Date   

2221 100MHz Scope repair project

satbeginner
 

Hi all,

I am new to this group.

My name is Leo, I am a Dutch, early retired technician living in Spain. In the period '80-'89 I used to work in a fusion research laboratory in The Netherlands (FOM Rijnhuizen) as an electronic design engineer.
I worked with Tektronix scopes, mostly 475/485 models, however, also a lot of old 5xxs were used. Also with the Polaroid camera's on them..... ;-)

Now I own a working 455 for my little home-projects.
Just this weekend I bought a non working model 2221, 100MHz scope on Ebay.
I do not have it right now, it will be shipped from Germany to me this week.

Restoring this to a working order will be my project this summer.

The known problems so far are (based on the sellers info and pictures):

- there is no trace visible; there is text on the screen though, also there is a trace when beamfinder is pressed.
- Both the knobs for Time Base are damaged. The Cal button broke of completely, the main Time Base rotary button is damaged, but still there.

Even before the unit reaches me I have some questions:
- is a service manual available? (pdf?)
- are the knobs still available? (Scrap-unit, imitation, other sources ??)
- can somebidy explain the real difference between the models 2221(60MHz?), 2221A (100MHz?) and 'my' unit, marked 2221 100MHz? (without the 'A')

Looking forward to some feedback,

Un saludo,
Satbeginner


Re: Tek 485 - Intensity at Max, Cannot Adjust Down

Tom Gardner
 

Nothing below contradicts your presumption that the DC restorer circuit is faulty.

Note that a month ago I hadn't looked at a CRT since school (not university) physics lessons.

On 21/03/16 16:36, otrfan01@... [TekScopes] wrote:

and the meter moves a very slight amount higher when measuring the grid.
That phrase makes me suspect all your measurements are w.r.t. ground. The critical voltage w.r.t. intensity is the difference between the cathode and grid, not the absolute voltage above ground. I put my handheld DVM directly across the cathode and grid, so I measured Vkg and did not have to compute Vk-Vg.

Note carefully I am not saying going anywhere near -2450V is safe; use your judgement and exercise extreme caution. Not being near the DVM/scope/leads when the scope was on made me feel less apprehensive :)

I see in your tests that the grid was driven -50V wrt the cathode, which pointed towards the bad CRT. My results seem to lead me back to the DC Restorer circuit.
I am led to understand that is a traditional problem.

Since the caps are expensive and hard to find, I replaced the cheap easy HV diodes. When that didn't improve anything, I started looking at changes in Vkg (average measured with DVM not a scope!) with changed intensity control and changed timebase/blanking waveforms. Proportionate changes in Vkg lead me to suspect the DC restorer was working, and hence the problem had to be in the CRT itself.

Thanks again for your help!
This group was very helpful to me; I try to return the favour.


Re: 7854 RAM card upgrade only - any interest?

petepdx1955
 

Dave, Looking this over, it *can* be routed on two layers, so using 4 would be for noise control reasons I assume?
The internal costs were pretty much the same around the time the board was designed and having a
ground plane made the decision pretty easy. Toss in the BOM cost vs the sale price and that slight
additional cost was in the noise.


Ah the good 'ol days when profits where so good. But there were times I've seen a 4 layer board where
2 would have been close to an overkill.


-pete


Re: 11402s in Portland for a very good price

Ashley40@...
 

I'd be happy to pick them up. Local brew ?

Thank you




Thank You
Kiss-Electronics
Ms Ashley Hall
183 N 5th Avenue
Cornelius, Oregon
97113


W7DUZ


www.kiss-electronics.com

-----Original Message-----
From: petepdx1955@... [TekScopes] <TekScopes@...>
To: TekScopes <TekScopes@...>
Sent: Mon, Mar 21, 2016 10:35 am
Subject: [TekScopes] 11402s in Portland for a very good price

Just been going through a stack of oldies and in it is a pair of 11402's

Have not yet had time to plug them in, but like to know if anyone local as in no packing/shipping
is interested before I spend the time.


I'm asking one local brew each


-pete









------------------------------------
Posted by: petepdx1955@...
------------------------------------


------------------------------------

Yahoo Groups Links


11402s in Portland for a very good price

petepdx1955
 

Just been going through a stack of oldies and in it is a pair of 11402's

Have not yet had time to plug them in, but like to know if anyone local as in no packing/shipping
is interested before I spend the time.


I'm asking one local brew each


-pete


Re: Shootout of Substitute SG504 heads = a better head.

 

Hi Craig,
I don't expect to be able to measure such a small difference. I expect to
learn more about HOW to measure such small differences. I see it more as an
opportunity to sharpen my measurement skills and make them more accurate.
David and Ancel have the RF Measurement equipment and the RF experience
necessary to make these kinds of measurements. What I have is older Tek
equipment that can be used, with some cleverness on my part, that I hope
will approach their accuracy and precision and fill in the gaps between the
frequency test points they used.

When I get Ancel's head I will have all three heads to compare. I also have
a different measurement methodology than David and Ancel which you can read
about (with pictures) in the archives. You can get it at
https://groups.yahoo.com/neo/groups/TekScopes/files/015-0282-00/
Its title is 015-0282-00 Head Comparison.pdf.

So my ultimate hope is to make a small contribution to the results and learn
more about how to make RF measurements in the process.

Dennis Tillman W7PF

-----Original Message-----
Sent: Sunday, March 20, 2016 11:54 PM
Subject: RE: [TekScopes] Shootout of Substitute SG504 heads = a better head.

Some might wonder why it would be worth concerning ourselves with
differences that are less than 0.1dB. The answer is simple - because we
can.
Dennis Tillman W7PF
I agree with the sentiment Dennis - I'm just as anal about measurement
accuracy as most of us. But in practical terms, 0.1dB (even as a delta) is
(a) very hard to measure (b) you can easily see differences of this order,
and larger, by wobbling a BNC connection or inaccurately torqueing an SMA,
(c) is right on the hairy limit of insertion loss and VSWR for top flight
SMA's http://www.amphenolrf.com/connectors/sma.html and certainly on the
hairy edge for BNC's http://www.amphenolrf.com/connectors/bnc.html and (d)
is less than 0.1 division on an 8 division screen.

Craig
------------------------------------
Posted by: "Craig Sawyers" <c.sawyers@...>
------------------------------------


Re: Tek 485 - Intensity at Max, Cannot Adjust Down

wmactor
 

Thanks, Tom -

I read the thread that you referenced, and borrowed a high voltage meter . I'm seeing approximately +2700 volts on the cathode of the CRT, and the meter moves a very slight amount higher when measuring the grid. The intensity control has absolutely no effect on the grid voltage.


I see in your tests that the grid was driven -50V wrt the cathode, which pointed towards the bad CRT. My results seem to lead me back to the DC Restorer circuit.


I'll see if I can find replacement diodes and caps, and see if that resolves the issue.


Thanks again for your help!


Re: Shootout of Substitute SG504 heads = a better head.

mosaicmerc
 

Interesting bit of kit.

I am looking at modifying the TEK SG504 head with an add on instrumentation grade opamp to feedback into the AM input of the HP8753VNA and achieve an improvement on intrinsic the +/- 1dB flatness to 3Ghz.
In effect such a capability would almost make the SG504 unnecessary for me. But I like the instrument's serviceability so I'll keep it.

If that project works well I may just upgrade the jellybean LM358 in the SG504 to get improved SG504 leveling Spec.


Re: FET leakage question

 

@David: Thanks! The application I have in mind is less than 100 Hz, which I am hoping will help with stability.

Dave


Re: OT: underground pipe locating

Greg Muir
 

Stefan,

The local natural gas utility now buries a radiating coax along with their piping around here which is a close parallel to what you do. When tracing pipes, they turn on the RF source at the nearest pump/valve house and then grab their field strength meter.

Greg


Re: Shootout of Substitute SG504 heads = a better head.

Craig Sawyers <c.sawyers@...>
 

There was an earlier instrument - the 067-0650-00. This was based on a GR unit oscillator, with a
Tek designed levelling system. This fortunately was on a captive lead, so does not go missing.

It is specced to 900MHz, although mine goes to 950MHz. Levelling is specified to 5% with respect to
the reference frequency as compared to 4% for the SG504.

Having got one, I don't feel a burning urge to spring for an SG504.

If you use the GR 1263B amplitude regulating power supply with unit oscillators, you get 5% to
500MHz, and with a small correction factor extend levelling to 2GHz. It is a bit more of a faff to
use for sure and end up with basically a kit of GR bits to make it work, but it does go higher in
frequency than the SG504 by quite a margin.

