Re: HP 8594E low signal level indicated
Bruce, Can you see the 0hz LO level¡ªwhat is it¡ªshould be around 0-to -10db? ? Mixers are not as easily shot as in the 8591? SA so that is good-but it is a possibility The 70 db front end attenuators are quite strong ?,but if its set to 0db? and 0db REF then the cal output should ?be 300Mhz -20db approx..,bypassing the att block is easy then you can fire the cal signal straight into the first LP filter located on top of the RF cage.. This is your starting point.. ? Let me know how you get on ? 73s? mike g4wam ? ? ? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of Rik Rasmussen
Sent: 26 February 2023 02:31
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP 8594E low signal level indicated ? Thanks to both of you. I will open it up tomorrow and check out the attenuator. ? On Sat, Feb 25, 2023, 8:47 PM Bruce <bruce@...> wrote: You would be looking for transmission loss.? If there is a problem?
with the attenuator, it could present very high attenuation.
Safest way to do this is to set the SA for zero attenuation, connect a?
power meter or another SA to the OUTPUT of the 1st SA attenuator and a?
signal generator to the input of the attenuator.? Then cycle the?
attenuator through its various attenuation settings and record the?
results.
A bit more dangerous method would be to inject a signal AFTER the?
input attenuator (I would start quite low -30 or -40 dBm) and see how?
the SA performs.
If the measured attenuation differs from the actual attenuation?
significantly, you have a problem.
Cheers!
Bruce
Quoting Rik Rasmussen <rikrasmussen@...>:
> Are you talking about the HP 8594E's INPUT attenuator? What would I?
> be looking for?
>
>
>
|
Re: HP 8594E low signal level indicated
Rik,
?
It sounds like your problem is a little deeper inside the instrument? But if it ever turn out to be the step attenuator (HP part # 33321-60036) this experience might offer some past dealings with the 8594E.
?
A unit I initially obtained had problems with intermittent levels when switching gains.? It turned out that the tiny rubber O-rings on the ends of the step attenuator solenoid plungers had deteriorated over time and several simply fell off.? This resulted in the leaf contacts not securely touching the plated contacts on the glass substrate.
?
I started to seek replacement O-rings but then found out that selection of the material was somewhat critical to prevent out gassing of the synthetic rubber from compromising the conductivity of the substrate contacts.? This then led me to search for an agency who could repair the attenuator.
?
I eventually found a concern who specialized in repair of these attenuators ¨C Test Equipment Plus located in Arizona ().? But if you use them their prices are rather beefy for repair - $400 for this particular one.? But if this is (or will be) one of your mainline analyzers I highly suggest using them.? When the attenuator was returned it was accompanied with network analyzer plots showing the performance of the repaired item.
?
If you so choose to attack an attenuator problem yourself I will caution you to be extremely careful when disassembling these items.? Yes, simply getting them out of the spectrum analyzer is a little labor intensive but taking apart the attenuator can easily lead to possible damage.? I have the experience to do so but obviously defer repair to a fully capable company.
?
As for your problem since it has more or less determined that the attenuator has not sustained damage there may have been possible damage to the first converter from the input being overloaded with the attenuator in the 0 dB state (direct pass-thru).? Looking at the CLIP it appears that the levels found at each input stage are:
?
CAL input signal: -20 dBm @ 300 MHz
Attenuator setting: 10 dBm attenuation
Attenuator output (A4 first converter input): -30 dBm (obviously)
A4 first converter output (A5 second converter input): ~-41 dBm +/- 2 dB @2.1214 GHz
A5 second converter output (A9 third converter input): -44 dBm +/- 2 dB @ 321.4 MHz
?
All levels are referred to 50 ohms impedance.
?
Hope this helps.
?
Greg
|
Re: HP 8594E low signal level indicated
OK -
First confirm the gain of the YTF. There should be an expected signal level specified and the expected output (input to the next stage) - you should be able to calculate the stage gain; then measure it. If it is low, look ad the drive signals - the problem could be the logic signals to the YTF - quit a bit of work to do B4 you decide the YTF is the problem.
I have not worked on this particular device so someone with more experience on the 8594E might be a better resource.
The fun is starting :-)
Cheers!
Bruce
Quoting Rik Rasmussen <rikrasmussen@...>:
toggle quoted message
Show quoted text
When I said I was feeding the First Converter, that was apparently icorrect. I was looking at the wrong block diagram. I am now looking at the block diagram for the 8594E. The block diagram does include the tracking generator although my unit does not have that option.
