¿ªÔÆÌåÓý

  1. HP-Agilent-Keysight-Equipment
  2. Messages
Expanded Topics Messages
Date
Re: HP8566B - Making sense of the figures

 

Indeed, and I've just done it. No sign of any YTO warning and the FM stations are back where they should be, too. I think we can call that a successful outcome, Ozan!
I'm extremely grateful for your knowledge and persistence in getting this thing back into fully working order again. No way could I have done it without you, that much at least is certain. Do you have a Paypal account? If so, PM me your particulars and I'll send you a 'drink' as we say as a gesture of my sincere gratitude for your guidance and expertise. Well done!
Re: What happened to HP/Agilent detailed circuit schematics

 

Today¡¯s corporate focus on product manufacture and service is many fold.

Profits made from equipment sales are only part of the equation. There is (sometimes much more) money to be made through support. If you don¡¯t provide good documentation the customer is basically faced with having to send the item in for repair.

Second, in order to produce a very detailed document describing every nut and bolt on how to fix the item requires a gaggle of tech documentation people which costs $$$ to employ. Make manuals simple and you have fewer related employees to deal with.

Third, it has been shown that if users do choose to repair their own equipment that fixing things at the higher level (i.e. PC board, etc.) proves to be easier both for the customer and the factory service department where it reduces time to fix. Neither the customer nor the factory needs to dig deep to find a problem if they follow the simple diagnostic charts provided in the manuals. And factory service effort is reduced by simply plugging boards into test fixtures to localize the fault. Or, if more economical, they may opt to simply scrap the boards and present a new board fee for the user when they purchase through ¡°exchange.¡±

There is more but I will not continue. One must remember that we now live in a ¡°disposable society¡± where people need instant everything which includes immediate replacement if something breaks. As for those of us legacy equipment collectors/users down here in the trenches we gradually see our inventory becoming older and older in age as we either cannot or opt not to try to fix ¡°mystery boxes¡± now being manufactured. And as it has been said by others in previous posts, opening a box only to see a PC board with only one non-repairable module in the center of it certainly adds to the issue.

And even if that module can be replaced by ordering another from the manufacture time is of a limit given the shorter and shorter time replacement part inventories are held after the manufacture is discontinued for that device. It¡¯s expensive to store parts for old equipment since everyone wants the ¡°latest and greatest¡± products and will achieve that goal by perpetually buying more new ones. From that the importance of supplying parts for legacy equipment becomes less and less. It¡¯s the corporate ¡°bottom line¡± speaking here.

A note on sending equipment out for repair. Yes, there are those third party agencies that will offer service long after the factory does not. But what I have seen even through actual visits to these places is that they, too often have scanty data to use to fix stuff. From that many of them will tell the customer that the item cannot be repaired and if the customer opts to simply abandon the item with the repair agency that agency will relinquish it to a ¡°bone yard¡± where they can use parts out of it to fix other units.

Greg
Re: What happened to HP/Agilent detailed circuit schematics

 

Very eloquently stated, Chuck!? A fantastic summary of our (sad) times!
Re: HP 5086-7906 YTO

 

¿ªÔÆÌåÓý

Yves, Steve, thank you very much for the offer of help
I'll try to make it short:
A few years ago my 8595E had shown signs of "deafness" in band 1 and given the prices of a SYTF I decided to keep it as it was
?
Unfortunately, some time later the main oscillator (YTO) also stopped working
?
This component also had prices out of my reach and, with a lot of patience and little hope, I purchased a YTO at a reduced price and, months later, a SYTF both sold "as is"
?
Not having in my hobbyist laboratory everything I needed to test them (for example another working 8595E) I left the project pending until a tiny instrument (TinySA Ultra) appeared on the market that is allowing me to address the problem, at least at a diagnostic level
?
And so I started to test the measured YTO on the bench, verifying that it covers continuously from 3.06 to 7.52GHz at 23mA/GHz with an output (excluding the extremes) between 7 and 10 dBm with peaks of 13.4dBm (FM coil not tested)
?
It would therefore seem to work well
?
I then tested the SYTF: the switch section seems to work well and also the filter section seems to work well as a bandpass with an insertion loss on the tested frequencies (5 - 6GHz) up to 11 dBm
?
And so I installed the two components in my 8595E and first I tried to do the calibration but the sequence soon stops with "Cal signal not found"
?
Now, the calibration signal is present (and in specification) both on the BNC connector, and after the attenuator, and after the SYTF filter, and after the low-pass filter at 2.9GHz
?
And for now I have come this far The next step will be to proceed to verify the Dual Mixer, (but first I have to understand what to expect to see on each of its 5 ports)
?
Thanks for any suggestions
Adri
?
?
?

