HP-IL over Cat5 - Compatible?



#2

The HP-IL spec says cable can be up to 100 meters long and that it must have 100 Ohms impedance. I measure ~50 Ohm resistance on a single strand of Cat5 twisted pair cable that is many 10s of meters long. The Cat5 spec calls for 100 Ohm impedance, so I assume that's the sum of two strands or 1 pair. The HP-IL spec calls for sheilded twisted pair on lengths over 10 meters.

My question is, despite the HP-IL spec, has anyone had success using UTP (Cat5 or other) to connect HP-IL devices? I'd like to save money on the cable, but I'm not willing to go hacking on my existing cables without some idea the project may work. 8)


#3

The '100 ohms' impedance of the HPIL cable refers to the characteristic impedance of the cable, and not the DC resistance of one of the conductors. The chararacterist impedance is more a property of the dielectric (insulator) and the dimensions of the cable.

You cannot measure characteristic impedance with a DC ohmmeter (which is what I assume you used) -- about the only intstrument that will measure it is a time-domain reflectometer.

I can't remember the characteristic impedance of Cat5, but it should be a 'standard' value. For the use that Cat5 is normally put to, it would be a rather important parameter. I would guess it's somewhere around 100 ohms, but I wouldn't bet on it.


#4

Yes, the spec says 100 Ohms.

I'm a rank amateur when it comes to electronics. I read the wikipedia entry on impedence, saw where it mentioned that for a steady state circuit, the impedence == resistence (since inductors are closed and capacitors are open.) and assumed that if the Cat5 spec said 100 Ohms, it meant it was the same over the operating range of the cable. I tested the resistence for the only data point I could collect, since I don't have a TDR, or a function generator or a test oscillator. But I think the spec may save me trouble there.

The question I have is the STP vs UTP issue. 100 Ohm STP is scarce. Most STP is 150 Ohm stuff (for token ring I think.) There's 100 Ohm "Screened Twisted Pair" (ScTP) which has a shield around the cable bundle, but the pairs themselves aren't shielded. If the HP-IL connectors were commonplace, I wouldn't hesitate to hack a couple of cables to pieces and just give it a shot. But while you can scare them up from time to time, and while I do have more than I presently need, the fact that they aren't manufactured anymore makes me reluctant to start using the side cutters.

#5

I suspect that Cat 5 UTP should work fine, but I haven't tried it.

I think the requirement for shielding is a case of HP being overly conservative. UTP should be fine. Using non-twisted cable for more than 10 meters might well be problematic though.

Where are you going to get connectors for the ends? Or are you just trying to lengthen some existing cables?


#6

I was planning on hacking on two existing cables to produce a single lengthy one with two connectors on each end. The idea is to extend the loop from my computer room to the bedroom. Sort of an HP-IL LAN. 8)


#7

Hi Howard.

Tony is right. But all this is somewhat beside the point for you.

I would use flexible copper wire rather than CAT-5 which is semi-rigid single core wire and as the connectors may move around a bit you may end up with an open circuit and no loop. Thats not all that good for the HP-IL driver chip in the long run.

If you do not live near a radio station and you make sure your
MOBILE PHONE is NOWHERE NEAR the loop (!), then you can use very light speaker cable for the HP-IL. The impedance is not dead rightbut it will work fine over the length you are talking about.

The HP-IL is designed to use in the control of among other things, RF (radio frequency) measurement equipment. This equipment or more properly the stuff it tests, generates interference which would stop HP-IL dead in it's tracks and probably do damage if the power level was high enough.

Since I assume you are not testing RF equipment at home twisted pair or even speaker wire shuld be fine for you at home.

If you are nervous then you can use twin core shielded audio coaxial cable. But be sure the shield is connected to ground and both signals wires are not connected to anything but the HP-IL pulse transformers. If you have a technical (good electronics repairman) friend with a CRO they can probably help you if the
link doesn't work. Electrical "Reflections" can be seen (as ringing) using a CRO.

I would not use UTP CAT-5 cable for this app. myself.
Just my opinion.

Further quiries, just write to me.

Don Wallace
donwallace63@yahoo.com.au

P.S. Hope this helps.


#8

Hi Howard.

To clarify some omissions I made:

Very light speaker wire (instead of speaker wire).

RF equipment would stop HP-IL OPERATING OVER UNSHIELDED
CABLE dead in it's tracks. That's why the HP-IL ssytem uses
shielded cable.

HP-IL operates over a BALANCED LINE. In plain english this is two wires (shielded) which are both "floating" with respect to "ground" (a reference point) and "driven" with respect to EACH OTHER.

