Date   

Re: 465/475 Volts/Div range LED swap problem.

Dave Peterson
 

Ozan,

Ah, you're who sent the LTSpice reference. Yeah, thanks so much. It's simplistic, but does everything it needs to do. I can live with rotate-mirror-rotate vs. vertical-flip. BFD. I love simple.

It also correlates with measured very well.

I have not mastered concise. I admire terse that gets the message across. Your input is always spot-on. Keep it coming!

Dave

On Saturday, February 20, 2021, 03:09:56 PM PST, Ozan <ozan_g@erdogan.us> wrote:

Hi Dave,
I look at the messages between tasks and usually on a small screen so my replies are usually on the short side. I hope it doesn't come out as terse. I like these puzzles, keep them coming.

Wouldn't the low R win over the LED and keep it weak/off?
The bulb must be consuming >50mA and your LED is probably consuming ~ 20mA so we can burn another 10mA without overloading the Darlington. On the other hand R has to be small enough to keep the LED off. Exact value is not critical but 470-ohm would work.

I spent a good amount of time yesterday and this morning coming up to speed on
LTSpice. For example, I needed to figure out how to add a Darlington pair for
I remember sending you an example LTspice schematic. Looks like you find it useful too, it is a great free tool.

Ozan


Re: P6021 Noise Problem?

Jean-Paul
 

Send scope shots of noise with scale factprs and sketch of entirely setup eg 50 ohm term at scope or1 m?

probe tip open or shorted to grd ring etc.

Check service manual for noise test conditions

Could be bad FET or failed bypass caps.

Jon


Re: 465/475 Volts/Div range LED swap problem.

Ozan
 

Hi Dave,
I look at the messages between tasks and usually on a small screen so my replies are usually on the short side. I hope it doesn't come out as terse. I like these puzzles, keep them coming.

Wouldn't the low R win over the LED and keep it weak/off?
The bulb must be consuming >50mA and your LED is probably consuming ~ 20mA so we can burn another 10mA without overloading the Darlington. On the other hand R has to be small enough to keep the LED off. Exact value is not critical but 470-ohm would work.

I spent a good amount of time yesterday and this morning coming up to speed on
LTSpice. For example, I needed to figure out how to add a Darlington pair for
I remember sending you an example LTspice schematic. Looks like you find it useful too, it is a great free tool.

Ozan


Re: P6021 Noise Problem?

Tom Lee
 

Do you live relatively near to an FM radio station?

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 2/20/2021 13:08, Chappy wrote:
I recently acquired a P6201 active 1X FET probe which appears to have never been used. The inspection tag was still attached, the accessory pack was still sealed, and there was not a single mark on the probe's ground sleeve. All functions of the probe are working but it has a 2mVpp high frequency noise component riding on any signals that I measure.

The manual states a 'Noise Tangental' of 300uV (150uV RMS), a figure well below the noise signal that I am seeing.

I have tried the probe on both my 7403 through a 7A18, and a new Keysight DSOX1102G with the same results. I have tried powering the device with a Tektronix bench power supply, and a home made unit with the same results. Interesting thing is that the noise level riding on the signal is almost non existent when I feed it a supply of +/- 5V. As I increase the supply voltage above this the noise becomes more apparent. Based on this observation I believe that i can rule out noise picked up by the cable between the amplifier and the probe. I have also scoped the power supply rails with negligible abberations on the supply traces on a 2mv/div scale. With the Keysight scope (BW = 70Mhz) a screen capture shows that the lowest component of the noise starts at approx 100MHz.

I am starting to think that there may be components inside either the amp or probe that have degraded over time. Could a degraded or failed electrolytic capacitor contribute to this noise? Or could it be due to a damaged FET or BJT?

If anyone here has one of these I would welcome your input before I go chasing a problem which perhaps is completely normal for this probe. Once again, its physical appearance indicated that it has never (or hardly) been used.

Sincerely,
Steve








TM500 Tester Board Set Deliveries

Larry McDavid
 

I just checked the UPSP tracking info for all the tester board set shipments mailed on 2/11/2021 and find the following on 2/20/2021:

1. All but one USA shipments are delivered

2. One USA shipment to Florida is in Orlando, Florida today, 2/20/21

3. Two shipments to Alberta, Canada were in Seattle, Washington on 2/18/21.

4. One shipment to Quebec, Canada was delivered 2/16/21

5. One shipment to Frankford, Germany was in Frankford on 2/14/21

6. Four shipments to UK were delivered 2/17 and 2/18/21

7. One shipment to Kenilworth, UK is in transit 2/18/21

8. One shipment to Australia was delivered 2/19/21

9. Two shipments to Australia are in transit 2/16 and 2/19/21

10. Two shipments to France are in transit 2/15/21

11. One shipment to Ireland is in Dublin 2/14/21

12. One shipment to Spain is in Madrid 2/15/21

I don't know what the delay in the one shipment to Florida is. Maybe weather... Gosh, there has been a lot of cold and snow out there!