----- Messaggio originale -----
Da: Steve Hendrix via groups.io <SteveHx@...>
Rispondi a: <[email protected]>
Inviato: 26/08/2024 03:28:52
Oggetto: Re: [HP-Agilent-Keysight-equipment] HP 5086-7906 YTO
I have a fully working (AFAIK) 8595E, if you need anything I can read from the screen or via IEEE-488. Swamped with work for the next few weeks, so any disassembly is out of the question.

Steve Hendrix

On 2024-08-25 3:50 PM, Yves Tardif wrote:

Hi Adri,

?

I have owned a 8595E for a few years, but it recently broke down and I had to repair it, what is your problem exactly?

I can see that the RF part is disassembled, but please give more details.

?

Yves

De?: [email protected] <[email protected]> De la part de alfa beta
·¡²Ô±¹´Ç²â¨¦?: 24 ao?t 2024 09:50
??: [email protected]
Objet?: Re: [HP-Agilent-Keysight-equipment] HP 5086-7906 YTO

?

I am a hobbyist from Milan - Italy and I am trying to bring my 8595E back to its former glory
I have been following the HP-Agilent group for years and I have appreciated your frequent and intelligent contributions among which the expertise regarding the HP analyzer series stands out
I was wondering if I can ask you for some advice that can help me get my analyzer back on track
thanks in advance
adri

?


20240826_185711.jpg
20240826_185756.jpg

Re: HP8566B - Making sense of the figures

 

Again short reply to catch you in your time zone: No need to move out of service position. If you have a piece of SMA cable that would work for connecting the front end.?
Ozan
?
?
On Mon, Aug 26, 2024 at 09:58 AM, Jinxie wrote:

Well, well. Because it was dead easy to do and made a certain amount of sense, I tried your 'sanity check' suggestion first instead of last and......<drum roll....> The YTO unlock warning has gone out! I'm tempted to re-connect the LO and see if my local FM stations have gone back to where they should be instead of 20Mhz adrift. I have not seen a screen clear of that YTO Unlock message for 4 years now!
Is there a possibility you may have misunderstood some aspect of the function of U1 which initially led you to believe there was still a fault with this section, do you think? Anyway, regardless; if this turns out to have fixed the issue then there will be some $$$$$$$$ coming your way for your trouble, Ozan. I think I'll hook up the LO and see if I can get BBC radio 4 where it ought to be in the UK - instead of where it ought to be in Japan!!
Re: HP8566B - Making sense of the figures

 

Well, well. Because it was dead easy to do and made a certain amount of sense, I tried your 'sanity check' suggestion first instead of last and......<drum roll....> The YTO unlock warning has gone out! I'm tempted to re-connect the LO and see if my local FM stations have gone back to where they should be instead of 20Mhz adrift. I have not seen a screen clear of that YTO Unlock message for 4 years now!
Is there a possibility you may have misunderstood some aspect of the function of U1 which initially led you to believe there was still a fault with this section, do you think? Anyway, regardless; if this turns out to have fixed the issue then there will be some $$$$$$$$ coming your way for your trouble, Ozan. I think I'll hook up the LO and see if I can get BBC radio 4 where it ought to be in the UK - instead of where it ought to be in Japan!!
Re: HP8566B - Making sense of the figures

 

I know it is getting late in your time zone, I will send a longer reply but for now could you try connecting gray (8) cable back, undo forcing of tune, and see if YTO is still giving unlock message? Connecting YTO OUT on coupler is not needed.?
?
Ozan
?
?
On Mon, Aug 26, 2024 at 09:04 AM, Jinxie wrote:

