|
Subject: 4276A LCZ analyzer needs cal/adjustment,paper manual, standards.
Gentlemen (and Ladies),
I have a newly acquired 4276A LCZ analyzer in excellent mechanical shape and
a well-worn but functional 16047A test fixture.
I tried to do some serious capacitor measurements today of some 470uF
electrolytic caps, and after multiple open/short compensations, I am still
getting wrong values. I have a hand held C-meter and a Genrad bridge that
agree quite closely. An electrolytic that measures about 440uF on those two
units is measuring about 400 on the HP. I believe the HP is lying to me,
which makes it useless for what I need to do.
I have the download manual from the Agilent website, but the very long
fold-out troubleshooting charts and schematics are hopelessly chopped up in
the scan. I am hoping to locate a paper manual. Best price so far is $119
+ shipping :>(. Anybody have one for less?
I also need to buy/borrow/rent some 4-terminal pair calibration standards.
The manual calls for the following standard capacitors:
HP 16382A 10pF
HP 16383A 100pF
HP 16384A 1000pF
And the SHORT and OPEN from the HP 16074A standard resistor set.
Resistors from the 16074A standard resistor set are from 0.1 Ohm, 1 Ohm, 10
Ohm, 100 Ohm, 1kOhm, 10kOhm, 100 kOhm.
The resistors only have to be good to 20 kHz, so I believe that I can use
carefully measured metal film resistors for most of these values.
Anybody who can help with any of this, please let me know.
Steve
Steven A. Hogan
The Sound Steward
1606 West Southgate Avenue
Fullerton, CA 92833-3851
www.soundsteward.com
(714) 871-6636 phone
(714) 904-6636 cell
(714) 871-0640 fax
|
Hi Steve,
I have a HP 4274A and it needs to be "open" and "short" calibrated with the appropriate test fixture fitted before use, otherwise the results can be "interesting".
has a sales spec PDF scan.
That's the two buttons under ZERO just left of center.
Easy to use :)
1. Without anything fitted, press "open".
Unit goes through it's compensation frequency range steps.
2. Short out the 2 test terminals, press "short".
Unit goes through it's compensation frequency range steps.
3. Should now be ready to use.
Assuming the backup battery is okay, you should be able to store the setup.
Remember to recalibrate if you change the test method/jig.
I would try checking with non-electrolytic film capacitors first, as they can be a bit funny if not measured with a bias voltage.
Best wishes,
Susan.
|
Hi,
Agilent has the 4276A LCZ Meter Operation and Service Manual on their website.
or go to www.agilent.com and do a search (of electrical test stuff) for 4276A if the above does not work.
Best wishes,
Susan.
|
Susan,
With the 16047A test fixture fitted, I performed the open compensation.
Then I inserted a copper short (as shown in the HP manual) and performed the
short compensation. It was after that compensation that I was still getting
'interesting" readings.
I am really unclear, based on the schematic how the battery could cause
problems storing the open and short compensation constants in the permanent
RAM once the power is on. According to my understanding of the schematic,
when the 5 Volts is present (power good) Q3 saturates and shorts out the
blocking schottky diode CR1 and that puts 5 Volts on the U12 D4364C-15L
Memory chip. At the same time it charges up the battery thru the series 470
Ohm resistor.
The battery is 2.6 Volts with the unit off and 3.3V while it is on and
charging.
However, the gentleman who sold me the unit informs me that the battery can
cause problems with the memory holding the open and short calibration
constants. So maybe the battery gets replaced no matter what.
Steve
Steven A. Hogan
The Sound Steward
1606 West Southgate Avenue
Fullerton, CA 92833-3851
www.soundsteward.com
(714) 871-6636 phone
(714) 904-6636 cell
(714) 871-0640 fax
_____
From: hp_agilent_equipment@...
[mailto:hp_agilent_equipment@...] On Behalf Of Susan Parker
Sent: Sunday, November 21, 2010 8:38 AM
To: hp_agilent_equipment@...
Subject: [hp_agilent_equipment] Re: 4276A LCZ analyzer needs
cal/adjustment,paper manual, standards.
Hi Steve,
I have a HP 4274A and it needs to be "open" and "short" calibrated with
the appropriate test fixture fitted before use, otherwise the results
can be "interesting".
has a sales spec PDF scan.
That's the two buttons under ZERO just left of center.
Easy to use :)
1. Without anything fitted, press "open".
Unit goes through it's compensation frequency range steps.
2. Short out the 2 test terminals, press "short".
Unit goes through it's compensation frequency range steps.
3. Should now be ready to use.
Assuming the backup battery is okay, you should be able to store the setup.
Remember to recalibrate if you change the test method/jig.
I would try checking with non-electrolytic film capacitors first, as
they can be a bit funny if not measured with a bias voltage.
Best wishes,
Susan.
|
Hi Steve,
The battery on mine is a lithium, not a rechargeable.