The USPS tracking of international shipments has little international detail but at least some show delivered so USPS gets some feedback. I don't know if that is true for all countries.

I suggest all who have not yet received the board sets wait until mid next week. If you still don't have the package, send me a private email and I'll provide the tracking number.

The package is 13 x 10 inches in a flat mailer and won't bend in normal handling. Some have reported the postal carrier has rung the doorbell and left the package. So, this package may not fit in a standard mail slot or mail box.

Many have reported successful delivery to me and all have said the boards arrived in good condition.

A comment: Dennis Tillman, based on his experience shipping the many CRT books, suggested I ask for an emailed postal receipt with tracking numbers. That was a great idea! If you ship numerous packages at once, do use the USPS email receipt capability. The tracking numbers on that emailed receipt are actually hyperlinks, making tracking easier. I was unaware of that capability and I expect others are also unaware.

Weather can wreak havoc on all kinds of deliveries. Local COVID vaccination sites have been shut down for days due to non-local weather-induced vaccine deliveries.

One mailer was delivered in one day; many were delivered in two days. I believe we've got a great post office!

--
Best wishes,

Larry McDavid W6FUB
Anaheim, California (SE of Los Angeles, near Disneyland)


Rather than a TM500 module extender....

 

I just want to mention again that all the issues raised with regard to module extenders are important,
and no matter which route you go, the result is always a bit sub-optimal, and awkward on the bench

The magic answer is just to take a TM502A apart, removing the blue outer shell. What's left is a compact dual connector power supply (complete with power switch + line connector), and total access to the module under test. This is the simplest, easiest, and most trouble free way to work on a TM500 plug in, and takes up the least possible bench space. I highly recommend this path over all others. A similar trick is possible with a TM5003 for the TM5000 modules, but it's certainly bigger.

I should point out that I did not move to this approach until long after I already had many extenders, and it's only this specific frame that yields such good results. In any case, it is a magical fix, full of test bench happiness.

Test plug-ins to validate the frame are still needed, and you may still want to extend a plug in from a specific frame, so that's the best case for an extender.

All the best,
walter (walter2 -at- sphere.bc.ca)


P6021 Noise Problem?

Chappy
 

I recently acquired a P6201 active 1X FET probe which appears to have never been used. The inspection tag was still attached, the accessory pack was still sealed, and there was not a single mark on the probe's ground sleeve. All functions of the probe are working but it has a 2mVpp high frequency noise component riding on any signals that I measure.

The manual states a 'Noise Tangental' of 300uV (150uV RMS), a figure well below the noise signal that I am seeing.

I have tried the probe on both my 7403 through a 7A18, and a new Keysight DSOX1102G with the same results. I have tried powering the device with a Tektronix bench power supply, and a home made unit with the same results. Interesting thing is that the noise level riding on the signal is almost non existent when I feed it a supply of +/- 5V. As I increase the supply voltage above this the noise becomes more apparent. Based on this observation I believe that i can rule out noise picked up by the cable between the amplifier and the probe. I have also scoped the power supply rails with negligible abberations on the supply traces on a 2mv/div scale. With the Keysight scope (BW = 70Mhz) a screen capture shows that the lowest component of the noise starts at approx 100MHz.

I am starting to think that there may be components inside either the amp or probe that have degraded over time. Could a degraded or failed electrolytic capacitor contribute to this noise? Or could it be due to a damaged FET or BJT?

If anyone here has one of these I would welcome your input before I go chasing a problem which perhaps is completely normal for this probe. Once again, its physical appearance indicated that it has never (or hardly) been used.

Sincerely,
Steve


Re: 465/475 Volts/Div range LED swap problem.

Dave Peterson
 

There are many facets to this experience for me:

One is that I want this info in the groups messages in the hope that it might help others in the future. So pardon if this is rudimentary for the more experienced of the group.

Another is that I wanted to reeducate myself in BJT circuits. I haven't dealt with them since school. This has been a great experience.

It was actually the thought of adding a parallel resistor to the 10X led/resistor to provide an alternate current path that led me to using a NPN. I wasn't sure what was going to happen when Q382 tried to drive the 10X LED. Wouldn't the low R win over the LED and keep it weak/off? I didn't have LTSpice setup yet. So I thought, since I'd just gotten this load of NPNs, why not? And down the Rube Goldberg path I went.

