|
Hi all,
I've been trying to troubleshoot an amplitude problem on my new 492, which has had multiple other little problems along the way. At this point, I can get a stable display and frequency is OK, but amplitude is 20-30 dBm lower than it should be. Attenuator and first mixer seem good, but I'm missing the 110 MHz output from the 2nd converter (should be -37 dBm or -39 dBm depending on band), measuring with HP 8484A.
The front panel 2nd LO output is ~17.6 dBm on the first band, and I measure +9 dBM and -1.5 dBm directly on the 2nd LO cavity oscillator, which all seems in the ballpark (and at least verifies the oscillator is not dead).
Could the problem be something as simple as the oscillator being off frequency? My microwave counter has a power supply problem that I'm also trying to fix, and I doubt a 7S11/7T11 combo is precise enough to adjust it. I'm mildly tempted to just try turning the adjustment screw on the cavity oscillator and see if the display responds.
Any other ideas? I'm not enthusiastic about opening up the 2072 MHz 2nd converter, but that looks like the only thing in the signal path that would lose the 110 MHz IF on all bands.
thanks,
Adam
|
Actually, a 7S11/7T11 combo will give you pretty accurate measurements of RF
signals (typically better than a power meter). The signal needs to be
large enough though, if you talking 0dBm that's fine (0.224Vrms), but if the
signal level is smaller than that, you will have problems and below about
-7dBm (100mV rms), you'll probably be out luck.
David
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] < [email protected]> On Behalf Of Adam R. Maxwell via groups.io Sent: 13 March 2025 06:01 To: [email protected]Subject: [TekScopes] 492 has no 110 MHz IF Hi all, I've been trying to troubleshoot an amplitude problem on my new 492, which has had multiple other little problems along the way. At this point, I can get a stable display and frequency is OK, but amplitude is 20-30 dBm lower than it should be. Attenuator and first mixer seem good, but I'm missing the 110 MHz output from the 2nd converter (should be -37 dBm or -39 dBm depending on band), measuring with HP 8484A. The front panel 2nd LO output is ~17.6 dBm on the first band, and I measure +9 dBM and -1.5 dBm directly on the 2nd LO cavity oscillator, which all seems in the ballpark (and at least verifies the oscillator is not dead). Could the problem be something as simple as the oscillator being off frequency? My microwave counter has a power supply problem that I'm also trying to fix, and I doubt a 7S11/7T11 combo is precise enough to adjust it. I'm mildly tempted to just try turning the adjustment screw on the cavity oscillator and see if the display responds. Any other ideas? I'm not enthusiastic about opening up the 2072 MHz 2nd converter, but that looks like the only thing in the signal path that would lose the 110 MHz IF on all bands. thanks, Adam
|
If you want to measure frequency, then the 7D14 is probably the tool choice
(to get in the ball-park anyway), but GPS locked counter will be better.
David
toggle quoted message
Show quoted text
-----Original Message----- From: David C. Partridge <david.partridge@...> Sent: 13 March 2025 11:03 To: ' [email protected]' < [email protected]> Subject: RE: [TekScopes] 492 has no 110 MHz IF Actually, a 7S11/7T11 combo will give you pretty accurate measurements of RF signals (typically better than a power meter). The signal needs to be large enough though, if you talking 0dBm that's fine (0.224Vrms), but if the signal level is smaller than that, you will have problems and below about -7dBm (100mV rms), you'll probably be out luck. David -----Original Message----- From: [email protected] < [email protected]> On Behalf Of Adam R. Maxwell via groups.io Sent: 13 March 2025 06:01 To: [email protected]Subject: [TekScopes] 492 has no 110 MHz IF Hi all, I've been trying to troubleshoot an amplitude problem on my new 492, which has had multiple other little problems along the way. At this point, I can get a stable display and frequency is OK, but amplitude is 20-30 dBm lower than it should be. Attenuator and first mixer seem good, but I'm missing the 110 MHz output from the 2nd converter (should be -37 dBm or -39 dBm depending on band), measuring with HP 8484A. The front panel 2nd LO output is ~17.6 dBm on the first band, and I measure +9 dBM and -1.5 dBm directly on the 2nd LO cavity oscillator, which all seems in the ballpark (and at least verifies the oscillator is not dead). Could the problem be something as simple as the oscillator being off frequency? My microwave counter has a power supply problem that I'm also trying to fix, and I doubt a 7S11/7T11 combo is precise enough to adjust it. I'm mildly tempted to just try turning the adjustment screw on the cavity oscillator and see if the display responds. Any other ideas? I'm not enthusiastic about opening up the 2072 MHz 2nd converter, but that looks like the only thing in the signal path that would lose the 110 MHz IF on all bands. thanks, Adam
|
Could the problem be something as simple as the oscillator being off frequency? My microwave counter has a power supply problem that I'm also trying to fix, and I doubt a 7S11/7T11 combo is precise enough to adjust it.
