Hello Forum...I have an SX-100 Mark 2 that I am bringing back to life. I have the radio basically working pretty well right now. I replaced all the tubular paper caps and some off-value resistors. I've tested all the tubes and they are OK according to my tube tester. I then did an alignment per the manual and I can hear my signal generator making AM down to around 2uV. When using SSB mode it's a bit better. Tuning around the shortwave bands I hear plenty of foreign broadcast stations. The remaining issue is the signal strength meter. A few seconds after power up the meter swings over to the 0 mark like you would expect but it never moves off the 0 unless I use an extremely strong signal (like 1mV!) from my signal generator. The meter moves upscale when the sensitivity control is dialed back so I know the meter works. I've tested all the parts around the AVC detector and they are OK but the radio is not generating any negative AVC voltage out of the detector until I put in the really big signal. I've also tested every cap and resistor that's on the AVC line, all OK. The meter potentiometer on the back panel makes the needle move around, seems to work.

I am stumped. Any ideas are much appreciated. Thanks!...Mike W8MX

Hi Mike, Mike here too.

As long as the AVC acted normally during tune-up, all should be OK there. The SX-100 mkII drives the S meter off the plate of the 1650Kc I.F. amp, V4. It is not driven off the AVC as you would expect in most designs. I'd start by checking the voltages on the pins of V4 and the resistors and caps around the meter circuit.

Even though the meter is not actually on the AVC line it is responding to what the AVC line is doing. So, have you put a high impedance voltmeter on the AVC line itself to see what it is doing? You said the detector isn't actually putting out AVC, but was that measured with a meter or being inferred by the S-meter action?

Voltages while on no station versus the voltages while on a strong station should be measured. Expect 0 to -1 volts with just low level background. Expect -15 to -20 when on a very strong station. If these are not happening then something is loading the AVC line itself (multiple possible causes). Otherwise the problem is in the meter circuit.

If you are getting no actual AVC voltage some resistance measurements are probably in order. The AVC line should show a very high resistance to ground (2 or 3 megohms depending upon where you measure it). If not, there is a stray leakage problem.

Another thing that can happen is grid leakage in one of the AVC controlled stages. When this happens one or more tube grids are attempting to pull the whole AVC system positive and over ride the ability of the detector to produce an AVC signal.

Voltage measurements and resistance to ground measurements at point 1 and 2 below would tell a lot about the health of the AVC.Curtis Eickerman

That might be part of your problem as you should be able to hear a 30% modulated AM tone much lower than that as the SX-100 is a pretty sensitive radio. Id say that on 20M a 10dB SNR on AM should be around .5uV.

Carl

Hi Mike,

OOPS, I apologize, have to admit I didn't read your ENTIRE post, or the alignment section of the manual, before I started typing. Curtis and Carl gave you good advise to follow.

After you check the circuit voltages and meter amp tube, if the meter is acting like it is too sensitive, an internal series resistor may be shorted, shunt resistor may be open, or the magnet may be weak.
One radio I worked on needed changes in the meter amp to reduce the current needed to pin the backwards-reading-meter at zero.
Once the max current was set to what the meter wanted, the indications were corrrect.
Pat

Thanks to all for the help. Burnt Fingers comment got me looking more deeply. I agree the radio is lacking in sensitivity and that is probably the root cause of the S meter issue. All the tubes test at or near 100% but I need to have the Sensitivity and Volume controls at full gain to hear stations. I do hear a lot of stations but nothing is very loud. I am suspicious of the 2nd converter sub-chassis stage because when I did the alignment per the manual the slugs in T9 wound up adjusted pretty far from where they had been. The top side adjustment now sits about 3/4 inch out of the case where the setting when I got it was only 1/4 inch out of the transformer top. Also, the manual states that a 1/2 Watt audio signal should be produced by injection of a 20uV AM signal at the 1650 Hz IF frequency directly into the grid of the first mixer, V2. This radio gets nowhere near that.

It would be helpful to know how much gain or loss each stage of the radio should have. Does anyone know if I should expect the 2nd conversion stage to have gain? I was going to try and test it alone. How about the gain of the various amplifiers? I would like to prove that everything else is OK before I try to pull that sub-chassis out.

