开云体育

Merry Christmas all!
My friends and I have been comparing what we see from the S Parameter
Test Set on the two ports. These tests are with the VNA uncalibrated
and with proper opens or shorts (even from APC7 cal kits).
Typically we get completely different shaped curves ranging from -35
to -55 dB.
I get very similar results using the same S Parameter Test Set on an
HP8753B and an HP8753C so the curve shape is caused by the test set
not the VNA.
Typically we see one port 10 dB better and flatter than the other.
Any comments?

p.s. the cal kit discussion was very informative and timely. I still
don't understand the comment (that was later corrected) about the open
usually missing a part. My friend's Type N cal kit does not have extra
parts for the open. Also I have seen an open/short combo advertised
and wonder if this comes apart to use the open??? I have not used one
and find the concept of one end a short (OK) and the other an open
(how?) confusing.

You're welcome, and all, enjoy the holidays. It may pay to download the 1993 HP Catalog, that was the last year all in the 345x line were available for sale, so you can compare and contrast the specs, also against the 3478.



The HP manuals, Keithley handbook and the volt-nuts archives are free, just download and search as needed, so you can get all the basic cal and operations info without cost to your misses before you actually buy something ;)

If you intend to stack a lot of gear ontop the 3456a and not sure about weight, put a 2x4 plank so weight get distributed to the edges of the chassis, this is were the most strength lies, while the top and bottom chassis covers are the weakest as there is also no cross bar for structural support. The plank allows a gap for added cooling, even if vents on the 3456a are on the sides and rear, the plank method works for all rack mounted sized gear.

As always too, heavies are best at the bottom, light at the top, as often the boat anchors are also made with stronger chassis to take not just its own weight but others too.

Lastly, I'm not a fan of designs that use battery backed up SRAM for system data, I prefer EEPROM at the least. Since all your purchases are used, the state of the battery is unknown and should be replaced, ASAP. If it snafus, or worst just corrupted, such as in buying "CANNOT TEST" "AS IS" type eBay gear, you'll need budget the possibility of adding $100-140 to your purchase cost to get it formally recalibrated.

With the 3456a, the all pots method puts the user in full control of its cal state, you can cal any range, any time [ keeping metrology guidelines in mind], so if you get one working or not, its also much easier to repair to working order entirely by yourself.

--- In hp_agilent_equipment@... <mailto:hp_agilent_equipment%40yahoogroups.com>, Jeff Machesky <jeff@...> wrote:

I have to say this response has really given me the most food for
thought. It provided exact details about the models and the pro's and
con's of each. It also made me realize something important. There is no
reason I have to buy one of these devices and run out and get it
calibrated. I can make sure it's readings are within reason and take it
out later on to get it calibrated. My main goal is resolution, so I can
have that now and have the accuracy and the resolution later. It will
also give me time to clean up and gain confidence in the device before
spending more money on it.

Very good feedback, Thanks.

Jeff

Hi Richard,

Thank you for your reply.

As an example, suppose I am on band C, set to the highest frequency. Peak the amp tuning knob. Set the power output knob so the RF output meter reads +4 dBm. Set the attenuator knob to 0 dBm. Measure the actual output (ideally using an RF power meter, but in my case I use my spectrum analyzer and RF probe on DMM.)

Then, move to band D. Turn frequency knob to the low end, same frequency as previously used on band C. Peak the amp tuning knob. Set power output so the RF output meter reads +4 dBm. Leave the attenuator knob at 0 dBm. Measure the actual power output.

In my case, I find that the actual output will vary even though the RF output meter reading remains constant.

Maybe there is some weird capacitance coupling effect between the attenuator probe antenna and the final amp output coil, as the bandswitch moves from one band to the next?

Patrick Wong AK6C

Thanks, I guess I will have to just keep looking. For $50 I would love to have one of these, or one of its sister units. And being in Los Angeles, there is always a likelihood I can find one that doesn't have to be shipped.

