¿ªÔÆÌåÓý

  1. HP-Agilent-Keysight-Equipment
  2. Messages
Expanded Topics Messages
Date
Re: Fixing a sick 4332A LCR meter

 

Just a repair note:

I have seen a lot of TO92 packaged transistors where
one or more lead has become intermittent. I believe
this is caused by fiddle fingered folks that just have
to push the transistor around to match their ideal of
position, or to read the number.

It always seems to happen when the transistor is mounted
with the leads in a tripod position with the base lead
bent forwards, or backwards.

Touch, or otherwise move the transistor, and it quits...
touch it again, and it works.

-Chuck Harris

[email protected] wrote:

Joel,

Thanks for the detailed steps on that.

I believe I found the smoking gun, a cold solder joint. I removed A1Q1 to test it out of circuit, which really didn't lead to anything. But it started working when I re-soldered it. I'm going to carry out more testing now and see if I can do the alignment (as best I can anyway, as I don't yet have the standard cap and inductor that are called for.) I think it makes sense now...if this FET wasn't switching correctly I think it would impact the operation of the oscillator.

Sean

On Fri, Apr 3, 2020 at 12:39 PM, Joel Setton wrote:


Sean,

I think we're making progress !

In Capacitance mode, the oscillator should run at 100 kHz in the left-hand
side of the range switch, and at 1kHz on the right-hand ranges. The
frequency switch is driven by relay K1.
Your voltage measurements on both sides of R25 show that when the
oscillator is stopped, the AGC loop (IC2, IC3, Q1) is rightly "attempting"
to restart it by increasing the gain of IC1, i.e. driving A1Q1 to a low
resistance. But A1Q1 is in variable-resistance mode only if its gate
voltage is positive, hence the 2.4V ideal value.

One thing that's puzzling me is that the oscillator should run at 1kHz or
100 kHz, but never at 12 kHz.Your measurements show that it runs at 12 kHz
when it should be doing 1kHz, on the "right-hand" C ranges. Let's see what
failure could increase frequency from an ideal 1kHz to 12 kHz. The 1kHz
frequency is set by R2, R3, C3 and C5. Testing the resistors should be
easy with an ohmmeter, and I would think they are OK, because if one of
them had failed the oscillator would still work OK on 100 kHz, using R1
and R4 (but it doesn't).
Testing the capacitors is more difficult. I would think that C3 and C5 are
suspects. You may want to try this: set the meter in Capacitance mode to
the "right" ranges, when the oscillator runs at 12 kHz. Then shunt C3 with
a 1nF capacitor and look at the output frequency (12kHz or 1 kHz ?). Then
remove the 1nF shunt capacitor across C3, and put a 10nF capacitor across
C5. Again, look at the oscillator frequency (12 kHz or 1kHz ?). That will
tell us something.

If one of capacitors C3 or C5 has failed in "open" mode, it would explain
why the oscillator runs at 12 kHz instead of 1 kHz. It would also explain
why it doesn't run at all on the "left-hand" ranges because while it can
do 100kHz, it can't run at 1 MHz. IC1 and IC2 are just too slow for that.

And, as before, DO keep me posted !

Joel
Re: Fixing a sick 4332A LCR meter

 

Joel,

Thanks for the detailed steps on that.

I believe I found the smoking gun, a cold solder joint. I removed A1Q1 to test it out of circuit, which really didn't lead to anything. But it started working when I re-soldered it. I'm going to carry out more testing now and see if I can do the alignment (as best I can anyway, as I don't yet have the standard cap and inductor that are called for.) I think it makes sense now...if this FET wasn't switching correctly I think it would impact the operation of the oscillator.

Sean


On Fri, Apr 3, 2020 at 12:39 PM, Joel Setton wrote:
Sean,

I think we're making progress !

In Capacitance mode, the oscillator should run at 100 kHz in the left-hand side of the range switch, and at 1kHz on the right-hand ranges. The frequency switch is driven by relay K1.
Your voltage measurements on both sides of R25 show that when the oscillator is stopped, the AGC loop (IC2, IC3, Q1) is rightly "attempting" to restart it by increasing the gain of IC1, i.e. driving A1Q1 to a low resistance. But A1Q1 is in variable-resistance mode only if its gate voltage is positive, hence the 2.4V ideal value.

One thing that's puzzling me is that the oscillator should run at 1kHz or 100 kHz, but never at 12 kHz.Your measurements show that it runs at 12 kHz when it should be doing 1kHz, on the "right-hand" C ranges. Let's see what failure could increase frequency from an ideal 1kHz to 12 kHz. The 1kHz frequency is set by R2, R3, C3 and C5. Testing the resistors should be easy with an ohmmeter, and I would think they are OK, because if one of them had failed the oscillator would still work OK on 100 kHz, using R1 and R4 (but it doesn't).
Testing the capacitors is more difficult. I would think that C3 and C5 are suspects. You may want to try this: set the meter in Capacitance mode to the "right" ranges, when the oscillator runs at 12 kHz. Then shunt C3 with a 1nF capacitor and look at the output frequency (12kHz or 1 kHz ?). Then remove the 1nF shunt capacitor across C3, and put a 10nF capacitor across C5. Again, look at the oscillator frequency (12 kHz or 1kHz ?). That will tell us something.

If one of capacitors C3 or C5 has failed in "open" mode, it would explain why the oscillator runs at 12 kHz instead of 1 kHz. It would also explain why it doesn't run at all on the "left-hand" ranges because while it can do 100kHz, it can't run at 1 MHz. IC1 and IC2 are just too slow for that.

And, as before, DO keep me posted !

Joel
Re: Fixing a sick 4332A LCR meter

 

Sean,

I think we're making progress !

In Capacitance mode, the oscillator should run at 100 kHz in the left-hand side of the range switch, and at 1kHz on the right-hand ranges. The frequency switch is driven by relay K1.
Your voltage measurements on both sides of R25 show that when the oscillator is stopped, the AGC loop (IC2, IC3, Q1) is rightly "attempting" to restart it by increasing the gain of IC1, i.e. driving A1Q1 to a low resistance. But A1Q1 is in variable-resistance mode only if its gate voltage is positive, hence the 2.4V ideal value.

One thing that's puzzling me is that the oscillator should run at 1kHz or 100 kHz, but never at 12 kHz.Your measurements show that it runs at 12 kHz when it should be doing 1kHz, on the "right-hand" C ranges. Let's see what failure could increase frequency from an ideal 1kHz to 12 kHz. The 1kHz frequency is set by R2, R3, C3 and C5. Testing the resistors should be easy with an ohmmeter, and I would think they are OK, because if one of them had failed the oscillator would still work OK on 100 kHz, using R1 and R4 (but it doesn't).
Testing the capacitors is more difficult. I would think that C3 and C5 are suspects. You may want to try this: set the meter in Capacitance mode to the "right" ranges, when the oscillator runs at 12 kHz. Then shunt C3 with a 1nF capacitor and look at the output frequency (12kHz or 1 kHz ?). Then remove the 1nF shunt capacitor across C3, and put a 10nF capacitor across C5. Again, look at the oscillator frequency (12 kHz or 1kHz ?). That will tell us something.

If one of capacitors C3 or C5 has failed in "open" mode, it would explain why the oscillator runs at 12 kHz instead of 1 kHz. It would also explain why it doesn't run at all on the "left-hand" ranges because while it can do 100kHz, it can't run at 1 MHz. IC1 and IC2 are just too slow for that.

And, as before, DO keep me posted !

Joel
Re: 8510C questions and cabling

 

You can daisy chain the test sets IF the test sets have the proper internal connections. Look to see if both "D" connectors have coaxial fittings, if so you can daisy chain. Test sets with this option are relatively rare in my experience.

There is an HP note about using multiple test sets with an 8510 - I'll look in my library to see if I can find it.
Quoting Pete Manfre <pmanfre@...>:

Yes you can't daisy chain the but you need option 001 in one of the test
sets...I do believe.

Pete wa2odo

On Fri, Apr 3, 2020, 1:49 PM Reginald Beardsley via groups.io <pulaskite=
[email protected]> wrote:

None of the cabling is obvious from the Preflight Checkout diagram or the
entire document. At this point my concern is identifying specialty HP
cables I need to get.

I'd like to be able to switch the 83631B among an 8511A, 8512A and 8515A
so I don't have to recable the system to change modes of operation. That
seems pretty straight forward, add a ST3P RF relay. But I can't find
anything to tell me if I can daisy chain the 8512A and 8515A using a
08510-60102 cable between the two. The presence of the J10 and J11
connectors on both test sets suggest you can, but I've not yet found any
supporting documentation.

And then there are mysteries such as, "For 8360, a BNC open, 1250-0053
should be connected to the 85102 SWEEP IN 0-10V". How is that different
from the bare BNC connector or a BNC dust cap?

I'm installing everything in custom racks built above a pair of ESD
surface benches. It's a small room, so pulling the bench out from the
wall to recable is not attractive if it's possible to avoid it.

Reg
Re: 8510C questions and cabling

 

The statement about the BNC open is an awkwardly worded
statement, likely by an English challenged individual.

All they are trying to say is don't connect anything to
that BNC connector... especially don't short it, or use
a 50 ohm terminator!

The 83631 trades a bnc cable for one of those combo RF and
multiple pin "D" style connectors... only they are using
only the signal pins, not the coax pins.

Was the cabling notes I posted sufficiently clear?

It described the cabling from the point of view of each
instrument looking outwards. In other words, you should
be able to look at the back of your test set, and see
from my description where each cable goes... Or look at
the back of your display unit, and see from my description
where each cable goes...

-Chuck Harris

Reginald Beardsley via groups.io wrote:

None of the cabling is obvious from the Preflight Checkout diagram or the entire document. At this point my concern is identifying specialty HP cables I need to get.

I'd like to be able to switch the 83631B among an 8511A, 8512A and 8515A so I don't have to recable the system to change modes of operation. That seems pretty straight forward, add a ST3P RF relay. But I can't find anything to tell me if I can daisy chain the 8512A and 8515A using a 08510-60102 cable between the two. The presence of the J10 and J11 connectors on both test sets suggest you can, but I've not yet found any supporting documentation.

And then there are mysteries such as, "For 8360, a BNC open, 1250-0053 should be connected to the 85102 SWEEP IN 0-10V". How is that different from the bare BNC connector or a BNC dust cap?

I'm installing everything in custom racks built above a pair of ESD surface benches. It's a small room, so pulling the bench out from the wall to recable is not attractive if it's possible to avoid it.

Reg
Re: 8510C questions and cabling

Pete Manfre
 

Yes you can't daisy chain the but you need option 001 in one of the test sets...I do believe.?

Pete wa2odo?

On Fri, Apr 3, 2020, 1:49 PM Reginald Beardsley via <pulaskite=[email protected]> wrote:
None of the cabling is obvious from the Preflight Checkout diagram or the entire document.? At this point my concern is identifying specialty HP cables I need to get.

I'd like to be able to switch the 83631B among an 8511A, 8512A and 8515A so I don't have to recable the system to change modes of operation.? That seems pretty straight forward, add a ST3P RF relay.? But I can't find anything to tell me if I can daisy chain the 8512A and 8515A using a 08510-60102 cable between the two.? The presence of the J10 and J11 connectors on both test sets suggest you can, but I've not yet found any supporting documentation.

And then there are mysteries such as, "For 8360, a BNC open, 1250-0053 should be connected to the 85102 SWEEP IN 0-10V".? How is that different from the bare BNC connector or a BNC dust cap?

I'm installing everything in custom racks built above a pair of ESD surface benches.? ?It's a small room, so pulling the bench out from the wall to recable is not attractive if it's possible to avoid it.

Reg
Re: 8510C questions and cabling

 

None of the cabling is obvious from the Preflight Checkout diagram or the entire document. At this point my concern is identifying specialty HP cables I need to get.

