Date   

Re: 7854 ROMs: the whole story

Dan G
 

On Mon, Apr 27, 2020 at 11:37 PM, Paul wrote:

1) What were all of the ROM versions? Mine end in -00, and I've seen others
end in -01. I've also seen reference to version 1.00, 1.01, and 1.03. How do
the part #s map to ROM version #s?
I have firmware version 1.03, which is made up of the following ROMs:

U100 160-0408-01
U110 160-0409-01
U120 160-0445-02 (Signetics 3-state FPLA)
U200 160-0410-01
U210 160-0411-01
U400 160-0466-02 (8x2k EEPROM)
U410 160-0467-02 (8x2k EEPROM)

You can find these ROM images on TekWiki.

I have also mapped the p-terms of the FPLA, so I can provide this information
if anyone is interested in checking the correctness of their FPLA, or if they
want to program a new one.

I have measured data access times of the original ROMs and EEPROMs
on my 670-5847-03 ROM board:

Mask ROMs: approx 160 ns from falling edge of _CS to stable data asserted.
EEPROMs: 280 ns

One slightly curious detail is that the FPLA contains a few terms that will
patch sections of the mask ROMs with exactly the same data from the
EEPROMs. Now, it could be that this was just an oversight at Tektronix,
or because the EEPROMs were also supposed to patch slightly
different ROM versions. However, another possibility is that this was an
intentional and subtle way of introducing extra wait states during
timing-critical code paths. I haven't yet disassembled the firmware
to check this possibility.

By the way, I had a problem with my ROM board years ago.
It turned out to be a completely missing solder joint on one of the EEPROM
socket pins: it was simply touching the PCB, and I guess that was good
enough to keep it the scope working for a number of years after it left
the factory.


dan


Re: 7854 ROMs: the whole story

David DiGiacomo
 

On Mon, Apr 27, 2020 at 9:37 PM Paul <devpool0@...> wrote:
1) Replace the original ROMs with pin compatible EPROMs. Image versions must match the FPLA, and burning requires a 25 V programmer.
This is not quite correct. If you use 24 pin EPROMs with the combined
images, you don't need the FPLA.

Also, you can use the CY7C264 EPROM, which programs at 12.5V.
(Unfortunately it's still not supported by the TL866.)


Schematics for 4041 DDU Disk Drives

Gary Robert Bosworth
 

Does anyone know where I can get the schematics for the 4041 DDU disk drive unit? I am interested in the power supply and interconnect SCSI cable.
Gary


7854 ROMs: the whole story

Paul
 

Greetings -
I've spent much of the weekend crawling through all of the posts related to the 7854 ROM issue & replacements, in an attempt to capture all of details and possible repair techniques, and compile a canonical reference for the Wiki.
In short, I understand 3 options for units up to serial # B100000:
1) Replace the original ROMs with pin compatible EPROMs. Image versions must match the FPLA, and burning requires a 25 V programmer.
2) Replace with more modern EPROMs. Requires some rework, may or may not leave room for the GPIB board, bypasses the FPLA. This approach has been well documented.
3) Build a new board. In various posts over the years, Nathan Johnson, David Hess, and others talk about their progress in designing boards, but I haven't been able to dig up any design files or examples of these in existence.
A gentleman named Klaus on the facebook page says he and a friend have recently completed a working design for a combo ROM/Diagnostic board. I have yet to hear how much they're looking to sell them for, or if they will share the designs. Are you on here, Klaus's friend? I just haven't seen any traffic on this issue in the last few years.

My outstanding questions:
1) What were all of the ROM versions? Mine end in -00, and I've seen others end in -01. I've also seen reference to version 1.00, 1.01, and 1.03. How do the part #s map to ROM version #s?
2) What are the differences in ROM versions? Are later ones noticeably better?
3) Are their any design files out there for replacement boards, either the combo ROM/Diagnostic, or the combo ROM/RAM?