The NPN does end up with an elevated base voltage when 10X is on and 1X is off. It doesn't effect the 1X diode, but it is dangerously close to turning on.

Now that I have the simulation set up I was able to test the 470 ohm resistor, and sure enough, it works great. See, I knew I was making it more complicated than need be. Simplicity is definitely what I was trying to accomplish.

I spent a good amount of time yesterday and this morning coming up to speed on LTSpice. For example, I needed to figure out how to add a Darlington pair for Q382. As well as the simple things too. In fact I started with two 2N2907's as my Q382 at first. Worked pretty well, and got me up to speed enough to tackle the next level exercise of adding a third party model (and make it portable too).

I'm not very good at learning a new tool in a purely academic way - even if it is very similar to what I've used before. I need something practical to apply it to. This was a perfect circuit to cut my teeth on. Thanks to all who suggested it before. Was that you Michael? It has now also been a good exercise in finding minimal solutions. I also now have a better model to test on than a bread board with too few components.

I don't think I'll ever be up to speed on these scopes and circuits as so many of you. But I am learning, and very much enjoying myself.

Thanks to all who tolerate me and help.
Dave


On Saturday, February 20, 2021, 11:14:07 AM PST, Ozan <ozan_g@erdogan.us> wrote:


decided to wire one parallel to the 10X LED. I was really trying to minimize
the complexity of this solution. This was the best I could come up with.
See my reply https://groups.io/g/TekScopes/message/179044
A 470-ohm from collector of the Darlington to ground (across the LED+current limiting resistor) is all you need. No need for ten LEDs in parallel, and it is not the most robust solution either.

Ozan


Re: 2467B TEST 04 FAIL 02

Jean-Paul
 

Tom V

the steps are divided into sections and each type of CAl is independent of the other sections.

Certain steps near the end require a very precise and symmetric test wave in both amplitude and frequency.

If you dont generate that précise signal, the CAL of that step will fail.

I think Chuck will recall previous discussions, but I forgot which CAL it was, perhaps the HOR.

It was the only step that is is so picky.

Jon


Re: Failed Transistor with Low h(FE)?

Ed Breya
 

Chuck wrote: "No wonder HP only used 1% or better parts"

Haha - I wouldn't be too surprised if there's a hole in the middle of those too, where the 0.1 % tolerance grade parts are "pre-removed" for your convenience. It depends on how far you need to, and can, grade parts from a given process, and on the economic value of doing so.

Ed


Re: 465/475 Volts/Div range LED swap problem.

Ozan
 

decided to wire one parallel to the 10X LED. I was really trying to minimize
the complexity of this solution. This was the best I could come up with.
See my reply https://groups.io/g/TekScopes/message/179044
A 470-ohm from collector of the Darlington to ground (across the LED+current
limiting resistor) is all you need. No need for ten LEDs in parallel, and it
is not the most robust solution either.
I now see the solution, 10x was just the label, not the number of LEDs. No need for extra transistor, 470-ohm is all you need. Sim might have worked OK but in the proposed solution base of the new NPN is not pulled to ground strongly so noise could couple to the base easily.

Ozan


Re: 465/475 Volts/Div range LED swap problem.

Ozan
 

decided to wire one parallel to the 10X LED. I was really trying to minimize
the complexity of this solution. This was the best I could come up with.
See my reply https://groups.io/g/TekScopes/message/179044
A 470-ohm from collector of the Darlington to ground (across the LED+current limiting resistor) is all you need. No need for ten LEDs in parallel, and it is not the most robust solution either.

Ozan


Re: TM500 Mainframe Tester construction master thread.

Andy Warner
 

So far, I have one observation, and one question:
1. observation - digikey BOM has a typo in the reference designators for the 1N4007 diodes. "CR11" should read "CR101".
2. question - if I insert the stops in S3 where the manual suggests on page 2, the index on the knob does not align with the front panel when using the flat on the shaft with the grub screw.
I plan on buzzing the switch out to see what the correct location is, but I thought I would ping the list about the inconsistency as well.

--
Andy


Re: 465/475 Volts/Div range LED swap problem.

Chuck Harris <cfharris@...>
 

When faced with that problem, I usually add a diode
in series with the emitter of the transistor switch.
That allows a "0" to turn the switch off better.

-Chuck Harris

Dave Peterson via groups.io wrote:

Sorry Dave.

The link was to the thumbnail. The full picture can be expanded by finding it in the thread:

https://www.eevblog.com/forum/repair/tektronix-465/msg697587/#msg697587


Not so much theatrics as impatience on my part. I'm cleaning up my LTSpice simulations to upload to the files area, and need to build up a photo album. In the interests of saving on the TekScopes group space usage I need to resize all my pictures by quite a lot.

No need to apologize. Pardon my impishness and impatience.

Yes, the 10x diode stays on weakly. I tried all manner of remedies, but the basic issue is that Q386 needs enough base current to drive the 1X LED, and that current is always enough to cause the 10X LED to start illuminating. I'd just ordered a batch of 100 2N3904's to replace one missing in the scope. So decided to wire one parallel to the 10X LED. I was really trying to minimize the complexity of this solution. This was the best I could come up with.

Posting the LTSpice files now. Want to/will add a writeup PDF too.

Dave

On Saturday, February 20, 2021, 10:11:48 AM PST, Dave Voorhis <voorhis@gmail.com> wrote:

On 20 Feb 2021, at 17:54, Dave Peterson via groups.io <davidpinsf=yahoo.com@groups.io> wrote:

Here's a hint:

https://www.eevblog.com/forum/index.php?action=dlattach;topic=45906.0;attach=157509;image

What's wrong with this picture?
It’s too small to see?

Sorry to be grumpy, but I’m not finding your guessing-game approach much fun. Could you share the problems you’ve found and the solution you’ve developed and, with respect — and unfortunately there’s no write to say this as politely as I would say it in person — please spare us the theatrics?












Re: Help replacing a resistor/inductor in SG503

TomC
 

The ones that failed were rated at 15 volts and were on the 15 volt power lines. Obviously a poor choice as they only lasted 40 or so years before failing. Well, maybe not so bad after all.

I think later revisions went to a 25 volt rating. That's what I used to replace the originals.

Tom

On 2/19/2021 6:54 PM, Harvey White wrote:
GIven that the recommendation of the manufacturers was to minimize the headroom between the working voltage and the rated voltage (to the point where things fail.....)
That's a good idea to replace with higher voltage ratings, but I'm dumping this in here in case someone only hits this thread and doesn't research the older (and similar) ones.
but I can't figure out how to shoehorn FOOM! in here.....
Harvey
On 2/19/2021 9:37 PM, TomC wrote:


On 2/19/2021 5:57 PM, Roy Thistle wrote:
On Fri, Feb 19, 2021 at 04:02 PM, TomC wrote:


To fix, I took a closeup photo of one that hadn't been damaged and counted
turns. There are 47 turns. Found the wire gauge with a micrometer on the wire
from one that was destroyed. Then ordered wire and a supply of carbon comp 10
ohm resistors.

Made up a jig to wind the 47 turns on new 10 ohm resistors. Then soldered the
ends and painted with dope.
As Stan Lee would say, "nuff said!"
Actually, what I failed to say was that I replaced those 4 tantalums, using a higher voltage rating, in every 7A26 I own. Both the ones that had failed short, and the ones that hadn't failed yet.  But that is probably obvious to the people here.







Re: 465/475 Volts/Div range LED swap problem.

Dave Peterson
 


Re: 465/475 Volts/Div range LED swap problem.

Dave Peterson
 

Sorry Dave.

The link was to the thumbnail. The full picture can be expanded by finding it in the thread:

https://www.eevblog.com/forum/repair/tektronix-465/msg697587/#msg697587


Not so much theatrics as impatience on my part. I'm cleaning up my LTSpice simulations to upload to the files area, and need to build up a photo album. In the interests of saving on the TekScopes group space usage I need to resize all my pictures by quite a lot.

No need to apologize. Pardon my impishness and impatience.

Yes, the 10x diode stays on weakly. I tried all manner of remedies, but the basic issue is that Q386 needs enough base current to drive the 1X LED, and that current is always enough to cause the 10X LED to start illuminating. I'd just ordered a batch of 100 2N3904's to replace one missing in the scope. So decided to wire one parallel to the 10X LED. I was really trying to minimize the complexity of this solution. This was the best I could come up with.

Posting the LTSpice files now. Want to/will add a writeup PDF too.

Dave

On Saturday, February 20, 2021, 10:11:48 AM PST, Dave Voorhis <voorhis@gmail.com> wrote:

On 20 Feb 2021, at 17:54, Dave Peterson via groups.io <davidpinsf=yahoo.com@groups.io> wrote:

Here's a hint:

https://www.eevblog.com/forum/index.php?action=dlattach;topic=45906.0;attach=157509;image

What's wrong with this picture?
It’s too small to see?

Sorry to be grumpy, but I’m not finding your guessing-game approach much fun. Could you share the problems you’ve found and the solution you’ve developed and, with respect — and unfortunately there’s no write to say this as politely as I would say it in person — please spare us the theatrics?


Re: Failed Transistor with Low h(FE)?

Chuck Harris <cfharris@...>
 

I have often wondered if many circuit designers
take these "holes" caused by factory robbery into
account when they specify their parts... I never
did, but perhaps the production guys would have
liked me better if I had?

You think you are using a 1K nominal part, +/-5%,
but in reality, your part is really:
[1.05K|0.95K] +/-2.5% of 1K

You thought you bought a one humped dromedary camel,
but Amil the camel dealer came rolling up to your
camp with a two humped bactrain camel, and none of
your saddles will fit.

It puts a completely different spin on a Monte Carlo
analysis on part tolerances.

No wonder HP only used 1% or better parts.

-Chuck Harris


Ed Breya via groups.io wrote:

Chuck wrote: " What I generally
find is that out of 1000's of 5% resistors, not a single one
was within 1% of its nominal value. But, all were within 5%."

These two-lobed distributions are common. The ones that fell in the middle range near 1% were likely separated out and sold as a better grade part, leaving the rest of the distribution (which should be normal) with a hole in the middle.

Ed







Re: 2467B TEST 04 FAIL 02

Chuck Harris <cfharris@...>
 

Sadly, there is no sanctioned way of telling what step
has failed.

You can tell with unsanctioned methods by reading the
NVRAM, and looking at the data bits. The location and
layout of the calibration data is only loosely described
in the maintenance section of the manual. There are 170
14 bit calibration constants in the NVRAM. Each of those
constants has its own checksum, and status bits to tell
if it is in or out of range. All left to us as a mystery
by tektronix.

If you look at the calibration data using the EXER02 function
in the calibration/test menu, you will see each of the
256 RAM locations you can read has the format:

AA DDDD P, where "AA" is the address, "DDDD" is the hex
data for the constant, of which 14 bits are valid, and P is
the parity bit X is even, blank is odd.

The actual locations of this data in the NVRAM have been
found and discussed by others on this group.

I meant one day to take a virgin NVRAM, and read it after
every step of the calibration routine is completed, and
to note the differences in the calibration section of the
NVRAM. I will probably never do so, because I, well, I have
a life, and I just don't care that much....

However, there is nothing stopping one of you from doing
the heavy lifting. I would suggest using what is called
an EPROM emulator, or RAM emulator to save wear on the
NVRAM socket. There will be at least 170 times the data
after a calibration step needs to be compared with the
"virgin" NVRAM data.

Please note that the NVRAM has only 170 calibration constants,
and the entire rest of the NVRAM is used as program storage,
and as a result, is reset every time you fire up the scope.

This is your task should you choose to accept it, good luck
with it! International fame and no fortune await!

-Chuck Harris

tvoshell@cableone.net wrote:

My 2467B >B052xxx suffered two nasties! (The processor board caps exploded and subsequently ate up some of the copper traces near them underneath the solder mask layer. (tough to see)
Also my sram backup died. I was able to get a new recent date code SRAM and it works ok but no cal data now. Also I was able to clean up the cap mess and replace the crummy electrolytic caps with tantalum caps.
I went thru the tek "ADJUSTMENT PROCEDURE" and set everything I could: the really fast horizontal cal's I didn't have gear for. I approximated the top end and pressed the trigger coupling button. I was able to program everything else, or so I thought, but now I am getting a cal limit error TEST 04 FAIL 02. Since the procedure is quite lengthy, with a lot of steps stored, it would be great to know which step of which cal is out of limit.

Is there a way to determine which cal/step is responsible for the error?

All responses are welcome, Thanks
TomV






Re: 465/475 Volts/Div range LED swap problem.

Ozan
 


A properly sized resistor in parallel with the X10 resistor should fix it.
I meant a properly sized resistor in parallel with the X10 LED should fix it.

Ozan

--------

Here's a hint:

https://www.eevblog.com/forum/index.php?action=dlattach;topic=45906.0;attach=157509;image


What's wrong with this picture?

Dave

On Saturday, February 20, 2021, 09:46:09 AM PST, Michael W. Lynch via
groups.io <mlynch003=yahoo.com@groups.io> wrote:

I am sure that more than a few people here and elsewhere would like to hear
more about this.  I am a devotee of Mr. Goldberg myself.

--
Michael Lynch
Dardanelle, AR





9221 - 9240 of 188004