S-parameter test sets

[Terry Given]

Joerg wrote:

Hello Terry,

done, thanks, delete at will. and yes please wrt hp-ib codes, although
now I have to save up and buy a usb-hpib adapter. I did some web
browsing, and it looks like it will cost me US$400-$500, the latter
price being the agilent box of tricks (which I will probably get).
Looks like I'd better do some more work (in my now
much-better-equipped lab) to earn enough to pay for it and the
s-parameter test set.


If you don't need full speed it can be cheaper. I don't know whether
this stuff is available in the US but in Europe they used to be cheaper
than NI:

http://www.plug-in.de/gpib/

Seems prices have gone up since the 300k GPIB is now a whopping 285
Euros. These used to be around $100 or so. They also have a USB-GPIB but
that's really expensive.

Regards, Joerg

thats consistent with what I found - perhaps a dozen unknown brands, all
around the US$300-$400 mark, and NI & agilent at $500.

Cheers
Terry

 

[Mike Monett]

Terry Given wrote:

Joerg wrote:

[...]

Seems prices have gone up since the 300k GPIB is now a whopping 285
Euros. These used to be around $100 or so. They also have a USB-GPIB but
that's really expensive.

Regards, Joerg

thats consistent with what I found - perhaps a dozen unknown brands, all
around the US$300-$400 mark, and NI & agilent at $500.

Cheers
Terry

Terry,

The GPIB bus is really simple. You can ignore most of the commands and just use
the ones you need. There are plenty of examples that use the parallel port on
your pc and a bit of glue logic to interface. Here's one, and google has lots
more:

http://www.cl.cam.ac.uk/users/djg/wwwhpr/gpibpage.html

The trick in working with the gpib is to have a simple way to monitor the bus
and set the data and command bits. Ziatech used to sell an inexpensive box
for about $250 with a bunch of leds and switches that allowed you to act as a
talker or listener and walk through all the bus transactions. It was very easy
to use. You could make one yourself for the cost of the leds and switches. It
is invaluable for debugging the interface.

Mike Monett

 

[Terry Given]

Mike Monett wrote:
[...]

Seems prices have gone up since the 300k GPIB is now a whopping 285
Euros. These used to be around $100 or so. They also have a USB-GPIB but
that's really expensive.

Regards, Joerg

thats consistent with what I found - perhaps a dozen unknown brands, all
around the US$300-$400 mark, and NI & agilent at $500.


Terry,

The GPIB bus is really simple. You can ignore most of the commands and just use
the ones you need. There are plenty of examples that use the parallel port on
your pc and a bit of glue logic to interface. Here's one, and google has lots
more:

http://www.cl.cam.ac.uk/users/djg/wwwhpr/gpibpage.html

The trick in working with the gpib is to have a simple way to monitor the bus
and set the data and command bits. Ziatech used to sell an inexpensive box
for about $250 with a bunch of leds and switches that allowed you to act as a
talker or listener and walk through all the bus transactions. It was very easy
to use. You could make one yourself for the cost of the leds and switches. It
is invaluable for debugging the interface.

Mike Monett

Hi Mike,

thanks for the link. I have also toyed with the idea of using a TUSB3210
(A USB micro I have used for other projects) to roll my own usb-to-HPIB
converter. Then I have to write software to make it all go :{

If I could get the agilent USB interface for, say $200 it would be a
no-brainer. Even at $500, it takes less than a days work to pay for it.
I doubt I could get a system up and running in less time, so I'll
probably just buy one.

Cheers
Terry

 

[Terry Given]

Fred Bartoli wrote:

"Terry Given" <my_name@ieee.org> a écrit dans le message de
news:KtSue.9967$U4.1309297@news.xtra.co.nz...

I was pretty happy with it. I figured out your last point while reading
the 3577 datasheet, so I spent $1000 and bought an HP3585A 20Hz - 40MHz
speccy. Which I brought home at about 10pm on wednesday night. Then
spent until 7am thursday morning playing with it, before driving up to
auckland for a customer meeting *grin* - kid in a candy store.....