I'd like to be able to switch the 83631B among an 8511A, 8512A and 8515A so I don't have to recable the system to change modes of operation. That seems pretty straight forward, add a ST3P RF relay. But I can't find anything to tell me if I can daisy chain the 8512A and 8515A using a 08510-60102 cable between the two. The presence of the J10 and J11 connectors on both test sets suggest you can, but I've not yet found any supporting documentation.

And then there are mysteries such as, "For 8360, a BNC open, 1250-0053 should be connected to the 85102 SWEEP IN 0-10V". How is that different from the bare BNC connector or a BNC dust cap?

I'm installing everything in custom racks built above a pair of ESD surface benches. It's a small room, so pulling the bench out from the wall to recable is not attractive if it's possible to avoid it.

Reg
Re: HP 3455a DVM repair

 

On 4/3/20 12:33 PM, Bob Albert via groups.io wrote:
After replacing the resistor all seems okay in that area.? No other
failures.? Just a bad resistor that went up in value and then overheated.
Went *up* and overheated? That doesn't sound right. I'd bet that it
overheated and went up in value as a result.

The latest is that a new failure has surfaced.? Now the unit blows
fuses.? One of the regulators gets hot so I have to stop with the
original issue and get into the power supply.? Suddenly the voltages are
out of specification.
Are there tantalum capacitors in a 3455A? (I had one for years, long
ago, but never had to work on it.)

I am getting a bit annoyed with this repair job and so may put it back
into the closet for the nonce.
You're making progress, though!

-Dave

--
Dave McGuire, AK4HZ
New Kensington, PA
Re: Fixing a sick 4332A LCR meter

 

On Fri, Apr 3, 2020 at 09:37 AM, Joel Setton wrote:
Ok, then I'd suggest we focus on getting the oscillator running again !
Since it used to be OK, you may want to start by looking for any broken wires, etc ... Maybe it's just a broken wire.

The problem could logically be a power supply fail, or in the oscillator itself which is on A1.
First check the power supplies; then if they're OK, look at what's happening in the oscillator.
The oscillator only depends on power supply lines, and needs no other signals to operate properly.
The troubleshooting tree (figure 5-8) provides some help in troubleshooting the oscillator, if you need it. It's a conventional Wien oscillator, so you may want to start by checking the DC levels on A1TP1 (approx -0.6V), then on either side of A1R25 which will tell you what the AGC loop is doing.

And please... keep me informed !
Have fun,
Joel
Joel,

I have been following the troubleshooting tree. With the settings suggested (C, 100 pF range, 100 pF cap across the unknown) So far, it's gotten me basically across the bottom of the tree, though I haven't tried disconnecting the feedback and inserting my own bias yet.

A couple things I have observed since the last message:

1) sometimes the K1 reed relay doesn't seem to be switching, other times it does.
2) The oscillator runs in whatever state it is in, on R, in every range.
3) The oscillator stops in C, when the range is in the left hand side of the knob range; it starts (again with the 12 KHz 70 mVrms) in the right hand side of the switch range. The DC offset on A1TP1 is +6 Vdc when the oscillator is off; when it is running it is -0.9 Vdc. Pin 6 of A1IC3 (connected to one side of A1R25) is -12.3 Vdc (supposed to be +2.4 according to schematic???) and the other side of A1R25 is -0.68 Vdc.
4) The behavior in (3) is apparently identical in the L function.

Sean
HP58503A GPSDO changing firmware to remove 1PP2S (Option H14)

 

I found an HP58503A GPSDO locally, with a faulty psu, but very cheap (?50). The PSU is a quick repair, using a Meanwell PT-65S (5V/1A, +/-15V 0.5A)
Unfortunately my version had RevC "Option H14", giving out 1 pulse per 2 seconds (1PP2S) for CDMA synchronsation, not 1PPS which is usually needed for oscillator disciplining.

In case anyone finds one of these machines and needs 1PPS like me, here is the method:
Download the RevA binary files from here:
Blow a new set of four AM29F010-120JI chips.? I kept the old ones in case I managed to brick the machine with the older firmware. Make sure you note (or photo) which chip goes where.
The older firmware works fine and gives a solid 1PPS, and operates the optional display buttons front panel perfectly.? Otherwise all is as before.