After looking at the output of the attenuator and checking that it is stepping correctly, I then inserted the CAL OUT signal directly to the next stage after the attenuator. According to the block diagram that is a "SWITCHED YTF". Feeding at that point I am seeing a low signal on the display and reading is very low.
|
Re: HP 8594E low signal level indicated
When I said I was feeding the First Converter, that was apparently icorrect. I was looking at the wrong block diagram. I am now looking at the block diagram for the 8594E. The block diagram does include the tracking generator although my unit does not have that option.
After looking at the output of the attenuator and checking that it is stepping correctly, I then inserted the CAL OUT signal directly to the next stage after the attenuator. According to the block diagram that is a "SWITCHED YTF". Feeding at that point I am seeing a low signal on the display and reading is very low.
|
Re: HP 8594E low signal level indicated
Reasonable - 1st converter would be suspect, but can you inject signal into the 2nd converter (at appropriate level) to confirm low 1st converter gain? The CLIP should give expected signal level at interstage points.
Cheers!
Bruce
Quoting Rik Rasmussen <rikrasmussen@...>:
toggle quoted message
Show quoted text
I am feeding the CAL OUT (300 mHz -20 dBM) directly to the input of the First Converter, using the hardcable that normally connects to the output of the attenuator.
The HP-8594E is showing the signal level as -50 dBm.
If I attempt to run amplitude cal I get that same error message CALSIGNAL NOT FOUND
So, looks like my loss of gain is somewhere after the attenuator.
|
Re: looking for HP85101C display keys plastic frame
On 2/26/23 06:17, tom_iphi via groups.io wrote: the resin printed parts from JLCPCB have arrived and I am positively surprised about the nice surface finish, see picture. Nice 3D model too!
|
Re: HP 8594E low signal level indicated
I am feeding the CAL OUT (300 mHz -20 dBM) directly to the input of the First Converter, using the hardcable that normally connects to the output of the attenuator.
The HP-8594E is showing the signal level as -50 dBm.
If I attempt to run amplitude cal I get that same error message CALSIGNAL NOT FOUND
So, looks like my loss of gain is somewhere after the attenuator.
|
Re: HP 8594E low signal level indicated
Sounds reasonable - I would "work" the attenuator a bit to be sure it is reliable - these attenuators can be erratic.
Cheers!
Bruce
Quoting Rik Rasmussen <rikrasmussen@...>:
toggle quoted message
Show quoted text
I have the CAL OUT connected to the front panel INPUT, with the 300 mHz -20 dBm signal.
I have my tinySA Ultra connected to the output connector on the attenuator.
The signal is reading -24 dBm on the tinySA Ultra when I have attenuation at 0dBm. As I step through the attenuator steps, the signal drops exactly as it should.
So, I conclude that the attenuator itself is OK.
Next I will insert a signal at the output of the attenuator. This is the input to the First Converter according to the block diagram.
|
Re: HP 8594E low signal level indicated
I have the CAL OUT connected to the front panel INPUT, with the 300 mHz -20 dBm signal.
I have my tinySA Ultra connected to the output connector on the attenuator.
The signal is reading -24 dBm on the tinySA Ultra when I have attenuation at 0dBm. As I step through the attenuator steps, the signal drops exactly as it should.
So, I conclude that the attenuator itself is OK.
Next I will insert a signal at the output of the attenuator. This is the input to the First Converter according to the block diagram.
|
Re: HP 8594E low signal level indicated
If you elect to inject a signat AFTER rather that THROUGH the attenuator be very careful - you risk damaging the first mixer.
It is much safer to disconnect the attenuator input (sig gne side), inject a signal there (or at the output) and measure power at the output (or input if you inject at the output connector).
Cheers!
Bruce
Quoting Mark Bielman <mbielman@...>:
toggle quoted message
Show quoted text
You clearly have a problem.
You are gonna have to open it up and get to the input attenuator.
Not sure how... these things are packed!
As mentioned, you must inject a signal AFTER the attenuator.
That's the 'best' way.
If you're lucky, the attenuator is the problem.
You might try a CAL using an external sig gen
Mark
________________________________
From: [email protected] <[email protected]> on behalf of Rik Rasmussen <rikrasmussen@...>
Sent: Saturday, February 25, 2023 3:50 PM
To: [email protected] <[email protected]>
Subject: Re: [HP-Agilent-Keysight-equipment] HP 8594E low signal level indicated
When I connect the CAL OUT to the input it does appear on the spectrum analyzer but at a much lower level than -20dbm.
If I connect CAL OUT to Input on the HP and press the CAL button a message on the screen says CAL SIGNAL NOT FOUND.
If I connect the CAL OUT to my tinySA Ultra it is indicated as -22.8 dbm. So, the CAL OUT signal appears to be OK. But the input to the HP seems to be attenuated.
|
Re: HP 8594E low signal level indicated
Good luck !
Cheers!
Bruce
Quoting Rik Rasmussen <rikrasmussen@...>:
toggle quoted message
Show quoted text
Thanks to both of you. I will open it up tomorrow and check out the
attenuator.
On Sat, Feb 25, 2023, 8:47 PM Bruce <bruce@...> wrote:
You would be looking for transmission loss. If there is a problem
with the attenuator, it could present very high attenuation.
Safest way to do this is to set the SA for zero attenuation, connect a
power meter or another SA to the OUTPUT of the 1st SA attenuator and a
signal generator to the input of the attenuator. Then cycle the
attenuator through its various attenuation settings and record the
results.
A bit more dangerous method would be to inject a signal AFTER the
input attenuator (I would start quite low -30 or -40 dBm) and see how
the SA performs.
If the measured attenuation differs from the actual attenuation
significantly, you have a problem.
Cheers!
Bruce
Quoting Rik Rasmussen <rikrasmussen@...>:
Are you talking about the HP 8594E's INPUT attenuator? What would I
be looking for?
|
Re: 9000 series 700 with basic in lieu of series 300 ?
I cannot recall exactly at what point you could not install HP Basic on the series 9000/300 series or the 9000/700 series, as the operation system (OS), but beyond that point you simply installed HP-UX as the OS, and HP Basic as an app on the system, then 11990 app on that, given sufficient memory, drivers, and necessary interfaces. So I think you might be able to install HP Basic on your 9000/715¡¯s, if not you install HP-UX, then HP Basic, then 11990.
HP repair and calibration used the 9000/300 systems up to the mid 80¡¯s and migrated onto the 9000/700¡¯s (712¡¯s ->720¡¯s->750¡¯s) after that, then finally to HP C110 workstations. Everything after that was on PC laptops using HT Basic as an app.
Don Bitters
|
Re: HP8510 push-button spring
I had a 3456A missing five springs, four were recovered from inside
the case but the fifth had on AWOL, a bit of 2 thou shim steel
(El-cheapo Chinese feeler gauges are a good source) and some tin
snips is how I made one.
I then some years later I had to open it up again and there was
number five right at the back of the case stuck on the rear panel
just smirking at me.......
Adrian
On 26/02/2023 11:56, tom_iphi via
groups.io wrote:
toggle quoted message
Show quoted text
Hi,
while cleaning the keyboard of my 85101C display section, I have
noticed that one key is missing a spring.
It seems to have disappeared. It is not found in the unit.
Does anybody know where I could obtain one of these or does
someone have a spare he would sell?
Thanks,
Tom DG8SAQ
|
Re: looking for HP85101C display keys plastic frame
Hi Tom,
You are correct, I'm in Norway. But we are not fully EU member, only 'EU-light'. Where are you located? Thanks, that would be very kind of you. We could take that off list I guess.
I guess the push button spring, is the flat thin springs that can jump out of the switches? You can make new ones from a feeler gauge, just find the right thickness and cut with scissors. You can search the group archive, I think it has been discussed many times.
Regards, Askild
Hi Askild,
I have asked, they don't combine shipment of PCB and 3d print orders, they said the taxes are different.
I got four made and that was 12 Euros including shipment. I can send you one of them if you like.
You are in Skandinavia, I think, EU country?
The only problem left is to find a replacement for a missing pushbutton spring...
Best regards,
Tom DG8SAQ
|
Re: looking for HP85101C display keys plastic frame
Hi Askild,
I have asked, they don't combine shipment of PCB and 3d print orders, they said the taxes are different.
I got four made and that was 12 Euros including shipment. I can send you one of them if you like.
You are in Skandinavia, I think, EU country?
The only problem left is to find a replacement for a missing pushbutton spring...
Best regards,
Tom DG8SAQ
|
Re: looking for HP85101C display keys plastic frame
Hi Tom,
Yes they really look great. Much better than my FDM printed one.
I might just have to get some printed at JLC as well. Maybe I can combine it with PCBs next time I order.
Hope it fits well.
Regards, Askild
Hi folks,
the resin printed parts from JLCPCB have arrived and I am positively surprised about the nice surface finish, see picture.
Thanks again for all the great advice!