I've compiled a more exhaustive (or exhausting) version of my notes here:
https://paulcarbone.com/blog/tektronix-7854/

Thanks!
Paul


Re: TEK 5054 CPU Fan

Harold Foster
 

Yes, I left the old one in... no reason to move it and I look at it as a built-in backup - one which already came in handy after learning not to jump in updating drivers. As for me, I'm going to leave all the OS elements just as they are; everything is working perfectly and apparently in calibration - just faster. I have inquired with Tek's support about the availability of the original OS and other software that came with the scope but haven't heard anything back yet; I'll pass on anything if they get back to me. All in all I really don't need it on the network since I tie all my TE together via GPIB with a NI USB-HS feeding to my computer - that lets me do anything remotely that I want without the risks inherent in having it live on the network.

Hal


Re: A bad CRT in 465?

Brendan
 

On Mon, Apr 27, 2020 at 11:43 AM, Timothy Freeze wrote:


$100 calibration is a bargain. I'll definitely look at that option. I'll
have to do the basic performance test before sending it in. I still want to
try some calibration checks for the educational value.

I see the PG506 is the right tool. I understand your very clear explanation
and appreciate you sharing your knowledge. If I can get mostly calibrated and
learn the principles, heck, I'm way ahead.

It will help me understand the scope more and that's good too.

Kindest regards,
Tim
I just looked on the auction site and PG506's look like they are commanding a pretty penny right now. When I was in the market I searched for weeks until one came up in my price range. Same with the aforementioned tunnel diode pulser. If you are not in a hurry and don't mind spending time searching one will pop up for a reasonable price. A one time calibration is nice until you get another scope and another and so on.
The 500 series mainframe can be found pretty cheap. Once you get the mainframe you can start collecting tons of plugins and drive your significant other insane.

Brendan


Re: TEK 5054 CPU Fan

David Kuhn
 

" On thing, though: do NOT connect a lan cable and update any drivers -
many are modified versions and are required for everything to work
correctly"

Thank you for that information and link. I did hook a network cable (after
imaging) and tried to get it to work on my network (to save screen shots to
a network drive perhaps) but I could not get it to work. I think there is
some more modern protocols missing from Windows 2000. It has been too damn
long since I worked with 2000. I can remember upgrading computers from
Windows 2000 to Windows XP at work.

What would be really cool is if we could install Windows 7 on it. I doubt
the Tek software would work. Would it work under Windows XP? I don't have
the TEK software and drivers to try or I would take a few hours to try.
Windows XP works in the network just fine.

Right now, waiting for the heatsink to show up. The SSD does boot now and
the scope software comes up and works. I shut it down right after that as
the heatsink does get toasty without the fan working.

Did you leave the old IDE drive in scope? I'm not sure how to get at the
lower screws that hold it to its mounting plate.

Dave

On Sat, Apr 25, 2020 at 9:05 PM Harold Foster <@HalFoster> wrote:

David-

It's a Celeron / Pentium Socket 478. Seems like the heatsink had a
different mounting arrangement, but I'm not completely sure. I just popped
it off long enough to change the processor and I don't remember the details.

If you want to upgrade the memory, this is what I got:


https://www.ebay.com/itm/2GB-2x1GB-PC2700-DDR333-333Mhz-184pin-Desktop-Memory-NON-ECC-Low-Density/221428970247?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

Four 1GB sticks installed and it "just worked".

I just upgraded my TDS5054B-NV-AV to the aforementioned 4GB, a Pentium 4
2.8GHz and a Sandisk 240GB SSD - no issues, and it makes for a noticeable
change. On thing, though: do NOT connect a lan cable and update any
drivers - many are modified versions and are required for everything to
work correctly.

Hal




Re: A bad CRT in 465?

Tom Gardner
 

On 27/04/20 18:19, Timothy Freeze wrote:
The schematic I've seen puts five (inverters?) in parallel with with each one in series with a 220 ohm resistor. This feeds a 50 ohm load. I'm not sure but it sounds like it might work.
That seems plausible, indicating the gate has an output impedance of 30ohms.

I fully understand your comments about inductance being critical at such a fast rise time. I'll buy the PCB available for this little project and it connects with the board directly to the scope.
Sounds good, particularly if it has a solid ground plane and the capacitors have short leads.


I'm a little over my head, but I want to learn, so thanks for the help.
Yes, pushing the envelope is fun, isn't it.

If you understand the fundamentals (and it sounds like you do), you have a better chance of working out the "suboptimalities that you will correct when you have the time and inclination :)

The other thing that is extremely useful is an RF 50ohm attenuator.