Craig

-----Original Message-----
From: TekScopes@... [mailto:TekScopes@...]
Sent: 21 March 2016 10:25
To: TekScopes@...
Subject: Re: [TekScopes] Re: Shootout of Substitute SG504 heads = a better head.


2445a power supply repair

Max Vlasov
 

Hello,

I've got a 2445a, which was burned by the previous owner (he was debugging the mains power supply without an isolation transformer). After repairs, where I've changed the channel #1 hybrid, TL072, TL074 and MC1458, now it seems, that the scope can be calibrated.
Before starting the calibration, I've checked the voltages on the power supply connector. All are nominal for exception of the analog +5 and -5 volts, which read well before the minimum.
I've got 4.78 and -4.76 volts. Voltages don't drift. Ripple is within the spec. The scope traces are stable before and after the warmup.
I looked at the PSU schematics and +5VA/-5VA are sourced via the linear regulators from the rectifier/filter. Could it be that the C1115 and C1114 caps have "dried out"? I don't see any leaks. Checked the capacitance, seems ok. Don't have a possibility to check the ESR.
Can I substitute a different capacitors instead of 250uFx20V?
Can I use for example, 330uFx25V?
Do I have to use the "low ESR" caps, or I can buy the standard "heavy duty" caps?

Than you!


Re: Shootout of Substitute SG504 heads = a better head.

Mike Usher
 

I would also like to be considered for an assembled head.
Would you be able to ship to the UK?
Thanks,
Mike


Re: Shootout of Substitute SG504 heads = a better head.

mosaicmerc
 

Now you guys are making me feel like spending more time do a VNA test jig for the SMA and BNC connectors & adapters. Just to quantify the 0.1dB variance vs frequency.

I have some short semi flex 26.5Ghz SMA cables and a couple longer Megaphase 4Ghz cables- new stock I can use to do a reliable jig. Probably with a pair of 6dB SMA pads bracketing the DUT insertion point.

Then Cal the VNA & test jig, insert the DUT and see what happens with the S parameters.

Unfortunately my EXCEL VBA addin: Agilent /HPIB linkup with the VNA to grab screens and data isn't working properly. I'll have to rustle up an old floppy to save and xfer screens and data. Looks like a timing error in the USB linkup. I have basic connectivity and can get the VNA to respond properly with single manual GPIB codes sent over the interface, but the automated stuff is breaking down.


Re: Shootout of Substitute SG504 heads = a better head.

mosaicmerc
 

The Stable CW was from an HP8753D VNA, not the SG504. It had about a +/-1dB over the range, but the Boonton allowed me to match peaks and null out errors. Took a while. But I could digitally dial in precise frequencies pretty quick on the VNA as well as scroll thru or auto sweep at slow rates like over a minute....while watching the SA match sweep with a peak tracking marker and pick out any odd wobbles in the levels after each pass of entering cal constants. After several passes I did narrow, slow sweeps and optimized the SA cal constants to reflect the Boonton levels. It took about 30 manual constant entries and a couple hours tweaking to get it right; could have been worse.

I went thru the trouble of calibrating the SA manually with the Boonton so I could dial in relatively precise frequencies on the SA from the SG504, also I could see harmonics. I could record the traces and export the evaluations to a PC as well If need be. Also, it's good to have an SA with decent Cal. for any other purpose.

I am looking at a project to nearly double the SA range so I can use it up to 2.5Ghz (Wifi /bluetooth). Just ordered a good VCO chip and mixer chip that seem to have the specs to do it. Looks like I'll have to do two mixes. One up for the TG and one down for the rcvr. I have a couple mini circuits 2Ghz mixers on hand for the basic tests. Once I can get some confidence in the results I'll buy another mixer chip and build out the system. I can use the VNA for a LO signal and my TPI synthesizer for the RF as it goes to 4.4 Ghz. Test out the concept and then do some RF design to do it at the chip level for a standalone device. The Boonton can validate the results being a 12.4Ghz capable unit. No doubt I'll have some more manual cal. constant fun to make that a useful result.
I can sync all the devices to a 10Mhz source including the VCO chip. My OCXO Trimble isn't here yet though.


