Re: HP 8644B Option 2 Upgrade
Well, thought Colby had the answer but when I looked and the I/O board does not have U32, I had to look at the service manual again. This is only for 8664/65 models. So stuck again.
Mark
|
Ok, but if the generator goes to 3GHz, you might just be within the high band. My 8562A has high band from 2.75GHz (low band is up to 2.9GHz). It has coax relay between high and low, so it doesn't sweep across, you have to select either low or high band, so that's why there is an overlap..
You might also use another spectrum analyser,? if it has 1st lo-out, that will most likely be in the 3-8GHz range 0, so then set it to zero span, just be aware, the level will be around 15dbm.
Askild
On Sun, 14 Apr 2019, 12:01 Dave Ireland, < mirrors@...> wrote:
Askild
Unfortunately the signal generator I
have goes to 3GHz and also the Spectrum Analyser also only goes to
3GHz (found an Anritsu Spectrum Master I forgot was a SA as well).
I did inject 310.7MHz at -31dBm into
the third mixer I get a trace on the screen at -32dBm.
Amir doesn't think that is correct but
I can't find anything in the manual to what it should be, rather a
complicated process of measuring RF in circuit levels which I
don't have the way to do accurately at the moment. As mentioned
earlier the 100MHz oscillator driver has died so I have it set up
on the bench injecting 100MHz into oscillator output (50ohm point)
and 600MHz into the second converter.
I think my focus at the moment is to
get the 100MHz back and then find out why the 600MHz after the
trippler and doubler is low output so I can fault find the rest of
the unit with out all this other gear hanging on it. Then see If I
can borrow a decent active probe.
Dave
On 14/04/2019 2:04 am, Askild wrote:
Hi
Do you have access to a signal generator >3GHz?
It would help a lot, because that can help you to figure
out where the problem is.
So if low band is low and high band is ok, the failure is
most likely low band first mixer or second converter, or any
signal/power feeding them.
If both low and high band is low, the you should check the
IF path, by injecting a signal into the third mixer and
checking what level you get on the display, compared to what
it should be compared to the service manual.
You might of course have more than one failure...
I once fixed a 8563A that had low level on only the low
band, problem was in the second converter, on the pcb.
When this converter was made, the one soldering the trough
hole components cut the leads after soldering, and cutting
down into the solder, over the years the stress on the
solderings made them break.. Would think that HP would know
better.. At least when I was trained for soldering
certificate, I was told this was a big NO-NO. If you did do
this, you should reflow.
Best of luck
Askild
oh,
was it heating up? had you measured the bias voltage and DC
current draw in that MMIC?
that usually (not always as I learned recently in repairing a
R&S SME) gives you enough clue about the state the MMIC is
in when you compare the values with datasheet
On Sat, Apr 13, 2019 at 11:02 AM, Dave Ireland wrote:
Amir
?
Haven't
done much to day but I am starting to get obsessed again
about the 100MHz level as when looking at the circuit
third converter for the reason that might be 20dB down I
realised that the local oscillator for the third converter
is 300MHz derived just from the 100Mhz tripler.
?
I
was just about to try injecting the 600Mhz from a sig gen
and try selecting SIG ID which appears to switch in a
298MHz LO instead of the 300MHz to see if it makes up the
24dB but as fate would have it the MMIC in the 100Mhz
oscillator has died and now the 300MHz has disappeared.
?
Off
to order the MMIC.
?
Dave
?
?
?
On
13/04/2019 7:44 pm, amirb wrote:
When working on RF circuits only a proper RF
active probe such as 85024A (ideal) or 1152A or 54701A or
etc....will work
forget about passive probes unless accurate amplitude
measurement is not your concern and you just want to see
the signal
even for that purpose sometimes passive probes load your
oscillator (depending on where you probe it) such that it
stops oscillating!
The total system bandwidth (probe + scope or SA) is the
key parameter.
Anyways, how do you measure the 600MHz? Do you have the
proper connector (SMB or SMC i forgot) and do you measure
it with 50 Ohm SA?
There are a couple of MMIC ampifiers in the final stages
of the 600MHz REF, I would check them for DC bias voltage,
DC current draw and? temperature to see if they are dead
of course with an active RF probe you could check their
gain.
I would still focus on the mixers, for now you can just
inject a 600MHz into the second converter
because if the low band mixer or the second converter is
dead they are very expensive to repair or replace
specially the second converter used to cost a fortune as I
recall
and you might find it beyond economic repair. the 600MHz
ref is easy to repair and its cheap. those MMICs can be
found on ebay and they are cheap
or you might conclude that the 1st and 2nd mixers are fine
and there is a problem in the IF and log amp, that is
usually easy and not expensive to repair.
On Fri, Apr 12, 2019 at 08:11 PM, Dave Ireland wrote:
Amir
?
I think you are correct about the 100Mhz. I have
become obsessed with the level without a proper method
of measuring it. All the 300Mhz stuff following is
working so can't be much wrong with it.
?
I am still however worried about the 600Mz level as
this is being measured at the proper alignment point
but another day dawns here in Australia and I have to
leave it for a while.
?
Thanks for your help to date.
?
Dave
?
?
?
Sent
from my Samsung GALAXY S5
-------- Original message --------
From: amirb <amir.borji@...>
Date: 13/04/2019 00:00 (GMT+08:00)
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8560E Low
Level
I still doubt that there is anything wrong with the ref
osc and 600MHz lo, really. it could be just a
measurement trap.
how are you measuring those 9dBm and 17dBm in-circuit
amplitudes? remember that only and only a proper RF
active probe?
with careful grounding can give you the correct value.
Any passive probe even the best 500MHz passive probe
will give you totally out of whack values. I have been
there...
even with an active probe (cap <1-2pF) still the
precise point at which you measure it, is very critical.
If the 100MHz is like 10dB low I am sure
you would be getting errors because the PFD will
probably fail to work and the loop becomes unlocked.?
Moreover, as you explained, even when you inject 600MHz
Lo from the outside your second IF is still 24dB low.
That has nothing to do?
with the ref osc anymore.
another thing that puzzled me is that when you inject
310.7MHz to the third converter you get a trace at
-32dBm? it should be much higher, close to -10dBm I
think.
(adjust your atten to 10dB and ref level to 0dBm) How is
the level of 300MHz third converter LO drive??
your 'default' noise floor is too low (it must be around
-65 to -70dB as I recall from the units that I had in
the past)
usually that means a mixer is not working but it can
have other reasons further down the signal path
Groups.io Links:
You receive all messages sent to this group.
View/Reply Online (#95837) | Reply To Group | Reply To Sender | Mute This Topic | New Topic
Your Subscription |
Contact Group Owner |
Unsubscribe [mirrors@...]
|
Askild
Unfortunately the signal generator I
have goes to 3GHz and also the Spectrum Analyser also only goes to
3GHz (found an Anritsu Spectrum Master I forgot was a SA as well).
I did inject 310.7MHz at -31dBm into
the third mixer I get a trace on the screen at -32dBm.
Amir doesn't think that is correct but
I can't find anything in the manual to what it should be, rather a
complicated process of measuring RF in circuit levels which I
don't have the way to do accurately at the moment. As mentioned
earlier the 100MHz oscillator driver has died so I have it set up
on the bench injecting 100MHz into oscillator output (50ohm point)
and 600MHz into the second converter.
I think my focus at the moment is to
get the 100MHz back and then find out why the 600MHz after the
trippler and doubler is low output so I can fault find the rest of
the unit with out all this other gear hanging on it. Then see If I
can borrow a decent active probe.
Dave
On 14/04/2019 2:04 am, Askild wrote:
toggle quoted message
Show quoted text
Hi
Do you have access to a signal generator >3GHz?
It would help a lot, because that can help you to figure
out where the problem is.
So if low band is low and high band is ok, the failure is
most likely low band first mixer or second converter, or any
signal/power feeding them.
If both low and high band is low, the you should check the
IF path, by injecting a signal into the third mixer and
checking what level you get on the display, compared to what
it should be compared to the service manual.
You might of course have more than one failure...
I once fixed a 8563A that had low level on only the low
band, problem was in the second converter, on the pcb.
When this converter was made, the one soldering the trough
hole components cut the leads after soldering, and cutting
down into the solder, over the years the stress on the
solderings made them break.. Would think that HP would know
better.. At least when I was trained for soldering
certificate, I was told this was a big NO-NO. If you did do
this, you should reflow.
Best of luck
Askild
oh,
was it heating up? had you measured the bias voltage and DC
current draw in that MMIC?
that usually (not always as I learned recently in repairing a
R&S SME) gives you enough clue about the state the MMIC is
in when you compare the values with datasheet
On Sat, Apr 13, 2019 at 11:02 AM, Dave Ireland wrote:
Amir
?
Haven't
done much to day but I am starting to get obsessed again
about the 100MHz level as when looking at the circuit
third converter for the reason that might be 20dB down I
realised that the local oscillator for the third converter
is 300MHz derived just from the 100Mhz tripler.
?
I
was just about to try injecting the 600Mhz from a sig gen
and try selecting SIG ID which appears to switch in a
298MHz LO instead of the 300MHz to see if it makes up the
24dB but as fate would have it the MMIC in the 100Mhz
oscillator has died and now the 300MHz has disappeared.
?
Off
to order the MMIC.
?
Dave
?
?
?
On
13/04/2019 7:44 pm, amirb wrote:
When working on RF circuits only a proper RF
active probe such as 85024A (ideal) or 1152A or 54701A or
etc....will work
forget about passive probes unless accurate amplitude
measurement is not your concern and you just want to see
the signal
even for that purpose sometimes passive probes load your
oscillator (depending on where you probe it) such that it
stops oscillating!
The total system bandwidth (probe + scope or SA) is the
key parameter.
Anyways, how do you measure the 600MHz? Do you have the
proper connector (SMB or SMC i forgot) and do you measure
it with 50 Ohm SA?
There are a couple of MMIC ampifiers in the final stages
of the 600MHz REF, I would check them for DC bias voltage,
DC current draw and? temperature to see if they are dead
of course with an active RF probe you could check their
gain.
I would still focus on the mixers, for now you can just
inject a 600MHz into the second converter
because if the low band mixer or the second converter is
dead they are very expensive to repair or replace
specially the second converter used to cost a fortune as I
recall
and you might find it beyond economic repair. the 600MHz
ref is easy to repair and its cheap. those MMICs can be
found on ebay and they are cheap
or you might conclude that the 1st and 2nd mixers are fine
and there is a problem in the IF and log amp, that is
usually easy and not expensive to repair.
On Fri, Apr 12, 2019 at 08:11 PM, Dave Ireland wrote:
Amir
?
I think you are correct about the 100Mhz. I have
become obsessed with the level without a proper method
of measuring it. All the 300Mhz stuff following is
working so can't be much wrong with it.
?
I am still however worried about the 600Mz level as
this is being measured at the proper alignment point
but another day dawns here in Australia and I have to
leave it for a while.
?
Thanks for your help to date.
?
Dave
?
?
?
Sent
from my Samsung GALAXY S5
-------- Original message --------
From: amirb <amir.borji@...>
Date: 13/04/2019 00:00 (GMT+08:00)
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8560E Low
Level
I still doubt that there is anything wrong with the ref
osc and 600MHz lo, really. it could be just a
measurement trap.
how are you measuring those 9dBm and 17dBm in-circuit
amplitudes? remember that only and only a proper RF
active probe?
with careful grounding can give you the correct value.
Any passive probe even the best 500MHz passive probe
will give you totally out of whack values. I have been
there...
even with an active probe (cap <1-2pF) still the
precise point at which you measure it, is very critical.
If the 100MHz is like 10dB low I am sure
you would be getting errors because the PFD will
probably fail to work and the loop becomes unlocked.?
Moreover, as you explained, even when you inject 600MHz
Lo from the outside your second IF is still 24dB low.
That has nothing to do?
with the ref osc anymore.
another thing that puzzled me is that when you inject
310.7MHz to the third converter you get a trace at
-32dBm? it should be much higher, close to -10dBm I
think.
(adjust your atten to 10dB and ref level to 0dBm) How is
the level of 300MHz third converter LO drive??
your 'default' noise floor is too low (it must be around
-65 to -70dB as I recall from the units that I had in
the past)
usually that means a mixer is not working but it can
have other reasons further down the signal path
Groups.io Links:
You receive all messages sent to this group.
View/Reply Online (#95837) | Reply To Group | Reply To Sender | Mute This Topic | New Topic
Your Subscription |
Contact Group Owner |
Unsubscribe [mirrors@...]
_._,_._,_
|
Re: hp8640B_with_rough_sound
Yes, the problem is solved. The Problem was a short circuit inside the main oscillator to the Varactor lines.
Initially I checked all supplies >> No Problem. Then the voltage that goes to the Varactor.?
This voltage should be about -14V. >> It was NOT correct.
The PLL Filter was correct, all checked. >> Problem in Varactor Unit.
So very important check Varactor and Transistor connections at the Main Oscillator unit BEFORE unmounting.
|
Re: HP 8644B Option 2 Upgrade
I know that unit shares a lot with 8664/8665 units, and if the attenuator is the same, your attenuator cal ROM only has cal data up to ~1GHz. If memory serves me right, it's U32 on the backplane/IO board.
Nothing you can do about it without replacing the attenuator assy with a matching cal ROM, OR getting a copy of one from someone else....but it won't be for your attenuator assy. You MIGHT be able to use one from an 8664 or 8665 just to see if the error goes away.
Back yours up, though.... It is a EEPROM, so you can reuse it.
Colby
On Sat, Apr 13, 2019, 5:34 PM Mark Bielman < mbielman@...> wrote: Hello group. I have an 8644B (standard, upto 1030 MHz) that I wanted to upgrade to 2060 MHz. Managed to locate an upgrade kit w/cables (thanks Steve!) but after installing it I get the following error:
6,040,117. The "6,040" is the key here and the service manual says it's an attenuator problem. (?) The output attenuator is good (repaired and checked) and the doubler board has no attenuators, at least according to the manual.
I did notice that when the frequency goes above 1030 MHz (when the doubler kicks in) the output (mechanical) attenuator also clicks in. Seems odd unless it is adjusting for output levels from the doubler.
The output (when set to 0 dBm) drops about 1.5 dB when this happens.