Bob

--- On Sun, 12/23/12, marvgozum <mxg003@... <mailto:mxg003%40jefferson.edu>> wrote:

From: marvgozum <mxg003@... <mailto:mxg003%40jefferson.edu>>
Subject: [hp_agilent_equipment] HP 3456a was Re: HP 3478A general

questions

To: hp_agilent_equipment@...

<mailto:hp_agilent_equipment%40yahoogroups.com>

Date: Sunday, December 23, 2012, 5:20 AM


eBay, if you're in no rush, patience is rewarded by cheap
buys. I've tracked 3456a prices on eBay for 4 years,
and the price I posted are ranges I've found. The
killer is often S&H, as its a wide and heavy beast, but
luckily its short and flat. Its about the perimeter of
a small coffee table, which makes it easy to stack
instruments ontop of it.

Long ago, this website tracked prices for a lot of good
gear, but its seems to be orphaned. But it still gives
what hi-lo used to be before 2008, as well as Tektronix.





Many links are broken but you can drill down without a
problem.


--- In hp_agilent_equipment@...

<mailto:hp_agilent_equipment%40yahoogroups.com>,

Bob Albert <bob91343@...> wrote:

Where can you buy one of these for $50?

Bob

------------------------------------

Yahoo! Groups Links


hp_agilent_equipment-fullfeatured@...

<mailto:hp_agilent_equipment-fullfeatured%40yahoogroups.com>

From: marvgozum <mxg003@...>
Subject: [hp_agilent_equipment] HP 3456a was Re: HP 3478A general questions
To: hp_agilent_equipment@...
Date: Sunday, December 23, 2012, 5:20 AM


eBay, if you're in no rush, patience is rewarded by cheap
buys.? I've tracked 3456a prices on eBay for 4 years,
and the price I posted are ranges I've found.? The
killer is often S&H, as its a wide and heavy beast, but
luckily its short and flat.? Its about the perimeter of
a small coffee table, which makes it easy to stack
instruments ontop of it.

Long ago, this website tracked prices for a lot of good
gear, but its seems to be orphaned.? But it still gives
what hi-lo used to be before 2008, as well as Tektronix.





Many links are broken but you can drill down without a
problem.


--- In hp_agilent_equipment@...,
Bob Albert <bob91343@...> wrote:

Where can you buy one of these for $50?

Bob

------------------------------------

Yahoo! Groups Links


? ? hp_agilent_equipment-fullfeatured@...

Correction! We are seeing this with a good load attached, no calibration.
Same load gives different curve for each port.
I would expect the reflected power to be much lower than -30 dBm.
Of course with the open or short the reflected is close to 100%.
Too much egg nog.

--- In hp_agilent_equipment@..., "Peter Bunge" <bunge@...>
wrote:

Merry Christmas all!
My friends and I have been comparing what we see from the S Parameter
Test Set on the two ports. These tests are with the VNA uncalibrated and
with proper opens or shorts (even from APC7 cal kits).
Typically we get completely different shaped curves ranging from -35 to
-55 dB.
I get very similar results using the same S Parameter Test Set on an
HP8753B and an HP8753C so the curve shape is caused by the test set not
the VNA.
Typically we see one port 10 dB better and flatter than the other.
Any comments?

p.s. the cal kit discussion was very informative and timely. I still
don't understand the comment (that was later corrected) about the open
usually missing a part. My friend's Type N cal kit does not have extra
parts for the open. Also I have seen an open/short combo advertised and
wonder if this comes apart to use the open??? I have not used one and
find the concept of one end a short (OK) and the other an open (how?)
confusing.