That is, you don't have one wire connected to ground.

When a "1" and then a "0" is sent, you get, on the cable:
_
WIRE#1 ----------/ \--\_/------------
_
WIRE#2 ----------\_/--/ \------------

This is called a "differential signal" over a "balanced transmission line". It is inherently IMMUNE to NOISE.
The impedance you are talking about (cable impedance), is
the high frequency IMPEDANCE the two wires present to the
pulse transformers. It's dependent on the insulating medium
between them (which determines the CAPACITANCE) and the
INDUCTANCE of the wires themselves...

Try the Wikipaedia on these terms, as you already know,
it's really cool.

Please forgive my rough ASCII ART... ;-)

Don W


When a


#9

Thanks for the information, Don.

I have some questions.

  • The short haul (<10 meters) HP-IL cable isn't shielded, is it? The cable feels very flexible, and the wires are quite thin.
  • Your opinion on CAT-5 is that the solid core makes it unsuitable for mating to the HP-IL connectors? Or is there something else that makes it problematic?
  • What is the issue in the gauge of the wire? Why would thin speaker wire be more suitable?

Thanks for the pointer to Wikipedia on impedence. It was helpful to reread the entry. I was already familiar with "balanced" circuits from using microphones with that feature. Also I read the discussion in the HP-IL spec that emphasized the good immunity to noise this offers.

Wikipedia to the rescue on "CRO"! In the states, we tend forget the "Cathode Ray" part and just say "Oscilloscope". If we want to shorten it, "scope" is often used.


Regards,
Howard


Edited: 24 Aug 2005, 10:35 a.m.


#10

Hi Howard.
You wrote:
(my replies interspersed like so)

I have some questions.

The short haul (<10 meters) HP-IL cable isn't shielded, is it?
(Don't know, don't own one. Perhaps not but I'm guessing)

The cable feels very flexible, and the wires are quite thin.
(It could still be very light coax)

Your opinion on CAT-5 is that the solid core makes it unsuitable for mating to the HP-IL connectors?
(yes)

Or is there something else that makes it problematic?
(no)

What is the issue in the gauge of the wire?
(convenience and it's light current use so thin gauge is fine)

Why would thin speaker wire be more suitable?
(It's CHEAPER!...)

Thanks for the pointer to Wikipedia on impedence. It was helpful to reread the entry. I was already familiar with "balanced" circuits from using microphones with that feature. Also I read the discussion in the HP-IL spec that emphasized the good immunity to noise this offers.
(Cool! I'm into audio too.)

Wikipedia to the rescue on "CRO"! In the states, we tend forget the "Cathode Ray" part and just say "Oscilloscope". If we want to shorten it, "scope" is often used.
(ROFLMAO. I must be showing my "age" ;-)

Regards, Howard
(Cheers, happy HP-ILing... I'm a smidgen jealous...)


#11

Thanks, Don! You've been a big help.

#12

I've used unshielded HPIL in the control of an RF test setup where we were doing repetitive production testing power amplifiers at 175MHz. No problem at all. Cell phones start at about 900MHz IIRC, and the HPIL transformers and circuitry would be much too slow to respond to, or interfere with, the cell phone stuff.

Characteristic impedance (Zo) is the impedance a cable will appear to have for RF when the load is at that impedance, regardless of length. IOW, for 50-ohm load out at the end of a 50-ohm cable will still appear to be 50 ohms, no matter how long the cable is. If the cable is sufficiently long, there will be a substancial amount of power lost in it, and of course it will introduce the propagation delay; but you'll still see 50 ohms looking into it. If the load is something other than the cable's Zo, some of the power fed into it will be reflected. For steady-state RF, you can use the Smith chart to determine the various effects like SWR, return loss, and the complex impedance looking into the cable. If the cable is long enough to present substancial losses, you'll need the attenuation data to get the right results with the Smith chart. What's more important with the fast digital work however is the reflections you get when the transmission line's Zo does not match the load. Those reflections will cause unwanted triggering of registers or otherwise be erroneously interpreted as data and cause problems. The cable would have to be pretty long to cause this kind of trouble with the slow edge rates of HPIL.

I have not tried to measure HPIL's voltages, rise times, or load impedances, but I don't think it would be difficult to get an extension cable to work. The transmission-line losses of 20 meters at the frequencies of interest will be minimal. IIRC, cat-5 cable is 100 ohms, which I think is what someone said HPIL wants. (I can't see the post now while writing the response.)


#13

Hi all. Hi Garth.

In my posts to Howard I was being *exceedingly* cautious as
I don't own any HP-IL gear.

The cellulars are very very (ultra ;-) high frequency.
However, I have noticed serious interference with computer
monitors (internal frequency response in the low MHz range for
the gun drivers and deflection circuit operating frequency
range of up to only say 40 kHz) from cell phones at close range.

Cell phones, as they are a spread-spectrum device and they frequency hop slowly, do give some LOW FREQUENCY MODULATION
of the RF.

Also, the cell phones use data compression and idle their transmitters when there's no speech to save power, so the transmitters operate in a burst mode most of the time.
The resultant "keying" of the RF on and off does generate a
lot of hash. Also, they generate surprising pulsed levels
of power, maybe half a watt or so?

I would imagine that one would not have trouble unless the mobile was within six inches (maybe three?) of the HP-IL cable.
The wire conductors are close together in the cable so the differential signal would be quite small.

I also forgot about the effect the pulse transformers would have
as effective RF chokes in limiting the transfer of power to the
HP-IL chips. However I'm not sure about the possibiliy of some capacitive coupling... but the calc is not grounded so it shpouldn't matter...

Once in the HP-IL driver chip output driver, any RF would get recitifed by the junctions and you would need a bit of power to do damage, but how much power would get generated I am not sure.
By that time of course, your link would not be still working as
the data bits would be too noisy.

But yeah Garth you are dead right of course. You used it there so you would know. I was thinking of you and your worklab as I wrote that post.

Don W


#14

I forgot something:

Speaker wire should wr\kr just fine.

On a different subject related to SHIELDING,
many years ago as a fellow student
one my my mates at uni told me that while working in the high voltage lab his hp41 lost all memory contents while in it's pouch. he asked me what could done about it.

I told him to line the pouch with Comalco ALFOIL.
He did so and reported compete success (!)

The alfoil cooking foil acted as a simple Farady Cage ;-)

Don W

#15

Well, it works.

To all you hard core EE types, let me just say, I'm glad you're here. I'm a software guy. I figured it would be fun and interesting to hack on hardware as a hobby. But I lack skills in that area. Now, I'm not talking board design or anything like that. I'm talking basic electronic tech manual skills, like soldering. I've actually gotten fairly OK for crude work, once I got myself a temprature controlled iron and figured out that it has to be really hot to do any good. But I still appreciate the help and input I get from the more experienced folks here.

I happen to have a spool of Cat5 here in my work cabinet. That, and the fact that the specified impedence of Cat5 matches the HP-IL's spec, and that for lengths over 10 meters, that same spec calls for twisted pair are what motivated me to consider this approach. Don's warnings regarding the problem of mating the solid core Cat5 wire with the stranded HP-IL cable gave me pause, however. So I figured I'd solder decapitated HP-IL connectors onto Cat5 cable, and cover them with shrink wrap in an attempt to protect the awkward mechanical mating of the dissimilar wires. Through trial and error, I found that the following process worked pretty well:

  1. Cut the heads off an HP-IL cable. (This is one of those "puckering" moments for an inexperienced electronic tech who is jumping into the unknown!)
  2. Strip about 3/4" off the ends of a CAT5 pair and both wires of one end of the decapitated HP-IL cable.
  3. If this isn't your first end, mate the other end of the HP-IL to the one you are about to solder. This makes it easier to be sure you solder the right wires.
  4. Mount the wires to be soldered in some sort of frame. I use one of those el-cheapo soldering stations with the magnifying glass and alligator clips on rails. I remove the glass (since I use a larger illuminated one) and put each end of the work into one of a pair of clips. On the HP-IL side, the hard plastic connectors are a good place to appy the clip. On the Cat5 side, the cable itself is OK.
  5. If this isn't your first end, check the end you already did to see which one of the Cat5 wires to solder to which HP-IL cable. The connections should be "straight through" as the mated cables will show.
  6. Cut a two inch piece of 3/16" shronk wrap tubing and pass the Cat5 pair through it.
  7. cut two one inch sections of 1/16" shrink wrap and thread one of the wires through each.
  8. The bundle of shrink wrap should be able to move at least an inch away from the work, because you are about to apply lots of heat there. 8)
  9. Maneuver the stripped ends so that they overlap each other on the same plane
  10. Bend the Cat5 wire around the HP-IL wire. This is the opposite of what the wires "want" to do. The reason is, the solid core Cat5 wire will retain its shape, which the soft HP-IL cable will conform to. This means that it will stay in place, which allows you to solder with both hands free.
  11. Make only two to three revolutions around the HP-IL cable depending on the length of the stripped wire. This makes a nice flat arc to place the soldering iron tip on.
  12. Heat your iron to above 600 degrees, Fahrenheit. I use 640. Tin the iron.
  13. Place the soldering tip underneath the arc of the entwined wires at one end of the arc.
  14. Count to three, moderately slowly
  15. Apply the solder to the top of the wires, above where the iron is placed. It should immediately melt.
  16. draw the iron along the underside of the wire bundle while dragging the (shortening line of) solder along the top.
  17. Remove the iron and solder from the work.
  18. After cooling and inspection, trim off excess wire at both ends of the new join.
  19. If this is your second connector, you have reached your earliest opportunity to test the cable. If you find a problem here, you'll save youself some work compared with finding it later. Test continuity through both conductors. You got the right ones? Good. Be sure the cable doesn't short at either end and plug it in to a test bed. I used EMU71 running on a PC to CAT a real 9114 on a loop that included my cable.
  20. Slide the small shrink wrap sections over the soldered wires and heat them until snug. (I use a cheap heat gun which works a lot better than the iron.)
  21. If you don't get all the conductor covered with the shrink wrap (oops) dress it with electrician's tape.
  22. Slide the larger shrink wrap over the two wrapped joins and shrink it into place.
  23. Test the cable again!
  24. You probably want to HP-IL out and back with the same cable. If so, do a second pair like the first on the same cable. (But with the gender switched, of course. You should have both a male and a female HP-IL connector at each end.)

And it works with about a 45' cable!

But that joint is still brittle. The shrinkwrap doesn't seem to do the trick. The stress occurs in the Cat5 copper just on the edge of the solder joint. Any ideas?

Update: It was brittle while I was trying to appy the shrink wrap. After I got that applied, I had no further trouble. It's undoubtedly still brittle, but protected somewhat by the plastic tubing. A radical bend at the end of the solder joint could still ruin my whole day, so I'm careful with it.

Edited: 26 Aug 2005, 8:56 p.m. after one or more responses were posted


#16

USE THIN TWIN PLASTIC COVERED SPEAKER CABLE-!

Don
P.S.
CAT-5, for HP-IL, INDEED! (just fooling about the mock indignation)
Sounds like someone doesn't trust me!


#17

Well, I understood your objection, and tried to design around it.

The main deal for me is the fact that it's twisted pair, and the spec calls for that over 10 meters. The fact that the impedence is right also influenced me. And also there's the fact that Cat5 packs eight conductors - four pairs - into the cable. That means I only have to run one 50' cable instead of two. Plus, I have that big 'ol spool. 8)


#18

Hey, man, it's cool ;-)

dw

#19

Hi Howard.

General info for intrepid hobbyists like yourself (good onya!):

FYI, a soldering iron should be set at very close(+/- 5deg.)
to 310 degrees celcius (deg.F? where's my calculator ;-)

The drama is measuring a temperature that "high" if you don't have
any industrial thermometer gear...

The main rules for soldering are:
1: Cleanliness IS godliness: (keep your wires, iron [and for Mil-Spec work, the solder as well] clean)

2: DO NOT MOVE the wires before solder has set.
3: Avoid opens and shorts if at all possible!
4: Avoid wiring errors (!)
5: Avoid over-heating semiconductors (!)

Cheers,

Don


#20

Thanks for the info, Don.

Quote:
FYI, a soldering iron should be set at very close(+/- 5deg.)
to 310 degrees celcius (deg.F? where's my calculator ;-)


That works out to 590 degrees F. (599 F for 315 C.) So I'm "a bit" high in my estimate. 8)

I assume this has something to do with the melting point of standard rosin-core solder? Do different solders call for different temperatures, or is it more material dependent?

Quote:

The drama is measuring a temperature that "high" if you don't have
any industrial thermometer gear...


I use a soldering station from Radio Shack with temperature sensing built in. It was under $100.00, so I imagine it isn't the highest quality obtainable.

Quote:
The main rules for soldering are:

1: Cleanliness IS godliness: (keep your wires, iron [and for Mil-Spec work, the solder as well] clean)

2: DO NOT MOVE the wires before solder has set.

3: Avoid opens and shorts if at all possible!

4: Avoid wiring errors (!)

5: Avoid over-heating semiconductors (!)


With the procedure I outlined here, my soldering tip is nowhere near an IC, thank goodness! I'm working my way up to dealing with ICs and PCBs slowly.

Quote:

Cheers,

Don


Thanks for the tips!

Edited: 1 Sept 2005, 2:53 p.m.


#21

Hi Howard.

In answer to your last post:

H.O.:
That works out to 590 degrees F. (599 F for 315 C.) So I'm "a bit" high in my estimate. 8)

D.W: That's ok. Better a ittle hot than a little cold...

H.O.:
I assume this has something to do with the melting point of standard rosin-core solder? Do different solders call for different temperatures, or is it more material dependent?

D.W.: Without boring people here, Material Science is interesting if you don't have to study to pass an exam on it.
Solder (for electronics) is an alloy of lead and tin.
When mixed these metals change both the melting point and the
characteristics of the alloy when "freezing".

The solder for electronics work is about 60% lead 40% tin. Ratios vary a bit but not much. The reason for this is that this ratio
constitutes a "eutectic mix" of the two elements. A eutectic mix has the special property that the two components (lead / tin)
both "freeze at the same time" put simply. This means the sloder will solidify rapidly and the chance of slight movement while this is happening causing "crystalline" or "cold" solder joints is minimized.

For plumbing you want the solder to go plastic for a while so it can be "wiped" (c.f. lead wiping from fifty years ago).

Brazing is similar to soldering but the metals are different
(copper/tin=bronze [from memory]) but the process is very similar just at much higher temperature.

H.O.:
I use a soldering station from Radio Shack with temperature sensing built in. It was under $100.00, so I imagine it isn't the highest quality obtainable.

D.W.: That's fine. Once you get your "chops" you shuld be able to do anything with that iron...

H.O.:
With the procedure I outlined here, my soldering tip is nowhere near an IC, thank goodness! I'm working my way up to dealing with ICs and PCBs slowly.

D.W.: Suggestion; get some old electronics and try removing components. This is good practice and you will improve with
time.

Cheers,

Don W


#22

I keep forgetting things...

Plumbers lead: a different ratio of lead and tin...

Brass=copper+tin?
Bronze=copper+zinc?

For practice soldering easy stuff (cable terminbations and
audio attenuators with resistors) is good to start with.
That hadrware is very forgiving and if you use your fingers to hold the parts you will rapidly get an idea f what "too much heat for too long" means...

Non-destructive desoldering can be quite hard to do when you are beginning...

Final note: When soldering or desoldering, the thing is
JOINT TEMPERATURE. This is dependent on good thermal conduction from iron tip to joint metal (both parts). So have enough solder
to ensure the joint gets hot (the solder blob conducts the heat...)
but not too much so you're swimming in it...

Have fun...

Don W


#23

Thanks for all the good advice, Don.

I do have plenty of old PCBs around that I can practice on. I just haven't squared the 'round tuit' up to now. I have practised desoldering using braid. I took apart a 41C single memory module and desoldered the PCB from the contact fingers. That went pretty well. And I've done several cables and odds and ends. My intermediate goal is to be proficient enough to do a 2X memory module using Cristoph Klug's instructions. That requires soldering jumpers between PCB's, on parts that average $50.00 apiece on eBay, and I need two! So I'm not going to attempt that until I've got concrete evidence I have a good chance of succeeding. 8)


Possibly Related Threads...
Thread Author Replies Views Last Post
  How to move lexfiles from PC to 71 w/o HP-IL? Joe Horn 9 442 10-18-2013, 03:50 PM
Last Post: Christoph Giesselink
  Hand Held Products RS232 to HP-IL aj04062 11 467 08-31-2013, 07:12 PM
Last Post: Paul Berger (Canada)
  HP IL over wifi ... (ILPer & go71b) Olivier De Smet 12 537 08-20-2013, 05:44 AM
Last Post: Olivier De Smet
  Stupid idea of the day: 41-compatible watch bhtooefr 0 125 03-31-2013, 08:23 AM
Last Post: bhtooefr
  HP-IL beginners question John Abbott (S. Africa) 4 235 05-28-2012, 05:45 PM
Last Post: Geoff Quickfall
  HP-IL SCSI Interface Richard Wagoner 15 504 02-19-2012, 07:39 PM
Last Post: Richard Wagoner
  HP 82165A HP-IL/GPIO Interface VS HP 82164A HP-IL/RS-232C Interface VS HP 82166A HP-IL Converter Fouad M. Kaadou 9 349 01-15-2012, 03:45 AM
Last Post: yves
  HP-IL ftp files robertoataulfo 4 221 12-04-2011, 04:34 PM
Last Post: robertoataulfo
  200LX Ram cards compatible with 48GX? mr-scorpio 6 316 11-11-2011, 07:40 PM
Last Post: Raymond Del Tondo
  HP7470A plotter with the HP-IL interface Richard L. Taylor 2 157 08-03-2011, 03:20 AM
Last Post: Juergen Keller

Forum Jump: