开云体育

If I'm not mistaken, 1/2" 6061-T6 aluminum tubing with two slots cut in it 180-degrees apart would do just about as nicely as anything for that.

Thanks,
Barry - N4BUQ

Yes, that BNC T-wrench on ebay is yet another form of installer. I have two different kinds from long ago, both commercial tools, but I have no ID or source info. One is made just like a nut driver, but the business end is a hollow cylinder with slots, that fits the outside. The other is a strange looking socket that has the same type features, but is made to go on a 1/4" square driver. I also built one way back, before I acquired the "real" ones. I took a male cable-end type BNC, and soldered it into a stubby-handled nut driver, in such a way that the outer ring was locked to the core. The original driver tip was whatever hex size that the BNC end fit closely enough to solder in. In use, it just goes on the female BNC as if it was a cable, but allows holding the position or slight torquing for installation, then unplugs just like a cable.

Be aware when using these things, that the BNC bayonet pins are not very strong, and will easily deform or even shear off if over-torqued. The real compression force and torque for installation should be applied to the nut in the back, while the tool just keeps it from rotating out of the desired pin position in the early tightening phase.

Ed

Hi Peter,

Besides the check to the power supply's capacitors, as Chuck noted and, as you will notice follwoing up the threads, is a kind of "standard" advice for all equipment of this vintage...
It may be helpful (for further troubleshooting by the group or yourself) if we can differentiate if the wide trace is due to problem with focus, or if it's wideband noise related..

Since you mentioned that readout is fine, there's less likelihood of being focus related, but still, not being familiar to the 2445's focus circuitry, there can still be an explanation for a focus problem while drawing the actual signal trace on screen, as many scopes have "auto-focus", circuitry that dynamically adjusts the focus for the often different intensity of the readout.

To rule-out focus from noise, it usually suffices to display a good square wave on the screen (e.g. calibrator).
At moderate sweep speeds, such as enough to display a single cycle on screen, the rising and or falling edge should look cleaner and sharper (while the tops and bottoms still look fat) if the issue is noise related. At slower sweep speeds it won't be much conclusive as you would barely see the rising and falling edges.
However, if the "fat" is due to focus, the rising or falling edges are likely to show just as fat.

Other resource to help rule-out noise as a source, is turn on the Bandwidth limit (usually to 20MHz)... this usually makes a "fat" noisy baseline to look sharper (if the issue is noise).
If, however, engaging BW limit doesn't improve matters, than it may tend more to be focus related.

Rgrds,

Fabio

On Wed, 25 Apr 2018, Larry McDavid wrote:

Here you go, works like a charm:



I bought this exact wrench couple of years ago and very happy with it.

Someone here mentioned a BNC installation tool. I took that in context to mean a tool that holds a chassis-mount female BNC jack while tightening the backside nut but I don't find any such tool when searching, only tools to disconnect mated BNC connectors.

Did I misunderstand what was suggested for straightening bent BNC connectors or can someone point me to such a tool?

---
*
* KSI@home KOI8 Net < > The impossible we do immediately. *
* Las Vegas NV, USA < > Miracles require 24-hour notice. *
*

Ok, cool. So I need to isolate whether the input transistors are leaky or whether there's a problem with the chip.

The transistors are socketed but the chip is not. So it's not practical to casually take the chip out. So I'll try removing the transistors and put in the 51 ohm on 2 and 3 as you suggest. Even if I can't take the chip out, this should tell me something. If the voltages are all nominal then, that would point upstream to the transistors.

While I have the transistors out I can ohm them out to check for a leak. The B side transistors should be identical and for the most part the B side works, so I could also swap those transistors and/or use them for comparison when ohming.

I'm still suspicious about the readings on pins 1 and 16. It will be interesting to see if they change when biasing 2 and 3 with the resistor.