Thanks, Ozan. I don't think there is an intermittent issue at all, on reflection, because I'd never measured the output of U9 before and that is what caused the confusion, I believe. Fingers crossed that's what happened, anyway.
One good thing is if U1 has failed, I have actually got at least one new, unused replacement 'in stock' here according to my parts inventory (which although I do say so myself is never wrong). So that's some comfort!
I would have already supplied all the voltage readings around U1, but access to those pins is particularly tricky and I didn't want to risk it if it wasn't strictly necessary.
Before I leap in and undertake your suggested tests, can I just confirm from you I have a valid set-up to eliminate that possibility? I'm using 4Ghz/zero span/sweep-off and the gray PD wire (8) is disconnected, as is the hard line connection to 'YTO out' on the -15dB coupler/isolator which I've terminated at the coupler/isolator end with a 50ohm sma load. All the other wires are connected. Just thought I'd better check that lot with you to make sure I haven't created a 'fault' myself which wouldn't otherwise be present.
Re: E4418B

 

I had a similar problem with mine, the power button barely worked at all.
?
It turned out to be a worn out mylar button membrane under the rubber button caps.
?
You can buy them relatively cheaply on Aliexpress and Ebay if you shop around a bit and they aren't too hard to replace.
?
There are a couple tricks to do the replacement though, so report back if you do decide to replace it and I can help you through it before you accidentally break something. :)
?
It's also worth replacing the internal battery too before your instrument gets amnesia and forgets what it is (model and revision data etc is kept in battery backed RAM), just be sure to take photos of all the serial number and revision number data in the menu there somewhere.
Keysight have simple instructions on how to reprogram the model data, as long as you know what it is... ;)
?
?
Regards
Jared
Re: HP8566B - Making sense of the figures

 

Thanks, Ozan. I don't think there is an intermittent issue at all, on reflection, because I'd never measured the output of U9 before and that is what caused the confusion, I believe. Fingers crossed that's what happened, anyway.
One good thing is if U1 has failed, I have actually got at least one new, unused replacement 'in stock' here according to my parts inventory (which although I do say so myself is never wrong). So that's some comfort!
I would have already supplied all the voltage readings around U1, but access to those pins is particularly tricky and I didn't want to risk it if it wasn't strictly necessary.
Before I leap in and undertake your suggested tests, can I just confirm from you I have a valid set-up to eliminate that possibility? I'm using 4Ghz/zero span/sweep-off and the gray PD wire (8) is disconnected, as is the hard line connection to 'YTO out' on the -15dB coupler/isolator which I've terminated at the coupler/isolator end with a 50ohm sma load. All the other wires are connected. Just thought I'd better check that lot with you to make sure I haven't created a 'fault' myself which wouldn't otherwise be present.
Re: What happened to HP/Agilent detailed circuit schematics

 

inline.

On 8/26/2024 11:48 AM, Dave B wrote:
HI Harvey,
Your creations sound far superior to my KISS approach to the issue.
Sometimes I think more may be less...
I have been designing my replacement 'module' with the idea it just 'drops in' in place of the existing inverter.
this would fit in the same space as the inverter board.? An 8 or 14 pin microprocessor (depending on innards), a very tiny regulator, a few SMT resistors and two transistors (one a DPACK).
I have watched an interesting Youtube video of a complicated attempt to solve this problem and gleaned the control voltages and supply data from this.
I have not ventured inside the instrument I have on my bench yet, as I only want to go inside it once!
It looks like there is a 3 pin connector with +5V, GND and a variable 'Dim' control voltage.
Then the 5 volt can be a bit of a problem.

Think 20 ma/LED, and 9 LEDS fit in a strip (I have such strips, pc board wise).? So maximum non-PWM controlled current is about 180 ma/strip, and that, for 2 strips, is 360 ma.? The problem is that the LEDS may not have enough voltage for 2 in series.? Hence parallel.

What's done in some LED backlit displays is to put the LEDS in series (say needing about 18 to 20 volts), and then run them with a small inverter.? That inverter might just have brightness control....

They make chips for that.? So perhaps that chip, a few transistors, a small inductor might just do the job for you.