If you want to finely adjust a 110MHz signal, a 7T11 time base would be useless with its very limited frequency accuracy and resolution. Raymond
|
On Mar 13, 2025, at 04:21 , Raymond Domp Frank via groups.io <hewpatek@...> wrote:
Could the problem be something as simple as the oscillator being off frequency? My microwave counter has a power supply problem that I'm also trying to fix, and I doubt a 7S11/7T11 combo is precise enough to adjust it.
If you want to finely adjust a 110MHz signal, a 7T11 time base would be useless with its very limited frequency accuracy and resolution. Sorry, I should have been more clear. I was looking at the manual step "Check/Adjust the Cavity 2nd LO" which says to connect a microwave counter to the 2nd LO output and measure 2182.0 ±0.5 MHz, and to loosen a lock nut and adjust the fine tune adjustment in the cavity. (p. 345 of S/N B029999 and below service manual). You confirmed my belief on the 7T11. There is no 110 MHz frequency to adjust directly with a counter, because if I understand correctly (big if!), that intermediate frequency is produced by mixing the 2nd LO 2182 MHz output with the 2072 MHz input signal from the 1st converter. I'm trying to understand why the 110 MHz IF is dead, even though the LO signal is present at approximately the right level. thanks, Adam
|
if I understand correctly (big if!), that intermediate frequency is produced by mixing the 2nd LO 2182 MHz output with the 2072 MHz input signal from the 1st converter. I'm trying to understand why the 110 MHz IF is dead, even though the LO signal is present at approximately the right level. Very small “if”, I’d say.” -;( You may look at those two ~2GHz signals with your 7T11/7S11, if you have a suitable sampling head, like e. g. an S2 or S4 and probing setup. Same for the 110MHz mixing result, if even to see if any of it is there. Not sure what Dave P. means about signal levels. Operating input level <= 1Vpp or 350mV RMS. Safe PP 5Vmax. Recommended input level between say 50mV RMS and a few 100mV RMS at most, ISTR. You’ll need a faster counter plug in than either a 7D14 or 7D15 to measure their frequency. Take the impedances of the signals you look at and your test setup into account but I guess you know all that. Raymond
|
Hi all,
I had a breakthrough over the weekend with my balky 492P. Rapping the VR module with a rubber screwdriver handle brought back the missing ~20 dBm of amplitude in fits and starts, so I replaced it with a spare and it seemed fine, working as expected with reference levels.
I was super excited until I started it up the next day and it had crapped out again: on the -20 dBm reference level, a -20 dBm signal is 2-3 divisions down from the top. Sadly, I did not check 110 MHz IF output power while it was working, so have no idea if I was on the right track with that.
Having resurrected (for now) my new HP 5350B counter,[1] I was curious to measure the LO frequencies.
The 1st LO on manual sweep can vary from 1.894 to 6.680 GHz. The service manual says the YIG has a tuning range of 2.072 to 6.35 GHz,[2] so I guess this is okay?
The 2nd LO maxes out at 2178 MHz in band 1. This is an old 492 with the cavity 2nd LO that has a tuning slug in the side, but it cannot be adjusted to 2182 MHz per the service manual.[3] If I disconnect the cable on the +10 dBm output of the 2nd LO, though (which goes to the four-cavity filter of the 2nd converter), the front panel output of the LO is at 2182 MHz as expected.
At this point, my guess is a bad 2nd LO, but apparently the Shaper and Bias board next to it tunes and sweeps(?) the 2nd LO. I left out a cal step of adjusting 1st LO and 2nd LO Sweep,[4] so maybe that warrants a closer look, as I thought the 2nd LO operated at a fixed frequency. None of this explains why it worked for a while on Saturday, though; some temperature dependency is all I can think of.
thanks,
Adam
[1] The counter reads 999.9998 MHz on the 1 ns output on my TG501
[2] page 5-5 of 492P service manual for SN B029999 and below
[3] page 3-45 ibid
[4] page 3-44 ibid
|
On Mar 19, 2025, at 00:03 , Adam R. Maxwell via groups.io <amaxwell@...> wrote:
The 2nd LO maxes out at 2178 MHz in band 1. This is an old 492 with the cavity 2nd LO that has a tuning slug in the side, but it cannot be adjusted to 2182 MHz per the service manual.[3] If I disconnect the cable on the +10 dBm output of the 2nd LO, though (which goes to the four-cavity filter of the 2nd converter), the front panel output of the LO is at 2182 MHz as expected.