The doubler board looks used as do the cables so that is most likely the issue, but no idea what to look at. The manual is a bit confusing, especially for HP.
If anyone has ideas, I would appreciate it!
Mark
PS - I have this unit for sale on ebay and was hoping to have option 002 as an added bonus, but perhaps not. :-(
|
Re: 141T varying CRT intensity?
Can you set it up to measure the bias voltage (cathode to control grid)
WHILE OPERATING - be good to find out what is changing.
toggle quoted message
Show quoted text
On Sat, April 13, 2019 16:37, saipan59 wrote: I suggest a cold/cracked solder joint could be the issue - influenced by
flexing/shifting when heat builds up. If it "jumps around" between good
and bad, that might rule out a component that has drifted out of spec or
is temp-sensitive.? And if trace brightness is the only thing affected,
then there are only a very limited number of parts that are in the right
area. I doubt that the CRT or electrolytics are responsible for the
behavior you describe.
Pete
|
HP 8644B Option 2 Upgrade
Hello group. I have an 8644B (standard, upto 1030 MHz) that I wanted to upgrade to 2060 MHz. Managed to locate an upgrade kit w/cables (thanks Steve!) but after installing it I get the following error:
6,040,117. The "6,040" is the key here and the service manual says it's an attenuator problem. (?) The output attenuator is good (repaired and checked) and the doubler board has no attenuators, at least according to the manual.
I did notice that when the frequency goes above 1030 MHz (when the doubler kicks in) the output (mechanical) attenuator also clicks in. Seems odd unless it is adjusting for output levels from the doubler.
The output (when set to 0 dBm) drops about 1.5 dB when this happens.
The doubler board looks used as do the cables so that is most likely the issue, but no idea what to look at. The manual is a bit confusing, especially for HP.
If anyone has ideas, I would appreciate it!
Mark
PS - I have this unit for sale on ebay and was hoping to have option 002 as an added bonus, but perhaps not. :-(
|
Re: HP 85044A T/R test set part - Cabinet Trim "zipper" - 08502-20007 - source or 3D printed copy?
Dave, Is it possible for you to share a photo of the case where the zipper locks fit? ?The service manual on the Keysight site is scanned from a poor microfiche.
toggle quoted message
Show quoted text
On Apr 13, 2019, at 16:31, Tam Hanna < tamhan@...> wrote:
Hello, just send what you need by registered mail to my office in Budapest. But: A) my oven can not handle large pieces B) I am semi-retired. So things can take time, as I have so many jobs...
With best regards Tam HANNA (emailing on a BlackBerry PRIV) Enjoy electronics? Join 14k other followers by visiting the Crazy Electronics Lab at Am 13. April 2019 23:22:10 MESZ schrieb Harvey White < madyn@...>:
On Sat, 13 Apr 2019 22:20:01 +0200, you wrote:
Harvey,
try to "temper" the PLA. Use a reflow oven.
I can try that. Right now, I'm working on making a board positioner
for the binocular microscope that reads positions of parts from the
board file.
I don't have any cases that need work until I start to make cases for
stuff myself.
I have heard of smoothing the PLA by putting it in acetone vapors for
a bit. That always sounded rather dangerous to me, and then again,
how long to leave it in is another matter. Hmmm, acetone vapors in a
glass jar....
I may forego that experiment.
I'll wait and see what I need, and thanks for the offer.
Harvey
If you want, you can send me some samples to HU and I will use my
"Wassulmayer" oven.
- - - - -
With best regards
Tam HANNA
Enjoy electronics? Join 13500 followers by visiting the Crazy Electronics Lab at
On 13.04.19 21:55, Harvey White wrote:
On Sat, 13 Apr 2019 10:58:33 -0700, you wrote:
I don't know. I'm fairly sure the OEM part was plastic. The parts don't seem to be under much stress, and the stress is distributed along the length. I've never done 3D printing, but a few components I've encountered that are printed seem to be made of material that would be strong enough. There doesn't seem to be any other practical way to keep the cabinet halves together, so I'm pursuing duplication of the zippers. Any other idea would be appreciated!
3D printer materials vary widely in strength.
PLA is pretty generic, ABS is available, some other varieties are
available as well as flexible filament.
You'd need a heated bed printer for some of them, though.
I'd be tempted to try a good ol' cheap print out of PLA to get the
sizes right, then see how that works.
Infill, wall size, and layer thickness also change strength of the
part as well.
PLA is cheap, but not very rugged, and definitely not as heat
resistant as you'd like, perhaps.
Harvey
Links: You receive all messages sent to this group.
View/Reply Online (#95841): /g/HP-Agilent-Keysight-equipment/message/95841
Mute This Topic: /mt/31070989/102401
Group Owner: [email protected]
Unsubscribe: /g/HP-Agilent-Keysight-equipment/leave/defanged [tamhan@...]
|
Re: 141T varying CRT intensity?