According to the video, the control voltage varies from 0.7V (display off) to 4.7V (full brightness) set by rotating the front panel adjustment.
So...
I'm assuming that the HV inverter would be taking a much greater current from the 5V supply than a few LED's to generate the same light as a CCFL, so that's the supply sorted?
Depends on number of LEDS, and hookup.? So not quite so easily sorted.


My concern was that without a schematic I don't know what the source impedance of the control voltage was, leading to my 'missing schematics' discovery!
As you hinted, I am powering the LEDs off the 5V and using a high impedance analogue input device to exacerbate any loading issues.
That works.
Assuming the voltages are the same on this instrument as measured on the video, then I should have a CCFL replacement backlight module for a great deal less
than an OEM module.
Which is a good idea.? Multiple ways of getting there.
It all works on the breadboard, I'm just waiting for the pcb's to arrive to start manufacturing and carry out a real life test.
Ah, then you've gone further in this design than I thought.

Best of luck, then, and let me know if I can contribute more ideas.

Harvey


Dave
Re: What happened to HP/Agilent detailed circuit schematics

 

HI Harvey,
Your creations sound far superior to my KISS approach to the issue.
I have been designing my replacement 'module' with the idea it just 'drops in' in place of the existing inverter.
I have watched an interesting Youtube video of a complicated attempt to solve this problem and gleaned the control voltages and supply data from this.
I have not ventured inside the instrument I have on my bench yet, as I only want to go inside it once!
It looks like there is a 3 pin connector with +5V, GND and a variable 'Dim' control voltage.
According to the video, the control voltage varies from 0.7V (display off) to 4.7V (full brightness) set by rotating the front panel adjustment.
So...
I'm assuming that the HV inverter would be taking a much greater current from the 5V supply than a few LED's to generate the same light as a CCFL, so that's the supply sorted?
My concern was that without a schematic I don't know what the source impedance of the control voltage was, leading to my 'missing schematics' discovery!
As you hinted, I am powering the LEDs off the 5V and using a high impedance analogue input device to exacerbate any loading issues.
Assuming the voltages are the same on this instrument as measured on the video, then I should have a CCFL replacement backlight module for a great deal less
than an OEM module.
It all works on the breadboard, I'm just waiting for the pcb's to arrive to start manufacturing and carry out a real life test.
Dave
Re: What happened to HP/Agilent detailed circuit schematics

 

Dave,
?
Not to be rude, but it does appear that you are trapped in the mists of time as you put it.? As a current lab instrumentation engineer and self-taught electronics engineer/repair specialist, there is no company that I am aware of today (at least in the biotech industry manufacturing lab equipment) that is providing schematics for their internal circuitry.? They don't even provide service manuals anymore (no I don't mean user manuals).? Service manuals as in manuals with troubleshooting, circuit descriptions, calibration procedures, and PM procedures.??
?
I tend to agree with Chuck.? And this is a topic that could be discussed for hours.? But this is what the whole 'Right to Repair' movement is about which I am huge proponent of.? It is one of the reasons I started my own YouTube channel to thwart the manufacturers and share with scientists how they can repair their equipment on their own.
?
Also this comment below is probably the most short-sighted and completely ridiculous.? There is still much repair you can do without having access to the firmware.? Although I believe companies should offer that, too.??
?
"Another topic related to the new designs is that they use a lot of
micro-controllers and FPGAs and even if you would have the schematic and
let's say you can get the exact part number to replace the
micro-controller or the FPGA you will not get the microcode/software to
write it back. So it is pretty useless for the new designs to get a
schematic if you cannot get the complete firmware for individual components."
?
-Frank
Re: HP8566B - Making sense of the figures

 

