Re: Testing dummy loads with a VNA
Thanks, Robert, I'll take a careful look at that.
|
Re: Testing dummy loads with a VNA
Thanks, Stephen. I don't seem to be able to find the user guide for this instrument. I do have the full original service manual, though, which is invaluable in getting the device fully functional again. It was a stretch for me, though. A bit above my pay grade as they say. But at least I did it, it works and I've learned a lot in the process - not about the process of vector network analysis but rather about working on something I'm not familiar with and gaining confidence in so doing.
|
Re: Testing dummy loads with a VNA
Thanks, Dave. Very instructive and useful response!
|
Re: Testing dummy loads with a VNA
I've also just discovered that any members' PMs to me on this subject have been accidentally deleted after being flagged as spam by my mail provider. If any of these are important enough, can the sender please re-send them? Many thanks and apologies.
|
Re: HP 8644 Phase Noise Problem and Repair
Nice schematic, easy to follow.
I dont know this unit exactly, but some general suggestions for this very sensitive piece of gear: 1. This is an OLD (35-40 yrs.?) unit. The HP doc (Memory Project) indicates its an early SMT construction.I'd go around that board and check, at random or in suspect circuits, value of resistors. Modern SMT resistors tend to stay put. I dont know about those from back in that era. "28.7 ohms" is probably a laser-trimmed value and certainly very critical. With old SMT, ID also look thru a good magnifying glass at component end-terminations for de-lamination- caused by mechanical stress. again, an old unit. 2. Remove the 'suspect' board and clean it thoroughly with a TCE solvent.? That will 'draw moisture' especially UNDER IC packages, so then dry it for an hour in a 110-150 degree F oven. When dealing with levels in the - 100dB level, even trash conducting across a circuit board can be a problem.Dont use organic solvents, they can lead to micro-growth- there was a Navy document years ago about that happening in a space application because of circuit board cleaning chemicals. If memory serves, Home Depot or an auto part had a TCE based spray but I dont recall where it was. Try GC or NTE? 3. Approaching such a job as a REPAIR is a mistake. It MUST be approached as a RESTORATION and that generally means: a. TOTAL power supply rebuilds. New filter caps b., check connectors for oxidation and noise, especially any that carry relatively high currents, c. on any PS regulators that dissipate enough heat to need heat-sinking, remove or loosen them and replace heat sink compound, if used. Check heat sink greaseless washers for shrinkage a/o tighten heat sink screws- they loosen over the years. d. PS noise checks. This is a piece of equipment that deals with noise on the order of -150 dB. Like you said, the 'scope cant even read it. The FIRST thing that will corrupt such circuits is POWER SUPPLY NOISE. At such low levels, use a good spectrum analyzer on the power supplies, especially supplies to this circuit board. This design revolves around control circuits. Control circuits are feedback paths to issue correction to some circuit function. Power supplies are outside the control loop, therefore, the control loop may not be able to compensate for PS problems. 4. The document, p. 15, describes the Mode 3 as a " frequency discriminator detects FM noise on the internal oscillator and cancels it in a feedback loop."? Also note "drawback...frequency switching time is longer..." p. 15. That description sounds as if its a frequency problem, but it is not. Theres no such thing as corrupting an oscillator with a FREQUENCY. Interference is in the amplitude domain, see the SSB noise on Fig. 4 p. 16. F-M signals, noise come from amplitude-modulating a variable frequency oscillator.... Id not go any further before checking, with a good spectrum analyzer that can read as far down as the spec on this unit, for PS noise to (or from!)? that circuit(s). If thats the cause, then youd never find it by troubleshooting as if it just a 'broken circuit...'. Back to the 'restoration' thing.... There are two kinds of PS noise- noise FROM the PS, and noise coupled from one circuit(s) to another (others). Either and both can be caused by leaky decoupling capacitors or resistive connector terminals OR... the sneaky one- excess current draw that destabilizes voltage or current regulators. For example, modern 3 terminal regulators require filter caps on the input and output to STABILIZE the control circuits in side..That implies a certain load current.
The schematics wont help in determining if excess current is being drawn, its a tough thing to deal with, but it may show up in excess heat somewhere.
The observation of 'longer switching time' might help. Thats probably a longer 'settling time' for the system to come to stability - they may be blanking the output till the system reaches lock. Is there a spec on that switching time? That might be a clue.
Realize this unit was designed and built 'ahead of the curve' using new SMT technology a LONG time ago... there have been YEARS for circuit boards to collect microscopic amounts of contamination, for components to age and drift and filter caps to leak.
Contrary to the opinions of our well meaning Hams here who routinely play amateur games with old equipment without a second thought to things like power supply purity, you're dealing with a very sensitive piece of equipment that can not tolerate any of these problems.
I got burned by this very problem by ignoring PS purity in a radio transmitter- there was - yes - phase noise on the carrier signal. Or so I thought...the scope trace was thickened when it should not have been. The power supply had been totally replaced (all new except power transformer). 100%. Or so I thought...
The problem had to be solved with a spectrum analyzer - the carrier had two spurious sidebands that were so close to the carrier the scope couldnt discriminate them. The sidebands turned out to be from an unbalanced power transformer (the only part I didnt replace) - 120 Hz modulation leaked from the PS into the carrier/demodulator circuit and mixed in the amplifiers and created the spurious sidebands. And those sidebands were just strong enough to make it to the TX output. I had torn that circuit and all the circuits around it to shreds looking for a problem and never found any. It was power supply!
|
Re: Testing dummy loads with a VNA
Thanks to all for the responses since I last posted. I'm obviously in a different part of the world to most of you so apologise if it looked like I was ignoring any of you - I'm simply on different hours!
The storage normalizer I have is the 8750A. AFAIK that's the matching one for this VNA. Likewise, the S-parameter set I have in mind to buy is the 8748A.
Clearly there's a lot more to this high-end RF malarky than I imagined. I'm going to have to do some serious revision. I did study much of this stuff in detail about 20 years ago but have forgotten a lot of it so I'm going to have to dig out my old notebooks and knuckle down!
|
Re: Testing dummy loads with a VNA
This application note, while for the 8505A, Covers RF test sets contempary to and usable with the 8754A
It aslo covers the basic measurement and a bit on the Storage normaliser.
Robert G8RPI.
|
Re: Testing dummy loads with a VNA
Suggesting other VNA's to the OP is not helpful when they hasve already stated that they don't want to go there.
Jinxie,
I don't know what RF test set you were offered for $450 but for 90% of work a 8504A will do what you need. One limitation i that you will have to reverse the item under test (if its two port) for some measurements. Don't buy a 8502B? as it is 75 ohm (unless you are doing 75R stuff like CATV).
Dr Dave Kirby has some links to good reference documents from basics to advanced.
The Storage normaliser (which one do you have, there are two?) allows you to remove test set up losses "calibration" etc and compare different traces amongst other things. Very useful when you get a bit more advanced.
Robert G8RPI.
|
Re: Testing dummy loads with a VNA
I very much agree:? The 8712x, 8714x, and 8753x are all much more capable and convenient than the 8754A.? The ability to use any (reasonable) length cables makes a wide range of measurements much simpler to accomplish.? The effects of both cabling and adapters can be largely calibrated away.? Be aware that the 8712 and 8714 have built-in test sets, while only some of the 8753x models do.
--John Gord
toggle quoted message
Show quoted text
On Sun, Mar 8, 2020 at 06:19 PM, nj902 wrote:
A storage - normalizer would also be a good addition to your 8754A.
However, between the cost of the test set and a normalizer, you might be approaching the cost of moving up a notch or two - say to an early 8753 or maybe one of the 8711/12/13/14 series.
You might be able to sell the 8754A to get a little seed cash to get the process started.? A newer instrument will have a lot of attractive features - greater dynamic range, calibration capability, digital trace storage, hard copy output capability, connection to John Miles software, etc.
|
Re: Testing dummy loads with a VNA
Hmmm, I see what you mean. So here is my two cents worth.
With a scalar analyzer (8756 or similar): you will need a dual directional or two single directional coupler for the band in question(ideally with know coupling data). ?I like 4 known attenuators 1, 3, 6, 10 dB. ?Hooking the ¡°R¡± port to the forward coupled port and the A to the re-elected coupled port will give you the return loss. ?The 8756, if calibrated will give you fairly good data. ?If not then put each of the attenuators, UNTERMINATED, on the output of the coupler. ?Note the reading. ?The trace will show 2X the value of the attenuators as return loss then you can interpolate as required.
With a VNA/ S parameter test set: ?you need to normalize the scope at the end of your test cable (take readings with known load, open, and short). ?Then just take the S11 reading.
As has been mentioned it¡¯s not hard but can be daunting the first time. ?Be careful of warm-up times, my 8753E is very odd until it warms up for half an hour and is then normalized.
The manual will have more complete information. ?This in my my idea is what you do when someone asks for help Regards, ? Stephen Hanselman Datagate Systems, LLC
toggle quoted message
Show quoted text
On Mar 9, 2020, at 05:40, Jim Ford <james.ford@...> wrote:
? I thought the OP said he had an 8754.? Correct me if I'm wrong.?
Jim Ford?
Sent from my Verizon, Samsung Galaxy smartphone
-------- Original message -------- From: "Stuart Landau via Groups.Io" <stuartl73@...> Date: 3/7/20 11:06 PM (GMT-08:00) Subject: Re: [HP-Agilent-Keysight-equipment] Testing dummy loads with a VNA
I think that an honest question regarding measuring SWR or return loss has gone off on a strange tangent. The answer to the question is relatively easy to answer if we know what frequency range is involved, and perhaps the type of connectors will be used and what the person currently has on hand. What we don't know is what sort of equipment is available to the person who asked the question. The S11 measurement is a very common one and is not "black magic" to perform. Usually the results are expressed as return loss (in decibels) or as voltage standing wave ratios. Since there are several ways to make the measurement, what is needed to describe the procedure to the person asking for help, is to know what he currently has available to make the measurement. I assume that he doesn't want to spend a lot of money buying additional equipment.? He asked for assistance, and that is what I offered. Possibly free advice is worth what you pay for it. Stuart K6YAZ Los Angeles, USA
|
Re: 693B sweep oscillator restoration
I worked in Electronics for over 50 years. I had to show more than one EE where they screwed up a design. The concept of using a boost or buck transformer to adjust the line voltage has been around longer than I have. It doesn't have to fit inside the instrument. I use one for the entire workbench. Some of my designs ended up as part of the ISS. Do whatever you want. I really don't care. I recently was called to a friends new location for their factory. A large machine tool wasn't working right, and didn't have enough torque to operate. The industrial electrician who wired it installed the proper 240V outlet, but wired it to 120V The building as three phase. He didn't install a boost transformer to take the 208 to 240 He was too damned lazy to go back to their office to get the proper breaker, or transformer.VAC. The business owner had to threaten not to pay the $3000 bill for wiring up their machines, until he replaced the single pole breaker wth the right dual pole breaker. I picked up the transformer at Grainger's, and brought the line voltage up to 245, during off peak hours. It ran like a champ, after that. My last new product rollout was an $80,000 Dual receiver DSP based Telemetry package for the Aerospace industry. Engineering had completed several prototypes, but there were no test fixtures, or usable test procedures. There were many problems with the design, when it was dumped into my lap. My first challenge was forcing the MEs to completely overhaul our Reflow process. It was developed for the 1206 SMD components ten years earlier. It couldn't handle the fine pitch ICs, or the 0402 resistors. There were nine custom VXI cards we built in house. Each board required test software to be loaded, but oru Programmers didn't see the need to streamline the process. It took eight or more hours with the bits and pieces of code, and four types of interface. Once I got them off their asses, it was a five to 15 minute job. Once the units were full assembled, you plugged a 3.5" floppy disk int the embedded controller and connected it to the Engineering server. Select the firmware version, and let it download the OS and firmware. I have also built TV stations, and serviced obscure industrial electronics with no manuals to get factories back into production.
M ichael
youre also lecturing me with false statements, off topic and making false statements. On Forums, we should discuss with OP on the topic, not go into lecturing side-bars.
But since youve launched false criticisms I will respond in kind.
Youre advice is false. You know nothing about the scope I worked on, just like Pete who knows nothing about it, and youre lecturing an experienced Design engineer.
The resistor in THIS application does NOT generate heat. It didnt even get warm. I tested both the dropping resistor and heat sinks, no heat in the dropping resistor to speak of.
In any case, why would it make any difference if hte resistor dissipated heat? It doesnt...
Your comment implies that a transformer is regulating whereas the resistor is not, and thats a false statement.
There is no space in the scope in which to put a transformer.
And youre lecturing an EE who can design the scope and sorry, but looking foolish in teh process.
Too many Trolls on this forum that have no idea what they are talking about, and not following Forum Etiquette in discussing what OP has estabilshed as the topic.
Moderator, where are you?
|
Re: HP4328A-4338A-B HP16143A or B Mating cable socket pins
On 3/8/20 10:20 PM, abalogh@... wrote: Looking to find the manufacturer name for the socket pins and
receptacles used the? HP16143A-B mating cables and test leads, also the
the part # on the Hitachi 5 conductors cable?
Have looking at all the part list, there is no mention at all. I don't know anything about the 16143A-B, but the 4328A's front panel connector is a Hirose RM12BPE-5PH(71). These are readily available from DigiKey, for example. -Dave -- Dave McGuire, AK4HZ New Kensington, PA
|
HP4328A-4338A-B HP16143A or B Mating cable socket pins
Looking to find the manufacturer name for the socket pins and receptacles used the? HP16143A-B mating cables and test leads, also the the part # on the Hitachi 5 conductors cable?
Have looking at all the part list, there is no mention at all.
Any help would very much be appreciated , thank you??
|
Re: HP1700 Series Rear Cord Wrap Feet
You might consider a trip to the hardware store. My Tek 453 had the same problem with his feet. I found some metal drawer pull knobs that looked like they might work. I drilled them out for 10/32 size screws, which is what the original feet used. They may not look like Tek parts, but they don't look outta place.
Burt, K6OQK
--
Sent from my Android thingamabob with K-9 Mail. Please pardon the spelling errors as the dog can't spell so good.
--
Burt I. Weiner Associates
Broadcast Technical Services
Glendale, CA 91201 U.S.A.
K6OQK
|
Re: Testing dummy loads with a VNA
On 3/8/20 9:31 PM, Jinxie wrote: I do actually have the storage normalizer for this VNA. But I have no
idea what it's for or what it does. Ahh, you do have one! Excellent. Basically a storage-normalizer provides digital storage, and simple (but specific) math capability for the VNA, providing a great improvement in general usability. First, digital storage. You'll note that you can usually get better RF performance from the VNA if you slow the trace down, but then you get a slowly moving dot rather than a quickly-refreshing trace, and it's much harder to visualize the curve. Digital storage preserves the waveform on the screen and allows it to be slowly replaced with new data as the next trace progresses. Normalization is a much bigger deal. It allows you to subtract one response curve from another. This is useful in the real world of imperfect cabling and connectors. Let's say you have two cables connected to ports 1 and 2 of your VNA, and you've connected an attenuator between them to measure its response over a frequency range. What you get isn't only the response of the attenuator, but also of the cables and connectors. Their contribution to the displayed data can be very significant, depending on several factors. The normalizer allows you to, say, connect those cables together in the middle first, without the attenuator, measure the response of that cable and the connectors, and store it in memory. Next, you insert the attenuator, and the normalizer subtracts the previously-stored response of the cable from the displayed measurement, giving you much closer to correct results. More modern VNAs (say from the early 1980s on) do this in a very different way, and that's a capability that your 8754A lacks, so that's outside the scope of this discussion. But you can see what the storage/normalizer unit does and how it would be useful. Do you have everything required to connect your storage/normalizer to your 8754A? What model is it? I don't want to go down the route of a newer instrument. With gear of
this vintage I can keep them alive and fix them up when they go wrong. I'm right there with you on that. But in the case of VNAs, there's very real functionality that has been added, real game-changing stuff, that you may end up wanting. Just keep it in mind for the future. It's not really like the oscilloscope world where display voltage against time has been a solved problem for decades, and the main "improvements" are production cost and profit margin. (except for high-double-digit-GHz bandwidth that almost nobody actually needs) -Dave -- Dave McGuire, AK4HZ New Kensington, PA
|
HP 8644 Phase Noise Problem and Repair
Greetings,
I have an 8644B that does not perform correctly in the low phase noise mode (mode 3).? I have measured the phase noise using an 11729C and the 8663A 640 MHz low noise reference output.? The Comm Discr module was intermittently failing the MLD test, so I decided to replace it.? I purchased two additional modules and neither of the two replacements worked.
It became apparent I was going to have to dig-in and troubleshoot this module.? I ended up chasing my tale without a good schematic, so I bit the bullet and created one.? The schematic has been uploaded to the files area.?
/g/HP-Agilent-Keysight-equipment/files/8644%208645%208664%208665%20Comm%20Discr%20Module%20Schema%20and%20Service
With the help of the schematic, I was able to fully repair one module to original specs.? The base-emitter junction of Q9 was open on this module.? I was also able to find multiple faults on the other two modules.? Both of these modules now pass all ILD, MLD, and calibration routines, however they do not meet the phase noise spec for mode 3.? I suspect there are noisy components in the phase detector, U4 has been replaced with no effect.? The circuit is extremely sensitive to noise and uses an ultra low noise LT1028 op amp.? The noise levels are well below what can be seen on a scope trace.