...And it occurs to me that the Brightness pot would be the first thing to look at, since it has 'moving parts', a wiper, etc.
If the pot is proven to be OK, then use it as a starting point, and investigate anything that is closely connected to the pot in the schematic.
Pete
|
Re: 141T varying CRT intensity?
I suggest a cold/cracked solder joint could be the issue - influenced by flexing/shifting when heat builds up. If it "jumps around" between good and bad, that might rule out a component that has drifted out of spec or is temp-sensitive.?
And if trace brightness is the only thing affected, then there are only a very limited number of parts that are in the right area.
I doubt that the CRT or electrolytics are responsible for the behavior you describe.
Pete
|
Re: hp3586a selective voltmeter jammed
Yes, my copy is poor as well.? I suppose you can check the startup circuit near U5 to start.
Maybe Artek has a decent scan.
Peter
toggle quoted message
Show quoted text
On 4/13/2019 6:02 PM, cc@... wrote: Yes, I just fixed the power supply some days ago and the three voltages (+5, +12V, -12V) sounds good even though I didn't check the ripple. I have the service manual the microprocessor parts unfortunately are a little bit blurred.
Thank you,
Ciro
|
Re: hp3586a selective voltmeter jammed
Yes, I just fixed the power supply some days ago and the three voltages (+5, +12V, -12V) sounds good even though I didn't check the ripple. I have the service manual the microprocessor parts unfortunately are a little bit blurred.
Thank you,
Ciro
|
Re: HP 85044A T/R test set part - Cabinet Trim "zipper" - 08502-20007 - source or 3D printed copy?
Hello, just send what you need by registered mail to my office in Budapest. But: A) my oven can not handle large pieces B) I am semi-retired. So things can take time, as I have so many jobs...
With best regards Tam HANNA (emailing on a BlackBerry PRIV) Enjoy electronics? Join 14k other followers by visiting the Crazy Electronics Lab at Am 13. April 2019 23:22:10 MESZ schrieb Harvey White <madyn@...>:
toggle quoted message
Show quoted text
On Sat, 13 Apr 2019 22:20:01 +0200, you wrote:
Harvey,
try to "temper" the PLA. Use a reflow oven.
I can try that. Right now, I'm working on making a board positioner
for the binocular microscope that reads positions of parts from the
board file.
I don't have any cases that need work until I start to make cases for
stuff myself.
I have heard of smoothing the PLA by putting it in acetone vapors for
a bit. That always sounded rather dangerous to me, and then again,
how long to leave it in is another matter. Hmmm, acetone vapors in a
glass jar....
I may forego that experiment.
I'll wait and see what I need, and thanks for the offer.
Harvey
If you want, you can send me some samples to HU and I will use my
"Wassulmayer" oven.
- - - - -
With best regards
Tam HANNA
Enjoy electronics? Join 13500 followers by visiting the Crazy Electronics Lab at
On 13.04.19 21:55, Harvey White wrote:
On Sat, 13 Apr 2019 10:58:33 -0700, you wrote:
I don't know. I'm fairly sure the OEM part was plastic. The parts don't seem to be under much stress, and the stress is distributed along the length. I've never done 3D printing, but a few components I've encountered that are printed seem to be made of material that would be strong enough. There doesn't seem to be any other practical way to keep the cabinet halves together, so I'm pursuing duplication of the zippers. Any other idea would be appreciated!
3D printer materials vary widely in strength.
PLA is pretty generic, ABS is available, some other varieties are
available as well as flexible filament.
You'd need a heated bed printer for some of them, though.
I'd be tempted to try a good ol' cheap print out of PLA to get the
sizes right, then see how that works.
Infill, wall size, and layer thickness also change strength of the
part as well.
PLA is cheap, but not very rugged, and definitely not as heat
resistant as you'd like, perhaps.
Harvey
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#95841): /g/HP-Agilent-Keysight-equipment/message/95841
Mute This Topic: /mt/31070989/102401
Group Owner: [email protected]
Unsubscribe: /g/HP-Agilent-Keysight-equipment/leave/defanged [tamhan@...]
|
Re: hp3586a selective voltmeter jammed
It looks like a digital control problem.? Check the power supplies for voltage and ripple first.? Do you have the service manual?
toggle quoted message
Show quoted text
On 4/13/2019 4:42 PM, cc@... wrote: Hello,
have you ever experienced this fault?
This unit is now stuck at this leve. Earlier it was jammed as well but only with annunciator LEDs (no LED display) then I moved and disconnected the front panel A98 and I don't know why something changed like shown in this video.
I tried with an oscilloscope but in the most of cases I only notice a single change of a digital state level and nothing else¡ it is so hard to troubleshoot this type of signal :-( and also very hard to find spare parts.
Thank you
Ciro
|
Re: HP 85044A T/R test set part - Cabinet Trim "zipper" - 08502-20007 - source or 3D printed copy?
