Author Topic: blown input on HP 8553B SA? [Solved]  (Read 6757 times)

0 Members and 1 Guest are viewing this topic.

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
blown input on HP 8553B SA? [Solved]
« on: June 30, 2017, 10:06:33 pm »
All,
I recently received an old HP 141T/8553B/8552B spectrum analyzer mainframe and booted it up for the first time today. Everything seems to be working alright, except when I connected the 30MHz cal signal up to the input, the peak was way lower than it should have been-- barely above the noise floor.

I have already checked the PSU, and all rails seem fine. The log reference level switch seems to be working correctly, and the input attenuator also seems to be working correctly (the signal moves up and down exactly 10dB, even within the limited dynamic range in which I can see the signal). The fuses on the back of the plugins all seem to be okay, and I don't think I noticed any intermittent switches on any of the controls.

It seems like the most likely scenario is that the input was overloaded at some point, so where would be the first place to check to see if the input is blown? the input mixer?

From there, what are my odds of being able to repair this unit? This is only a 100MHz SA unlike the massively delicate and likely much harder to repair 8555A 18GHz SA, so I would think it would be repairable, but I just don't know. I appreciate any advice or anyone who has experience with these units.
« Last Edit: July 18, 2017, 12:17:48 pm by JLNY »
 

Offline orbanp

  • Regular Contributor
  • *
  • Posts: 168
Re: blown input on HP 8553B SA?
« Reply #1 on: June 30, 2017, 11:07:07 pm »
The service manual describes the DBM mixer test for the fist converter subassembly, in test sheet 4.
The diodes can be replaced if they are faulty.
If the transformers are blown, that is a different story. then I would probably look for a complete DBM subassembly.

Good luck, Peter
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #2 on: July 01, 2017, 01:12:57 am »
Okay, So I opened up the first mixer sub-assembly, and so far nothing is jumping out at me. The two limiter diodes at the input both seem to test good, and I appear to be getting a voltage drop around .18-0.3V across all of the mixer diodes, although it is difficult to tell with them in circuit, as the transformer baluns are low enough resistance that they are bridging across the diode positions. I have not tested continuity across all of the baluns, but visually, I see nothing that would suggest that there is anything wrong with them. I suspect that the diodes would blow long before the transformers would.

 I'm not sure that I want to hazard unsoldering any of them unless I can be sure that is the issue. For now, I think that my plan may be to try injecting some signals after the mixer to see if that has any effect. I think I may be able to borrow some of the needed SMC adapters for connecting to those coax lines from work, and I have an 8640B sig gen which I might be able to try sending it some low-level signals. Unfortunately, I will have to wait until Wednesday at the earliest to try it due to the 4th of July Holiday.
 

Offline orbanp

  • Regular Contributor
  • *
  • Posts: 168
Re: blown input on HP 8553B SA?
« Reply #3 on: July 01, 2017, 10:50:52 am »
You could also check if that converter subassembly works as a DBM mixer should.
If you have proper equipment at work and they would tolerate that you check out the mixer there, take the DBM with you, feed the LO and RF ports, and see on an SA if you get proper output signals.
Those SMA to BNC adapters (I think they are SMA, but could I be wrong) are cheap:

http://www.ebay.com/itm/12-8-RF-Pigtail-Cable-SMA-Male-to-BNC-Male-Adapter-Connector-LW-/171865292481?hash=item2803f83ec1:g:628AAOSwkZhWTte-

http://www.ebay.com/itm/BNC-female-jack-to-SMA-male-plug-RF-connector-straight-gold-plating-Adapter-tbca-/201731426213?hash=item2ef82147a5:g:4gwAAOSw~otWcZ7g

Good luck, Peter
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #4 on: July 01, 2017, 02:29:50 pm »
Unfortunately, the connectors used in the 8553B are not SMAs. They are actually SMC (Sub-Miniature C), a slightly more obscure RF connector that is basically the bastard child of a push-on SMB and an SMA. It has an overlapping insulator and similar dimensions to an SMB but a threaded coupling nut like an SMA. If they were SMAs I would have the adapters for that on hand.

