stock Pyramid brand cap and I think it may have been leaky. I had already changed out the old 1N1692 AVC clamp diode because it shorted and I'm wondering if maybe the bad cap caused that. When you switch to SSB it connects C130 and every once and a while it would come up with no audio and I had to switch it several times before the audio would come back on SSB. The old cap measures 1uF, but I'm guessing its probably leaky. The switch itself is fine and the contacts were already cleaned.

    I also came across some old notes someone had sent me a few years back when I had my first NC-303 and supposedly they came from the old 1960 CQ magazine Hints and Kinks section. I had already changed C55 to a 50pF cap to get wider SSB audio, but I installed a 100pF cap from that point to ground for attenuation instead. I'm wondering if the 5pF mod would work better than what I already did or if that mod was more geared towards CW and not SSB? I need to find the 1960 CQ article, but I'm currently not sure which months issue it was in. It also looks like they are changing the coil voltage on the CWO/PD tube as well by making R46 270-ohms instead of 120-ohms. The last change deals with R72 and R73 which goes to the noise limiter tube.

    1. Change C55 from 33 to 5 pf.

    2. Change R46 from 120 to 270 ohms.

    3. Interchange resistors R72 and R73 (Make R72 100K & R73 470K).

  25. W5HRO

    W5HRO Administrator

    I picked a different place and installed the other regulators. With all 5 regulators in place the input is around 17.4V so I will need to add the dropping resistor to get it down to around 10V so they run cooler, but everything is working now. I also have a temporary 0.4-ohm series resistor between the output of the 5th regulator and the lamps. It put the lamps right at 6.00V, but I want them a hair bit lower. Maybe around 5.7V or so. That way they will last longer.