On Mon, Aug 26, 2024 at 04:18 AM, Jinxie wrote:
Well, I stand corrected! This thread is the gift that just keeps giving for our resident tech addicts. My fears in my last post proved unfounded. With +4.5V applied to Tune, I still have -6.6V on the output of U9, but +6.6V on phase detector output! How curious. Will this ever get solved? Hmmm....
?
====
Either there is an intermittent contact (on A11A4 or your tune input), or some failure in U1 that gives exactly 6.6V but it is not clear why U1 would fail with that voltage?
?
I recommend connecting two meters, or two probes of a scope, one to U9 output, the other to U1 output and look at them at the same time while +4.5V is applied to tune input. If you really see two different voltages check control voltage of U1 at pin 8 to be above 2.5V. (sample and hold is in sample mode). If input at pin3 is -6.6V, pin 8 is above 2.5V, and output is still +6.6V, U1 failed. I find it unlikely though.?
?
If you find U9 pin 6 to be intermittent, look at collector of Q4B again to see if the intermittent issue is before or after U5.?
?
As a sanity check regardless of above findings: Connect A11J1 to the coax cable again removing any other connection to tune input (battery etc). Check if YTO unlock message is still there, ?A11J1 is still +6V, and whether right side of R34 (unlock detect) is high or low.?
?
Ozan
?
?
?
?
?
?
?
Re: What happened to HP/Agilent detailed circuit schematics

 

I've considered such replacements as I do design microprocessor boards with (replaced by LED) CCFL displays.

Many of the inverters have a DC control voltage as you know.? I'd be tempted to have the LED string capable of working off the dc supply voltage for the inverter, and I'd bag the whole idea of worrying about the DAC.? Feed a very small (8 pin or so) microprocessor's A/D input through a divider and let the processor generate the PWM.? That, with a transistor level shifter allows you to use a P channel power FET to control the LEDS.

IF you had access to the software and IF it had an I2C interface you could control on the main board, something like a PCA9634 FM LED controller would be overkill, but would work.? I use that LED controller on my generic boards and just reserve one channel of 8 for the display backlight.

Harvey

On 8/26/2024 7:59 AM, Dave B wrote:
HI, well I suspected that may be the case and in the digital side of the instruments, (as you say) one wouldn't have a chance of a repair.
However I would have imagined that just about every possibility of RF circuitry would have been explored by now so wouldn't particularly be top secret?
And, even in my HP 1725A 'scope they used 'hybrid' (secret?) modules, and just depicted them as a block but showed the rest of the analogue workings as a 'proper' schematic.
It's frustrating as I'm developing a tiny solid state (plug and play) LED backlight 'module' to replace the CCFL inverter and tube.
The backlight is borne to fail as the CCFL lamp dies after a while. The inverter and lamp were (I believe) made easy to replace as they are pretty accessible without having to
access any of the electronics within the sig-gen. (It's just behind the front panel).
However, without knowing what is supplying the +5v and the dimming control voltage could easily be answered if a schematic was available.
Questions like:-
Does the control come from a DAC (I think it does) but what drive capability does this have. Is the output of the DAC protected by a series resistor?
Whilst I could start to take the instrument internal covers off to trace this out...........sigh!
Dave
Re: HP8510C VNA, read LIF DD floppy, initialize HD floppy? Using FlashFloppy or other emulator

 

Hi Staffan,
?
what if you copy the floppy with a real floppy drive inside the 8510 to a GPIB drive attached to the 8510 and then copy it back from there to the GOTEK?
Then the 8510 should at least be able to read the proper data from the GOTEK and you may be able to analyze what is different from your original setting.
Btw, I have configured an old WinXP machine to serve as GPIB drive, quite handy.
?
Best regards,
Tom
Re: What happened to HP/Agilent detailed circuit schematics

 

Consider that we live in an amoral litigious society, where
companies have patented just about every conceivable
combination of thoughts, algorithms, or circuits, that have
happened in the past, present, or future, and covetous lawyers
roam the planet sniffing out shares of any monies that can
be taken from the legally weak.

In such a society, it is impossible to design circuitry when
you have to challenge your every thought, and trace the origin
of your every idea just to see who owes what and to whom...

So why, in such a society, would you put your company in the
sights of marauding lawyers by freely revealing exactly what
you have done inside of your circuits, ASICS, ROMS, and FPGA's?

Make the b**tard lawyers work for their plunder!

Component level servicing manuals are a relic from a much
friendlier time; a time when most people had moral values,
and lawyers were few, and traded their craft for livestock and
produce.