Thanks!
Lorn
KK4KRI

snippet of the previous measurements:

U7331

Pin, expected Vdc, actual
1. -0.8, -1.1mV
2. 0, .761 V
3. 0, .761 V
4. 4.9 actual (tied to +5 through 1k resistor)
5. -8, -7.95
6. -1.0, -.959 (close enough)
7. ground, -.5mV (effectively zero)
8. 10.6, 10.72
9, 10.6, 9.73
10. 0, -39.7 mV
11. 10.5, 10.68
12. -0.9, -.852
13. 7.1, 7.18
14, 0, .7 mV
15. 0, .7 mV
16. -0.8, -204.6 mV
<< end snippet


On Wed, Apr 25, 2018 at 11:46 am, Fabio Trevisan wrote:

Hello Lorn,
I also own a 464 and mine also had a problem triggering... And while looking
at your description I looks like deja vu to me.
On mine one of the FET transistors of the input circuitry were leaky...(it had
about 200kOhms impedance between source and gate, while reverese biased. The
good ones measure Infinite, or very high Megohms.
The leak developed a voltage drop on the 1M resistor at their gates and make
the output voltage to have a huge offset of... wow! 70mV...(that's 1/10th of
what you're getting).
70mV offset was enough to take the preamplifier IC so off that trigger would
only work by adjusting the level potentiometer near its end of range and,
still, after warming up, the offset would grow just a bit more and then even
taking the Level potentiometer to the end was not enough to make it trigger.
On yours it seems the offset is some 0.7V, so no wonder why it doesn't
trigger.
Your course of action is exact and removing the FET transistors will allow you
to rule-out if the positive offset is coming from the IC, or if it's coming
from the buffer stage into the IC.
It's likely, however, that if you just leave pins 2 and 3 opened, as they will
get when FETs are removed, you will be fooled by false voltage readings at
pins 2 and 3, due to the input pins need some biasing.
So, after you remove the FET ICs, connect a resistor of - say - 51R between
pins 2.3 of the preamp IC to ground.
The impedance at the other side of this balanced input differential amplifier
is roughly 50Ohm, so biasing pins 2,3 with 50Ohms will make the input
perfectly balanced.
In this condition, pins 2 and 3 must read a low voltage, close to 0V. By no
means it can be more than 50mV... as I told, my 464 had 70mV and that was
enough to screw up the triggering.
Alternatively, you can check the FET input buffer's offset, by taking the IC
out and measuring voltage directly at pins 2.3 of the IC socket...
Don't worry, taking the IC out is safe... nothing that's past of the IC will
burn because the IC isn't present. (I had this same doubt back then).
Set the Trigger input to EXT, coupling to DC, and ground the Ext. input, so
that you're sure that you're not feeding DC into the input buffer, and its
output should be very close to 0V.
If the transistors are perfectly matched and thermally even, in theory, the DC
offset of this input buffer is exactly ZERO.
Good luck,
Rgrds,
Fabio

Someone here mentioned a BNC installation tool. I took that in context to mean a tool that holds a chassis-mount female BNC jack while tightening the backside nut but I don't find any such tool when searching, only tools to disconnect mated BNC connectors.

Did I misunderstand what was suggested for straightening bent BNC connectors or can someone point me to such a tool?

--
Best wishes,

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

Hello Lorn,
I also own a 464 and mine also had a problem triggering... And while looking at your description I looks like deja vu to me.
On mine one of the FET transistors of the input circuitry were leaky...(it had about 200kOhms impedance between source and gate, while reverese biased. The good ones measure Infinite, or very high Megohms.
The leak developed a voltage drop on the 1M resistor at their gates and make the output voltage to have a huge offset of... wow! 70mV...(that's 1/10th of what you're getting).
70mV offset was enough to take the preamplifier IC so off that trigger would only work by adjusting the level potentiometer near its end of range and, still, after warming up, the offset would grow just a bit more and then even taking the Level potentiometer to the end was not enough to make it trigger.
On yours it seems the offset is some 0.7V, so no wonder why it doesn't trigger.
Your course of action is exact and removing the FET transistors will allow you to rule-out if the positive offset is coming from the IC, or if it's coming from the buffer stage into the IC.
It's likely, however, that if you just leave pins 2 and 3 opened, as they will get when FETs are removed, you will be fooled by false voltage readings at pins 2 and 3, due to the input pins need some biasing.
So, after you remove the FET ICs, connect a resistor of - say - 51R between pins 2.3 of the preamp IC to ground.
The impedance at the other side of this balanced input differential amplifier is roughly 50Ohm, so biasing pins 2,3 with 50Ohms will make the input perfectly balanced.
In this condition, pins 2 and 3 must read a low voltage, close to 0V. By no means it can be more than 50mV... as I told, my 464 had 70mV and that was enough to screw up the triggering.
Alternatively, you can check the FET input buffer's offset, by taking the IC out and measuring voltage directly at pins 2.3 of the IC socket...
Don't worry, taking the IC out is safe... nothing that's past of the IC will burn because the IC isn't present. (I had this same doubt back then).
Set the Trigger input to EXT, coupling to DC, and ground the Ext. input, so that you're sure that you're not feeding DC into the input buffer, and its output should be very close to 0V.
If the transistors are perfectly matched and thermally even, in theory, the DC offset of this input buffer is exactly ZERO.
Good luck,
Rgrds,
Fabio