At this point, my guess is a bad 2nd LO, but apparently the Shaper and Bias board next to it tunes and sweeps(?) the 2nd LO. I left out a cal step of adjusting 1st LO and 2nd LO Sweep,[4] so maybe that warrants a closer look, as I thought the 2nd LO operated at a fixed frequency. None of this explains why it worked for a while on Saturday, though; some temperature dependency is all I can think of. Reading TFA and TFM again, the 2nd LO SWEEP is only sent at spans ≤50 kHz/div in bands 1-3. Probing the shaper board, I could indeed see a ramp at those spans, so clearly that wasn't what was pulling the 2nd LO frequency down at full span. Since the cavity oscillator itself is obviously capable of 2182 MHz, I figured the control voltage must be wrong…and happened to have another bias/shaper board to swap in. LO (ha!) and behold, it was now running at ~2190 MHz, and I was able to adjust it down to 2182. As a bonus for 2nd LO frequency correction, the Phase Lock now works properly, which makes me happy. I assume a resistor has drifted or maybe a diode is dead in the nonlinear amplifier. The replacement doesn't have the fancy dual zener reference, so maybe I'll look at it again someday. Another problem possibly found: it's been starting up deaf when cold (no zero spike, just baseline noise), and suddenly comes alive. Tonight I plugged the 100 MHz output into a 7104 when I started the 492 up, and the 100 MHz oscillator was dead. Also swapped that out, but the replacement doesn't quite make -20 dBm. The original has an air variable cap in the middle fully meshed, which seems fishy. In spite of this progress of sorts, a -20 dBm signal is still 30 dB below the top of the display, so I'm still missing something. thanks, Adam
|
Continuing my diary of 492P woes in case it's helpful to some other poor schlub reading the archives in future. On Mar 20, 2025, at 22:00 , Adam R. Maxwell via groups.io <amaxwell@...> wrote:
Another problem possibly found: it's been starting up deaf when cold (no zero spike, just baseline noise), and suddenly comes alive. Tonight I plugged the 100 MHz output into a 7104 when I started the 492 up, and the 100 MHz oscillator was dead. Also swapped that out, but the replacement doesn't quite make -20 dBm. At least one version of the service manual says you can replace select resistor A34A1R1018 if output of the 100 MHz reference oscillator is low. Schematic shows it as 309, and this one had over twice that. Adjusted for correct amplitude by setting the 100 MHz peak to a -20 dBm peak from my 8640B (set using a power meter). The original has an air variable cap in the middle fully meshed, which seems fishy. "Tuning capacitor C3031 in the collector circuit serves to adjust for maximum output." It was adjusted for maximum output, so no problem there. Still takes ~5 minutes to start. In spite of this progress of sorts, a -20 dBm signal is still 30 dB below the top of the display, so I'm still missing something. I left it running for a couple hours yesterday while cleaning up the shop, and noticed that the 100 MHz reference had suddenly regained 30 dBm. Checking power input to VR module, it was -25 dBm. This morning, it was once again down 30 dBm on the display, and power to VR module was about -26 dBm, so I'm thinking all the RF plumbing on the bottom is fine. At this point, I'm thinking it's probably temperature-related, but I'm stuck until I can fabricate extenders to poke at the VR while it's operating. Ideas welcome. Adam KK7ASV
|
Adam,
Could the transistors be bad? The 2N5179 replaces ones on this board of the p/n, 151-0282-00. If you have component cooler, use that to briefly spray on the transistors to see if cooling them some makes a difference. If any transistors are bad, Mouser has the transistor in stock. The two 68mfd decoupling condensers for the B+ supplies on this board could be high in ESR.
Mark
|
Hi Mark, On Mar 23, 2025, at 11:04 , Mark Vincent via groups.io <orangeglowaudio@...> wrote:
Could the transistors be bad? The 2N5179 replaces ones on this board of the p/n, 151-0282-00. If you have component cooler, use that to briefly spray on the transistors to see if cooling them some makes a difference. At this point, anything could be bad as far as I'm concerned, including my reasoning to get to the VR as a potential problem. Without extenders, I don't think there's enough room to spray anything inside the VR assembly. I forgot to mention that two VR modules have the same gain issue, which might point a finger at what's controlling them (the microcomputer). thanks for the ideas, Adam
|
On Mar 23, 2025, at 13:15 , Adam R. Maxwell via groups.io <amaxwell@...> wrote:
At this point, anything could be bad as far as I'm concerned, including my reasoning to get to the VR as a potential problem. I ran the 492 long enough for it it wake up again, and tried swapping a cold VR module in…and it kept working with a full scale peak, which ruled out the VR (and calls into question my notes on power levels). After powering off for a 1/2 hr or so, it was back to showing the 100 MHz reference as -50 dBm instead of -20 dBm. I'm now fairly sure the 829 MHz 2nd converter is the culprit: More careful power measurements (paying attention to ref level setting), the 110 MHz IF out from the 829 MHz second converter is -55 dBm when working, and -70 to -65 dBm when not working. Both seem low; I have -40 dBm out from the 2072 MHz 2nd converter IF output, in the working and non-working cases, and my reading of the schematic (drawing <14>) is that it should not be losing that going to P232. I think this points back to the IF select section inside the 829 MHz 2nd converter, which was my original hypothesis at the start of this dumb thread. For grins, I heated the converter up gently with a heat gun and actually saw the peak jump 30 dBm. Cooled the converter down with cold packs and lost that 30 dBm again, then heated back up with a heat gun and gained it back, so the experiment is repeatable. What I don't know is if this points to a bad solder joint? I thought semiconductors generally behave badly when heated. thanks, Adam
|
I'd say it's most likely a bad solder joint or broken part - something in series with the signal path such that it goes open, but still has some capacitance to couple an attenuated version of the signal. A short of signal to ground can act that way too, say from a shield braid strand, solder glob, or loose piece of hardware or wire clipping, but I think it's less likely than an opening scenario. It doesn't necessarily have to be right in the signal path - a bad connection in a circuit upsetting operation can do it too. I think you're getting very close, at least location-wise. Your descriptions have jogged my memory a little. I believe the "IF select section" you mentioned routs everything ultimately to the single 110 MHz IF out. I assume the "2072 MHz 2nd converter IF output" you mentioned is its version of the 110 MHz IF. The 829 and 719 MHz stuff does not participate in the low band operation, but apparently the 110 MHz IF stuff is built into the same module. Right?
I don't recall the RF deck construction, but I remember everything is jammed in tightly. I hope you can get enough access to figure it out without too much grief. Good luck.
Ed
|
On Mar 27, 2025, at 17:00 , ed breya via groups.io <eb@...> wrote:
I'd say it's most likely a bad solder joint or broken part - something in series with the signal path such that it goes open, but still has some capacitance to couple an attenuated version of the signal. A short of signal to ground can act that way too, say from a shield braid strand, solder glob, or loose piece of hardware or wire clipping, but I think it's less likely than an opening scenario. I think this is entirely correct. Much to my chagrin, I could reproduce this by heating a replacement 829 MHz 2nd converter, which also explains why I couldn't find a problem in the IF select board: it's something next to the converter, either the bias return or the transfer switch. <headdesk> Sadly, I am now in need of a bias return. The SMA connector on it was loose (the one connector I apparently never tightened), and unfortunately the barrel on the brass device was loose in its locknut. In attempting to get the barrel at the right depth and snug the locknut down, I broke the center contact holder inside the bias return, and it looks exotic and irreparable. I'll see if I can get some photos so others know what not to do. Anyone have a 119-1164-00 Bias Return they want to part with? thanks, Adam
|
On Mar 27, 2025, at 20:41 , Adam R. Maxwell via groups.io <amaxwell@...> wrote:
In attempting to get the barrel at the right depth and snug the locknut down, I broke the center contact holder inside the bias return, and it looks exotic and irreparable. I'll see if I can get some photos so others know what not to do. Actually not sure if I broke this or if it was broken before, since it was loose, but there are signs of scarring inside. Anyway, it sure is broken now. The series of photos shows how the bias return is assembled, and I see no room for error in setting the depth of the barrel connector and tightening things down, since there's no spring in the system. /g/TekScopes/album?id=301504Construction is similar to the S-4 sampling head 3.5mm connector, but the collar seams are hidden by the outer plastic sleeve, so it wasn't immediately obvious how it should come apart or be reassembled, since the steel locknut is fully recessed. /g/TekScopes/photo/301504/3900054?p=Name%2C%2C%2C20%2C1%2C0%2C0Anyone have a 119-1164-00 Bias Return they want to part with? For now, I've bypassed Option 3 phase lock, and it's reassembled and working, so I guess this is progress of sorts. Adam
|
David C. Partridge
Raymond Domp Frank