Comments? Ideas? Thanks...Mike W8MX

Mike,

There should definitely be conversion gain in the second mixer and low injection level is the most likely cause of low gain/loss in this stage. It is certainly suspicious that you had to adjust the slug that much. If the output of the crystal was somewhat low at the peak then setting it for half output is going to be a problem. This is a somewhat odd instruction as typically the goal is to ensure that the crystal will always start reliably and this condition should be met by tuning for only a SLIGHT decrease on the gradual side of the peak. The later SX-101A instructions were changed to adjust the 1700 KC oscillator coil for a 6db drop and then to adjust the 1600 KC coil to match this level. Although the instructions clearly say half output I have to wonder if the original intent was 3 db instead of 50% of the total output as measured on your output meter (assuming, for example, that your output meter scale covers a full range of 30db).

Does your gain/sensitivity change when you change from the USB to LSB setting?

Also, there were originally a number of tubular caps in the selectivity switching circuits of the 50.75 KC IF. You might want to recheck your replacements to make sure that you didn't make a mistake in value with these. When you switch through the selectivity choices from .5 to 5Khz. does the bandwidth seem to widen normally without radical changes in gain?

Thank you for your comments. Yes, I too find the 2nd mixer adjustment instruction perplexing. I thought they might have been trying not to overdrive the mixer with the 1600/1700 Hz LO. I wonder what others think. Comments??

The radio's sensitivity does not seem different on the USB/LSB settings as long as I adjust the 1600 and 1700 Hz signals to the same level.

I will double check those caps in the selectivity switch but I think it sounded normal to me when I switched through the settings.

I think my next test will be to measure the conversion gain of the subchassis. I might also unsolder the coax from the subchassis output and inject a 50.75kHz signal and measure the 50kHz IF gain.

I have fixed dozens of radios in my time as a ham but this has been the most troubling. I suppose it will also be the most rewarding when I finally find the problem.

Thanks for the help...Mike

Did you replace the tubular paper caps inside the subchassis? My SX-100 contained two of them which were both leaky.

Dave

If you haven't pulled a Hallicarfter's Dog House before, its not nearly as formidable as it first appears.



I took a couple pics as I pulledl it, to help when I put it back in. I also suggest you replace the original coax with some RG-174. It makes the installation process a lot easier.



This straight 101 was my first Hallicrafters in 50 years or so and I was impressed with the receiver's design and performance.

Thanks Mike and Dave...I lifted the subchassis output coax and injected a 50.75kHz signal from my generator, with about 30% AM. I could run it down to a little less than 1mV and still hear a decently strong signal in the speaker. I think that proves the 50.75kHz IF, the AM detector, and the audio amp are all working although I wish I knew if 1mV was a reasonable level. Seems reasonable to me.

I guess I will bite the bullet and take that subchassis out of there. Good suggestion on the coax. What's in there looks like RG58, pretty fat and stiff. I replace all the tubular paper caps as a routine thing. I often measure them after I get them out and they are usually out of tolerance at a minimum. I check all the resistor values and replace as needed.

I'll keep the group posted on the progress. Thanks to all...Mike

I have extracted the sub-chassis and opened it up. I was hoping to find some tubular paper caps but I only see a handful of disk ceramic and a couple rectangular micas. I will replace all the caps and measure the resistors. I notice that T9 apparently has a couple 300pF caps inside it. I wonder if it's worth taking it apart to replace those. Anyone ever see these go bad? They are probably micas as well.

Ooops...sorry...I must have been thinking of the BFO subchassis in an S-16 that I recently restored.

Dave

When you adjusted the oscillator coils in the sub-chassis were you using an output meter (calibrated in db) or a regular scale measuring AC volts? If using an output meter adjust for about a 3 db drop (AVC off of course), it using an AC voltmeter adjust for half of peak indicated voltage.

I would certainly retest your radio with the oscillator injection set closer to the peak to see how that impacts overall gain. Given what you have found so far, insufficient second oscillator injection could certainly be the problem.

As for the 300 puff caps, you're rite there so why not just replace them. The slugs seemed to adjust "normally" on my radio so I left them.

Its strange that since they have a block diagram, they didn't show the stage gains.

He has a SX-100, not a 101. While similar in some areas there are differences.

Carl

Understood, but the question is, does the SX100 manual show stage gains,



apparently not, or he wouldn't be asking about stage gain numbers.

A lot of the small value caps used in this radio are small tube shapes with the wires wrapped and soldered around each end. The tube appears to be coated with silver and they have painted color code markings. These are especially prevalent in the bandswitch wiring. Does anyone have any experience with the failure rate of these parts? Obviously the paper caps had to go but what about these and the "domino" micas?

Tubular caps, they are probably glass, but may be ceramic. Failure rate, "0" I suspect. As for the molded mica's, I've never had any problems with them either, but others may have.