UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Test Gear and Workshop Equipment

Notices

Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders.

Closed Thread
 
Thread Tools
Old 18th Dec 2014, 2:16 pm   #41
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: HP8568B calibrator noise / spurious signals

Thanks David. Not having much luck sourcing a DC block that goes down to 136kHz with little loss. Any ideas where and how much? Running without one gives me the willies.
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 18th Dec 2014, 3:31 pm   #42
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
Default Re: HP8568B calibrator noise / spurious signals

If you don't need to go very high in frequency, make one. You can also use the BNC input on the 8568 with it, so a little box with male and female BNCs fitted at opposite ends will do.

Xc=1/(2*Pi*f*C)

Pick Xc=50 Ohms to choose the 3dB down point. Say 3dB down at 100kHz is what you want..... so 50=1/(2*Pi*100000*C) so C would be 32nF

You can scale thhis directly. 10kHz would mean 320nF and so on.

Just pick the nearest E12 value. Stray L will spoil the RF response, but may be OK, and you could stitch to the highband DC block.

David

(calculation done on a vintage HP calculator, by the way)
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 20th Dec 2014, 3:40 pm   #43
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

I've built similar things in the past and I include a fuse and a set of limiter diodes to protect the analyser in some of them.

Here's an old image of my range of homebrew limiters and/or DC blocks that I use to protect my analysers.

The lowest frequency one in the big box with the BNC connectors works down to AF. You can see the circuit drawn on the outside of the box. The fuse is a little 20mm 160mA jobbie from Maplin. The limiter diodes are back to back 1N4148 type. The cap is a 22uF electrolytic so I do have to be careful with the polarity of any external DC.

The idea is that the fuse will blow before the limiter diodes if the analyser gets hit by several Watts or more of power.

The other big one in the image (but with N type connectors) uses a similar fuse/block/limiter circuit and it uses a 100nF SMD cap on a little internal PCB and this works down to about 100kHz with a -3dB rolloff at about 16kHz. It has low loss and very low VSWR up to about 300MHz.

This does mean that limiting takes place at just over 0dBm so I try and keep the normal levels into the analyser below 0dBm in order to prevent clipping or distortion in the limiter diodes. But the circuit does give very good peace of mind. You could try building a copy of the low frequency limiter but use a non polarised cap at maybe 100nF or 220nF instead of the big clumsy electrolytic. The bare project box with the BNC connectors is a bit expensive to buy new but it does mean you can build it quickly
Attached Thumbnails
Click image for larger version

Name:	Limiters4.jpg
Views:	144
Size:	83.5 KB
ID:	101824   Click image for larger version

Name:	AF_Limiter.jpg
Views:	128
Size:	110.4 KB
ID:	101825  
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 20th Dec 2014 at 3:48 pm.
G0HZU_JMR is offline  
Old 2nd Jun 2015, 8:13 pm   #44
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: HP8568B calibrator noise / spurious signals

Quote:
Originally Posted by G0HZU_JMR View Post
I'm not sure how hard/soft you are tapping the RF unit but it's unrealistic to not expect some microphony effects on the first LO when you tap the RF unit. The loop bandwidth is quite wide but I don't think this will make it immune to microphony.

However, in your video I can see a few spurious terms that appear even when you are not tapping the unit. Presumably these are due to the original spurious issue.

So most of what you see in your tapping test might be 'normal' LO1 microphony and not the same thing that is causing the original spurious issue.
I am back on this, as I need to use it as a poor man's VNA with a directional coupler and the HP8444A tracking gen. I am feeling bolder and pulled the covers on the RF section. There are 4 tantalum axial caps in each of the output rails form the PSU, and I have heard talk of how they give trouble, so....

I pulled the 4 tantalum caps heard spoken of, tested them, and they were all OK on a Peak ESR+ ESR70 ESR and capacitance meter. I renewed them though, as a matter of course. I also pulled the very big electrolytics off the motherboard. All tested fine except for a 50,000uF 15V DC one, which showed "in circuit or leaky". I thought I had it, but no, it seems this meter will not handle electrolytics of greater than 22,000uF. It looked fine, no swelling or leakage, so I put it back.

Testing the supply voltages with a meter showed them to be all correct, and a scope showed no real ripple or noise.

Much tapping and removal of boards for finger cleaning got me nowhere, save the feeling that die cast oscillator region at the side of the RF unit sometimes made the 20 MHz test signal distort and waver, and sometimes it did not.

