Re: Agilent E8357A Network Analyzer Phase Lock Lost Error
Forgot to add the Block Diagram.
|
Agilent E8357A Network Analyzer Phase Lock Lost Error
I have been troubleshooting my Network Analyzer using the E8356A, E8357A , E8358A Service Guide June 2004. This 8357 has an upgraded 700 MHz CPU.
After reading and Option 015.
My problem is Phase Lock Lost error in a number of bands. All would seem to relate to the A8 Fractional-N Synthesizer Board Assembly.
Fortunately I could partially run the Serial Bus Test where I found wildly out of spec voltages on the board.
The Serial test node numbering on my application do not coincide with the manual Nodes but the description places me in similar territory as the Manuals Node descriptions.
The test has a minimum and maximum range for the Node as I will describe.
??????????????????????????????????????????? Min????????????? Result?????????? Max
+5Vdc Supply Voltage???? 4.28.?????????????? 5.0?????????????? 5.78.???? OK
Band ALC Level???????????????? -8.27?????????? 14.0169????????? 1.0??????? Bad
VC01 Tune Voltage???????? ?? 6.3???????????? -1.345 ?????????? 14.3???? Bad
VCO Compensation ? ?? ?? 2.07??????????? 13.748??????????? 4.07????? Bad
????????
?
Node descriptions A11 Phase Lock Board. ( the signals to the A11 board 30kHz block come from the A8 Fractional-N board.
30kHz Phase Detector???? -0.33.???????? -5.088 ? ? ? ? ? 0.33. ? ? Bad?
30kHz Phase Detector Integrator
?????????????????????????????????????????? -2.86?????????? -13.827?????????? 0.44.???? Bad
?
10MHz Frequency.? 10.000000
5Mhz Frequency?????? 4.999756
?
A10 Phase Lock Test??
Network Analyzer Setting?? W39? From A10J5
Band 0???????????? Setting 9.9MHz???????????? Result should be? ??? 9.9MHz ??????????????????? Actual Measurement ??????? 9.87??????? (Very weak signal)
Band 1 - 6.?????? Setting. 1 GHz?????????????? Result should be ?? 1.04166MHZ ??????????? Actual Measurement ?????? 1.041616MHz
?
A8 Fractional -N
Band 0 Output????? Setting 5MHz.?????? Result Should be 120.16 MHz.??? Actual Measurement?????????? No distinct signal
???????
I have added pics from CW Testing showing the frequencies and Bands affected.
And the 20MHz - 550Mhz segment.
?
My questions: 1. Would you agree with A8 Fractional board being bad?
???????????????????????? 2. Are there A8 Boards available in the < = 500 dollar range.
?
?
|
HP 8902A Measuring Receiver
To anyone with HP 8902A expertise,
My HP 8902A is in need of repair and some expert advice would be most helpful. Before going into details, I would like to hear from someone who is willing to assist.
73
Raymond ON4DBV
|
On 5/23/25 20:46, Philip Freidin via groups.io wrote: I'm not a group manager, but since you have posted a message,
I am pretty sure you have now joined the group. Welcome.
You may want to select subscriptions on the left side of the web
page, an adjust settings to match the way messages are delivered I had given him this info offline. Nice to see you here Philip. -Dave (group owner) -- Dave McGuire, AK4HZ New Kensington, PA
|
I'm not a group manager, but since you have posted a message,
I am pretty sure you have now joined the group. Welcome.
You may want to select subscriptions on the left side of the web
page, an adjust settings to match the way messages are delivered
?
Philip
?
|
Re: HP8566B with YTO unlock indication appears to have no YTO output at all
I think it's fixed and running beautifully. The other A19 board was functional but significantly different on top and none of the labeling on the plastic cover matched the board, so I decided I should fix the original. I quickly found that the -12.6V reference was out (only about -1.7V), but it took a while to figure out since there are a number of things that could do it. The culprit was perhaps the absolute last thing I would have expected. I had a couple of usual suspect Ta caps, a couple of high-K ceramic caps, a bunch of negative opamp rail circuitry to make -24V from the -40V supply, and of course the opamp itself that regulates the reference from a TC Zener.
?