The 2.5V output step is great for tweaking many parts of the scope, but the front-end attenuators require a smaller input when tweaking their frequency response.


Re: A bad CRT in 465?

Giovanni Carboni
 

[image: rejuvenator.png]
73 Giovanni IZ5PQT


Re: A bad CRT in 465?

Timothy Freeze
 

$100 calibration is a bargain. I'll definitely look at that option. I'll have to do the basic performance test before sending it in. I still want to try some calibration checks for the educational value.

I see the PG506 is the right tool. I understand your very clear explanation and appreciate you sharing your knowledge. If I can get mostly calibrated and learn the principles, heck, I'm way ahead.

It will help me understand the scope more and that's good too.

Kindest regards,
Tim


Re: A bad CRT in 465?

Eric
 

Be very sure what is riding on the filaments voltage. In some Tek get the -HV supply DC is on the filaments. I think it is this way is some of the 7K frames and I am sure it is this way on the 576. You also don’t want to turn the bonding wires in to a fuse that will make the tube terminal for sure.

Eric

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Timothy Freeze
Sent: Monday, April 27, 2020 1:55 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] A bad CRT in 465?

So I'm thinking after reading a few CRT rejuvenation links that I may put 6.3 volts DC or maybe a little more on the filament. And tie the grid to the cathode. Sort of a quick and easy attempt at bringing the tube back to life.

Nothing really to lose here and pretty safe I think.

Kindest regards,
Tim


Re: A bad CRT in 465?

Eric
 

To do the high frequency compensation of the vertical what is needed is a clean square wave with a nice level top. This is for the bulk of the adjustments. The high frequency part of this needs a square wave with a fast rising edge with as square a corner as you can get (as little ringing / overshoot / undershoot as possible). The sharper the corner and the better. The bandwidth of the scope dictates the speed of the edge. In working on 485's I needed 350 Mhz this is attained with a tunnel diode pulser. But this is only good for the 1st 40ns of the wave then it Is unleveled again. The pg506 gets the 1ns rise from a -1V signal shorted to ground which gives you a "rising edge". The falling edge is a +1V fall to ground also if you are sure you can get it close Tektronix will actually still factory calibrate a 465 for $100.00 at least that is what they quoted me when I called. The 485 was $850.00 and the 7000 frames were over a grand for a factory cal. Needless to say I worked through those myself. Learned a lot too.

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Timothy Freeze
Sent: Monday, April 27, 2020 1:19 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] A bad CRT in 465?

Tom,
The schematic I've seen puts five (inverters?) in parallel with with each one in series with a 220 ohm resistor. This feeds a 50 ohm load. I'm not sure but it sounds like it might work.

I fully understand your comments about inductance being critical at such a fast rise time. I'll buy the PCB available for this little project and it connects with the board directly to the scope.

I'm a little over my head, but I want to learn, so thanks for the help.

Kindest regards,
Tim


Re: A bad CRT in 465?

Timothy Freeze
 

So I'm thinking after reading a few CRT rejuvenation links that I may put 6.3 volts DC or maybe a little more on the filament. And tie the grid to the cathode. Sort of a quick and easy attempt at bringing the tube back to life.

Nothing really to lose here and pretty safe I think.

Kindest regards,
Tim


Re: A bad CRT in 465?

Timothy Freeze
 

Tom,
The schematic I've seen puts five (inverters?) in parallel with with each one in series with a 220 ohm resistor. This feeds a 50 ohm load. I'm not sure but it sounds like it might work.

I fully understand your comments about inductance being critical at such a fast rise time. I'll buy the PCB available for this little project and it connects with the board directly to the scope.

I'm a little over my head, but I want to learn, so thanks for the help.

Kindest regards,
Tim


Re: A bad CRT in 465?

Timothy Freeze
 

Sounds like a good option for me and good enough for hobby electronics.

Kindest regards,
Tim


Re: A bad CRT in 465?

Timothy Freeze
 

I tried tying them together and no improvement.

Kindest regards,
Tim


Re: A bad CRT in 465?

Timothy Freeze
 

Very good thought and I was able to adjust both cathode and grid to the same voltages, but still no brighter.
Bright scope Cathode= -2459, G=-2536, and Vkg=-77 volts. Dim scope Cath=-2452, G=2529, and Vkg=-77 volts.