See the attached pictures. I also included a lineup of other connectors for size reference. Sadly, I don't have any SMBs on hand for comparison.

https://en.wikipedia.org/wiki/SMC_connector

I might be able to test it at work, but it'll depend on how busy I am and whether or not the equipment is in use. All I really need are the adapters and from there I can test it at home. SMC connectors aren't used too commonly these days, so I don't think that anyone has used them in a while. Maybe I'll see if I can order some to keep at home later on.

I actually have an 8555A SA that I can use to check the mixer. I am not unfamiliar with the 141T series, but this will be the first time I have worked on the RF sections and not just the power supply/display unit.
« Last Edit: July 01, 2017, 02:31:45 pm by JLNY »
 

Offline orbanp

  • Regular Contributor
  • *
  • Posts: 168
Re: blown input on HP 8553B SA?
« Reply #5 on: July 02, 2017, 10:47:16 am »
Unfortunately, the connectors used in the 8553B are not SMAs.
...
Hmm, interesting!
Similar vintage HP equipment that I had the pleasure of repairing (8601A, 8640B) are choke full of them!
I guess it shows that I have not yet opened up my 8553B/8552A!

Peter
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #6 on: July 06, 2017, 01:41:18 am »
(EDIT: it turns out that I did this test wrong. future readers should disregard this post)

Okay, progress!
I didn't have time to test the mixer at work today, but I was able to bring home a few SMC adapters to test things on my 8555A SA back at home, and I got some interesting results.

First, I checked the LO output from the 8553B, and everything looked good there. This makes sense seeing as the zero Hz peak was showing up as normal.

Next, I checked the mixer using an 8640B sig gen. I input a -30dBm 200MHz signal at the LO in and 250KHz at the RF in with a simple audio sig. gen, and the mixer seems to be unbalanced. (first image: 200MHz at -30dBm at LO in, no signal at RF in. second image 200MHz at -30dBm, 250KHz at an unknown arbitrary signal level)

Last, I set the 8640B sig gen to exactly 200MHz at -20dBm and injected it directly into the IF amplifier at the connection normally connected to the mixer IF out, and it caused the display to read a signal across the entire sweep as expected (see image 3).

I think this is pretty strong evidence. What do you guys think? I think the next thing to try may be to try replacing the diodes in the mixer.

From my reading, it looks like the diodes in the first mixer are hand-selected as a matched quad, so it probably wouldn't make sense to replace them one at a time. My plan for the moment is to do some reading to figure out what kind of diodes are used in the mixer and where/if I can find some.
« Last Edit: July 19, 2017, 12:08:06 pm by JLNY »
 

Offline orbanp

  • Regular Contributor
  • *
  • Posts: 168
Re: blown input on HP 8553B SA?
« Reply #7 on: July 06, 2017, 12:00:44 pm »
...
Next, I checked the mixer using an 8640B sig gen. I input a -30dBm 200MHz signal at the LO in and 250KHz at the RF in with a simple audio sig. gen, and the mixer seems to be unbalanced.
 ...

Good to hear about the progress you are making!

I think you are not driving the DBM with the proper LO level when you are testing it!

The service manual, Service Sheet 4 (pg. 151 in my pdf file, document number: TM 11-6625-2781-14 & P2), mentions 1Vrms level at the LO port, that is +13dBm according to my conversion chart, assuming 50 Ohm impedance.
That LO level sort of makes sense!

Also, do you have the +13dBm LO signal in the 8553, you never mentioned the level?

Good luck, Peter
« Last Edit: July 06, 2017, 12:34:37 pm by orbanp »
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #8 on: July 06, 2017, 12:35:51 pm »
Yes, that's true, I probably could have been driving the LO much harder, but even so, the LO-IF isolation seems to be virtually nil, so I think the issue is with the mixer. I can double check that, especially considering that the uncal light was on on my 8555 looking at it again.

I was getting fairly high signal level from the LO, I think I measured roughly +18dBm IIRC. I had to make sure the input attenuator was turned all the way up on my 8555A lest I blow out the mixer on that. Would there be any possibility that the LO level might actually be too high?
 

Offline orbanp

  • Regular Contributor
  • *
  • Posts: 168
Re: blown input on HP 8553B SA?
« Reply #9 on: July 06, 2017, 01:41:51 pm »
I would repeat the DBM test with the proper LO level.
Low LO drive level severely limits the performance of a DBM, the diodes in the ring just would not switch, would not turn on.

I do not think that +18dBm instead of +13dBm would damage the mixer. but this is just IMHO!

If push comes to shove, you could probably just replace the conversion assembly with a commercial +17dBm level DBM.
The only snag would be that the input RF frequency would not go down to the lowest frequency of the 8553B!
Wes Hayward in his home-built SA overcame this by swapping the RF and LO ports, the LO ports are directly coupled to the diode ring.
Read that paper, it is very interesting:
http://www.qsl.net/n9zia/wireless/pdf/9808035.pdf

Peter
« Last Edit: July 06, 2017, 01:44:20 pm by orbanp »
 

Offline WastelandTek

  • Frequent Contributor
  • **
  • Posts: 609
  • Country: 00
Re: blown input on HP 8553B SA?
« Reply #10 on: July 06, 2017, 02:00:10 pm »

Wes Hayward in his home-built SA overcame this by swapping the RF and LO ports, the LO ports are directly coupled to the diode ring.
Read that paper, it is very interesting:
http://www.qsl.net/n9zia/wireless/pdf/9808035.pdf

Peter

Isn't that where the zero spur comes from?   It has been quite a few years since I built up the W7ZOI system but I remember something about that.
I'm new here, but I tend to be pretty gregarious, so if I'm out of my lane please call me out.
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #11 on: July 06, 2017, 11:36:44 pm »
All,
I went and double-checked my measurements from yesterday.

I remeasured the LO output and actually paid attention to the power level this time. It seems to vary between about +13 and +16dBm along the sweep. I agree that this seems fine to me. This was just measured on the 8555A, so it's not super exact. I have a proper RF power meter that I could pull out if we want more accurate values.

I remeasured the mixer, this time using a +13dBm LO signal from the 8640B.

Image 1: IF output with +13dBM @ 200MHz LO, no RF input.

Image 2: +13dBM @ 200MHz LO, approx -19dBm (~35mVp) @250KHz RF input.

Image 3: +13dBM @ 200MHz LO, approx -2dBm (~250mVp)@250KHz RF input.

I'm not quite sure what to make of that. I'm not super familiar with double balanced mixers, but are we supposed to be seeing that much LO signal at the IF output?
 

Offline orbanp

  • Regular Contributor
  • *
  • Posts: 168
Re: blown input on HP 8553B SA?
« Reply #12 on: July 11, 2017, 01:57:23 pm »
I do not think that you have all the data listed (or I just looked it through too quickly).
A good way to see if the data makes sense is to compare it to commercial DBM mixers.
Looks like the HP mixer is roughly equivalent to a level 17 (+17dBm LO level) DBM mixer.

For example Mini-Circuits makes DBM mixers: https://www.minicircuits.com/WebStore/Mixers.html
e.g. the "ADE-1H+" is such a 15 - 17 level mixer: https://www.minicircuits.com/pdfs/ADE-1H+.pdf

DBM mixers have a conversion loss, for the ADE-1H it is about 6dB. It means that if driven with the required LO level, the converted IF output is lower from the RF input level by the the conversion loss. So if you have -30dBm RF input signal, the converted outputs (LF+RF, LF-RF) at the IF port would be about -36dBm.

The LO to IF isolation data gives the "leak-through" of the LO signal to the IF port. It is typically 30dB or better for DBMs, depending on the frequency. So for driving the mixer with +17dBm LO signal, you should see no more than -13dBm LO signal at the IF port (it is then rejected by the IF-filter).

There is also the LO to RF port isolation which is not always given. The LO to RF leak-through is similar in magnitude to the LO to IF leak-through.

There are lots of application notes on the MC site, look them through! The following is on mixers:
https://www.minicircuits.com/app/AN00-009.pdf

The above mentioned commercial mixers start at about 0.5MHz RF signal, the HP SA goes down to a lower frequency. Probably that is the reason HP makes (made) their own mixers. It is the transformers that set this lower frequency (probably the upper too), so the HP conversion module probably contains specially made transformers.

Peter
« Last Edit: July 11, 2017, 05:59:35 pm by orbanp »
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #13 on: July 11, 2017, 08:59:30 pm »
The data is just listing the power levels and frequencies. The Exact SA settings will have to be read from the images. The major factor for reading the displays is to note that the log reference level is 0dBm.