I was lucky enough to get hold of an HP3585B for 400Euros.
One point I didn't expected is that the fans noise (bearings or something
like this) is, well, pretty present...

Does yours have the same 'feature' or is it just that mine needs some work?

I am (was), from time to time, looking for a used 3577 or 4195 VNA too, but
these are rare and pretty expensive.
I think you've got a good price for yours. Were did you get it?

Being desperate I finally decided to build my own one but, you know, so many
things to do and not enough time...

Hi Fred,

thats a good price. The fan on my one blows a howling gale, too. But its
quiet compared to the 3577

the 3577 came from Iomega corporation, via ebay. Its still got their
property tag on it (#4043).

Cheers
Terry

 

[Mike Monett]

Terry Given wrote:

[...]

Hi Mike,

thanks for the link. I have also toyed with the idea of using a TUSB3210
(A USB micro I have used for other projects) to roll my own usb-to-HPIB
converter. Then I have to write software to make it all go :{

If I could get the agilent USB interface for, say $200 it would be a
no-brainer. Even at $500, it takes less than a days work to pay for it.
I doubt I could get a system up and running in less time, so I'll
probably just buy one.

Cheers
Terry

Hi Terry,

I usually find problems with outside hardware/software can take more than a day to
solve. Intermittent failures, random software bugs, installation problems, data
transfer errors, problems saving files and having them disappear, ad infinitum.
When this happens, good luck getting support from the vendor.

The gpib interface is not that difficult, and is really easy to code. When you
have done it yourself, you can easily fix any problems that might come up, and
duplicate the system at very low cost.

A major problem with gpib is the interface is wire-or, so when you have an
intermittent cable problem, it can be very difficult to troubleshoot. You don't
know if it is in the equipment, or in the hardware or software, so you don't know
where to start looking.

I have never seen diagnostics in commercial code that looks for any failure
conditions, but you can easily add them to your own code.

YMMV, so best of luck. Let us know how it turns out.

Mike Monett

 

[Joerg]

Hello Mike,

A major problem with gpib is the interface is wire-or, ...

The main problem I found with GPIB are those unwieldy garden hose type
cables. Once I unhooked one and sure enough it recoiled, sending my half
filled coffee mug sailing, crashing and spilling the contents all across
the lab floor. On top of that it was my favorite mug.

I think Terry's path towards a TUSB uC may be a good one. Just imagine,
replacing the garden hose with a USB cable from Walmart. Also, the cost
for a 6ft cable would go from $100 or so to less than $10.

The only downside with USB is that its EMI performance is, well, not the
best at times. RS232 would be better but many "modern" PCs don't have
serial or parallel ports anymore. Oh well, out here we just keep a stack
of ferrites handy.

Regards, Joerg

http://www.analogconsultants.com

 

[Mike Monett]

Joerg wrote:

Hello Mike,

A major problem with gpib is the interface is wire-or, ...

The main problem I found with GPIB are those unwieldy garden hose type
cables. Once I unhooked one and sure enough it recoiled, sending my half
filled coffee mug sailing, crashing and spilling the contents all across
the lab floor. On top of that it was my favorite mug.

You are missing most of the fun. In the olden days, the cables connected
various 19 inch rack-mounted equipment.

When you had to move stuff to a demo, various ill-formed and smelly people
arrived with dollies to take your precious and delicate instrumentation down 40
flights of stairs into a waiting van. From there, up 10 flights to the demo
area.

It never made it. Vibration turned the 1 inch GPIB cable into a torture rack,
severely distorting any connector that happened to be in the way, and rendering
useless any equipment mistakenly connected during the transit.

I think Terry's path towards a TUSB uC may be a good one. Just imagine,
replacing the garden hose with a USB cable from Walmart. Also, the cost
for a 6ft cable would go from $100 or so to less than $10.