I have found several references to assist with troubleshooting.? The first is in the HP Journal issue on the 8644/45/64/65? PSGs? HP Journal, October 1989 linked below:.? Page 30 has a general introduction to the delay line discriminator.? The second reference for understanding theory of operation is one of the HP product notes for the 11729C carrier phase noise test set (PN 11729C-2).? This note introduces the math involved with determining the sensitivity and S/N ratio of the delay line discriminator.
I will continue to report on my progress with the remaining modules, but if anyone is interested they can study the schematic and offer suggestions.? I hope this information and schematic will be useful to others.? If you discover any errors, report them on this thread and I will correct them.
Best,
Ted S
|
Re: Testing dummy loads with a VNA
I've been piecing together an 8754a VNA set. I recently picked up the
8748a S param set for around $100. It only works with the 8754a and
that's about as obsolete a VNA as you can get but still be usable.
When you point out to sellers how useless it is, you can sometimes
get them down to a reasonable price point.
It also needs the reference plane set which generally does not come
with it.
I found a storage normalizer for $35 and the VNA itself was around $150.
So I'm scraping the bottom of the barrel, but it's all I can justify on
RF work since it's not in my main line interests (but projects are always
an excuse to buy new equipment). It gets worse when you start buying all
the cables, adapters and connectors you need to do anything.
Paul
toggle quoted message
Show quoted text
On Sun, Mar 08, 2020 at 08:35:11PM -0400, Dave McGuire wrote: On 3/8/20 8:10 PM, Jinxie wrote:
Yeah, but I don't like the Nanovna type stuff; it's not a/proper/ VNA in
my view. How about I buy the matching S-parameter test set for my VNA?
I've been offered one for USD450 Would that solve all my
problems? And is there anything else I might need in the future? $450 seems a bit steep, but admittedly I've not been watching that
market. It would certainly give you good functionality, including the
specific functionality that you need for the task at hand. I can't help
but think you should be able to get away a good bit cheaper than that,
but I could be wrong.
The only other thing you might want is a calibration kit, but as the
8754A has no automated correction capability, it would be of limited use
there. Our own Dave Kirkby makes and sells decent calibration kits at
good prices.
I agree about the NanoVNA. I did pick one up recently, though...it
was on sale, and I want to put it next to my HP 8510C for some
side-by-side comparisons. For what it is (a hobbyist toy) it is
surprisingly capable and a very nice piece of work. But a piece of lab
equipment "it ain't".
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
!DSPAM:5e658f4b63082044810264!
--
Paul Amaranth, GCIH | Manchester MI, USA
Aurora Group of Michigan, LLC | Security, Systems & Software
paul@... | Unix & Windows
|
Re: 693B sweep oscillator restoration
Does anyone have a line on manuals?
Sean
|
Re: 693B sweep oscillator restoration
My thoughts exactly, Dave McGuire!? David C, I don't know why you thought you had to take on at least 3 members of this group.? ?That is not wise and shows insecurity.
Imparting ones knowledge to others with humility is encouraged; flames, boasting, and personal attacks are not.? Let's keep it civil, please.
Jim Ford?
Sent from my Verizon, Samsung Galaxy smartphone
toggle quoted message
Show quoted text
-------- Original message -------- From: Dave McGuire <mcguire@...> Date: 3/8/20 4:06 PM (GMT-08:00) Subject: Re: [HP-Agilent-Keysight-equipment] 693B sweep oscillator restoration
On 3/8/20 2:22 PM, David C wrote: > Moderator, where are you? ? Here's one!? I'm wearing my moderator hat, and I'm telling you that you're officially on thin ice.? Tone it down and speak respectfully. ? And for heaven's sake, PLEASE start checking your facts. ??????????? -Dave -- Dave McGuire, AK4HZ New Kensington, PA
|
Re: HP1700 Series Rear Cord Wrap Feet
On 2020-03-08 8:59 PM, Sam Reaves wrote: I have a virtually mint HP1740A that was willed to me 30 years ago. I
just took it down off the shelf and all 4 cord wrap feet disintegrated.
Does anyone have a set or know where I can find them? eBay turns up nothing.
Fred Usack can machine high quality rubber feet, with the bolts if needed; I've ordered several sets from him for HP 17xx and Tek scopes. Just tell him thoe model # or give specs. --Toby (just a satisfied customer) Thanks,
Sam
W3OHM
|
Jinxie
Michael A. Terrell
Dave McGuire
Burt K6OQK