Merry Christmas all!
My friends and I have been comparing what we see from the S Parameter Test Set on the two ports. These tests are with the VNA uncalibrated and with proper opens or shorts (even from APC7 cal kits).
Typically we get completely different shaped curves ranging from -35 to -55 dB.
I get very similar results using the same S Parameter Test Set on an HP8753B and an HP8753C so the curve shape is caused by the test set not the VNA.
Typically we see one port 10 dB better and flatter than the other.
Any comments?

p.s. the cal kit discussion was very informative and timely. I still don't understand the comment (that was later corrected) about the open usually missing a part. My friend's Type N cal kit does not have extra parts for the open. Also I have seen an open/short combo advertised and wonder if this comes apart to use the open??? I have not used one and find the concept of one end a short (OK) and the other an open (how?) confusing.

It appears that there are several members with 608s responding to this string.

With regards to the HP 608D, there are at least two attenuator circuit designs and therefore two schematics that can be found on line via Google search.

One design, the newer one, has two capacitors, and therefore resistance measurement at the RF output connector will normally show infinity.

The other has one capacitor and therefore a normal resistance measurement at the RF output connector will show ~53 ohms.

It is relatively easy to remove the attenuator probe for inspection, no special tools are needed, and it is not very hard to reinsert the probe although obviously care is needed to avoid damage to the grounding fingers surrounding the probe end.

I've never been able to figure out why the RF power meter on my 608D is not very accurate, at least based upon comparisons with my 8642B. The error is +/- 4 dB or more.

What is really weird is that if I produce a signal at the top frequency of one band, supposedly producing 0 dBm as measured per the meter, then switch to the next higher band, turn the tuning capacitor all the way to the low end of the range to produce the same frequency, and set output to 0 dBm per the meter, I find the actual output as measured by either my 8591E SA or Fluke 87 with RF probe is much different. The probe shouldn't care what frequency range the 608D is set to.

I tried cleaning the bandswitch contacts using electrical contact cleaner. The attenuator resistors measure correctly and show no signs of physical damage. The attenuator had been replaced at some point by HP service as a sticker on the chassis shows a notice to that effect.

My 608D has way too much frequency drift for my liking as well as tremendous incidental FM when using AM modulation. I posted about this a few months ago along with photos showing the incidental FM.

I'd be interested if other members owning 608-vintage equipment as well as spectrum analyzers can look at the quality of their AM signals with regards to incidental FM, and post photos showing good or bad results.

I recognize that if you are serious about using that generator, you have to leave it on 24 hr/day but I don't need to use a signal generator enough to have it running continuously, consuming 400W or whatever its power requirement is. So it sits around taking up space and serves mostly as a cabinet to put other stuff on.

Good luck,

Patrick Wong AK6C

I have to say this response has really given me the most food for
thought. It provided exact details about the models and the pro's and
con's of each. It also made me realize something important. There is no
reason I have to buy one of these devices and run out and get it
calibrated. I can make sure it's readings are within reason and take it
out later on to get it calibrated. My main goal is resolution, so I can
have that now and have the accuracy and the resolution later. It will
also give me time to clean up and gain confidence in the device before
spending more money on it.

Very good feedback, Thanks.

Jeff

Where can you buy one of these for $50?

Bob

A DC measurement across the output connector should show
53 ohms. I know this is not a standard value but that is what the diagram
says. My 608 shows infinity. That's not a good sign.

Regards.

Max. K 4 O DS.

It's really very simple. The center conductor of the connector goes through
a 53 ohm resistor to one side of the pickup coil, the other side of the
pickup coil goes to the sleeve. The center conductor also goes through the
series combination of a 53 ohm resistor and a variable capacitor to the
sleeve. The capacitor is designated as c37. The two resistors are
designated as R58 and R59. I can't find these listed in the parts list I
suspect that HP probably sold the entire attenuator assembly as a single
replacement part. A DC measurement across the output connector should show
53 ohms. I know this is not a standard value but that is what the diagram
says. My 608 shows infinity. That's not a good sign.

Regards.

Max. K 4 O DS.

The 3456a was the most accurate in the HP fleet when introduced in 1981. The model line was 345x, so in its heritage was the 3455, 3457, and finally the ersatz standard DMM today, introduced in ~1989, the 3458. HP introduced a high accuracy ADC method, multislope II, with 3456a, you'll see marked change in posted accuracy versus the 3455a. The 3458a further improves that technique. In summary, if you're looking for accuracy these series of DMMs are one to shoot for.