I am in great favor of such moves. The next best thing would be a battery
powered wireless connector that mounted on the rear of a gpib connector. Speff
described such a device recently. The Zigbee. Perfect!

The only downside with USB is that its EMI performance is, well, not the
best at times. RS232 would be better but many "modern" PCs don't have
serial or parallel ports anymore. Oh well, out here we just keep a stack
of ferrites handy.

Regards, Joerg

Keep looking. There will always be someone who makes AT-compatible interfaces.

I invested in 10 AT keyboards about 20 years ago, and they are still working.

So it is logical to assume I will still be able to buy motherboards that will
connect to them. Otherwise my investment was for nought.

Hehe

Mike Monett

 

[Joerg]

Hello Mike,

Keep looking. There will always be someone who makes AT-compatible interfaces.

Maybe but when you have to use a laptop it's not that way. All but a few
very high end (and thus high $$) laptops have done away with legacy
ports. All they have is a few USB ports. You can buy cheap USB-parallel
adapters but when it comes to two-way communication their performance
sometimes becomes a bit rocky.

I invested in 10 AT keyboards about 20 years ago, and they are still working.

Did they appreciate in value? ;-)

Regards, Joerg

http://www.analogconsultants.com

 

[Joerg]

Hello Terry,

Now that I have the 3585, I can do all of my conducted emissions stuff.
When I eventually get an HF speccy, it only needs to go down to ~ 10MHz,
which (hopefully) means it'll cost less.

If it's for radiated EMI it won't need to go much past 1GHz either and
the LO can be fairly loose.

What I am surprised about is that nobody seems to have come out with a
"faceless" analyzer. Just a USB pod with a BNC jack would be fine. Or a
glorified WLAN card with pre-converter and a BNC jack.

Regards, Joerg

http://www.analogconsultants.com

 

[keith]

On Mon, 27 Jun 2005 00:47:37 +0000, Joerg wrote:

Hello Terry,

Now that I have the 3585, I can do all of my conducted emissions stuff.
When I eventually get an HF speccy, it only needs to go down to ~ 10MHz,
which (hopefully) means it'll cost less.

If it's for radiated EMI it won't need to go much past 1GHz either and
the LO can be fairly loose.

Huh? The FCC requires testing well above 1GHz. WHen I was doing
pre-compliance testing in the late '90s we had to show up to 6GHz (ten
times our fastest clock).

What I am surprised about is that nobody seems to have come out with a
"faceless" analyzer. Just a USB pod with a BNC jack would be fine. Or a
glorified WLAN card with pre-converter and a BNC jack.

There is no substitute for an analog spectrum analyzer. Any digital link
would simply add a perceptable delay in the control response.

--
Keith

 

[Mike Monett]

Joerg wrote:

Hello Mike,

Keep looking. There will always be someone who makes AT-compatible interfaces.

Maybe but when you have to use a laptop it's not that way. All but a few
very high end (and thus high $$) laptops have done away with legacy
ports. All they have is a few USB ports. You can buy cheap USB-parallel
adapters but when it comes to two-way communication their performance
sometimes becomes a bit rocky.

Yes, I bow to the inevitable. We can still by motherboards with
At-keyboard interface and serial and parallel ports at the local computer
store. When that ends, I will have to scour the neighbourhood on garbage
days and grab all the old computers that get thrown out.

So you see, it's a good thing I have learned to run on Win 3.11 and DOS. I
can run on anything anyone wants to throw out.

I'm not kidding. My main computers died a while ago, and I am currently
running on a 200MHz machine, with Linux and Win 95 on other ex-discarded
machines. I also have a 66MHz 486 I'm looking forward to placing in
service to do some control work.

I invested in 10 AT keyboards about 20 years ago, and they are still working.