-Jeremy
Re: Fixing a sick 4332A LCR meter

 

Ok, then I'd suggest we focus on getting the oscillator running again !
Since it used to be OK, you may want to start by looking for any broken wires, etc ... Maybe it's just a broken wire.

The problem could logically be a power supply fail, or in the oscillator itself which is on A1.
First check the power supplies; then if they're OK, look at what's happening in the oscillator.
The oscillator only depends on power supply lines, and needs no other signals to operate properly.
The troubleshooting tree (figure 5-8) provides some help in troubleshooting the oscillator, if you need it. It's a conventional Wien oscillator, so you may want to start by checking the DC levels on A1TP1 (approx -0.6V), then on either side of A1R25 which will tell you what the AGC loop is doing.

And please... keep me informed !
Have fun,
Joel
Re: Configuring an 85301C test range using an 8530A and 83631B

 

On 4/3/20 12:11 PM, Reginald Beardsley via groups.io wrote:
I've been studying the Agilent 85301B/C configuration guide, 5967-6042E, with the aim of being able to set up an automated anechoic test range for the 2.4 & 5 GHz ISM bands.

My 8530A does not have option 005, so I can't use an 85370A position encoder, however, it doesn't seem to me that presents a significant obstacle. I should be able to hack together a controller and stepper motors for the small antennas I'd be playing with and run it all from a PC.

Figure 17 & 20 appear to suggest that with a directional coupler and an 8511A, I'll be good to go except for cables , antenna positioning system and the anechoic chamber.

Can anyone with experience with the 85301C configuration confirm this?
Yes, I've done exactly this, minus the PC. I threw together a simple
stepper controller, got a cheap stepper-driven rotary table, and tied it
into my 8530A. It worked fine, though I'd really like to get a better
rotary table. And yes, you will need a directional coupler.

This was a few years ago and I don't really remember many of the
details, but my little stepper driver board connected to the 8530A via
two BNCs for the synchronization protocol. I hacked together a little
ARM microcontroller and wrote the firmware in C.

I was able to sweep the radiation patterns of ~10GHz horns with this
setup with ease.

It's no longer all put together, but it's still here. If you want to
try to use what I built as a starting point, I can dig up my notes and
other info more info on that setup.

-Dave

--
Dave McGuire, AK4HZ
New Kensington, PA
Re: HP 3455a DVM repair

Bob Albert
 

After replacing the resistor all seems okay in that area.? No other failures.? Just a bad resistor that went up in value and then overheated.

The latest is that a new failure has surfaced.? Now the unit blows fuses.? One of the regulators gets hot so I have to stop with the original issue and get into the power supply.? Suddenly the voltages are out of specification.

I am getting a bit annoyed with this repair job and so may put it back into the closet for the nonce.

I want to thank all of you for your guidance.

Bob
On Friday, April 3, 2020, 06:05:12 AM PDT, paul swedberg <paulswedb@...> wrote:


When I was repairing the 3455 I checked Artek and did not see the 3455 manual so I used the very poor quality HP.
Will have to recheck?Artek as there manuals are very good and very reasonable. I have purchased numbers of them. Sometimes everything doesn't have to be free. Its if you see honest value. Artek gives that value to me at least.
Regards
Paul
WB8TSL
Re: 58503A GPSDO PSU details/availability

 

Hi Andy,
I asssume you sorted this out after all this time, but if not, I can send you the wiring details. And it's just a very ordinary PSU.? 5V 1A and +/-15V 0.5A.
Look on EEVBLOG and you will find various comments about psu replacement.
-Jeremy
Configuring an 85301C test range using an 8530A and 83631B

 

I've been studying the Agilent 85301B/C configuration guide, 5967-6042E, with the aim of being able to set up an automated anechoic test range for the 2.4 & 5 GHz ISM bands.

My 8530A does not have option 005, so I can't use an 85370A position encoder, however, it doesn't seem to me that presents a significant obstacle. I should be able to hack together a controller and stepper motors for the small antennas I'd be playing with and run it all from a PC.

Figure 17 & 20 appear to suggest that with a directional coupler and an 8511A, I'll be good to go except for cables , antenna positioning system and the anechoic chamber.