After some test point wire tacking and operating measurements I found that the input to the opamp U3 was shorted to ground by C18, showing about 18 ohms. I couldn't believe this high grade plastic cap could have failed - it's not even a high stress spot at all, just extra filtering at the input against 5.11 kR from the Zener. C18 and its counterpart C22 in the +10V reference apparently needed to be high grade (I'm virtually certain these are PPF = polypropylene and foil) types for very low ESR, ESL, and DC leakage. I have some of these salvaged from other gear over the years, so I immediately recognized the type, and I can often tell by the density, which usually "feels" higher than film type caps. This is a relatively fancy and expensive part, and physically big (board space-wise), so there must have been good reason. A cap like this failing in a spot like this has to be an extremely rare random failure.
?
I found an equivalent one in stock, popped it in, and voila, just like new. All the original boards are back in now. I assumed the original setup for A19 is still good enough, so I only did the A20 adjustments, and this time the procedure worked as it was supposed to. It seems to be all working now, but will need some more checkout to be sure. I am happy for now.
?
Ed
|
Re: HP8566B with YTO unlock indication appears to have no YTO output at all
I tried to do the adjustment procedure for A20, but could not get the YTO frequency above about 2.1 GHz, and likewise with any settings of the SA. Fortunately I have on hand a nearly full parts carcass, so I found by board swapping the actual fault appears to be on A19, the DAC pre-tuning board. Now the unit has original A21 and A20 (but all messed up from trying to tweak it before), and a different A19, that all now seems to work. The A19 is a different board version - I don't know if it's newer or older than the original here, but it should be OK.
?
I'll have to try the whole procedure for tweaking it up, unless I can quickly find and fix the fault in the original. Now that I know where to look, I will poke around in there and see what's up.
?
Ed
|
Thanks Don.
I noticed that when I turn on the device the red LED on the A6A7 (YTX current driver) board lights up for a few seconds, after which it turns off completely and the green LED seems to indicate that everything is correct: is this normal?
|
Re: HP8566B with YTO unlock indication appears to have no YTO output at all
The original YTO is indeed shot. I swapped in the one from the other A13, and it now runs and the YTO locks, but on a way wrong frequency. I'm assuming it has slipped a cog in the synthesis, that should be cured by proper adjustment, unless there's a circuit failure there too. The SA is running with external 10 MHz reference from the counter so I can get "exact" readings of the 1LO. I need to review the frequency plan and YTO adjustment process in the manuals.
?
Ed
|
Re: BNC that does not have the bayonet lugs
Mike,
On the BWD instrument these connectors are recessed deep in an insulator where the rotating part of the mating BNC connector is nestled down inside the same insulator. This prevents any ability to rotate it in the normal locking sense because you can't get to it. The recess obviously keeps all metal portions of the connector away from grabbing probably due to the high voltage rating of this instrument and related safety factors.
Since the BWD instrument inputs are purely differential and each channel has two inputs the shields of the probe cables are at ground level for their normal noise shielding purpose. Since this instrument is designed for connection to the AC mains for analysis purposes the inputs are rated to 1000V. Even with the BNC shells at ground BWD still apparently designed in the extra level of protection by recessing the connectors.
Greg
|
Re: Off Topic - Looking for assistance with power sensors
"Razvan Popescu via groups.io" <yo8ryr@...> writes: Hi,
On 22/05/2025 22:11, Sven Schnelle via groups.io wrote:
This is likely just coincidence because they are all calibrated by
keysight - it might be at any location in the EEPROM.
Yes, that is correct. The start address formula x*8+9 applies and you
can use any address as long as you are following that formula.
Practically you can start with your table at address 9.
Without updating the checksum of the table the power meter will not
accept the new data... I attached a program i used to rewrite the
calibration data of my sensor. Keysight doesn't do any business with
private persons, so there was no other way i guess.
I didn't have any issues with the old HP branded E4413A. I didn't see
anything that would look like a checksum. The end of the EEPROM in my
case had many bytes written with value 255. Maybe the first 9 (0 to 8)
addresses that cannot be used for calibration table are used for
checksum and other data?!
Is it possible that the old E441XA sensors don't have a checksum?