Best regards,
Tom
|
Re: looking for HP85101C display keys plastic frame
Hi folks, the resin printed parts from JLCPCB have arrived and I am positively surprised about the nice surface finish, see picture. Thanks again for all the great advice! Best regards, Tom 
|
HP8510 push-button spring
Hi,
while cleaning the keyboard of my 85101C display section, I have noticed that one key is missing a spring.
It seems to have disappeared. It is not found in the unit.
Does anybody know where I could obtain one of these or does someone have a spare he would sell?
Thanks,
Tom DG8SAQ
|
hello group,
here I have just transformed my HP 8593e by equipping it with the LODA and its control card for me to be able to use my HP 85644A tracking with my analyzer is it necessary or is there a setting to be made for the LODA voltage because I have accidents on my frequency response curve, should the voltages written on the LODA (bias and LO SENSE) be adjusted??.
Thank you very much ONN4LJ Joel
|
Re: HP 83595A Repair Advice
Hi Mark,
Glad to help, hope you get it up and running soon. Just guessing here, but maybe your A6 board has at some point been next to another A5 board where a tantalum capacitor has decided to become a rocket engine.
I'm also looking for a reasonably priced plugin for my 8350 / 8757 combo that go higher in frequency than my current one, 8.6GHz 83525A-H03. But I'm in Norway, and usually what I find on ebay does not ship here, or the shipping is very high. I'm not in a hurry, so I will just keep looking.
Regards, Askild
On Sat, Feb 25, 2023 at 7:19?PM Mark Bielman < mbielman@...> wrote:
Askild,
You were right! That cap is clearly bad. And the associated inductor looks damaged as well.
Surprised I missed it. Not sure how the black stuff made its way to where it did.
New parts ordered and on the way.
Thanks!
Mark
Hi Mark,
Yes, the one just next to the orange tantalum.
Just measure them with an ohm meter. Usually they will be short circuit or very low ohms if they are bad.
Regards,
Askild
On Sat, Feb 25, 2023 at 1:42 AM Mark Bielman < mbielman@...> wrote:
Hi Askild,
Do you mean the brown one? (left of the inductors)
There are many of these and are different values (and color) than the orange ones.
Also, the burn marks are on the opposite corner of the board.
I'm really confused by this one. I cannot find anything that could have caused this.
Maybe something fell into the unit and caused a short? Really do not know.
Have some new caps coming and I will replace those orange ones just in case.
(maybe the brown ones too)
Mark
Hi Mark,
I did see this listing on ebay.
I did notice in one image from the listing, what looks as a bad tantalum capacitor
The top one in the lower left corner, looks as it has failed and been hot.
Regards,
Askild
On Fri, Feb 24, 2023 at 2:30 AM Blue Smoke Lee < pedlowl@...> wrote:
It¡¯s not that tantalum¡¯s are unreliable, to the contrary, they are very reliable and have long service lives.? Instead, it¡¯s all about HOW they fail when they do (eventually) fail.? Most aluminum electrolytics fail by having an increasing esr
value over time up to the ultimate end of life state of an open circuit.
A tantalum capacitor¡¯s failure mode is most often as a short circuit.? If the tantalum cap is in a circuit capable of delivering a reasonable current, that results in fire.? I¡¯ve got a field returned product in front of me right now with exactly that situation.?
It¡¯s a game of probabilities and that¡¯s why some manufacturers have banned them.?
In the older HP gear I restore and collect for a hobby, I see lots of the egg and epoxy axial tantalums go poof. It¡¯s one reason I can get an expensive instrument so cheaply.? In those cases, the current is limited so there is no roaring bonfire, but the
tan egg does let out the blue smoke even though the supply voltage and ripple is entirely within spec. I just repaired a HP-8663 that had a tantalum filter cap in a 15 volt internal module supply bus go poof.?
? "Blue Smoke Lee" <pedlowl@...> writes:
Tantalum caps are notorious for going into ¡°kingsford mode¡±, where they fail exothermically.? Your photo shows characteristics of a tantalum cap failure and it¡¯s one of the
more mild failures I¡¯ve seen.? I recall a case where the tiny tantalum cap burned through multiple Fiberglas layers and torched the opposite side of the board. Many companies
have rules prohibiting the use of tantalum caps in product designs for that same reason.
Tantalums should be very reliable if treated with respect. Specifically,
follow Dave's advice and derate it sufficiently.
I found the following two posts very helpful:
-
/g/HP-Agilent-Keysight-equipment/message/122066
-
/g/HP-Agilent-Keysight-equipment/message/97724
Matt
|
Rik Rasmussen WA4BAN