Hello Vaclav.
This forum doesn't take attachments...It's sort of controversial, but bottom line is that it doesn't.
You need to post the pictures to the photos section and... since there's no way to "search" on the photos section, you better copy the link to the album and post it on the message thread, so that anyone can quickly jump to the photo.
Rgrds,
Fabio

I would appreciate some assistance to find out why my 464 has this problem.
I do have a shop manual.

PS this is? my first post here so please no flames if I did it wrong.
Vaclav in Houston TX

I have two 465s. I have repaired one. They way it is constructed does not lend itself to be easily repaired. One must be patient with repairs on a 465. Also be very careful about moving boards around. I love these scopes when they work. I use at least one daily to repair electronic equipment. It was a great functional design, not the best design to allow access to parts.

It's not the 190A either, it is much too small and has a big vertical dial that would be easy to see (I have one); I'd assume the 190B would be similar.
-Dave From: Jerome D Leach <jeromeleach17@...>
To: [email protected]
Sent: Wednesday, April 25, 2018 12:40 AM
Subject: Re: [TekScopes] Another Tek in media sighting

Type 130 L,C possibly? Wasn't a lot of Tek stuff with meters. 105 square
wave gen., and the 190 A and B constant amplitude signal gen.

added pictures of waveforms of 2n3055 and 2N5881. At least they are now labeled. the earlier ones numbers are all FPJ13007 with the poor thermal mounting per that picture. that is no more, just not enough thermal mass or xfer. still signals really have not changed.
looking at them together I am thinking the issue must be prior the Q1174/84 chopper/switch.

Adjusting the Osc never manages to give me a leading edge just changes the overall Freq.

The waveforms in the book are all at correct supply (aprox 28VDC) so not a whole lot of help while running at 11.5 to 12VDC input. wish the book had a few waveforms on the other side of T1171 to verify operation at 11.5V

I may try to put the good original 2n1899 back in circuit (original location Q1184) and pair it with a TO-3 and see what happens.....when I have time later.
搁别苍é别

glad to hear the DC/DC converter is working that should be a relief.
so why is the Ac-Dc linear causing an issue?? bad rectifier? or one/both of the filter caps bad?

it is 9am here, had a chance to try a pair 2n5881, 2n5039, 2n3055, and Mj10000 earlier this mourning.
11.5VDC input and looked at the waveforms junction D1176/77
I had attached sockets and left them dangle off the side for testing...now where did i put that heatsink??
the MJ10000 (Darlington) would not give stable waveform- kinda expected it may not work.
it does not seem to matter much between the rest. The venerable 2N3055 drew the least current (3.5A) and was stable, 2n5039 a bit more current and stable same with the 2N5881. not having heat sink limited the "on" time. Conclusion- The Chopper xsistor is not too important which were my original thoughts prior to over thinking a replacement. It appears the venerable 2N3055 should be fine. I wonder why the used the 2n1899??...I may find out later when i get the rest of this darned thing working.
I still do not understand why I do not have a leading edge anywhere close to "per the book" . I think the "on" time is too long. I may have to go back prior the choppers....I only have the PDF of the AC/DC supply and the waveforms are really hard to see...they were in blue in the original and obviously did not xfer well.