It appeared totally random as to whether this occurred.

I have possibly discovered something of note though. There's a board designated A4A2, Comb Bias. It has an LED on it, DS1, marked 2nd LO HIGH. Now, if I have any sort of understanding this LED should be either on or off. I believe its state is changed by SHIFT UP arrow and SHIFT DOWN arrow keys. It does not go on and off but SHIFT UP arrow
sets it bright, whilst SHIFT LOW arrow sets it dimmer. This does not seem a sensible way for such an LED to behave.

For ease I have done screen shots of the two repair manual pages that seem to discuss this little board, maybe someone far better informed can tell me in simple terms if this LED is behaving incorrectly, and if so, what it implies please? I have also put screen shots of the block diagram for this region, and the schematic of this converter bias board.

http://www.gatesgarth.com/converter-bias1.jpg

http://www.gatesgarth.com/converter-bias2.jpg

http://www.gatesgarth.com/converter-bias3.jpg

http://www.gatesgarth.com/converter-bias4.jpg

Thanks
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 2nd Jun 2015, 10:03 pm   #45
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

Hi Chris

First thing... I don't have a service manual for the HP8568B here but it looks like you do!

Can you send me a copy please?

I did study the synthesiser system used in the HP8568B about 25 years ago at my place of work and the 20MHz comb is part of a clever 3 loop system that produces an overall conversion system with very low close to carrier phase noise at the lower end of the tuning range. This gave a much lower synthesiser noise floor (close to carrier) than its competitors and the HP8568B was king here for many years.

However, I suspect that the 2nd LO High LED is nothing to do with this and the LED reflects the status of the LO2 frequency. I think this analyser can sometimes sidestep the LO2 frequency by a few MHz (to help with spurious performance?) and maybe they use this LED as a flag to show an engineer if LO2 has been deliberately shifted or not. So you may see cases where it winks on and off as it (dynamically) tries to dodge any spurious terms at certain tuned frequencies. Presumably the SHIFT key function provides a user override feature that lets you shift LO2 manually although I've never played with this myself.

But that's just a guess.
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 2nd Jun 2015 at 10:14 pm.
G0HZU_JMR is offline  
Old 2nd Jun 2015, 10:19 pm   #46
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

In normal operation the LED may wink on and off a lot across a wide sweep but I've never looked at it in mine to confirm this. I'm guessing that the LED is meant to be used as a status confirmation when the engineer tries to do the manual override with the SHIFT key function. Presumably it lights up when LO2 is shifted to its UP state.

I'm not too keen to take mine apart to have a look but if a service manual appeared suddenly then I might be persuaded otherwise...
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 2nd Jun 2015, 10:34 pm   #47
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: HP8568B calibrator noise / spurious signals

Check your PM's Jeremy!

The LED is just either full bright or quite dim, it does not appear to flicker at all, unlike another one in there.
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 3rd Jun 2015, 5:37 pm   #48
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

Thanks!

I've set my 8568B up on a work bench and I've removed the bottom cover from the RF unit. Lots of thick wooly dust in places! I don't think this 8568B has ever needed a repair as I've known its history from nearly new (approx 1990) and apart from the dust it looks to be in great condition with no evidence of the internal screws ever seeing a screwdriver.

I have just booted it up and from a fresh start the LED glows green for a second or so and then it goes out and stays out.

If I press SHIFT, UP it glows a steady green and if I press SHIFT, DOWN it turns off again.

If I hit PRESET (to start afresh) and if I then set it to 100kHz span and tune to 1144MHz the LED is out.
if I keep it to 100kHz span and tune to 1142MHz the LED is on.


i.e. if I set it to 100kHz span and tune to 724MHz the LED is out.
if I keep it to 100kHz span and tune to 722MHz the LED is on.

Does this help? Note: I think you need to be running the same firmware version as me to get similar results as this behaviour will be set in the firmware.

Mine is dated 7.4.87 on the CRT on bootup.
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 3rd Jun 2015 at 5:50 pm.
G0HZU_JMR is offline  
Old 3rd Jun 2015, 5:59 pm   #49
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: HP8568B calibrator noise / spurious signals

Hmm, well unless I have hit upon some msetting that gives a third state for the LED mine is different. Either bright or dimmer, never seemingly off. Can you think of a definitive, none invasive test to see if whatever should switch the state of the circuitry the LED is connected to is actually functioning correctly? On the HP / Agilent forums someone has suggested some voltage measurements, but I have nailed the damned thing back together now, as seeing it naked gave me the willies in case I dropped something in it

Thanks Jeremy! I'll check out my firmware, not sure without looking just what version it is, but I find it hard to believe that LED should be dim, and not OFF...
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 3rd Jun 2015, 9:10 pm   #50
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

Are you absolutely certain that the LED is staying on dimly? The LED type they have used is very reflective and I can look at mine and think it is still lit dimly if there is a lot of light hitting it at the right angle. Even if I turn off all power to the RF unit.

