Re: Kludging together a TDR or similar?
It would be help full to know how long the cable is
and how close to the problem you need to get
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] [mailto: [email protected]] On Behalf Of cheater cheater Sent: Thursday, June 28, 2018 7:32 AM To: [email protected]Subject: [TekScopes] Kludging together a TDR or similar? Hi everyone, I have a DOCSIS 3 150 Mbps cable modem which just stopped working after a substantial storm (i replaced it to make sure). Apparently the cable is broken. While I'm waiting for the technician to show up next week, I was wondering if anyone had an idea how I could kludge together a TDR to see if I can find where the break might be? Am I right to think that you just inject a pulse from a sig gen and measure the delay time to find out where the break is? Do I need to use any specific bandwidth equipment, or will anything do? Is there an easier way to get this done? I don't have a TDR, but I have some 7000 series scopes and some non-tek sig gens that go up to iirc 20 MHz. The tek mainframe's pulse gens will go higher (IIRC I have a 7904), but not sure if I can use those for this purpose. Thanks
|
Re: Kludging together a TDR or similar?
As Stefan said, how fast a step pulse you need and how much scope bandwidth you need both depend on how much resolution you need. Do you happen to have sampling plug-in set for your 7000 series mainframes? If not, how much scope bandwidth do you have available? Also, how fast a risetime can you generate in a pulse?
Also as Stefan mentioned, if you can get down to a nS, you should be able to get pretty close!
I'd recommend (for the long term) that you keep an eye on ebay for a good deal on a Tektronix 1502 TDR. One of the handiest units I have. A person can't have too many TDRs though! Check out a 7S12 when you have time. Absolutely delightful! :-)
Best of luck with the cable!
Tom
|
Re: Kludging together a TDR or similar?
Thanks. Does the propagation time depend on the coax type?
i don't have the parts to build the circuit right now - could I use the
mainframe's pulse generator?
toggle quoted message
Show quoted text
On Thu, 28 Jun 2018, 19:50 stefan_trethan, <stefan_trethan@...> wrote: Depending on the length you need a reasonably fast pulse, there are
example circuits on the web that use standard logic drivers (extra
fast AC stuff I believe) with a bunch of outputs tied together.
<>
With a 100+ MHz scope you should see something, if the cable is a
reasonable transmission line.
A foot per nanosecond for the speed of light, (only use I ever had for a
foot).
In coax the foot is shorter, only about 2/3rds or so.
ST
On Thu, Jun 28, 2018 at 4:31 PM, cheater cheater <cheater00@...>
wrote:
Hi everyone,
I have a DOCSIS 3 150 Mbps cable modem which just stopped working after a
substantial storm (i replaced it to make sure). Apparently the cable is
broken. While I'm waiting for the technician to show up next week, I was
wondering if anyone had an idea how I could kludge together a TDR to see if
I can find where the break might be? Am I right to think that you just
inject a pulse from a sig gen and measure the delay time to find out where
the break is? Do I need to use any specific bandwidth equipment, or will
anything do? Is there an easier way to get this done? I don't have a TDR,
but I have some 7000 series scopes and some non-tek sig gens that go up to
iirc 20 MHz. The tek mainframe's pulse gens will go higher (IIRC I have a
7904), but not sure if I can use those for this purpose.
Thanks
|
Re: Kludging together a TDR or similar?
i probably only need to be within a meter or two.
toggle quoted message
Show quoted text
On Thu, 28 Jun 2018, 20:04 cheater00 cheater00, <cheater00@...> wrote: Thanks. Does the propagation time depend on the coax type?
i don't have the parts to build the circuit right now - could I use the
mainframe's pulse generator?
On Thu, 28 Jun 2018, 19:50 stefan_trethan, <stefan_trethan@...> wrote:
Depending on the length you need a reasonably fast pulse, there are
example circuits on the web that use standard logic drivers (extra
fast AC stuff I believe) with a bunch of outputs tied together.
<>
With a 100+ MHz scope you should see something, if the cable is a
reasonable transmission line.
A foot per nanosecond for the speed of light, (only use I ever had for a
foot).
In coax the foot is shorter, only about 2/3rds or so.
ST
On Thu, Jun 28, 2018 at 4:31 PM, cheater cheater <cheater00@...>
wrote:
Hi everyone,
I have a DOCSIS 3 150 Mbps cable modem which just stopped working after a
substantial storm (i replaced it to make sure). Apparently the cable is
broken. While I'm waiting for the technician to show up next week, I was
wondering if anyone had an idea how I could kludge together a TDR to see if
I can find where the break might be? Am I right to think that you just
inject a pulse from a sig gen and measure the delay time to find out where
the break is? Do I need to use any specific bandwidth equipment, or will
anything do? Is there an easier way to get this done? I don't have a TDR,
but I have some 7000 series scopes and some non-tek sig gens that go up to
iirc 20 MHz. The tek mainframe's pulse gens will go higher (IIRC I have a
7904), but not sure if I can use those for this purpose.
Thanks
|
Re: Kludging together a TDR or similar?
Hi Damien,
The faster the edge of the pulse the finer the discontinuity you can resolve. Light travels 299,792,458 m/Sec in free space but in a cable it goes slower at about 66% to 70% of the speed in free space. For simplicity assume it goes 200,000,000 m/Sec, or 20 cm/nSec in your cable. Since the information about the break has to return back to you the pulse's round trip travel time is twice the one way distance so the correct figure is 1/2 * 20cm/nSec or 10cm/nSec.
To resolve a discontinuity 10cm long would take a pulse with a 1nSec rise time (approximately).
To resolve a 1cm long discontinuity requires a pulse with a 100pSec rise time.
The 7T12 TDR using an S-52 pulse generator head, which puts out a <=25pSec rise time pulse, would be capable of resolving 2.5mm discontinuities (approximately). In reality you have to take into account the S-6 sampling head speed (<=30pSec) as well. The combined system (S-52 and S-6) speed is Sqrt(25^2 + 30^2) = 39pSec so the 7S12 resolution is limited to approximately 4mm.
It is relatively easy to find pulse sources that put out sub-nanosecond rise time pulses. Unfortunately your scope will probably be the limiting factor in your ability to locate the fault. A 100MHz scope (risetine = 3.5nSec) will only be able to resolve to about 40cm.
Dennis Tillman W7PF
toggle quoted message
Show quoted text
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of
cheater cheater
Sent: Thursday, June 28, 2018 7:32 AM
To: [email protected]
Subject: [TekScopes] Kludging together a TDR or similar?
Hi everyone,
I have a DOCSIS 3 150 Mbps cable modem which just stopped working after
a substantial storm (i replaced it to make sure). Apparently the cable
is broken. While I'm waiting for the technician to show up next week, I
was wondering if anyone had an idea how I could kludge together a TDR
to see if I can find where the break might be? Am I right to think that
you just inject a pulse from a sig gen and measure the delay time to
find out where the break is? Do I need to use any specific bandwidth
equipment, or will anything do? Is there an easier way to get this
done? I don't have a TDR, but I have some 7000 series scopes and some
non-tek sig gens that go up to iirc 20 MHz. The tek mainframe's pulse
gens will go higher (IIRC I have a 7904), but not sure if I can use
those for this purpose.
Thanks -- Dennis Tillman W7PF TekScopes Moderator
|
Re: Kludging together a TDR or similar?
Since this is with respect to coax, it probably won't work very well; however, I found that an inexpensive little cable tracker (I got mine at Harbor Freight) is quite useful for finding breaks in open wire. It might be the cheapest "long shot" you could try.
Thanks,
Barry - N4BUQ
toggle quoted message
Show quoted text
----- Original Message ----- From: "cheater cheater" <cheater00@...>
To: [email protected]
Sent: Thursday, June 28, 2018 9:31:55 AM
Subject: [TekScopes] Kludging together a TDR or similar?
Hi everyone,
I have a DOCSIS 3 150 Mbps cable modem which just stopped working after a
substantial storm (i replaced it to make sure). Apparently the cable is
broken. While I'm waiting for the technician to show up next week, I was
wondering if anyone had an idea how I could kludge together a TDR to see if
I can find where the break might be? Am I right to think that you just
inject a pulse from a sig gen and measure the delay time to find out where
the break is? Do I need to use any specific bandwidth equipment, or will
anything do? Is there an easier way to get this done? I don't have a TDR,
but I have some 7000 series scopes and some non-tek sig gens that go up to
iirc 20 MHz. The tek mainframe's pulse gens will go higher (IIRC I have a
7904), but not sure if I can use those for this purpose.
Thanks
|
Re: Kludging together a TDR or similar?
If you have access to both ends, you can just use capacitance measurement if the center conductor is open - not shorted. Just disconnect both ends so there's no load shunting it, check for DC open-ness with a DMM. Assuming your cable is about 75 ohm type, the C should be around 25 pF/foot - I don't recall the exact figure, but you can it look up, and maybe get the exact spec for your cable type if you can identify it. If you can measure at both ends, you can get a better estimate of the location. Chances are that the problem is at a splice/connector, unless the cable is actually cut or burned open somewhere.
If you don't have a C-meter, trust me, it's a lot easier and cheaper to just get one, than to set up a TDR rig, especially for this kind of one-time use.
Ed
|
Re: Kludging together a TDR or similar?
Well, light in free space goes at about 30cm per nanosecond. It will go somewhat slower in the coax,
but to locate the break to 3cm accuracy you will need a pulse rise time of about 100ps, or less.
Suggest you have a look at Tek's Time Domain Reflectometry Measurements which you can download from
here
Craig
toggle quoted message
Show quoted text
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of cheater cheater
Sent: 28 June 2018 15:32
To: [email protected]
Subject: [TekScopes] Kludging together a TDR or similar?
Hi everyone,
I have a DOCSIS 3 150 Mbps cable modem which just stopped working after a substantial storm (i
replaced it to make sure). Apparently the cable is broken. While I'm waiting for the technician to
show
up next week, I was wondering if anyone had an idea how I could kludge together a TDR to see if I
can
find where the break might be? Am I right to think that you just inject a pulse from a sig gen and
measure the delay time to find out where the break is? Do I need to use any specific bandwidth
equipment, or will anything do? Is there an easier way to get this done? I don't have a TDR, but I
have
some 7000 series scopes and some non-tek sig gens that go up to iirc 20 MHz. The tek mainframe's
pulse gens will go higher (IIRC I have a 7904), but not sure if I can use those for this purpose.
Thanks
|
Re: Kludging together a TDR or similar?
Depending on the length you need a reasonably fast pulse, there are
example circuits on the web that use standard logic drivers (extra
fast AC stuff I believe) with a bunch of outputs tied together.
<>
With a 100+ MHz scope you should see something, if the cable is a
reasonable transmission line.
A foot per nanosecond for the speed of light, (only use I ever had for a foot).
In coax the foot is shorter, only about 2/3rds or so.
ST
toggle quoted message
Show quoted text
On Thu, Jun 28, 2018 at 4:31 PM, cheater cheater <cheater00@...> wrote: Hi everyone,
I have a DOCSIS 3 150 Mbps cable modem which just stopped working after a
substantial storm (i replaced it to make sure). Apparently the cable is
broken. While I'm waiting for the technician to show up next week, I was
wondering if anyone had an idea how I could kludge together a TDR to see if
I can find where the break might be? Am I right to think that you just
inject a pulse from a sig gen and measure the delay time to find out where
the break is? Do I need to use any specific bandwidth equipment, or will
anything do? Is there an easier way to get this done? I don't have a TDR,
but I have some 7000 series scopes and some non-tek sig gens that go up to
iirc 20 MHz. The tek mainframe's pulse gens will go higher (IIRC I have a
7904), but not sure if I can use those for this purpose.
Thanks
|
Re: Help needed with no trace no beamfinder on 465 (not b)
I will be selling or parting out this 465, and a 464. Anyone who wants them or parts should contact me NOW via PM or email.
Keith Ostertag
stuff@...
|
Kludging together a TDR or similar?
Hi everyone,
I have a DOCSIS 3 150 Mbps cable modem which just stopped working after a
substantial storm (i replaced it to make sure). Apparently the cable is
broken. While I'm waiting for the technician to show up next week, I was
wondering if anyone had an idea how I could kludge together a TDR to see if
I can find where the break might be? Am I right to think that you just
inject a pulse from a sig gen and measure the delay time to find out where
the break is? Do I need to use any specific bandwidth equipment, or will
anything do? Is there an easier way to get this done? I don't have a TDR,
but I have some 7000 series scopes and some non-tek sig gens that go up to
iirc 20 MHz. The tek mainframe's pulse gens will go higher (IIRC I have a
7904), but not sure if I can use those for this purpose.
Thanks
|
On Thu, 28 Jun 2018 at 07:53 Albert Otten <aodiversen@...> wrote: BTW: I had no patience to view the whole video. Bu I saw at 32-33 minutes
that the AC input voltages to the bridge rectifiers are measured. For +15
that was 1.3 V. I find this hard to believe. Apparently then there is
nearly a shortcut loading the transformer winding. I guess this would
immediately blow the main fuse?
Yeah, I rewound to watch this again. I don't think he was making contact, as you see the voltage he measures spike to ~1.5V with only one of the probes touching, and basically wandered around that value while he was poking at it.
|
I'm working on a 2215 with a current limit board. The HV section is very clean so Im questioning if the CRT is actually bad. I know about the CRT over voltage issues with this scope. Here is a picture of one of the waveforms from the auto intensity and z-axis board. It does not look correct to me but I dont know if this is actually a problem. This scope came to me with the fuse on the current limiting board blown. Thanks guys. /g/TekScopes/photo/61679/0?p=Name,,,20,1,0,0 This is a copy of the waveform from the manual. /g/TekScopes/photo/61679/1?p=Name,,,20,1,0,0
|
On Thu, Jun 28, 2018 at 04:53 am, Albert Otten wrote: BTW: I had no patience to view the whole video. But I saw at 32-33 minutes that
the AC input voltages to the bridge rectifiers are measured. For +15 that was
1.3 V. I find this hard to believe. Apparently then there is nearly a short circuit
loading the transformer winding. I guess this would immediately blow the main
fuse? I too struggled to watch the video, a bit too laconic for my taste I guess. I was somewhat put off by the ham-fisted removal of the bridge rectifier and the filter cap. Damaging the PCB because you didn't disassemble the 'scope to get at the PCB properly, why? The suggestion that a MJE3055 would be a suitable replacement for the series pass darlington seems somewhat far fetched to me, with no attempt to look at the specs to find a close replacement. (And is it just me or was the darlington probably physically damaged during its removal?) As a study of methodical troubleshooting and repair the video is a long way of the mark IMHO. Cheers, Brian.
|
(continued)
+22V will be OK otherwise +55 would be far off
Well, unless it happens to be say +14 V. That's enough for U1524A (+55) but not for U1524B (+15), see also post by Dave Hills.
|
On Wed, Jun 27, 2018 at 07:11 pm, musicamex wrote:
This REALLY helps as it is the only detailed repair dealing with almost
exactly my issues.
May I disagree with the end of the sentence? The only common diagnosis so far is that +55 and +110 are slightly low and that +15, +5 and -8 are close to zero. You measured +22.5 V or so after the +15 rectifier bridge. So probably nothing wrong there, while the video starts with a problem there. Your slightly low +55 and +110 might both be caused by an improperly set or dirty adjustment R1538. This can't explain the other faults, the other reg voltages would just be slightly too small. +22V will be OK otherwise +55 would be far off. The dead +15 will always cause a failing +5 and -8 and might simply be the only cause here of these failures -- even likely in my opinion. Your dead +15 could be due to just one cap (C1559) or diode (CR1559), or an open R1553. I think you would have made much more progress already if you had done the suggested measurements in earlier posts (you didn't report any.) In the best case you would have been more optimistic than in your last post! BTW: I had no patience to view the whole video. Bu I saw at 32-33 minutes that the AC input voltages to the bridge rectifiers are measured. For +15 that was 1.3 V. I find this hard to believe. Apparently then there is nearly a shortcut loading the transformer winding. I guess this would immediately blow the main fuse? Albert
|
Re: 2465DMM repair and a $1000 US fine or one year year in prison?