Re: Tek 2215

 

I think reliability was the reason Tektronix used carbon composition
and then special film resistors for the focus resistor chain; normal
thin film parts would be been questionable unless significantly
voltage derated. Other oscilloscopes made during this time used a
custom thick film network which was only inexpensive because the same
network was used to regulate the cathode voltage which was not
necessary in the 2 channel 22xx oscilloscopes. I have never heard of
these thick film networks failing.

Carbon composition (and thick film) resistors are bulk devices so they
do not suffer from surface effects like in a thin film resistor. They
also tend to fail open rather than closed although at high voltages, I
suspect that is no advantage.

As far as I have been able to tell, the film resistors Tektronix used
later were specifically designed for high voltage operation and had
some type of special processing done after being trimmed to their
proper value to prevent the surface effects like electromigration
which would normally be a limitation.

On Mon, 21 Mar 2016 00:14:28 -0400, you wrote:

Ha. They probably didn't think someone would be using it 30+ years later. High voltage does do strange things over the long term.

t


Re: Shootout of Substitute SG504 heads = a better head.

Craig Sawyers <c.sawyers@...>
 

Some might wonder why it would be worth concerning ourselves with differences that are less than
0.1dB. The answer is simple - because we can.
Dennis Tillman W7PF
I agree with the sentiment Dennis - I'm just as anal about measurement accuracy as most of us. But
in practical terms, 0.1dB (even as a delta) is (a) very hard to measure (b) you can easily see
differences of this order, and larger, by wobbling a BNC connection or inaccurately torqueing an
SMA, (c) is right on the hairy limit of insertion loss and VSWR for top flight SMA's
http://www.amphenolrf.com/connectors/sma.html and certainly on the hairy edge for BNC's
http://www.amphenolrf.com/connectors/bnc.html and (d) is less than 0.1 division on an 8 division
screen.

Craig


Re: OT: underground pipe locating

stefan_trethan
 

I made a similar noise source for detecting wires in the wall. I used
1MHz then and similar 555 based modulation. I believe I used just a
simple squarewave oscillator, one of these 5V metal box jobs, and a
small AM radio.
It worked very well (too powerful even) but I think the high frequency
would not go through soil much.

In hindsight I should have just hammered nails into the wall. Because
of Murphy's law it is certain any nail hammered into any wall will hit
a wire. Murphy is much stronger than dousing even.

Goes to show why politics and religion is forbidden on these lists for
good reason. It always turns out the man you thought you had so much
in common with is really crazy, and you only had a very narrow subject
in common after all. Best not to stray off topic really, and stay
friends respecting each other.

ST

On Sun, Mar 20, 2016 at 10:17 PM, Dwayne Verhey tekscopes@...


Tracking the wire for the fence, I generated an RF frequency of around
600KHz with a signal generator and modulated it with a 4 or 500Hz AF 1/2
wave generated by a real down-and-dirty 555 based circuit, connected one
end to the wire and followed the sound with a cheap $10 transistor
radio tuned to my carrier. A bit of time was spent fine-tuning the
output power so I was following the wire instead of listening to the
source but once I had that figured out, I was able to locate the break
within a foot or two in short order.


Re: OT: underground pipe locating

Greg Muir
 

Once a skeptic, always a skeptic...

I think someone should talk to some of the boys at Verizon who use this technique for locating their cables. They will be interested in knowing that "...it doesn't work..."

Greg


Re: Tek 2215

 

Ha. They probably didn't think someone would be using it 30+ years later. High voltage does do strange things over the long term.

t

----- Original Message -----
From: David @DWH [TekScopes]
To: TekScopes@...
Sent: Sunday, March 20, 2016 11:55 PM
Subject: Re: Fwd: Re: [TekScopes] Re: Tek 2215



In a pinch that is a great idea. Using 1/2 watt resistors would be
even better.

I really think Tektronix ran the Allen-Bradley carbon composition
resistors too close to their voltage rating but apparently it worked
out well enough given that they lasted so long. Maybe they had
experience showing that they did not need to be voltage derated.

On Sun, 20 Mar 2016 23:38:57 -0400, you wrote:

>Or you could use a pair of ~ 500k - 1/4 watt film resistors standing off the board for each of the 1 meg resistors. That would spread out the HV stress a bit more.
>
>T