Maybe the checksum was added later in E93XXA series?! Can't say because i don't have many power sensors (namely an E9304A, and two a ECP-18A + E4412A). The eeprom contains an eeprom header in the following format (everything in big endian format): struct header { uint8_t version; /* 0x00 */ uint16_t unknown_0x01; /* 0x01 */ uint16_t tablecount; /* 0x03 */ struct tabledesc tables[0]; /* 0x05 */ } __packed; tablecount is the number of calibration tables, version the overall version of the eeprom. The table descriptors are starting at offset 5 in the eeprom, and each table header looks like this: struct tabledesc { uint16_t offset; uint8_t type; uint8_t version; uint16_t size; uint8_t checksum; } __packed; 'offset' is the absolute start location in the eeprom, 'type' what kind of table this is (cali factors, temp compensation, linearity, etc). 'version' which format the table has - for most of the tables at least two versions exist. 'size' is how many bytes the table contains. 'checksum' is the sum of all bytes in that tables. I haven't seen that this is optional, but can't say for sure. An entry in the calibration factor table is: struct calfactor_entry { uint32_t freq; /* 0x00 in khz */ uint16_t calfactor_low; /* 0x04 */ uint16_t calfactor_high; /* 0x06 */ } __packed; which is the 8 bytes in length you mentioned.
|
Re: BNC that does not have the bayonet lugs
Maybe for a quick connect/disconnect? - Mike
Mike B. Feher, N4FS
89 Arnold Blvd.
Howell NJ 07731
908-902-3831
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] < [email protected]> On Behalf Of Greg Muir via groups.io Sent: Thursday, May 22, 2025 6:30 PM To: [email protected]Subject: Re: [HP-Agilent-Keysight-equipment] BNC that does not have the bayonet lugs Beyond conjecture..... BNCs used in the BWD 880 PowerScope. Greg -- This email has been checked for viruses by Avast antivirus software. www.avast.com
|
Re: HP8566B with YTO unlock indication appears to have no YTO output at all
I swapped out the entire A13 (I think that's the right number) YTO/PLL assembly with one from an old 8566A carcass that I was using for another project. The YTO output was good and strong, but way off in frequency since the YTO is now different and the coil driver system is at the original's tweak settings, which I don't want to mess with unless I have to. It still shows YTO unlock warning - not surprising but it would have been nice if it came out close to right on and started working. Oh well. Also, some of the mounting screws were missing, so someone had been looking in here before.
?
So, it looks like it's narrowed down to something in A13. I still have some swapping options to try out yet, then better measurements on the supplies and support circuitry, then probably a bunch of adjustments, depending on the true cause.
?
Ed
|
Re: HP 478A Thermistor Sensor. Input return loss and matching
I spent some time looking at my early 478A sensor again this evening and removed the ferrite (that I had added) and also bypassed the 2.7R damper resistor R1 fitted by HP. I then looked at the return loss on a VNA and also looked at the frequency response when the 432A + 478A was used to level a HP 83752A sweeper.?
?
The plots below show that with no damping the input return loss shows a classic zigzag shaped resonance just above 50 MHz. The VSWR is actually slightly better at 50 MHz without the damper but the VSWR obviously gets much worse at 53.5 MHz.?
?
With the 2.7R resistor R1 removed, the frequency response using the sweeper shows a narrow upwards blip of almost +0.3dB at 53.5 MHz and this is due to the undamped resonance here. The resonance draws energy away from the thermistors and this degrades the efficiency. The blip is positive and this shows that the efficiency must be lower in the dip as the sweeper has to put out a bigger signal to get the same output from the 478A sensor.?
?
The response from 10 MHz to 50 MHz is exactly as a simulation with about 0.1dB change from 10 MHz to 50 MHz. If I add a ferrite, this completely removes the blip at 50 MHz but it does degrade the efficiency down at 10-20 MHz. So there is no free lunch with the ferrite. I still haven't decided on the best choice of ferrite here and I may also try increasing the capacitance after the ferrite to be similar to the later 478A sensor.
?