I also checked the calibration of my HV probe using measured 110 v source and it was ok.

So I guess I need to replace the CRT.

I guess a face shield and gloves are in order!

Thank you and kindest regards,
Tim


Re: A bad CRT in 465?

Tom Gardner
 

On 27/04/20 16:38, Timothy Freeze wrote:
Thank you.

Great ideas. I'm enjoying the education and learning stuff completely new to me. I'm looking at "Schmitt Trigger Oscillator / Tutorial / 74AC14 Inverter / squarewave generator" described at:

https://www.youtube.com/watch?v=NuXitMK3HSA

I'm hoping this is good enough for the fast rise time square wave. It looks like I'll have some fun with it.
That will be fine as an oscillator, but would not use that to directly drive the load. At a minimum you should use the other gates in the package to drive the load.

The logic gate will be swinging through 5V driving a 100ohm load, so the current change will be 50mA occurring during the transition. If the transition time is t, then di/dt=0.05/t, and since t is small, di/dt is large.

Now remember than the voltage induced across an inductor is v=L*di/dt, and realise that the rule of thumb is L=1nH/mm of wire length. Now measure the distance between the power pins and the center of the IC, calculated the inductance and hence the induced voltage. That voltage will "corrupt" the output you are after.

All that should indicate the benefits of small packages, and emphasise the necessity for excellent decoupling.

You will be driving a 50ohm cable terminated so you need the output impedance to be 50ohms - you can't simply connect the gate's output to the line and expect good results. Achieving the 50ohms output impedance is done by inserting a series resistor R between the gate output and the cable. The value of R is chosen so that the output impedance of the gate plus R is 50ohms. For 74LVC, that means R should be 43ohms. I don't know what it should be for 74AC.

For my generator I had three 74LVC logic gates each in a different small package. That minimises the inductance and allows three decoupling caps to be close to the chip. Each logic gate had a 143ohm series resistor so the output impedance wad 150ohms. Put three of those in parallel and you have 50ohms output, and each gate only has to supply 1/3 of the current. All of that maximises the chances of getting a clean step edge.

Now the 74LVC gates have sufficient output drive that they can nominally drive a 50ohm line directly, so only having to drive 1/3 of that load is well within their capabilities. I do not know the 74AC output characteristics.


Re: A bad CRT in 465?

Timothy Freeze
 

Oops, I shouldn't do arithmetic in my head when I'm not fully awake! You are so right about the numbers and I'm a goof, lol.

The focus is good, the only problem is the dim trace.

I'll try your suggestion adjusting the HV to 2450, using R1400 and then the CRT bias.

I'll go do that now.

Kindest regards,
Tim


Re: A bad CRT in 465?

greenboxmaven
 

You are speaking of a brightener. They would give an acceptable picture for maybe a years or so. They also were offered in versions that isolated the heater from the cathode to allow a tube with a H-K short to be used until it weakened with time or developed a more serious fault. They were designed for operation on 60 cycles from a filament transformer. The advent of TVs using the horizonal output or a switching power supply for the CRT heater pretty much precluded the use of brighteners. This was not much of a problem, however, because by that the time these filament supplies became popular, the lifetime and reliabilty of CRTs was greatly improved and TVs were discarded for other faults or replaced with a newer or different set. A rejuvenator was a test instrument that would deliver a heavy jolt of current between the cathode and 1st. grid. This would blast impurities off the cathode surface without ( you HOPED! ) damaging the layer underneath. It also tended to enlarge the aperature in the first grid, which also increased spot size on the screen. Brightness could be greatly improved, but resolution was likely to suffer. People searched for fifty years for a truly effective and long lasting way to revive weak CRTs, but no truly satisfactory method has yet been discovered. Meanwhile, the art of rebuilding them is on the thin edge of being lost.


Bruce Gentry, KA2IVY

On 4/27/20 11:16 AM, stevenhorii wrote:
I recall that in the days of CRT-based TVs, there used to be a device that
you would plug between the CRT and the socket for it (so, relatively easy
to install). I think it was called a “rejuvenator” or something similar.

Do any of you know if this thing used the method described by Jan?

Steve H

On Mon, Apr 27, 2020 at 10:30 Jan Weber via groups.io <jancarlweber=
googlemail.com@groups.io> wrote:

In the last few weeks there was a discussion in the HP group about
rejuvenating CRTs by running the filament at slight overvoltage and at a
higher anode current to break up the oxide layer on the cathode. Maybe you
can give this a try?