If the battery isn't doing it's thing then all that happens is that the calibration is lost if the power is turned off.
On mine I also have to do a specific save calibration sequence.
For yours I would have thought that 2.6 volts would be enough to hold CMOS backup.
As to the measurements, have you tested other types and values of capacitors? Even without accurate parts a 10uF film cap should still be something like 9 to 11uF in actual value.
The manual has the checking sequences, checking with better quality Rs and Cs would give you a in/out of ballpark tests. Inductors are simpler since one can make up an air core inductor and calculate the inductance from the physical properties (there are programs on the web to do this for one).
Otherwise you may need to go through the bench test and setup procedure. There are a couple of calibration voltages that I did decide to tweak a bit when I first got mine up and running. A good DVM here with an accurate reading is important as the cheaper hand held units (free with breakfast cereal types) can lose there accuracy very quickly, particularly if the battery is near end of life.
Sorry, not able to help much beyond this as you have a different model to mine.
Best wishes,
Susan.
|
The 4276A may be the one telling the truth. Electrolytic caps can read much different in value at different test frequencies. Try changing the frequency (controls on right side, if you're not familiar with it yet) and see how the values change - the lowest it goes is 100 Hz. Try to compare a mica around 1000 pF, or a plastic film around 1 uF to see if the different instruments agree.
When I first got my 4276A I thought it was broken. I couldn't get any proper readings until I figured out how to set up the controls and hook it up to the DUT properly.
I also looked into L/C/R standards but found they may cost more than I paid for the instrument. Then I studied the manual to see how it works, and found that the test method used is intrinsically very accurate, and uses (as I recall - it's been a while) only one or two high precision internal caps for reference, and likewise some high precision resistors, and there is not that much that can go wrong in terms of basic accuracy. The measurements are derived mathematically from full quadrature measurements of all the signals. I found a few 1% tolerance caps that I keep just to confirm it is about right, in case there is any doubt, and it has always been great.
It has turned out to be one of my favorite and most useful instruments. It can easily measure a few pF at 100 Hz, and with no setting changes (auto-mode), it jumps right up to measure any cap up to 19,000 uF whenever needed. The only drawback I found is that both sides of the DUT are "hot" with either the generator signal or the measurement input, so it has to float. I designed a ground-reference converter for it so one end of the DUT can be grounded - now it can measure virtually anything - like electrolytic caps in-circuit, or for estimating the length of installed coaxial cable without TDR.
So, unless you need very high precision for filters or something, you probably don't need to recalibrate, and there shouldn't be any need for extreme precision on electrolytics. Also note that cored inductors can measure differently at various frequencies.
Good luck getting it working - it is a fine instrument. Ed Breya.
toggle quoted message
Show quoted text
--- In hp_agilent_equipment@..., "Steve Hogan" <stevehogan@...> wrote:
Subject: 4276A LCZ analyzer needs cal/adjustment,paper manual, standards.
|
Hi Ed,
Just wondering if you can provide some detail of how you implemented the "ground-reference converter", and any affect it has on measurements, such as accuracy.
Thanks - John
toggle quoted message
Show quoted text
It has turned out to be one of my favorite and most useful instruments. It can easily measure a few pF at 100 Hz, and with no setting changes (auto-mode), it jumps right up to measure any cap up to 19,000 uF whenever needed. The only drawback I found is that both sides of the DUT are "hot" with either the generator signal or the measurement input, so it has to float. I designed a ground-reference converter for it so one end of the DUT can be grounded - now it can measure virtually anything - like electrolytic caps in-circuit, or for estimating the length of installed coaxial cable without TDR.
|
It's very simple - I don't have a schematic, so will have to look inside (it's been a couple years) and then write up a verbal description. The main thing you need to find is a dry type audio coupling transformer, the kind commonly used in older modems from the late 1990s - 2000s.
Ed
toggle quoted message
Show quoted text
--- In hp_agilent_equipment@..., "john.spuma" <john.spuma@...> wrote:
Hi Ed,
Just wondering if you can provide some detail of how you implemented the "ground-reference converter", and any affect it has on measurements, such as accuracy.
|
I'm pretty sure I can supply that! I have a couple of old modems.... for parts. Ed (another one) On Wed, 01 Dec 2010 15:43:44 -0000 "Ed Breya" <edbreya@...> writes: It's very simple - I don't have a schematic, so will have to look
inside (it's been a couple years) and then write up a verbal
description. The main thing you need to find is a dry type audio
coupling transformer, the kind commonly used in older modems from
the late 1990s - 2000s.
Ed
____________________________________________________________ SHOCKING: $100 Walmart Gift Card for $6.39 SPECIAL REPORT: High ticket items are being auctioned for an incredible 90% off!
|
I found my notes from back then, and looked at the insides to refresh my memory - I had forgotten exactly what I did and how it works. The circuit is deceptively simple, but took a lot of investigation and experimenting to make it work. I will sketch up some cleaner diagrams and info (for my own use as well - my notes were hard to decipher even for me), and either scan or photo it to post. For now, I can give a verbal description to contemplate, and rough parts list.