I think this experiment does indicate that the 2.7R resistor R1 was added by HP to try and tame this internal resonance. The combined capacitance value of C3 and C4 may well have something to do with the operation of the 431 meter, but I think HP added some damping resistance to one of the capacitors to tame the resonance at 53 MHz and this may be partly why these parts are marked as factory selected.?
?
?
?
?
--
Regards
Jeremy
|
HP8566B with YTO unlock indication appears to have no YTO output at all
I have never encountered a bad YTO in a 8566, but this might be one. I just got this unit and had some time to check it out, assuming typical small problems. I'll have to bring a scope out to my truck next - it's the only place left where I can park enough equipment and have some flat space to actually work on it. I figured the unlock was for common reasons, and looked first with a counter at the 1LO output and found nothing. Since microwave counters can't get a reading sometimes if the frequency is flailing around, I next put a 436A and 8484A on to see if there's any RF power at all regardless of frequency - nope.
?
I haven't exhaustively checked the possible reasons for no YTO signal yet, so all the supplies will be next up for ripple etc - the DC values seem OK but the scope will say for sure. I have to get familiarized again with this beast since I haven't worked on the main RF section guts in a few years. Then all the usual other stuff to check out in the connections, signal paths, and support circuits. So far everything looks normal control- and display-wise, except there's no input signals showing above the noise floor in any band, which makes sense if there's no 1LO at all. I think all the internal PLL and other indicator LEDs are showing normal conditions. The unit overall looks very clean inside, and complete. One odd thing is that someone in the past had put the little front panel SMA 2IF I/O link in the wrong spot - the IF out was tied to the 1LO and the IF in was terminated. I don't think this could have possibly hurt anything.
?
Should the worst be the case, I have spare YTOs, but I'm hoping it's something simple in the circuitry to be discovered.
?
Anyway, while I study and revisit the manual and start poking around, I figured I'd check to see if anyone has seen this sort of problem before.
?
Ed
|
Re: BNC that does not have the bayonet lugs
Beyond conjecture.....
BNCs used in the BWD 880 PowerScope.
Greg
|
Re: Off Topic - Looking for assistance with power sensors
Hi, On 22/05/2025 22:11, Sven Schnelle via groups.io wrote: This is likely just coincidence because they are all calibrated by
keysight - it might be at any location in the EEPROM.
Yes, that is correct. The start address formula x*8+9 applies and you can use any address as long as you are following that formula. Practically you can start with your table at address 9. Without updating the checksum of the table the power meter will not
accept the new data... I attached a program i used to rewrite the
calibration data of my sensor. Keysight doesn't do any business with
private persons, so there was no other way i guess.
I didn't have any issues with the old HP branded E4413A. I didn't see anything that would look like a checksum. The end of the EEPROM in my case had many bytes written with value 255. Maybe the first 9 (0 to 8) addresses that cannot be used for calibration table are used for checksum and other data?! Is it possible that the old E441XA sensors don't have a checksum? Maybe the checksum was added later in E93XXA series?! Regards, Razvan
|
Re: Optiion upgrade HP 8753C/B
Correction: when running service code 56 it is "OPTION COR FAIL".?
|
Re: Optiion upgrade HP 8753C/B
Hello All,?
?
Caesar graciously provided me with the option codes to install option 006 on my 8753C several months back. I have finally located an 85047A for a reasonable price, and I am trying to install the code for Opt 006 to enable it. After entering the code Caesar provided, I keep getting the error "CORRECTION CONSTANTS NOT STORED" while running service code 56, and I am still not able to select higher than 3GHz. I have moved the jumper on the A9 to the "ALT" position but it does not seem to be accepting the upgrade.?
?
Also, I am getting the same error when trying to run the service code "55" for serial number correction.?
?
I have been following the instructions here:? Instalation note 8753C.pdf?as well as the instructions beginning on page 123 of the HP Service Manual here:? ?
Has anyone run into this situation where the option codes are not accepted??
?
|
Re: BNC that does not have the bayonet lugs
Already tons of various and diverse messages that are ultimately just conjectures... Paul, please post a photo of the connector you're talking about. Indeed, we often say that a small drawing is worth far more than a long speech. In your case, a small drawing would truly be the best. Thanks in advance.
|
Dave McGuire
Philip Freidin