If it really is staying on weakly then maybe Q2 has gone leaky or the diode CR1 in its control line is faulty or the logic gate driving all this isn't pulling as high as it should when trying to turn off Q2 (unlikely). I think CR1 is there to provide a degree of immunity to this issue but maybe something has failed or aged.

As these parts are cheap it would be worth buying a couple of each.

However, if Q2 or CR1 has gone leaky then the LO2 frequency could become erratic.

Does your original fault change behaviour when the LED is on properly?

Note: I'm a bit confused by the orientation of diode CR2 on the schematic. I think this is drawn backwards? Also, I doubt that it will be a regular Schottky diode like CR1. But this is not related to your issue anyway...
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 3rd Jun 2015, 10:40 pm   #51
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

I measured a few voltages on mine with the LED on and also with it off.

When the LED is on the circuit passes about 10mA (via Q2 collector) into a diode Q3 in the 2nd LO cavity (not shown) and this will presumably shift it up in frequency by a few MHz..

When the LED is off the diode in the cavity gets reverse biased with the -14.9V at Q2 collector and the frequency of LO2 will fall back to normal

Note that I used a 10Meg Ohm DMM to take these readings and I think the voltage at the LED cathode in the OFF mode will be affected by the meter impedance.

See attached images.
Attached Thumbnails
Click image for larger version

Name:	LED_ON.jpg
Views:	113
Size:	82.1 KB
ID:	109091   Click image for larger version

Name:	LED_OFF.jpg
Views:	113
Size:	80.1 KB
ID:	109092  
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 3rd Jun 2015 at 10:49 pm.
G0HZU_JMR is offline  
Old 3rd Jun 2015, 11:23 pm   #52
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: HP8568B calibrator noise / spurious signals

I am 99% sure it's either full brilliance, or perhaps 50% brilliance, but I will check again. I was using the 20 MHz calibration signal input and with a 20MHz centre frequency, 1MHz span to do the test, then using SHIFT UP and SHIFT DOWN arrows to see the different states. I'll test again, but the background light was not very bright in the shack, I am sure as I can be that there were 2 states, normal LED full brightness,and a lesser brightness, not an OFF state for the LED.

I also see a change between less than 1MHz span and an over 1MHz span. The peak of the signal, in over 1MHz span, jitters frequency wise, half a graduation, left to right, at worst, when viewed with an over 1MHz span.

Less than a 1MHz span the frequency is 100% stable, no jitter. The original noise can appear in both a less than 1MHz span, or over a 1MHz span, however. RBW, VBW, input socket changes, and attenuation seem to make no odds to the noise, save VBW and attenuation taken to extremes can hide the noise, of course.

I have now spotted a spur when I change SHIFT UP to SHIFT DOWN with the 20MHz cal signal inputted (sp). Please see:

http://www.gatesgarth.com/shift-up.jpg and http://www.gatesgarth.com/shift-down.jpg


Not sure if that's helpful though.... I am wondering if a test can be devised to see if the noise is in the RF section, or in the display / IF box? Is such a test feasible? I have only recently become aware that IF stuff is in the display box.

Thanks!
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 4th Jun 2015, 8:12 am   #53
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: HP8568B calibrator noise / spurious signals

Seems the spur is a known artefact, one which nearly curtailed the release of this SA, and the reason was mentioned on the HP forum. Seems that I am beta testing 30 years late. A red herring, sorry, this spur a known trait, and not a fault as such.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

The spur is there "by design". There is a series of spurs which are related to the nature of the 8568 block diagram. Obviously, nobody wanted them but they were discovered after the project was started. Those spurs looked like a show stopper. They nearly killed the 8568 project. After a traumatic pause, a solution was found. It involved shifting the 2nd LO a few MHz back and forth, depending on the center frequency. That's what the SHIFT-UP and SHIFT-DOWN buttons do manually. Of course, in normal operation, this is transparent to the user. The processor does it as needed. So, there is some ancient history for 8568 owners.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 4th Jun 2015, 10:00 pm   #54
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

I think you need to establish the health of the LO2 shifter before anything else. if the LED is still weakly on (with SHIFT DOWN) then this implies that Q2 isn't switched off properly and there could be an erratic and weak (and unwanted!) bias fed to CR3 when it should be biased off. This could cause erratic modulation of the 2nd LO.

This oscillator is a cavity oscillator normally running at a nominal 1748.6MHz. When CR3 is biased on by Q2 the LO2 frequency will shift up about 5MHz.

The main PLL system will self correct for this shift in LO2 and it should also be able to clean up the close to carrier phase noise on LO2 when the whole system locks on narrow spans. This is because LO2 is inside the complex PLL system. But if there is a problem with Q2 leaking (so the LED is weakly on when it should be off) then you could get enough modulation and jitter on LO2 (via CR3) that the PLL can't fully cope with so the trace will show erratic jitter and noise. i.e. the PLL can't fully clean up the interference caused by the extra unwanted jitter/noise on the LO2 cavity oscillator.

So you need to check Q2 is OK. At the collector of Q2 it should have either 0.7V approx when the LED is on and it should have about -15V when the LED is off. It needs to be in the ballpark of -15V to heavily reverse bias CR3 off when SHIFT DOWN is pressed.

The metal can of Q2 is the collector so this is easy to probe with a DVM
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 4th Jun 2015 at 10:23 pm.
G0HZU_JMR is offline  
Old 6th Jun 2015, 1:35 pm   #55
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: HP8568B calibrator noise / spurious signals

OK, thanks Jeremy. I replaced Q2, which was leaky, and now the LED is either on or off. Sadly it's made no difference to the noise issue, which comes and goes seemingly randomly. One thing I do not believe I have mentioned is that noise is only ever there when a signal is input. Disconnecting the 20 MHz input from the test socket results in the noise floor showing but with no spikes. In case the 20MHz source was in itself noisy I tried a similar level signal from my sig gen, and the noise was exactly the same.

I also saw a test in the manual for an Average Noise Level Test. I ran it last night and attach or link to a screenshot. The test is detailed in the manual 8568B_PTA.pdf on page 44. The only noise spikes seem to be those at low frequencies on the LH end of the scan.

I need guidance as to some clever way to isolate a possible cause, I am getting annoyed at my inability to work out a way to divide and conquer this issue.

On the A22 Frequency Control board I measured some Tants and two showed as "in circuit or leaky". I pulled one leg of each and retested. One was 2.2uF and the other 1.2uF. Both showed correct capacitance, but both showed about 0.91 Ohms ESR. A similar value 2.2uF of similar voltage rating, on the same board, showed 0.0 Ohms ESR. My Peak Electronics ESR meter has a book with suggested values of ESR for various caps, and even 0.9 Ohms is WELL within accepted tolerances, according to that. I am a bit concerned an identical value and age one on the same board measures 0.0 Ohms though. Reason for concern and to change them?

http://www.gatesgarth.com/noise-test.jpg

Thanks all.
Attached Thumbnails
Click image for larger version

Name:	noise-test.jpg
Views:	119
Size:	85.5 KB
ID:	109198  
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 6th Jun 2015, 4:47 pm   #56
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

I can't really advise about the tant caps other than that some low cost ESR meters can give confusing results with certain cap values.

However, I'd expect a 2.2uF tant to have very low ESR. Much less than 0.9R but then it depends on the type of tant.

It's a shame that the change to the leaky Q2 didn't cure the issue.

You could try injecting a sig gen to the IF2 BPF stage at 301.4MHz and -20dBm. You would have to turn the sweep time right up to several seconds to let you watch the signal on the analyser as it will appear swept otherwise.

If you do this does the signal appear clean?

See the image below of my unit to show you where the relevant connection is. It's the yellow arrow. You will need a cable with an SMB connector to connect here.

When I had mine apart to look at the LO2 HI LED behaviour I noticed that quite a few screws were loose on mine. eg screws like the ones arrowed in red were sometimes loose by several turns.

It's worth checking yours although I now think my unit must have been apart for servicing/alignment at some point in the past because the screws were so loose. I can only guess that the tech forgot to tighten them up.

There is also an SMB connector for the LO2 signal. This will be approx 1748.6MHz (or +5MHz if LO2 HI LED lit) so your analyser won't be able to self monitor its LO2 signal but you could mix it down using a sig gen and mixer and feed it to the HP8568B.

The LO2 signal is a cavity oscillator and it runs freely inside the main PLL. So when you monitor it you will get to see the raw noise/jitter on this LO. It should look quite respectable.

In normal operation on narrow spans the noise contribution from this LO2 signal gets treated/removed as an error in the overall PLL system but you can monitor the uncorrected/raw LO2 signal at an SMB connector that lives on the RF converter that is very close to the PCB you swapped Q2 on.
You could also try mixing down the LO1 signal from the rear panel of the RF unit when on zero span to see how clean and jitter free the LO1 signal is. i.e. mix it down and look at it with the HP8568B.
Attached Thumbnails
Click image for larger version

Name:	hp8568B_RF.jpg
Views:	130
Size:	114.2 KB
ID:	109203  
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 6th Jun 2015, 5:27 pm   #57
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

Note that the ideal downmix method would be to use another analyser but if you have some patience and a sig gen that can go to maybe 2.2GHz you can also mix down to approx 20kHz and use a soundcard and some spectrum analyser SW to look at the LO jitter on a PC display. But you would need to know exactly where to set the LO from the sig gen because otherwise (obviously) the IF signal will fall outside the limited range of the soundcard.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 6th Jun 2015, 5:47 pm   #58
Chris Wilson
Heptode
 
Join Date: Nov 2011
Location: Whitchurch, Shropshire, UK.
Posts: 612
Default Re: HP8568B calibrator noise / spurious signals

Thanks Jeremy, I was wondering who I know that might lend me another SA for diagnostics, but came up blank. I guess a Signal Hound would be the ideal tool to borrow, as it's tiny and light, and should move around with no risk of internal damage, unlike the big beasts. I guess another SA is the ideal test instrument?

My sig gen goes no higher than 1024MHz (or thereabouts, going from memory right now), it's an elderly Marconi unit, and I do not have any commercial mixers. I am not sure if I have the skill to make a mixer that works at those sorts of frequencies. So I'm struggling bit here I do have a scope with 500MHz capability, and a counter that goes to 18GHz.

I believe I can do the IF2 BPF test though, so I will try that.

Can you please briefly explain why, when say in full scan, or in narrower scans, the noise never shows, it only shows when a signal is presented to the SA?

Thanks, I'll see what progress I can make.
__________________
Best Regards,
CW. 2E0ILY
Chris Wilson is offline  
Old 6th Jun 2015, 6:16 pm   #59
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
Default Re: HP8568B calibrator noise / spurious signals

Briefly is difficult.

There are two sorts of noise problem.

One is additive noise, which is just what it says. Your noise floor is increased, or jumps up and down. Because of the log display, big signals don't show much effect on their tops.

The other sort is modulation noise or multiplicative noise. This is where there is an iffy component, contact, joint somewhere and the gain of a system varies erratically - noisily in other words. It only really shows when there is a signal, and it shows on even the tops of big signals.

Mixers are multiplicative things, so jitter on the amplitude of any of your LOs will transfer modulation noise onto all signals. So your fault could be along the receiver strip, or from any LO.

The trick is to make a spectrum analyser diagnose itself. It can do a lot, though not everything.

1) Does the noise show up on the zero-hertz response - the blip you see at 0Hz even with no input applied? If so this eliminates the stuff before your first mixer.

Your sig gen may stop at 1024MHz, but there will be useful low level harmonics and you cn use this to make signals for the analyser IFs.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 6th Jun 2015, 9:58 pm   #60
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: HP8568B calibrator noise / spurious signals

Quote:
I believe I can do the IF2 BPF test though, so I will try that.
Thinking about it, I think the signal in will appear like it is in zero span mode if you inject 301.4MHz at IF2 so you could see if the signal gives a flat and steady line across the screen. If it is flat then there is no AM component to the interference. I'd expect that an intermittent connection would show some AM on the trace. You could try selecting linear display mode to get the best vertical AM sensitivity?

Also, play with the RBW setting as a wide RBW will mainly display AM changes but a narrower RBW will show both FM and AM effects.


But this test might not be as useful as I first hoped...

If you had another analyser you could look at the signal at the end of the cable feeding into this IF2 unit with a clean signal fed into the analyser front panel input. Or mix it down from 301.4MHz to 20kHz using your sig gen to look at it on a soundcard. But you would need to slow the sweep speed down on the analyser.
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 6th Jun 2015 at 10:17 pm.
G0HZU_JMR is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 8:40 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.