Did they appreciate in value? ;-)

Of course. To me, they are pure gold. I've been running this one for over
10 years. If the rest have the same reliability, I can expect another 100
years or so of excellent performance. Since I am somewhat over 60, that
should more than satisfy the requirement

Regards, Joerg

Mike Monett

 

[Keith Williams]

In article <1LFve.130$0V3.33@newssvr13.news.prodigy.com>,
notthisjoergsch@removethispacbell.net says...

Hello Mike,

A major problem with gpib is the interface is wire-or, ...

The main problem I found with GPIB are those unwieldy garden hose type
cables. Once I unhooked one and sure enough it recoiled, sending my half
filled coffee mug sailing, crashing and spilling the contents all across
the lab floor. On top of that it was my favorite mug.

Not to mention that they didn't always work. Some I had to wire in a
star, others daisy-chained to get the damned instruments to talk. I
always hated GPIB!

I think Terry's path towards a TUSB uC may be a good one. Just imagine,
replacing the garden hose with a USB cable from Walmart. Also, the cost
for a 6ft cable would go from $100 or so to less than $10.

I'd prefer Cat-5 and Ethernet. Indeed, I'd rather have Ethernet than
_any_ USB. Connecting a device across the Internet would be a little
easier too. ;-)

The only downside with USB is that its EMI performance is, well, not the
best at times. RS232 would be better but many "modern" PCs don't have
serial or parallel ports anymore. Oh well, out here we just keep a stack
of ferrites handy.

--
Keith

 

[Keith Williams]

In article <42BF56E8.B3E@spam.today>, no@spam.today says...

Joerg wrote:

Hello Mike,

Keep looking. There will always be someone who makes AT-compatible interfaces.

Maybe but when you have to use a laptop it's not that way. All but a few
very high end (and thus high $$) laptops have done away with legacy
ports. All they have is a few USB ports. You can buy cheap USB-parallel
adapters but when it comes to two-way communication their performance
sometimes becomes a bit rocky.

Yes, I bow to the inevitable. We can still by motherboards with
At-keyboard interface and serial and parallel ports at the local computer
store. When that ends, I will have to scour the neighbourhood on garbage
days and grab all the old computers that get thrown out.

I'm a little confused here. AT keyboards and PS2 keyboards are
compatible, electrically. ...just buy an adapter (either way) for $5
(I just threw a bunch out last week).

So you see, it's a good thing I have learned to run on Win 3.11 and DOS. I
can run on anything anyone wants to throw out.

I'm not kidding. My main computers died a while ago, and I am currently
running on a 200MHz machine, with Linux and Win 95 on other ex-discarded
machines. I also have a 66MHz 486 I'm looking forward to placing in
service to do some control work.

I invested in 10 AT keyboards about 20 years ago, and they are still working.

Did they appreciate in value? ;-)

Of course. To me, they are pure gold. I've been running this one for over
10 years. If the rest have the same reliability, I can expect another 100
years or so of excellent performance. Since I am somewhat over 60, that
should more than satisfy the requirement

Nothing but Model-M's here. This one is over 16 years old (Feb 17,
1989). ;-)

--
Keith

 

[Joerg]

Hello Keith,

If it's for radiated EMI it won't need to go much past 1GHz either and
the LO can be fairly loose.

Huh? The FCC requires testing well above 1GHz. WHen I was doing
pre-compliance testing in the late '90s we had to show up to 6GHz (ten
times our fastest clock).

Depends on the product. The run of the mills stuff is often tested 30MHz
to 1GHz radiated and 150kHz to 30MHz conducted. Most of the time there
isn't anything to write home about above 500MHz.

What I am surprised about is that nobody seems to have come out with a
"faceless" analyzer. Just a USB pod with a BNC jack would be fine. Or a
glorified WLAN card with pre-converter and a BNC jack.

There is no substitute for an analog spectrum analyzer. Any digital link
would simply add a perceptable delay in the control response.

True. However, for pre-compliance I often like small portable equipment.
It doesn't have to be fancy. I shoot for 10-20dB margins which typically
makes the product fly through the UL lab in a jiffy.