If you picture the front end of the 4276A, the device under test (DUT) is driven by "HC" - the ground-referenced 100Hz-20kHz source, and the other end goes into a virtual ground "LC" - the I/V converter, which measures the current. "HP" and "LP" are attached at the DUT, kind of like Kelvin sensing, measuring the voltage. This arrangement, and sending the signals through individual coaxial lines, eliminates parasitic effects except for the relatively small ones right at the DUT where the environment changes from 4/5-wire to 2-wire. These small parasitics are zeroed out when "OPEN" and "SHORT" checks are done, leaving only residual effects.
The ground-reference converter rearranges it so the source is in series with the LC terminal, rather than coming directly from HC. That's where the transformer comes in - the HC drives the primary, and the secondary connects between LC and one end of the DUT, so its other end can be grounded. The HP and LP signals still connect as before. However, this change causes a problem - the capacitance of the LP cable and any extension of it are no longer intrinsically canceled out, so must be compensated for. That's where the other parts come in - a fixed capacitor slightly smaller than the LP capacitance, a variable capacitor to fine tune it, and a resistor to compensate for phase shift due to winding resistance and other parasitics are needed. The total compensation capacitance (Ct) will be approximately equal to the total LP cable capacitance, and is specific to the fixture design. The OPEN and SHORT checks will work the same, eliminating the main parasitics, as long as they land within the automatic adjustment range.
So, those parts plus four BNC cables or pigtails, one BNC pigtail, a metal enclosure (for shielding), and various connectors and hardware is about it.
More later. Ed
toggle quoted message
Show quoted text
--- In hp_agilent_equipment@..., "Ed Breya" <edbreya@...> wrote:
It's very simple - I don't have a schematic, so will have to look inside (it's been a couple years) and then write up a verbal description. The main thing you need to find is a dry type audio coupling transformer, the kind commonly used in older modems from the late 1990s - 2000s.
Ed
--- In hp_agilent_equipment@..., "john.spuma" <john.spuma@> wrote:
Hi Ed,
Just wondering if you can provide some detail of how you implemented the "ground-reference converter", and any affect it has on measurements, such as accuracy.
|
Hi Ed:
The only place I can find ground on my 4-terminal pair LCR meters is at the Lp shield. All seven other DUT connection nodes are not at ground. See:
I'm looking forward to learning more about this.
Have Fun,
Brooke Clarke
hp_agilent_equipment@... wrote:
toggle quoted message
Show quoted text
Re: 4276A LCZ analyzer needs cal/adjustment,paper manual, standards.
|
The measurement ground becomes the place where the lines converge - as close to the DUT as possible. I'll cover this more, and the sources of error in the next part. Ed
toggle quoted message
Show quoted text
--- In hp_agilent_equipment@..., Brooke Clarke <brooke@...> wrote:
Hi Ed:
The only place I can find ground on my 4-terminal pair LCR meters is at
the Lp shield. All seven other DUT connection nodes are not at ground.
See:
I'm looking forward to learning more about this.
Have Fun,
Brooke Clarke
hp_agilent_equipment@... wrote:
Re: 4276A LCZ analyzer needs cal/adjustment,paper manual, standards.
|
It just realized maybe there is some confusion here - the ground-reference converter makes it so that one end of the DUT CAN be grounded, NOT that it HAS to be grounded. Ed
toggle quoted message
Show quoted text
--- In hp_agilent_equipment@..., "Ed Breya" <edbreya@...> wrote:
The measurement ground becomes the place where the lines converge - as close to the DUT as possible. I'll cover this more, and the sources of error in the next part. Ed
|
?
Might Not be relevant, ... since applies to the 4274a,? but:
?
Just noticed an Hp Service Note in the Hp 4274a LCR manual (4274a-12) :
?
" Improvement for Memory Backup Battery "
?
-- about swapping out the original Schottky diodes (1901-0518) for? more common(?)
Si diodes ( 1901-0025 ), due to degradation of the Schottky diodes over time, causing the Memory backup Battery to discharge faster.?
Found in the Sept. 1980 4274a Serv. Man., pg. 2 for instruments having Opt. 003.
?
Given that Hp used common designs in many instruments, ... might want to consider.
rick
?
On Sun, Nov 21, 2010 at 10:11 AM, tss_steve_990 wrote:
I am really unclear, based on the schematic how the battery could cause
problems storing the open and short compensation constants in the permanent
RAM once the power is on. According to my understanding of the schematic,
when the 5 Volts is present (power good) Q3 saturates and shorts out the
blocking schottky diode CR1 and that puts 5 Volts on the U12 D4364C-15L
Memory chip. At the same time it charges up the battery thru the series 470
Ohm resistor.
The battery is 2.6 Volts with the unit off and 3.3V while it is on ...
|
Susan Parker