Galaxy DX-2517 Mosfet version RX issue

[Hawkeye351]

Got a DX - 2517 Mosfet version on my desk as a return. Radio is practically brand new. No swing mods whatsoever, no volt mods whatsoever, no hacks at all. It's stock, other than 11m added.

The owner had just bought it from another local operator. After about 2 weeks of running it, he decided to have it checked out and aligned back to spec to be on the safe side.

Well, all I did was:
Adjust 10.240 Crystal adjustment per service manual.

Adjust VCO voltage per service manual.

Adjust VCO Buffer per service manual.

Adjust PLL and Carrier oscillators per service manual.

Adjust RF Stage per service manual and spectrum analyzer.

All voltages are correct
All frequencies are spot on

After I carried it back to him, he used it over night and then notified me the receive was down. I told him I don't mess with the receiver adjustments at all because my signal generator is down due to an inverter board in the display, therefore I never touch receive.

I drove over and checked it out while it was still hooked to his antenna. He was right, the receive was acting weird. It's as though the radio was on a beam and picking certain local stations up strong, but others in the opposite direction were non existent on signal, barely could hear them.

Well, another friend of ours brought his 2517 that I went through over there to compare the receive on the same antenna, our friends receive didn't act like that, receive was great.

So, I touched up on receive with antenna hooked up and had several locals key up to give me a signal to align to, I only chose those locals running radios I went through because I know they're dead on frequency. Well, same result, those in one direction came up a lot, whereas those in other directions dropped or completely vanished.

We're all stumped on this. We've never experienced a radio that acts as a directional receiver using an omni directional antenna.

Anyone ever run across the same weird issue before? How did you resolve it?

This incident actually makes me look bad to newcomers. The old timers know it's not me, they trust my work.

I've got the radio back on my desk again now, trying to figure this strange issue out, but like I said earlier, we are all stumped on this.

 

[Hawkeye351]

Hmmmm, odd...

Going back over my previous alignment on this 2517, it seems as though everything drifted off at his location for some reason.

PLL oscillators were off about 10hz each.
Carrier oscillators were off about 10hz each.
VCO voltage was down .06v
Buffer was still set where I previously set it.

I wonder why the oscillators shifted 10hz from my desk previously to his location...

I'm thinking these oscillators (PLL and Carrier) is what was throwing the receive off.

Gonna try it on air later when the locals get off work and start rag chewing.

I'll give an update as soon as I can test it on air.

I could really use a signal generator, lol...

 

[brandon7861]

How long did you let the radio and FC warm up before doing the alignment? Takes about an hour for both to be stable enough.

What probe did you use to measure the frequencies? Even a good 10x probe can load down a galaxy oscillator.

In my limited experience, I don't really recall having issues with other radios, just galaxy radios at TP6 doing the SSB side. My probes always loaded it down and either the FC would jump around, or it would be stable yet off a little due to the probe capacitance. I tried all sorts of stuff with my 10x probe, like adding a capacitor to the end and inline amplifiers with little success. When I went to a DIY active probe with virtually no input capacitance (because the amp is right at the tip), I have had no further issues at TP6.

Edit: 10Hz error is within spec and I doubt thats the issue. Most people wouldn't hear the difference. You could probably get by doing a quick and dirty RX tuneup without a signal generator. If your volt meter can do AC up to a Khz, have one of your friends transmit (with the lowest power possible after skip goes away) a constant tone and adjust the receive cans while watching the audio voltage going to the speaker.

 

[Hawkeye351]

I used a standard 10x probe.

The first alignment on this radio I did a few days ago, I did let it sit for over an hour before aligning.

But this time, it was on for only about 30 minutes before I aligned it.

Just checked it again since it's sat over an hour after realigning it, and everything is dead on, like I had set it over an hour ago.

Gonna look into that active probe trick. It makes a lot of sense with your explanation. Thank you for that tip.

Also gonna see if I can get one of the locals to try the poor man's receive alignment trick you shared. Thank you for that tip also.

Thank you for chiming in.

 

[nomadradio]

Unhooking a probe from an oscillator circuit can make the frequency "jump" from where it was set with the probe attached.

There is an easy dodge, so long as you have some sense of tone pitch. Tune in the carrier crystal with a receiver that covers the crystal frequency. Use sideband, either one and tune the receiver for a tone around 1 kHz. Doesn't need to be exact. Set the trimmer cap on the crystal with the probe attached for the desired counter reading. Now detach the probe. If you hear a significant shift in the pitch of the tone in the receiver, touch the probe back to the circuit. Note the pitch of the tone in the receiver. Unhook the probe and touch up the trimmer to match the tone pitch you had when it was attached.