Regards, Joerg

http://www.analogconsultants.com

 

[Joerg]

Hello Mike,

Of course. To me, they are pure gold. I've been running this one for over
10 years. If the rest have the same reliability, I can expect another 100
years or so of excellent performance. Since I am somewhat over 60, that
should more than satisfy the requirement

Not necessarily. I know some folks here in town who are at least 20
years older and are happily pounding the keyboard.

Regards, Joerg

http://www.analogconsultants.com

 

[Joerg]

Hello Keith,

I'd prefer Cat-5 and Ethernet. Indeed, I'd rather have Ethernet than
_any_ USB. Connecting a device across the Internet would be a little
easier too. ;-)

Agree. I always wondered why USB was conceived in the first place. It
does have one advantage: Lots of DC power. Then again, two of the four
CAT-5 pairs are usually vacant but the Ethernet standard lacked any
power specs for a long time.

Regards, Joerg

http://www.analogconsultants.com

 

[Terry Given]

Joerg wrote:

Hello Keith,

If it's for radiated EMI it won't need to go much past 1GHz either
and the LO can be fairly loose.


Huh? The FCC requires testing well above 1GHz. WHen I was doing
pre-compliance testing in the late '90s we had to show up to 6GHz (ten
times our fastest clock).


Depends on the product. The run of the mills stuff is often tested 30MHz
to 1GHz radiated and 150kHz to 30MHz conducted. Most of the time there
isn't anything to write home about above 500MHz.

What I am surprised about is that nobody seems to have come out with a
"faceless" analyzer. Just a USB pod with a BNC jack would be fine. Or a
glorified WLAN card with pre-converter and a BNC jack.


There is no substitute for an analog spectrum analyzer. Any digital link
would simply add a perceptable delay in the control response.


True. However, for pre-compliance I often like small portable equipment.
It doesn't have to be fancy. I shoot for 10-20dB margins which typically
makes the product fly through the UL lab in a jiffy.

Regards, Joerg

and best not to think about how the display appears on the 3577 or 3585.

Cheers
Terry

 

[Terry Given]

Joerg wrote:

Hello Keith,

I'd prefer Cat-5 and Ethernet. Indeed, I'd rather have Ethernet than
_any_ USB. Connecting a device across the Internet would be a little
easier too. ;-)


Agree. I always wondered why USB was conceived in the first place. It
does have one advantage: Lots of DC power. Then again, two of the four
CAT-5 pairs are usually vacant but the Ethernet standard lacked any
power specs for a long time.

Regards, Joerg

In the mid '90s I looked very closely at USB, thinking to use it for a
serial bus within high-power motor controllers, so all IO devices could
be separate cards and plugged in at will.

One detailed read of the protocol was enough to kill that idea. What a
dog! Trust computer scientists to pick the most complex method of doing
anything.

Cheers
Terry

 

[Joerg]

Hello Terry,

and best not to think about how the display appears on the 3577 or 3585.

The 3577 can't do spectrum display. Wish it could, that would free up
some lab rack space although its fan creates quite a racket.

Regards, Joerg

http://www.analogconsultants.com

 

[Joel Kolstad]

Terry,

"Terry Given" <my_name@ieee.org> wrote in message
news:jm1we.10830$U4.1396447@news.xtra.co.nz...

One detailed read of the protocol was enough to kill that idea. What a
dog! Trust computer scientists to pick the most complex method of doing
anything.

What didn't you like about it? I've used USB before, and while it is somewhat
complex, if you're actually trying to achieve everything that USB does (power
control, plug and play, guaranteed delivery as well as bulk delivery, "dumb"
modes for initial boot of PCs with keyboards and mice as well as the standard
"smart" modes, multiple speeds, etc.), I think that much of the complexity is
needed. (I'd say the one place they got carried away was with control
reads/writes... and once they let the "software guys" loose on defining stuff
like HID, but that's no an inherent problem with USB itself).

---Joel