LO-IF isolation is sitting at about -29 to -30 dB, so I think that checks out. RF-IF isolation can't be measured because the 250KHz RF signal would be below the 10MHz cutoff of my 8555A.

conversion loss seems to be somewhere around -10 to -12dB, with pretty noticeable harmonic distortion for RF levels above maybe -6dBm or so. I do not know if this would be normal behavior for this mixer or not.

It should be noted that my estimates for the RF signal levels from my audio sig gen are not very exact since they are being measured separately using an oscilloscope in 50 ohm input impedance mode.
 

Offline orbanp

  • Regular Contributor
  • *
  • Posts: 168
Re: blown input on HP 8553B SA?
« Reply #14 on: July 12, 2017, 09:52:20 pm »
Just curious, have you verified the if the "Cal" signal from the 8552B is indeed -30dBm?

What if you feed the SA from your 8640B? Does the indicated level on the SA match the output of the generator?

You mentioned that the "Cal" signal shows barely above the noise floor.
Noise floor of the SA is below -100dBm according to the specs.
Conversion loss of the mixer, you measured that about 10dB, so this does not mesh with the -30dBm signal just above the noise floor!

Reading the manual there are a couple of interesting facts (I never used my SA seriously, just played with it, it is a recent acquisition).
Max input signal to the mixer is only +13dBm!
They also suggest to have at the mixer input -40dBm RF signal level for minimum distortion and best dynamic range.
They also mention that RF signal level above -10dBm at the mixer input will compress, so your signal distortion above -6dBm is not out of character!

Peter
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #15 on: July 13, 2017, 12:09:37 am »
The -30dBm cal signal appears to be working correctly. I have used the 8552 IF plugin with other units and it operates normally.

However, it seems I have made a discovery of sorts. I tried inputting a -30dBm signal from my 8640B, and similar to the cal signal it was down near the noise floor. I then started to increase the power level, and suddenly when switching from -10dBm to 0dBm, the peak jumped up by almost 40dB to just a few dB below 0dBm. Interestingly, changing the log reference level seems to behave normally and changes the position on the screen by 1 division, but raising the input attenuator up to -10dB made the signal drop down by 40dB again, so it appear to be somehow related to the IF signal level at the output of the mixer. Perhaps the biasing to some amplifier stage has gone open-circuit, and only turns on once the IF signal level is sufficiently high?

Actually, thinking back on it, when I had tried injecting a 200MHz signal at the IF amplifier, I think I had noticed a similar nonlinearity where the trace jumped by multiple divisions when changing the power level, but at the time I hadn't been paying attention to it and had attributed it to the narrow bandwidth of the IF amplifier.  :palm: I definitely want to double-check that, but the SMC adapters are back at work, so it'll have to wait until tomorrow.

Since I had been seeing the 0Hz spur as normal, I had foolishly assumed that the problem had to be before the IF amplifier. However, it would seem that the LO-IF leakage was sufficient to appear normal on the screen.  |O

Stand by for further testing, but it seems that the mixer is working normally. Next step will probably be to check the signals before and after the IF amplifier and see what the deal is with that.
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #16 on: July 15, 2017, 03:52:12 pm »
I tried injecting a 200MHz signal directly at the IF input, and I was able to observe a similar oddity in the amplitude response.

However, I checked the signals before and after the IF amplifier, and it seems to be working correctly with a reasonably linear response all the way from 0 to -50dBm. (see images. I connected my 8555A to the output of the IF amplifier and set the 8555A to a slow sweep zero scan mode, then slowly stepped down the output power of my 8640B to measure the amplitude response)

From what I can tell, all of A9 is working correctly. My next step will be to try and check the A10 second converter, but that may be a bit more difficult given that A10 is located on the underside of the plugin for some reason, which makes it a lot less accessible and a lot harder to do live testing on...