A back door, but a cheap one if you have that extra receiver on hand.

73

 

[brandon7861]

An audio spectrum app would be helpful with Nomad's suggestion as it will show relative changes in that tone that might be too small for your ears to detect or to see how much it changes it.

Another phone app option to make the process of detecting probe effects easier is a tone generator. Instead of the spectrum app, beat that 1kHz SSB receiver tone with another tone of the same frequency and you will hear a 1Hz change from removing the probe (not that it has to be that accurate though)..

Edit: Realistically, if you can't hear the change by ear in the receiver, its well within that 20Hz spec. I just mentioned this as a way to know how far off (if its a lot) or to be able to hear tiny changes if curiousity demands it.

 

[Hawkeye351]

Got it,
Was a cold solder joint in the 1st IF section. Fiddled with the alignment and receive over and over again, checked voltages and frequencies over and over again.

Finally started proding around with my nylon tipped tool on all the solder joints in the receive section. BAM, 1 cold joint at the 1st IF section. Guess it came loose the first time I had it from moving it around so much. Removed old solder, applied new solder, she now received better than ever, although I did crack the receive back a little, I had it cranked up so much due to the issue that it had way too much hash.

Q17 was the part that had cold solder joints, if anyone ever runs into this issue.

Thanks everyone for all the suggestions.
Also, thanks for the poor man's receive alignment tip, worked great. I even did the noise blanket alignment with a volt meter, turned out perfect, thanks for that tip.

You guys are great.

 

[Eldorado828]

Got it,
Was a cold solder joint in the 1st IF section. Fiddled with the alignment and receive over and over again, checked voltages and frequencies over and over again.

Finally started proding around with my nylon tipped tool on all the solder joints in the receive section. BAM, 1 cold joint at the 1st IF section. Guess it came loose the first time I had it from moving it around so much. Removed old solder, applied new solder, she now received better than ever, although I did crack the receive back a little, I had it cranked up so much due to the issue that it had way too much hash.

Q17 was the part that had cold solder joints, if anyone ever runs into this issue.

Thanks everyone for all the suggestions.
Also, thanks for the poor man's receive alignment tip, worked great. I even did the noise blanket alignment with a volt meter, turned out perfect, thanks for that tip.

You guys are great.

It's absolutely crazy how often that's an issue in a ranger board and has been for years. Even their smt boards suffer the same issue though not as common as through hole. My x-9 brand new had a bad can and a bad solder joint too go figure but thankfully it was warranty repaired.

 

[Hawkeye351]

Well, I still have this radio and was listening just to be sure I got it right before I get it back to the owner. The only issue I see now is that any very strong signal on any adjacent channel bleeds over to the next channel as though that signal is on that adjacent channel.

Example:
The guys on the bowl (channel 6) put the same signal and sound exactly the same on channel 5.
Any strong signal on channel 5, sounds the same with the same signal on 6.
Any strong signal on channel 2 sounds exactly the same and with the same signal strength on channel 1.
Like it has poor adjacent channel filtering on receive.

That's with the noise blanket/ANL turned on. If you turn those off, it's worse.

I went back over solder work again, everything fits snug, no more bad solder, everything seems fine. Just the adjacent channel bleedover.

Radio is only about a year old.


AM Crystal filter?
IF/RF alignment?
Receive still too high?

Anyone ever run into this?

 

[nomadradio]

Like it has poor adjacent channel filtering on receive.

The AM has two filters. A broad one with only three inline legs at 10.695 MHz. The one that narrows you down to just one channel is the black plastic one with five legs, typically marked CFU455H.

That's the one that should suppress the adjacent channels.

73

 

[Hawkeye351]

Could that 455kz (cfu455) be bad in a practically new radio? Or one of the 455kz IF adjustments being off have that much of an effect on the adjacent bleed over? I have all four maxed out for signal.

Or, could L5 (typically L6 in most 3400 boards) being maxed out cause this issue?

On a quiet night there's always about 3 S-units of hash. Whereas my yaesu 950 and galaxy 949 is around .5 up to 1 S-unit on the same antenna.

I think I may still have the receive just a bit too high. But which one will effect adjacent bleed over the most?
L5,
L6,
L7,
L8,
Or one of the 455kz IF adjustments?

 

[nomadradio]

Could that 455kz (cfu455) be bad in a practically new radio?

Routinely. It's common to find that part off frequency by 1.5 kHz or more in a new radio out of the box. Yours is just worse than average.

73

 

[Hawkeye351]

Got a call from the owner today. He's extremely happy with how it turned out. I told him I had help from some of the best techs I know, you guys. He thanked me and asked me to thank all that helped.

Thank you guys, you made an old man very happy.