Hi all, I added a picture of a set of feet how they are supposed to look like. Pictures can be found here: /g/TekScopes/album?id=61275I looked at both Qservice and Sphere, but they are not on stock anywhere... Un saludo, Leo
|
Re: Tektronix 2440 bad power supply?
I don't know if this will help, but I used to have a lot of problems with new 16 bit MC68340 based embedded controller boards that used a pair of battery backed RAM modules. (It was used in multiple product lines at Microdyne.)
The OEM's testing left random data in them, and often the board wouldn't boot. Since they were socketed, I would swap them and that would usually allow the board to initialize them properly. Pairs where that didn't work had to have all the data wiped on a programmer. At one point, I would draw all the new inventory and wipe them before they were issued to the assemblers.
The problem was caused by data that was out of the normal range expected by the EPROM based operating system. The problem didn't start showing up until the board had been in production about four years, and there were too many different OS EPROMs to be modified, so we just noted the problem in the documentation.
Michael A. Terrell
toggle quoted message
Show quoted text
-----Original Message----- From: Szabolcs Szigeti <szigiszabolcs@...>
Sent: Jun 28, 2018 2:06 AM
To: [email protected]
Subject: Re: [TekScopes] Tektronix 2440 bad power supply?
Hi,
May be trivial, but this has happened to me. Check all the ribbon
cable connections between the boards. I had one that got pulled out
partially when I folded out the board, and was causing all kinds of
weird symptoms. Check that all all firmly seated in their connectors.
If it never completed self cal, there may be garbage in the nvram, so
once you get the scope running, I recommend to do a Cold Start to
clear it, otherwise it may get very much confused.
Also, check the ripple on the power lines, that may be out of spec.
While I'm no fan of the common Internet wisdom of 'recapping cures
everything', I had 3 24xx DSO, and all of them had bad/marginal caps
in the PSU, so I guess one can safely assume that those caps need
replacing. Don't forget about C244 in the control power supply.
Szabolcs
RichR via Groups.Io <choicetrans1@...> ezt ¨ªrta
(id?pont: 2018. j¨²n. 27., Sze, 23:53):
I recently purchased a 2440 with a 4000 code, battery weak. Everything else appeared to be working. That was the only fault code. I replaced the two ram modules with new. The original were obsolete (DS1235BWL-120) so I purchased DS1230-100+. When I first turned on the scope everything powered up and there were plenty of fault codes so I started a self calibration. It never completed the self calibration. The scope went blank. I was able to power it up only occasionally until it stopped completely. It no longer tries to test when it is powered on the only lights on are the three GPIB lights. The screen will start with some erroneous dots and then go blank. I have tested the low power board and have some readings that do not match the schematics The voltage reading according to the power supply overcurrent check page 478 of the manual do not match. The voltages on the side board are correct so I am not sure where to proceed. I have attached a - of the readings I got from the power supply. Any help is greatly appreciated. Thank you, Rich
|
Re: 2465DMM repair and a $1000 US fine or one year year in prison?
Hi all, pictures of the feet of a normal 2465 and a 2465DMM are in the album. I think if I got myself two pairs of feet I can cut them so the upper half would fit as well. When I cut the top corner of a normal foot, and than add a half normal foot, it will work? Pictures can be found here: /g/TekScopes/album?id=61275Un saludo, Leo
|
Re: 2465DMM repair and a $1000 US fine or one year year in prison?
Hello Fabio & Tam, I'd rather stay in rural Spain instead of going to prison, but if I must...... Anyway, I already added the original eBay picture from the add, to show how it was advertised, and I will add pictures of the feet later this morning. Basically the feet are very similar to the feet of an 2465, but extended with an extra section of rubber so they match the higher cabinet of a DMM version. Pictures can be found here: /g/TekScopes/album?id=61275Un saludo, Leo
|
Brian Cockburn
satbeginner
Michael A. Terrell