On Sat, 13 Apr 2019 22:20:01 +0200, you wrote: Harvey,
try to "temper" the PLA. Use a reflow oven.
I can try that. Right now, I'm working on making a board positioner for the binocular microscope that reads positions of parts from the board file. I don't have any cases that need work until I start to make cases for stuff myself. I have heard of smoothing the PLA by putting it in acetone vapors for a bit. That always sounded rather dangerous to me, and then again, how long to leave it in is another matter. Hmmm, acetone vapors in a glass jar.... I may forego that experiment. I'll wait and see what I need, and thanks for the offer. Harvey
If you want, you can send me some samples to HU and I will use my
"Wassulmayer" oven.
- - - - -
With best regards
Tam HANNA
Enjoy electronics? Join 13500 followers by visiting the Crazy Electronics Lab at
On 13.04.19 21:55, Harvey White wrote:
On Sat, 13 Apr 2019 10:58:33 -0700, you wrote:
I don't know. I'm fairly sure the OEM part was plastic. The parts don't seem to be under much stress, and the stress is distributed along the length. I've never done 3D printing, but a few components I've encountered that are printed seem to be made of material that would be strong enough. There doesn't seem to be any other practical way to keep the cabinet halves together, so I'm pursuing duplication of the zippers. Any other idea would be appreciated! 3D printer materials vary widely in strength.
PLA is pretty generic, ABS is available, some other varieties are
available as well as flexible filament.
You'd need a heated bed printer for some of them, though.
I'd be tempted to try a good ol' cheap print out of PLA to get the
sizes right, then see how that works.
Infill, wall size, and layer thickness also change strength of the
part as well.
PLA is cheap, but not very rugged, and definitely not as heat
resistant as you'd like, perhaps.
Harvey
|
hp3586a selective voltmeter jammed
Hello,
?
have you ever experienced this fault?
This unit is now stuck at this leve. Earlier it was jammed as well but only with annunciator LEDs (no LED display) then I moved and disconnected the front panel A98 and I don't know why something changed like shown in this video.
I tried with an oscilloscope but in the most of cases I only notice a single change of a digital state level and nothing else¡ it is so hard to troubleshoot this type of signal :-( and also very hard to find spare parts. Thank you Ciro
|
Re: HP 85044A T/R test set part - Cabinet Trim "zipper" - 08502-20007 - source or 3D printed copy?
Harvey,
try to "temper" the PLA. Use a reflow oven.
If you want, you can send me some samples to HU and I will use my "Wassulmayer" oven.
- - - - -
With best regards
Tam HANNA
Enjoy electronics? Join 13500 followers by visiting the Crazy Electronics Lab at
toggle quoted message
Show quoted text
On 13.04.19 21:55, Harvey White wrote: On Sat, 13 Apr 2019 10:58:33 -0700, you wrote:
I don't know. I'm fairly sure the OEM part was plastic. The parts don't seem to be under much stress, and the stress is distributed along the length. I've never done 3D printing, but a few components I've encountered that are printed seem to be made of material that would be strong enough. There doesn't seem to be any other practical way to keep the cabinet halves together, so I'm pursuing duplication of the zippers. Any other idea would be appreciated! 3D printer materials vary widely in strength.
PLA is pretty generic, ABS is available, some other varieties are
available as well as flexible filament.
You'd need a heated bed printer for some of them, though.
I'd be tempted to try a good ol' cheap print out of PLA to get the
sizes right, then see how that works.
Infill, wall size, and layer thickness also change strength of the
part as well.
PLA is cheap, but not very rugged, and definitely not as heat
resistant as you'd like, perhaps.
Harvey
|
Re: HP 85044A T/R test set part - Cabinet Trim "zipper" - 08502-20007 - source or 3D printed copy?
On Sat, 13 Apr 2019 10:58:33 -0700, you wrote: I don't know. I'm fairly sure the OEM part was plastic. The parts don't seem to be under much stress, and the stress is distributed along the length. I've never done 3D printing, but a few components I've encountered that are printed seem to be made of material that would be strong enough. There doesn't seem to be any other practical way to keep the cabinet halves together, so I'm pursuing duplication of the zippers. Any other idea would be appreciated! 3D printer materials vary widely in strength. PLA is pretty generic, ABS is available, some other varieties are available as well as flexible filament. You'd need a heated bed printer for some of them, though. I'd be tempted to try a good ol' cheap print out of PLA to get the sizes right, then see how that works. Infill, wall size, and layer thickness also change strength of the part as well. PLA is cheap, but not very rugged, and definitely not as heat resistant as you'd like, perhaps. Harvey
|
Hi Do you have access to a signal generator >3GHz? It would help a lot, because that can help you to figure out where the problem is. So if low band is low and high band is ok, the failure is most likely low band first mixer or second converter, or any signal/power feeding them.
If both low and high band is low, the you should check the IF path, by injecting a signal into the third mixer and checking what level you get on the display, compared to what it should be compared to the service manual.
You might of course have more than one failure...
I once fixed a 8563A that had low level on only the low band, problem was in the second converter, on the pcb.
When this converter was made, the one soldering the trough hole components cut the leads after soldering, and cutting down into the solder, over the years the stress on the solderings made them break.. Would think that HP would know better.. At least when I was trained for soldering certificate, I was told this was a big NO-NO. If you did do this, you should reflow.
Best of luck Askild
oh, was it heating up? had you measured the bias voltage and DC current draw in that MMIC?
that usually (not always as I learned recently in repairing a R&S SME) gives you enough clue about the state the MMIC is in when you compare the values with datasheet
On Sat, Apr 13, 2019 at 11:02 AM, Dave Ireland wrote:
Amir
?
Haven't done much to day but I am starting to get obsessed again about the 100MHz level as when looking at the circuit third converter for the reason that might be 20dB down I realised that the local oscillator for the third converter is 300MHz derived just from the 100Mhz tripler.
?
I was just about to try injecting the 600Mhz from a sig gen and try selecting SIG ID which appears to switch in a 298MHz LO instead of the 300MHz to see if it makes up the 24dB but as fate would have it the MMIC in the 100Mhz oscillator has died and now the 300MHz has disappeared.
?
Off to order the MMIC.
?
Dave
?
?
?
On 13/04/2019 7:44 pm, amirb wrote:
When working on RF circuits only a proper RF active probe such as 85024A (ideal) or 1152A or 54701A or etc....will work
forget about passive probes unless accurate amplitude measurement is not your concern and you just want to see the signal
even for that purpose sometimes passive probes load your oscillator (depending on where you probe it) such that it stops oscillating!
The total system bandwidth (probe + scope or SA) is the key parameter.
Anyways, how do you measure the 600MHz? Do you have the proper connector (SMB or SMC i forgot) and do you measure it with 50 Ohm SA?
There are a couple of MMIC ampifiers in the final stages of the 600MHz REF, I would check them for DC bias voltage, DC current draw and? temperature to see if they are dead
of course with an active RF probe you could check their gain.
I would still focus on the mixers, for now you can just inject a 600MHz into the second converter
because if the low band mixer or the second converter is dead they are very expensive to repair or replace specially the second converter used to cost a fortune as I recall
and you might find it beyond economic repair. the 600MHz ref is easy to repair and its cheap. those MMICs can be found on ebay and they are cheap
or you might conclude that the 1st and 2nd mixers are fine and there is a problem in the IF and log amp, that is usually easy and not expensive to repair.
On Fri, Apr 12, 2019 at 08:11 PM, Dave Ireland wrote:
Amir
?
I think you are correct about the 100Mhz. I have become obsessed with the level without a proper method of measuring it. All the 300Mhz stuff following is working so can't be much wrong with it.
?
I am still however worried about the 600Mz level as this is being measured at the proper alignment point but another day dawns here in Australia and I have to leave it for a while.
?
Thanks for your help to date.
?
Dave
?
?
?
Sent from my Samsung GALAXY S5
-------- Original message --------
From: amirb <amir.borji@...>
Date: 13/04/2019 00:00 (GMT+08:00)
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] HP8560E Low Level
I still doubt that there is anything wrong with the ref osc and 600MHz lo, really. it could be just a measurement trap.
how are you measuring those 9dBm and 17dBm in-circuit amplitudes? remember that only and only a proper RF active probe?
with careful grounding can give you the correct value. Any passive probe even the best 500MHz passive probe will give you totally out of whack values. I have been there...
even with an active probe (cap <1-2pF) still the precise point at which you measure it, is very critical. If the 100MHz is like 10dB low I am sure
you would be getting errors because the PFD will probably fail to work and the loop becomes unlocked.?
Moreover, as you explained, even when you inject 600MHz Lo from the outside your second IF is still 24dB low. That has nothing to do?
with the ref osc anymore.
another thing that puzzled me is that when you inject 310.7MHz to the third converter you get a trace at -32dBm? it should be much higher, close to -10dBm I think.
(adjust your atten to 10dB and ref level to 0dBm) How is the level of 300MHz third converter LO drive??
your 'default' noise floor is too low (it must be around -65 to -70dB as I recall from the units that I had in the past)
usually that means a mixer is not working but it can have other reasons further down the signal path
On Fri, Apr 12, 2019 at 11:08 AM, Dave Ireland wrote:
John
?
Just some feed back. Tracked down the manufacturer of the crystal, Croven Crystals in Canada who are very helpful, and found the replacement crystal is the exact unit and even has the the HP number 0410-4065 lasered on the side where I didn't notice it.
?
Might just go away now and brush up on my oscillator theory to try to get more level so I can drive U700 with +9dBm instead of the -2dBm I have now.
?
Thanks for you assistance.
?
Dave
?
On 12/04/2019 5:27 pm, Dave Ireland wrote:
John
?
I am still back at U700 which seems to be working but I can't get enough drive into it.
?
The oscillator circuit is not tuning properly and I suspect it is the crystal. It has been changed sometime in the past and I not sure characteristics of the crystal are the same. I will see if I can trace? details of what has been put in it.
?
Thanks
?
Dave
?
?
?
On 12/04/2019 1:55 pm, johncharlesgord via Groups.Io wrote:
Dave,
I think I had an HP856x with a similar 600 MHz problem.? It turned out to be a failed RF amp on A15. (U501?? U701?? U702?)
--John Gord
On Thu, Apr 11, 2019 at 06:16 PM, Dave Ireland wrote:
Amir
?
A lesson in the need for methodical fault finding techniques methinks.
?
Checking out the RF section block diagram, trying to follow your advice, I noticed the 600Mhz drive into the Second Converter. The 600Mhz is derived from the 100MHz VCXO that the regulator transistor had failed on. The 600MHz out of A15 is about 15dB low and when I injected 600Mhz at the right level into the second converter the level has come right up. Still about 20dB down but more promising.
?
I will go back now and go through the A15 board and get the outputs correct and then after that maybe check out A14 for the mixer and converter bias volts.
?
Thanks
?
Dave
?
?
On 11/04/2019 10:29 pm, amirb wrote:
I forgot to ask, what is your noise floor level at full span?
On Thu, Apr 11, 2019 at 10:27 AM, amirb wrote:
I assume there are no error messages appearing anymore, right? Any error message about Log amp or step gain amplifiers?
if there is no error and you are sure the attenuator and the input AC/DC coupling are ok (you can even check with oscilloscope by injecting a signal)
then the best and perhaps only way to isolate the problem is by injecting signal at various points and measuring power levels according to the service manual
I dont think this is due to low LO amplitude in any of the converters. otherwise you would get an error message and besides
90dB down is way too much for that. I suspect this is due to a fault in one of the 3 converters or in the Log amp (but not in any of the LO drives)
since you dont have a second spectrum analyzer, I suggest you inject a signal at say -20 or -21dB to the first mixer (make sure span is set to zero)
and then measure the 310.7MHz second IF with an oscilloscope (I hope you have a 400-500MHz scope?) or even a power meter.
if the amplitude is way off then the problem is either the first mixer or the second converter. By the way, you can check the mixer bias voltages coming from A14
In some versions even the first mixer has bias voltage.
Next you can inject a clean 310.7MHz at around -35 or -36dBm to the third IF input (you might have it on the front panel) (you must choose it first in the menus and also choose zero span)
and then you can measure the 10.7 Mhz final IF with an oscilloscope it must be around -15 to -17dBm I think.
On Thu, Apr 11, 2019 at 09:44 AM, Dave Ireland wrote:
I am repairing a HP8560E and until now have been doing well.
The unit on delivery had the classic distorted display and no apparent response to any input though it did sweep it was very slow. 12 Alarms were reported.
Choose the "600 MHz Reference Oscillator Unlocked" first and replacing failed 2N2222A regulator on the 100MHz oscillator which bought the display back and all the alarms disappeared. Best 40c I've ever spent.
Was it completely fixed? No unfortunately not, the input reads about 90dB low. Every thing else looks fine, the resolution band width, span and frequency accuracy all look good. Checked the attenuator and the filter and even went straight into the RF port on the first low band mixer but still 90dB down.
I am at an impass now until I can borrow a decent SA as I do not have anything to measure 3-6GHz.
I did the LO feed through test in the service manual ie centre freq 0Hz, Span 1MHz, input attenuator 0dB and it says the feed through should be between -6 and -30dBm. It says if it is inside this range the RF path following the Low band mixer is operating properly.
Mine is -26dBm, so in the range.
My question is does that test mean that the the Yig, the LO distribution amp, the first mixer LO and IF ports and all the following mixers are operating correctly albeit not 90dB out? If so, the only thing I can think is a faulty RF port on the mixer (sounds unlikely). Also, I suppose if the LO drive was low, maybe 10 dB or so, there would still be feed through but may be not enough level to bias the mixer to mix.
Any though while I wait for a better spec an?
Dave
?
?
?
?
?
|
Re: HP 85044A T/R test set part - Cabinet Trim "zipper" - 08502-20007 - source or 3D printed copy?
I don't know. I'm fairly sure the OEM part was plastic. The parts don't seem to be under much stress, and the stress is distributed along the length. I've never done 3D printing, but a few components I've encountered that are printed seem to be made of material that would be strong enough. There doesn't seem to be any other practical way to keep the cabinet halves together, so I'm pursuing duplication of the zippers. Any other idea would be appreciated!
|
Armin Roesch
saipan59 (Pete)