I also did look at the waveforms with and without scope attached same waveform just 0.5A more current draw.
nice square wave at the Base of choppers.
I will post some pictures later today of the 3055 after I get them off the tablet thing. lots of other stuff that needs be done prior my being gone for the next 10 days (performances/rehearsals/lots of driving- hate the last item)
搁别苍é别

What I have found so far on the 465B trigger problem:
Serial number B061387, service manual says B060000 and up.
Removed case and did a very brief physical inspection. Very clean, no dust. No obvious burned parts so far, no burn smell. Most transistors seem to be socketed.

Voltages: input selector is set to 104-126. I know the power company sometimes keeps switching between 120 and 125 because my UPS complains about that.

Used an old Radio Shack (yeah I know, can't afford a Fluke) DVM with true RMS capability.

Voltage test point, reading, ripple, comment
+110 unreg 164.7 Vdc .425 Vac dc voltage seems high?
+110 111.9 2.8 mV
+55 54.8 1.3 mV
+15 14.95 1.4 mV
+5 5.00 1.4 mV
-8 -7.98 1.4 mV

So that all seems nominal.

Schematic 6, A and B Trigger Generator board
Set the front panel controls for DC measurement per the schematic.
Test Point 43, which is not marked on the board layout picture, but is marked on the schematic:
Should read zero volts, reads -.1mV so I would call that zero (input to Q7324A and B).

U7331
Pin, expected Vdc, actual
1. -0.8, -1.1mV
2. 0, .761 V
3. 0, .761 V
4. 4.9 actual (tied to +5 through 1k resistor)
5. -8, -7.95
6. -1.0, -.959 (close enough)
7. ground, -.5mV (effectively zero)
8. 10.6, 10.72
9, 10.6, 9.73
10. 0, -39.7 mV
11. 10.5, 10.68
12. -0.9, -.852
13. 7.1, 7.18
14, 0, .7 mV
15. 0, .7 mV
16. -0.8, -204.6 mV

I'm thinking based on these numbers the answer must be staring me in the face. Pins 2, 1, and 16 aren't correct, so I'm thinking either a U7331 problem or possibly the A trigger view circuit which hangs off of 1 and 16 may have a problem. I'll check the voltages in the trigger view circuit next opportunity, and maybe pop out Q7424 and Q7422 to isolate the chip from the trigger view circuit and see if anything changes on 2, 1 and 16.

I didn't write down the numbers but a quick probe of some of U7131 on the B trigger side showed expected voltages, especially on 2, 1, and 16.

You guys are an amazing resource! I knew I came to the right place. I also hope this thread turns out to be useful to someone else.

Lorn
KK4KRI

What is like gold? There is nothing from the prior to know what your referring too
Dave
NR1DX

--
Dave
Manuals@...
www.ArtekManuals.com

Suhweet! Thank you so much! This is like gold!

Lorn
KK4KRI

It's 11:00 AM here and I just hooked it up. It is now running off the unfiltered battery charger into its own filters. It's got about 15 volts going in, best I can do without the variac. However, I think it will be enough to detect heat buildup in a few hours. It used to shut down in less than 2 hours. I am injecting the power right before the thermal cutout so if it does get hot it shouldn't let the smoke out. I can't sit here and watch it all day.

If there is no heat buildup in a few hours, I'm going to disconnect it and also disconnect the secondary of the AC power transformer and just let that run with no load. At that point I can be sure the DC/DC convertor is running properly and we can trust the waveforms.

Right now, the ripple is 3 volts P-P while it was only 1 volt with the 10,000 uF.

We'll see.

Hello Group,
Moving on to the next issue (this one is minor... but annoying) on my recently acquired 7623A.
When I select any scale (either from Vertical amplifier or from Time Base) that requires 2 zeroes, it displays only one zero while the scope is still cold.
After warming up for - say - 5 to 10 minutes - it becomes normal.
While it's still cold, also, if I press IDENTIFY, the word identify doesn't show and instead, it shows (IIRC) a u symbol, in place of one of the zeros that are currently on display.
I narrowed down the temperature sensitivity to U2232, the "zeroes logic & memory" IC.
However, I`m not sure if the problem is in it, or if one of the signals coming into it is at a marginal level, and maybe when U2232 is cold, it's just misinterpreting the inputs which are marginal.
Fiddling with the "Column Match", or "Row Match" adjustments doesn't have any impact on the problem... it still "swallows" the second zero while cold, no matter what position those 2 adjustments are.
So, my questions to the group are:
1. Is there any Tek semiconductor manual available that shows what it has inside? The "Tek_Made_sm.pdf" available at TekWiki doesn't show it.

2. Does any of you guys have a broken readout board and would be willing to part this IC with me? I know that Sphere and Qservice have them, so I already have that as a last resort.

3. On my suspicion that the problem may *not* be at U2232, looking at the Column decoded signals that enters to it (and which are ultimately sampled at Time Slot 1 and memorized, to define whether it should add one zero, or two, or none), I find a little bit odd that the signals there don't really resemble "logic" levels, each (C1, C2, C3, C4 and C10) are sort of different from each other, and some of them really look much more analog signals then a digital signal with clearly distinct discrete levels.
I will try to take pictures from each one probed on my other scope and post it.
From what I understood, although the original Row and Columns signals are analog (OK. Discrete 10 levels analog signals)... after they are processed by the Row and Column decoders, those signals should have been translated into 10 different "logic" outputs, with only one output showing an "active" level, and all the other ones showing an "inactive" level (I don't say high or low because I`m not really sure if they're active high or active low).
Still, those "Column Decoded" digital signals (that don't look like digital) are the same ones going to the character generator IC matrix... And there doesn't seem to be a problem for the character generation circuitry, and all the displayed characters are correct, leading me to conclude that despite the Column signals don't look like what I expect... they might be correct anyway.
So my 3rd question: Can any of you take a look at the column decoded signals (the ones entering at pins 10,11,12,13 and 14 of U2232) and tell me if they look like digital signals, with clearly distinct "active" and "inactive" levels?

Thank you in advance for any input...

Rgrds,

Fabio

Hi chuck,
Thanks for the reply, as far as I Know nothing has been changed, I have owned
The Scope since about 2002, Before that it was in the Navy.
Thanks............Peter

Tony, did you do the Power Supply test on page 5-8 of the manual : Adjust
Inverter Control R1293 for a meter reading on TP1625 of +40V? Maybe your
operating point for the Inverter Controller is way off. Mine was way off in
the negative range until I adjusted it.

Best,
John

On Tue, Apr 24, 2018 at 10:52 PM, Yiu On Tony C via Groups.Io <
tonycheung_hk@...> wrote:

Hi John :
It is overall power handling issue for the PSU , U1275 is the controller
for the shut down action, 5V light is NOT only output, there is feedback
to R1304 to controller too !
On the other hand the total current sampling from T1235 to U1275.
I did check C1316, C1317 and C1318 are normal .
My PSU was damaged by miss-align one pin of SENS input on LV Board ( P1483
) :
pin 1 ------- Beam I Sens input on LV board NO Connection.
pin 2 ------- GND Sens input on LV board connected to P83 pin 1
-------Beam I sens. --------------- Beam I sens short to GND.
pin 3 ------- +5V Sens input on LV board connected to P83 pin 2 -------
GND sens, -------------means +5V short to GND !
pin 4 ------- -50V Sens input on LV board connected to P83 pin 3 --------
+5V sens, ------------- short -50V to +5V !
pin 5 ------- +50V Sens input on LV board connected to P83 pin 4 --------
-50V sens -------------- short +50V to -50V
pin 6 -------- -15V Sens input on LV board connected to P83 pin 5 --------
+50V sens ------------- short -15V to +50V
pin 7 ------- +15V Sens input on LV board connected to P83 pin 6 -------
-15V sens ------------ short +15V to -15V