One thing I am noticing, however, is that the amplitude response also appears to have a time component, as changing the sweep time also appears to affect the amplitude of the signal and zero hz peaks on the display for the 8553B. for some reason a slower speed seems to result in a lower amplitude on the screen. I'm not sure what this means, and it doesn't have to do with the uncal light, as that does appear to be working correctly.
« Last Edit: July 15, 2017, 03:55:06 pm by JLNY »
 

Offline orbanp

  • Regular Contributor
  • *
  • Posts: 168
Re: blown input on HP 8553B SA?
« Reply #17 on: July 17, 2017, 02:13:59 pm »
Hi JLNY,

Looking at the manual, it looks that the second converter gets about the same level RF input signal as the first converter.
The first IF-amp has bout 14dB gain, the first IF-filter has about 3dB attenuation, and that with the 10dB conversion loss of the first converter gives about the same RF level at the second converter input.

Anything out of ordinary by visually inspecting the second converter assembly?
BTW, the DBM mixer uses the same diode quad as the first converter!

Interesting that time dependent behavior!
Do you see any such time dependent behavior at the first LO or second LO drive levels?

There are SMC connectors with pigtail cables available on eBay, those could be used possibly if access to the second converter is limited.

It also looks like that the second converter can be tested "out of the box" just by providing the -10V supply and driving and terminating it with the proper signals/loads.

Regards, Peter
« Last Edit: July 17, 2017, 02:16:22 pm by orbanp »
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #18 on: July 17, 2017, 05:25:09 pm »
Peter,
I have started doing some testing on the second converter assembly, and it seems to be the source of whatever odd behavior I am seeing with the sudden non-linear jumps in amplitude.

I did try to do some time-dependent testing on the first LO by itself, but it was tricky and I had to get a bit creative to do it. It's slightly difficult to explain:
I set both the 8553B and my 8555A mainframe to the same sweep time and both to "line" sweep triggering, then I set the 8555A to zero bandwidth mode centered on the 200MHz IF output. Because both units were triggering off the same mains cycle, the changes in amplitude were synchronised with the sweep and the display on my 8555A pretty closely matched what I would have expected to see on the 8553B if it were working. Let me know if that makes any sense.

As for the second converter, I am definitely getting 50MHz at the output, but for some reason I am not seeing a peak at 150MHz where I would expect to see the LO leakage? There seems to be a basic LPF at the output of the mixer, but not enough where I wouldn't expect to see any LO leakage. I wonder if the LO amplitude is low for some reason, or if the LO amplitude might be fluctuating? If I'm lucky maybe all it needs is a bit of re-tuning on the L3 adjust. I still need to check the LO level.

Visually, I can't see anything obviously wrong with the second converter. I checked the transistors in-circuit and they do pass a basic diode test. I also checked several resistors and the values measured good. The diodes in the mixer seem to measure good, with about 0.3 Vf across the board. I was already planning in the next day or so to do some out-of-box testing on the second converter as you suggest.

I ordered some SMC cable connectors a while back for future use (I like to assemble my own cables, so I ordered bare cable connectors), but they are shipping from China, so I expect that I won't have them in time to use them for testing.

Thanks for all your help, by the way!
« Last Edit: July 17, 2017, 05:29:40 pm by JLNY »
 

Offline JLNYTopic starter

  • Regular Contributor
  • *
  • Posts: 60
  • Country: us
Re: blown input on HP 8553B SA?
« Reply #19 on: July 18, 2017, 02:13:59 am »
Got it!

It turns out that adjusting L3 seems to have done the trick!

I tried hooking up the second converter outside the case, and the LO amplitude was very low, and seemed to fluctuate slightly when I prodded around the board. I had to adjust L3 quite a bit to peak the LO signal, but once I reinstalled in into the 8553B, the unit appears to be working normally now.

My only theory is that the LO must have been right on the edge of maintaining oscillation, so perhaps leakage from high RF drive levels was able to excite the oscillator enough to increase the LO level? This is the only explanation I can think of for the sudden increase in amplitude when the RF input was above a certain level. I'm not sure. In the end, the fact that the LO was still partially working seems to have made the problem much harder to debug.
« Last Edit: July 19, 2017, 12:07:17 pm by JLNY »
 
The following users thanked this post: alm

Offline WastelandTek

  • Frequent Contributor
  • **
  • Posts: 609
  • Country: 00
Re: blown input on HP 8553B SA?
« Reply #20 on: July 18, 2017, 03:26:54 am »
excellent, looks like you got it sorted   :-+
I'm new here, but I tend to be pretty gregarious, so if I'm out of my lane please call me out.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf