Error 1 and error 4 on the HP 4275A

[WayneL]

Hi

Could someone be so kind and lookup what eror 1 and error 4 is on the HP
4275A.

I am due to get manuals soon but need to understand what these errors are?


Thanks

Wayne

 

[Winfield Hill]

WayneL wrote...

Could someone be so kind and lookup what eror 1 and error 4 is
on the HP 4275A.

err1, Error in ZERO offset adjustment. The value of the
residual parameter present in the measuring circuit exceeds
the offset control limit.

I think that means the instrument needs servicing.

err4, Error in measuring circuit configuration. It goes on
to describe an open or short in the test fixture (i.e.
improper connections to the four BNC signal jacks), hinged
cover open on 16047B, or a DC-bias measurement-range conflict.

 

[WayneL]

Hi Win
on the issue of err1. I have a HP 4275A and HP 4274A both give
the same results. However, I am using four bit of wire.
The four inputs are Lcur, Lpot, Hpot and Hcur.
Am I correct in assuming that Lcur and Lpot are shorted and Hpot and Hcur
are also shorted at the point of measure e.g. at the DUT.


Thanks


Wayne

 

[Winfield Hill]

WayneL wrote...

on the issue of err1. I have a HP 4275A and HP 4274A both give
the same results. However, I am using four bit of wire.

They use coax, with all the grounds connected at the end.

The four inputs are Lcur, Lpot, Hpot and Hcur.
Am I correct in assuming that Lcur and Lpot are shorted and Hpot
and Hcur are also shorted at the point of measure e.g. at the DUT.

Yes.

 

[Chris]

Hi

Could someone be so kind and lookup what eror 1 and error 4 is on the HP
4275A.

I am due to get manuals soon but need to understand what these errors are?


Thanks

Wayne

In addition to Win's response, if I could ask: do you get this err
message with the factory zero meter test fixture plugged in? It
affixes right to the BNC connectors on the front of the instrument. If
you have that fixture, try that first as a reality check, and screw
both clamps closed so they touch. That might solve your problem. This
meter is picky about fixturing. The other factory fixture has white
coax cables that seem to get damaged easily, so it's better to do a
GO/NOGO check with the fixture without the coax cables -- less to
break.

If you are getting those error messages with the 0 m factory fixture
plugged in, and can't zero, it's possible you need repairs. Wait for
the manual and read it carefully first. In addition to a
troubleshooting flowchart, there's also a lot of useful measurement
information in there.

Good luck
Chris

 

[WayneL]

Hi Chris

Ahhhh. I do not have a factory zero meter test fixture. Do I need to get
one or can I live without one?
What are the part numbers of both attachments?

Cheers

Wayne

 

[Chris]

Hi Chris

Ahhhh. I do not have a factory zero meter test fixture. Do I need to get
one or can I live without one?
What are the part numbers of both attachments?

Cheers

Wayne

These meters always came with a standard test fixture, which plugged
directly into the 4 coax connectors on the front panel. It's a small
white box with 4 female connectors. You can use it to measure
components directly. It's not made specifically for zeroing the meter,
although you can use it for that. The meter won't zero out without
having some fixture capacitance there (and also having the high current
connected to high potential and low current to low potential too, I
believe).

Look around and see if you have this fixture. As I remember, it's a
small (approx. 6" X 3" X 2") white metal box with flat recessed
gold-plated clips on the top, and two knurled thumbscrews on the front
to lock in components or small conductive contact blocks.

The advantage of using the factory fixture is that it has a known
impedance which the meter _should_ be able to zero out. As I remember,
there is a switch on the front of your meter which has two positions --
0 meter or 0.5 meter (length of the fixture coax to compensate). I
preferred to use the 0 m setting and this fixture (you can't plug in
the fixture I'm talking about without switching the front panel switch
to 0 m -- it won't fit) because you didn't have to worry about
intermittent coax on a quick meter check/zero compensation.

When you get the manual, do the self-test sequence -- it may give you
more information on what's going on than a zero/fixture compensation
cycle can achieve.

A bit of bad news -- on the last one of these meters I used (about 8
years ago, all this is from memory), the symptoms you're reporting were
the initial indications of the bitter end. I hope you'll have better
luck.

Chris

 

[[email protected]]

Actually it's because your parasitics from the Open measurement exceed
a predefined usable dynamic range for the instrument. You're supposed
to be using a "standard" HP 160xx LCR cable. The open load is know and
within the dynamic range of the instrument. Using something else can
will result in a compensation value that puts the instrument out of its
calibration spec. Pretend this is actually complex impedance instead
but basically: C(compensated) = C(measured) - C(offset) where the
offset is so big you'll get round-off, etc. and C(compensated) won't be
in spec anymore.

Since the 4274 and 4275 are virtually identical (firmware-wise) and
both use only 8-bit microcontrollers, there's the same finite dynamic
range and round-off characteristics on measurements with the same
fixturing, thus the same error condition results.

M

 

[Winfield Hill]

[email protected] wrote...

Actually it's because your parasitics from the Open measurement exceed
a predefined usable dynamic range for the instrument. You're supposed
to be using a "standard" HP 160xx LCR cable. The open load is know and
within the dynamic range of the instrument. Using something else can
will result in a compensation value that puts the instrument out of its
calibration spec. Pretend this is actually complex impedance instead
but basically: C(compensated) = C(measured) - C(offset) where the
offset is so big you'll get round-off, etc. and C(compensated) won't be
in spec anymore.

Since the 4274 and 4275 are virtually identical (firmware-wise) and
both use only 8-bit microcontrollers, there's the same finite dynamic
range and round-off characteristics on measurements with the same
fixturing, thus the same error condition results.

I'm not sure what message you're responding to since it doesn't show
up on my server, but I'll comment that it's not unusual to get these
errors while using the proper standard hp fixtures. I don't know but
wonder if the fix can be as simple as an internal adjustment. If my
service manual, which has disappeared, turns up, maybe I'll find out.