-Chuck Harris


On Mon, 26 Aug 2024 04:59:29 -0700 "Dave B" <davebullockmbe@...>
wrote:
HI, well I suspected that may be the case and in the digital side of
the instruments, (as you say) one wouldn't have a chance of a repair.
However I would have imagined that just about every possibility of RF
circuitry would have been explored by now so wouldn't particularly be
top secret? And, even in my HP 1725A 'scope they used 'hybrid'
(secret?) modules, and just depicted them as a block but showed the
rest of the analogue workings as a 'proper' schematic. It's
frustrating as I'm developing a tiny solid state (plug and play) LED
backlight 'module' to replace the CCFL inverter and tube. The
backlight is borne to fail as the CCFL lamp dies after a while. The
inverter and lamp were (I believe) made easy to replace as they are
pretty accessible without having to access any of the electronics
within the sig-gen. (It's just behind the front panel). However,
without knowing what is supplying the +5v and the dimming control
voltage could easily be answered if a schematic was available.
Questions like:- Does the control come from a DAC (I think it does)
but what drive capability does this have. Is the output of the DAC
protected by a series resistor? Whilst I could start to take the
instrument internal covers off to trace this out...........sigh! Dave
Re: What happened to HP/Agilent detailed circuit schematics

 

HI, well I suspected that may be the case and in the digital side of the instruments, (as you say) one wouldn't have a chance of a repair.
However I would have imagined that just about every possibility of RF circuitry would have been explored by now so wouldn't particularly be top secret?
And, even in my HP 1725A 'scope they used 'hybrid' (secret?) modules, and just depicted them as a block but showed the rest of the analogue workings as a 'proper' schematic.
It's frustrating as I'm developing a tiny solid state (plug and play) LED backlight 'module' to replace the CCFL inverter and tube.?
The backlight is borne to fail as the CCFL lamp dies after a while. The inverter and lamp were (I believe) made easy to replace as they are pretty accessible without having to
access any of the electronics within the sig-gen. (It's just behind the front panel).
However, without knowing what is supplying the +5v and the dimming control voltage could easily be answered if a schematic was available.
Questions like:-
Does the control come from a DAC (I think it does) but what drive capability does this have. Is the output of the DAC protected by a series resistor?
Whilst I could start to take the instrument internal covers off to trace this out...........sigh!
Dave
Re: HP8566B - Making sense of the figures

 

Well, I stand corrected! This thread is the gift that just keeps giving for our resident tech addicts. My fears in my last post proved unfounded. With +4.5V applied to Tune, I still have -6.6V on the output of U9, but +6.6V on phase detector output! How curious. Will this ever get solved? Hmmm....

U1 sample _n_ hold.png
final stages.png

Re: HP8566B - Making sense of the figures

 

On Mon, Aug 26, 2024 at 10:32 AM, Dave_G0WBX wrote:
Dave B.
Q1 is not a suspect in this case, Dave. I did take a look at its connections just out of interest and in real life as opposed to the photo, the patina of them is identical to that of the surrounding components' connections.
Re: What happened to HP/Agilent detailed circuit schematics

 

Hello Dave,

You are not missing the schematic supplement, it is not published or
sold anymore.

I am not sure if this is because of IP inside or because they don't want
you to repair the product to force you to buy a new one. Or maybe both?!

Another topic related to the new designs is that they use a lot of
micro-controllers and FPGAs and even if you would have the schematic and
let's say you can get the exact part number to replace the
micro-controller or the FPGA you will not get the microcode/software to
write it back. So it is pretty useless for the new designs to get a
schematic if you cannot get the complete firmware for individual components.

Regards,
Razvan

On 26/08/2024 11:57, Dave B via groups.io wrote:
HI HP experts,
Maybe I am trapped in the mists of time, but what happened to 'proper'
schematic diagrams in HP service manuals?
I have been trying to work on an HP ESG-D3000 and scoured the web for
service information.
I found what appeared to be both operators and service manuals but the
service manual only has block schematics.
Whereas every nut/bolt/clip/washer/rivet and aluminium panel is detailed
in the minutest detail!
Am I missing a schematic supplement, or is servicing these days expected
to just be a 'board swapping' exercise?
I look forward to your comments.
Dave
5421 - 5440 of 151302