There are pros and cons to the discontinued models, 56a and 57a. For example, The 56a is as good as the 57a in DCV at 1 year ppm, while the 57a is 10x more stable in 1yr as a ohmmeter. The 57a uses all close case calibration and has more unobtanium in its parts list, versus the 56a, which was the last to use all pots based calibration. The basic 57a reads only to 300Vrms maximum, but also has a current mode.

As for prices, the low end is similar on both, about $50 without shipping, but on the high end with calibration, the 56a runs to $300 while 57a about $700.

The 3478a was a general purpose working meter, ~10x less accurate in DCV than the 3456a.

As mentioned earlier you can use a resistor network; dekaviders or KV bridges in DC, to generate other DC voltages needed to check ranges, but maintaining stability in the reference enough to check the meter's accuracy, is the challenge. Its enough a whole new forum discusses this issue alone, volt-nuts@... <mailto:volt-nuts%40febo.com>. A similar problem exists for AC.

To calibrate a meter to realize its best accuracy requires grasping key concepts in metrology. A simplest approach is to take your DMM and compare its readings against a known good DMM. The reference DMM should ideally be 4x to 10x more accurate than the DUT. This is the concept of TUR, test uncertainty ratio. When making comparisons, say voltages, the test volt must be stable and precise enough during the transfer, for example if you are checking the 1V range between DMMs, you need to generate a 1V voltage stable to 1uV resolution or less, 1.000 000 VDC. Without a high level of thoroughness and stability, your 6.5 digit DMM will have increasing uncertainty as you get to its LSD.

On the low end of the range, special conditions apply that are fully covered in Keithley's low level measurement handbook, and these need to be controlled once dealing with uV, mohms, and uA adjustments.

Lastly, one has to decide if the effort described in volt nuts is worth it, or should one just take your DUT to the calibrator periodically for $100-$150 to insure your 6.5 digit DMM is as good as you expect it to be.

--- In hp_agilent_equipment@... <mailto:hp_agilent_equipment%40yahoogroups.com>, Jeff Machesky <jeff@...> wrote:

So it's about the time of year where I'm allowed to buy one piece of
test gear and I'm thinking I want a bench meter with higher
resolution. The best I've got so far is the Tek 2465BDM which has a
20,000 count 0.03% accuracy DMM. It's however a little out of cal and
I just don't want to spend the cash to get it back in perfect spec.
It's also a bit crazy to have a scope running just to use the DMM.
This brings me to the 3478A. While I could technically afford to buy a
new 120,000 count DMM I just can't justify the cost for the hobby type
work that I do. Plus I love working with the older gear and saving

money.

Couple quick questions about the 3478A:

Can one range be calibrated at a time? This is not highly clear in the
service manual. Can I calibrate just the 3V range for example?
Are they reliable, or will I have to fix up most of what I would buy
on say ebay?

The problem I have with the meter on my Tek is that you have to run the
entire cal for each measurement type, If you start with DC you have to
go all the way up the range from millivolts up to several hundred
volts. As we all know it's easy to get a precision calibrated low
voltage references such as a DMM check board, it's another thing to get
500 volts or higher reference at the spec of the meter.

So is this really a meter to buy if your willing to send it off to be
calibrated or have a calibration transfer standard meter that's above
the specs of the 3478A? In other words if all you have is just a few
standards are you wasting your time?

Has anyone had any luck using a precision reference, some precision
resistors and an op-amp to generate other reference voltages with any
degree of accuracy? Also has anyone found a good source of say 500
volts 60 Hz AC? The right audio amp and a sig gen can get you some of
the other odd cal standards. High voltage DC is just so much easier to
obtain then a nice high voltage sine wave.

Sorry for the random questions, and thanks for any replies. The
54111D is still running great. My thanks again for everyone who helped
me fix the ROM's.

Jeff