I'm sorry, but I can't remember the link someone posted there, where the
procedure was explained. The 'patient' in said example was - IIRC - a CRT
out of a VNA, so the parameters are different, but nevertheless, it might
be interesting.

Regards

Jan

Tom Gardner <tggzzz@...> schrieb am Mo., 27. Apr. 2020, 15:45:

The scope's specs are 2%, and nowadays those are easily attainable with
very
cheap equipment - especially if it is for hobbyist use.

Vertical low frequency attenuation can be done with an accurate PSU
and/or
a
DVM. It helps to have a 10:1 attenuator available.

Horizontal calibration can be done with any crystal-controlled square
wave
generator.

That leaves vertical high frequency (transient) tweaking. That requires:
1 a square wave generator with a 1ns (not nS, my pet hate!) or less
risetime
2 an RC normaliser and 50 ohm load

You can make the latter reasonably easily, provided you don't use a
ceramic
capacitor. Ceramic caps have a voltage dependent capacitance, and at
medium
voltages the risetime can be visible non-exponential. My preference is
for
an
old-fashioned trimcap, but a C0G ought to sufficient (In my case it
wasn't
so
perhaps it wasn't actually a C0G capacitor!)

The former can also be homebrewed, since jellybean modern logic has sub
nanosecond risetimes. I've used 3*(74LVC1G04+143ohms series) to give a
50ohm
output of 2.5V into 50ohms, and a risetime of ~300ps.
You have to pay very careful attention to decoupling and layout; a
solderless
breadboard won't cut it :)

Perfect? No.
Cheap and sufficient for a 465? Almost certainly.


On 27/04/20 13:37, Eric wrote:
Timothy,

Changing the CRT will also require and entire re calibration of the
unit. If
accuracy of measure is important. Which is a very deep rabbit hole I
recently
crawled out of my self, but would do it again. I you need a PG506 an
SG503 and
a TG501 or equivalent to do the cal when you go to sub nS rise time the
cal
gear can get a little exotic. There is also and RC normalizer at the
correct
capacitance for the input. But a Type 184 is really useful for the
geometry
adjustment. This also assumes all cal signal sources are repaired and
in
spec.
A 465 is 100 Mhz so 3.5nS rise time. The PG 506 will get you there with
the
1nS outputs. you will also need a 10x 5x and 2x attenuator in BNC if
memory
serves correctly.

Eric

On 4/27/2020 6:48 AM, Timothy Freeze wrote:
I'm trying to rescue a Goodwill vintage 465 scope. I've repaired all
the
problems except for one.

The final symptom is the intensity control works normally till it gets
to
about a quarter turn, and then the display gets no brighter. The
display so
dim I'd need to put a shade on it to use it.
I'm thinking the CRT is bad. Perhaps the cathode is at the end of
life
emissivity ?

I got the following from Wiki:
"The emissive layers on coated cathodes degrade slowly with time, and
much
more quickly when the cathode is overloaded with too high current. The
result
is weakened emission and diminished power of the tubes, or in CRTs
diminished
brightness."

What I've done:

I set up my other functioning 465 with the same input signal and
measured
voltages and waveforms from the unblanker input (J6-10) through the
high
voltages on the CRT.

At TP1486 the dim scope voltage ranges from 14 to 94 volts as the
intensity
control is adjusted, and compares to the good scope measured at 16 to
95
volts, (both changing with intensity control position).

The CRT negative voltages are:

good scope Dim Scope

Cathode 2450 2440

Grid 2495 2505

Vkg -55 -35

Grid range
with intensity 2526 - 2461 2526 - 2450
change
(min to max)

High voltage 12,960 V 12,860 V

I've also tied the grid to the cathode and it doesn't get any
brighter.
I'm new to the vintage scope repair, but finding it great fun and
like
solving a puzzle. This is a great group and resource. I've learned
lots from reading the messages.. So I'd appreciate any advice before
proceeding with a CRT replacement. I'm planning to put used CRT in it
and have
never replaced one. So, I'd appreciate any tips on how to do it.
I've
found
the manual replacement procedure.

Kindest regards,
Tim