Hi Patrick,


I removed the 5mhz oscillator module and found that it is mounted with 4 pins. My option one TG501 has a 7490(U50) mounted on the bottom of the main pwd. The ground pin is under U50. The +15V pin is 1 inch in the direction of the rear edge connector. The output pin is 1 inch in the direction of the board edge. The 4th pin, a NC( co connect) pin is one inch toward the front of the TG501. After removing the oscillator, Y1, I powered up the oscillator module with a bench power supply and found it working with a 4V squarewave on the output pin. At this point I do not know why it did work in circuit. The 8V I found on U50 pin is a mystery now.


The Y1 osc module had been serviced before. The pads were damaged and may not have been making good contact. U50 is on the etch side of the main circuit board. The osc module is on the component side of the main board. The A2, multiplier circuit with all the fancy RF plumbing covers the main board component and must be removed to access the main board component side. U50 is mounted with 14 feedthru pins. This prevents the osc module from mounting flat on the main board(A1). The metal case of the Osc module may have been touching the U50 pins.


The A2 multiflier pwb mounts on the main pwb (A1) with 4 standoffs and can be unplugged for removal. Page 5-9 of the manual show jumpering for the oscillator options. My module had all the Option 1 Clock jumpers except J4 but that is an output and should not have caused my problem.


I do not remember checking the +15V regulator circuit which uses a power module(TM500) transistor. That may be bad. I think I can power up the Main board with the multiplier removed.


I fear U50, U150, U220, and U215 may be damaged by what ever caused the 8vdc on the clock line.


Jerry Massengale
jmassen418a@...




-----Original Message-----
From: Patrick Wong <patwong3@...>
To: TekScopes <TekScopes@...>
Sent: Thu, Jan 19, 2012 3:35 pm
Subject: [TekScopes] Re: TG501 option 1 connections





Hi Jerry,

I also could not find power connections for Y50 on the TG 501 schematic. Can this be determined via visual inspection? (My TG 501 does not have that option.)

You should see a 5 MHz signal at pin 1 of U50, and a 1 MHz signal at pins 11 and 14. Since you only measure a DC voltage at pin 1, perhaps Y50 has failed.

I do not understand your question, "Can the module be powered up with the A2 multiplier board?" Which module are you referring to?

Patrick Wong AK6C

--- In TekScopes@..., Jerry Massengale <jmassen418a@> wrote:

I have a TG501 with no output from the Option 1 oscillator. Pin1 of U50 has almost 8VDC. Where can I find the power connections for the Y1 circuit. Can the module be powered up with the A2 multiplier board?

They make low profile sockets which work as you describe. The
machined collets have no pin on the bottom and sink directly into
larger than normal holes in the board.

Message #70804 indicates that the socket adds not-insignificant L and C. If these are big enough, it may upset the operation if something was ringing and activating parasitic junctions - that was one of the theories discussed. TTL bangs the lines pretty hard, so the L can make a big difference. If the board is probed and seems OK, it doesn't mean the signal on the IC lead at the package is too, if there's enough inductance in between. ECL is more forgiving since it has smaller signal swings and is mostly differential.

Ed

--- In TekScopes@..., "phosphorphile" <gumbear@> wrote:

I doubt that the use of a socket would be the cause of what we've been talking about although it may cause some reduction of the maximum frequency margin. For the sake of that issue, what is the maximum frequency at which your circuit still functions with your socket in place?

I've created a "clean" socket by removing the machined contacts from a socket and inserted them into a board with an IC plugged into them for alignment. Removing the plastic lowers capacitance a bit. Sinking contacts into the board would increase the coupling between pins in the board so the goal should be to design circuits that function WITH sockets and then their eventual removal would favor guaranteed performance. Good design always allows for plenty of safe margin because Murphy is always at work looking for ways to crowd performance.

Arden

For this test circuit I intentionally used a socket, in spite of the
performance hit I knew it would give me, because I wanted to try several
different 74F86 brands and date codes. So I picked the socket with the
smallest pins hoping they would have the lowest capacitance.

One nice side effect of using a socket was that at one point it became
important to characterize the socket capacitance. I figured out a way (with
a TDR) to measure, and later to confirm, the socket capacitance as well as
the 2-3pF of input capacitance contributed by the actual gate itself.

But, I was really hoping that the untied inputs were the culprit.

They weren't. They made no difference tied or untied as far as I have been
able to tell so far.

Phosphorphile,

In the broadest sense it is similar to the 74F86 - .........

Phosphorphile posted

Ayup. I always wondered why Tek used the 8393, the 13 > volt filament version of the 7586. Anyone?

I dont know what instrument Tek used the 8393 in, so cant comment on an actual TEK unit, but running a13v volt filament on 12 v (provided the emission is high enough so the characteristics still suit the application ) will have two effects
1. All other things considered, it can lengthen the working life.
2. It will raise the input impedance much higher than when run at the specified filament watts.

I remember running 6.3 v fil s at 4,0v to get hi Z inputs for electrometer applications in the distant past.
Cant remember the tube type now, maybe an EF37.
Underrunning filaments in this way was a common thing to do in high input Z applications, but you had to select the right tube that could handle it, without poisoning etc.
Special electrometer tubes were made with deliberately underrun filament voltages and some were specially selected types from a production run of normally 6.3 V filament units tyen rebranded to a different type number.
The anode current was sigificantly reduced when this was done for the same grid base.

A somewhat parallel thing was done with the LM308A IC, which achieved a 40Mohm Dc Zin by running in "current starved " conditions

But maybe TEK had a diferent reason?
Jojn Byers

Hello Jimmy.
You were quite right. A flying lead connector had come off and I had
replaced it the wrong way round! The cathode of VR890 measured zero
volts and the anode measured -8.86volts. ( I am grounding my neg probe
to chassis ) Also, pin 6 of P901 is zero ohms with respect to my chassis
grounding probe.
David.

David,

Something seems very wrong with your V measurements, OR you have a
grounding problem. One end of VR890 (the cathode) [NOT VR980] should
be very close to 0 volts, referenced to chassis ground. Did you ground
one probe (the negative/black) of your DMM to the chassis when you
measured these two voltages?

P901-6 of the LV Power Supply Board is the GND SENSE FEEDBACK
connection to/from the Main Interface Board (service sheet 1). This
connection connects the reference ground on the LV PS Board (ground
symbol looks like a triangle) to the chassis ground on the Main
Interface Board (ground symbol looks like a 3-prong rake). This
prevents ground currents, thus mimimizing noise in the V reference.
With the scope OFF, do a continuity check to make sure that pin is
very close to 0 ohms, referenced to chassis ground (surely less than 1
ohm assuming the resistance of your DMM leads are zeroed out). Someone
else please chime in hear if that's not correct. Then with the scope
on, you should get very close to 0 volts on the cathode of VR980
referenced to chassis ground, and right around 9 volts on the anode.

Let us know what you find. If your measurements were in error, please
repost corrected Vs.

Jimmy

I have a TG501 with no output from the Option 1 oscillator. Pin1 of U50 has almost 8VDC. Where can I find the power connections for the Y1 circuit. Can the module be powered up with the A2 multiplier board?

?

?

For this test circuit I intentionally used a socket, in spite of the
performance hit I knew it would give me, because I wanted to try
several different 74F86 brands and date codes. So I picked the socket
with the smallest pins hoping they would have the lowest capacitance.

One nice side effect of using a socket was that at one point it became
important to characterize the socket capacitance. I figured out a way
(with a TDR) to measure, and later to confirm, the socket capacitance
as well as the 2-3pF of input capacitance contributed by the actual gate