The results is all ( +5V, -50V, +50V, -50V, -15V, +5V and Beam I sens
short to GND ). may be serious damage for the PSU or main unit ?
Anyway it is under repair for the damage !
RegardTony CheungAPR 25 2018



From: JJ <jajustin@...>
To: [email protected]
Sent: Wednesday, April 25, 2018 5:26 AM
Subject: Re: [TekScopes] 7904 Mainframe damaged

Tony,
If all you can draw is 100ma on 5v lights, you could have a problem with
that power supply's pi section filter - I had to change C1318 to get my PSU
up. It measured 40uf instead of 100uf. That cap may be right on the edge of
failure at certain loads. There are also 2 other caps in that filter -
C1316, C1317. Note that 5v lights is not regulated on the LV regulator
board - it just goes through the board to the connectors - so the problem
is localized to the cap rectifier board.

I would first change C1318 with any 100uf cap (doesn't have to be tantalum
to just test). And, do the same sequentially for the other two - swap out
then test. They're pretty easy to change.

Best,
John



On Tue, Apr 24, 2018 at 2:35 PM, Yiu On Tony C via Groups.Io <
tonycheung_hk@...> wrote:

Hi John :
Yes, I did put it back to 7904 , but find something NOT working !
1. Graticule Light always on ----------- I did check NO -15V at R1451,

but

PSU was tested with full loading pass .2. Front panel push button's light
not function.3. For Horizontal Mode Switch ALT , CHOP will shut down the
PSU. A or B is OK !
Regard
Tony Cheung


From: JJ <jajustin@...>
To: [email protected]
Sent: Wednesday, April 25, 2018 1:21 AM
Subject: Re: [TekScopes] 7904 Mainframe damaged

Tony, have you tried putting the PSU back into the mainframe to see if it
works in its normal load and interaction environment? You may be chasing
the proverbial wild goose! :)

Best,
John

On Tue, Apr 24, 2018 at 12:16 PM, Yiu On Tony C via Groups.Io <
tonycheung_hk@...> wrote:

Hi:
It is a good idea ! But I leave my workshop already ! May check day

after

tomorrow. It is midnight in Hong Kong !
One thing I find at least, I disconnect-15 V loads, it become normal.

So

that my next step to check the resistor!
Regard
Tony Cheung


從我的 iPhone 傳送

dadhills@... 於 2018年4月24日 下午11:40 寫道：

Have you checked R1286, 1K, 5%? It is a carbon composition resistor

and

if it has drifted high in value, it will lower the "total" load current
limit threshold.

Dave

On Tue, Apr 24, 2018 at 07:59 am, Yiu On Tony C wrote:


Hi Harvey :
My PSU are test alone and connected to Loading resister and " sens "

return to

LV board.
For the current limiter was controlled by U1275 , and I find the

zener

VR1297

---- 9V are opened , I replace it and now always shut down again ,

even

I open

the zener again , still shut down ! It means cannot return as

before ,

I did

check all the diodes on A12 , but no defective !
There may be something wrong again !
RegardTony CheungAPR 24 2018

From: Harvey White <madyn@...>
To: [email protected]
Sent: Tuesday, April 24, 2018 9:41 PM
Subject: Re: [TekScopes] 7904 Mainframe damaged

On Tue, 24 Apr 2018 13:27:01 +0000 (UTC), you wrote:

Hi :
My 7904 still have problem for PSU !
+/-50V, +/-15V, +5V , +130V are normal with loading power on !
Once I connect the loading to +5V light , it cannot load even 0.1A

!

it will

shut down PSU. ( If normal , It can output up to 3.5A ! )

The diode for +5V lights are defective ( shorted ) and replaced by

fast

recovery diode FR607 ( junction Cap 80pF )or HER208 ( junction cap

14pF ).

Sounds as if the supply is working normally for regulation, but the
shutdown circuitry is too sensitive. There's a current sense

resistor

in there, and some associated parts. Could the current sense

resistor

be too high in value? That will certainly cause this supply to shut
down on a lower current. I'd also check the amplifier for that

sense,

to make sure it is working properly.

Harvey