Can anyone with experience with the 85301C configuration confirm this?

Thanks,
Reg
Re: Fixing a sick 4332A LCR meter

 

On Fri, Apr 3, 2020 at 09:00 AM, @0culus wrote:
I think that something else has broken...the C function was somewhat functional before, and now it is not. Both C and L, the needle remains below zero scale and there is no oscillator activity.
OK, this is getting ridiculous. Now it's "working" again. Almost makes me think there's something mechanical going on too. I hate finding mechanical issues...I have a Tek 7104 that had the resistor in the fan circuit loose and randomly shorting the fan circuit to ground. Took forever to find it after I noticed that moving the SMPS in and out of the case sometimes let the scope start.
Re: Fixing a sick 4332A LCR meter

 

I think that something else has broken...the C function was somewhat functional before, and now it is not. Both C and L, the needle remains below zero scale and there is no oscillator activity.
Re: Fixing a sick 4332A LCR meter

 

Couple of further notes:

- The signal at A1TP1/A1Q2 emitter disappears entirely in L or C
- The frequency does not change no matter what range we're on
- The signal actually is correct at 70 mVrms...I had forgotten to compensate the probe I am using for the 7A22 vertical amp after using it with a different vertical amp in my Tek 7904A.


On Fri, Apr 3, 2020 at 08:41 AM, @0culus wrote:
Joel,

1) no worries
2) OK, I'm not sure how I didn't put 2+2 together on that but it makes perfect sense now. The 427A has a log scaled resistance scale!
3) A1TP1/A1Q2 emitter (in R mode, with a 1K resistor connected) shows a ~175 mVpp (~62 mVrms) clean sine @ 12.37 KHz.

Sean
Re: Fixing a sick 4332A LCR meter

 

Joel,

1) no worries
2) OK, I'm not sure how I didn't put 2+2 together on that but it makes perfect sense now. The 427A has a log scaled resistance scale!
3) A1TP1/A1Q2 emitter (in R mode, with a 1K resistor connected) shows a ~175 mVpp (~62 mVrms) clean sine @ 12.37 KHz.

Sean


On Fri, Apr 3, 2020 at 05:46 AM, Joel Setton wrote:
Hello Sean,

I'm answering your questions out of sequence, but I'm sure you'll be able to make some sense out of this !

* I wrote about the emitter of A2Q2 but I meant A1Q2, I'm sorry for the typo.
* The 4332 has linear scales, so it's normal for the ohmmeter function to "peg" the needle if the input is open-circuit. Of course, "infinite" is way above full-scale!
* Since the ohm-meter is *somewhat* functional but reads a resistance value which is much too low, I believe there's something wrong with the oscillator. My 4332A puts out a 77mV signal at the INPUT terminal, a point where you're seeing only 12mV. Of course your mileage may vary, but I'd suggest we first look at the oscillator. It's described in the manual on page 4-3, paragraphs 4-11(a) and 4-11(b). Please probe the signal at A1 TP1 or the emitter of A1Q2, which is the oscillator output. Paragraph 4-11(b) says there should be 70mV across A1R37. Is the amplitude OK ?

If the amplitude is too low but the signal still looks like a clean sinewave, the next step will be to look at the rectifier A1IC2 and the voltage comparator A1IC3 which control the oscillator amplitude via A1Q1 (a FET which is operated as a variable resistance to control oscillator amplitude). For example, a fault around A1IC3 would cause an error in oscillator amplitude, similar to what you're seeing. But I'm thinking aloud here ...

So, please let me know what you see. And, just as a matter of personal curiosity, in which timezone are you ?

Sincerely,

Joel Setton
Re: First Gen HP 54810A with dead HD, Got ISO 3.5, Any way to write it to a drive?

 

by the way, later you can download the V4.3 and V4.5 from Tony G depository (above) and update the software first to 4.3 and then to 4.5
for that you simply need to connect that CD drive to the IDE header again and put those upgrade files on a CD rom and then navigate to the CD drive in windows 98
and run the upgrade (forget what that HP manual says)
46941 - 46960 of 151433