Maker Pro
Maker Pro

HP 3310B Malfunction Generator

R

Robert Baer

Jan 1, 1970
0
rickman said:
Ok, I can get to that one. But what is it for? That is not the same unit
at all. It looks like a much more modern unit with wide use of op amps
rather than a bazillion transistors? I guess it might be a better unit
to buy?
Yes, i figured that the design was different even tho the function
seemed to be the same.
Been awhile since i looked at the incomplete Agilent schematic, but
the diode array and biasing should be the same.
In this schematic, except for Q306 and Q309, all transistors are
fully active (no reversed biased E-B or forward biased B-C junctions).
Maybe i should look at their crummy schematic..

...still looking at it; will respond later.
 
R

Robert Baer

Jan 1, 1970
0
*********** BIG SNIP
Yes, i figured that the design was different even tho the function
seemed to be the same.
Been awhile since i looked at the incomplete Agilent schematic, but the
diode array and biasing should be the same.
In this schematic, except for Q306 and Q309, all transistors are fully
active (no reversed biased E-B or forward biased B-C junctions).
Maybe i should look at their crummy schematic..

..still looking at it; will respond later.

See Agilent PDF page 71 TP5 is equivalent to junction on mine
described by CR322, CR321, R353, R351 and R350. Likewise, TP7 is
equivalent to junction on mine described by CR323, CR319, R352, R346 and
R344.
Lastly, the "center", (sinewave) of TP6 is my TP13.

Now on mine, it is mostly obvious that the incoming triangle wave via
R369 (from output of op amp U301) is/must be symmetrical about ground
and (going from left to right) each pair of diodes will "clip" that
waveform at a level set by their common resistor at the output line
TP13; each clipping point being more severe (lower common R value) and
at a higher level (further away from ground).
Finally, the peak is clipped at the top, set by R348 and R342.

On the Agilent, PDF page 18, note there is NO "high frequency" monkey
business WRT the sine forming, as i seem to have seen mentioned. Ditto
with mine. Exception may be small trim caps or such.
TP6 appears to be at ground with the waveform symmetrical, just like
mine.
So...at "low" frequencies (10-1000 cps) the triangle coming from the
triangle amp (Pg 18 again) better the hell be EXACTLY symmetrical and
well behaved.
And it should stay that way all the way to the high end on the top range.
I think that may be the key, and so there may be more than one
problem - the first being this triangle waveform ain't not nowhere near
symmetrical (to use bad Engrish).

TP5 should be as negative as TP7 is positive; i derived that from the
necessary biasing for that DFG, and the schematic voltages noted confirm
that.
Resistor and voltage values from one to the other in the DFG will be
different, but more-or-less on a ratio basis, they should track.

Once you get the triangle symmetrical, then it should be fairly easy
to find a bad resistor or diode in the DFG (am guessing this as being a
secondary problem).
Maybe one of the stared capacitors in the sine shaper amplifier is
bad, but i doubt it; if there is a problem there HF-wise, it would be
due to a transistor (Q2 or Q3).
 
R

rickman

Jan 1, 1970
0
See Agilent PDF page 71 TP5 is equivalent to junction on mine described
by CR322, CR321, R353, R351 and R350. Likewise, TP7 is equivalent to
junction on mine described by CR323, CR319, R352, R346 and R344.
Lastly, the "center", (sinewave) of TP6 is my TP13.

Now on mine, it is mostly obvious that the incoming triangle wave via
R369 (from output of op amp U301) is/must be symmetrical about ground
and (going from left to right) each pair of diodes will "clip" that
waveform at a level set by their common resistor at the output line
TP13; each clipping point being more severe (lower common R value) and
at a higher level (further away from ground).
Finally, the peak is clipped at the top, set by R348 and R342.

On the Agilent, PDF page 18, note there is NO "high frequency" monkey
business WRT the sine forming, as i seem to have seen mentioned. Ditto
with mine. Exception may be small trim caps or such.
TP6 appears to be at ground with the waveform symmetrical, just like mine.
So...at "low" frequencies (10-1000 cps) the triangle coming from the
triangle amp (Pg 18 again) better the hell be EXACTLY symmetrical and
well behaved.
And it should stay that way all the way to the high end on the top range.
I think that may be the key, and so there may be more than one problem -
the first being this triangle waveform ain't not nowhere near
symmetrical (to use bad Engrish).

Yes, that is exactly the problem. The triangle is not symmetrical. The
high speed circuit is not about clipping the triangle into a sine wave,
it "adjusts" the circuit that detects the triangle has reached +-5 volts
and reverses the direction of the waveform. The -5 volt detect seems to
work ok on the lower ranges. On the 10k and 100k ranges the high speed
circuit kicks in to modify the operation of the +-5 volt detect circuit
so that it triggers just early enough to give a proper +-5 volt triangle
in spite of whatever delays it has. This is when it malfunctions and
looks like it is not going far enough negative, which is confirmed on
the scope. The trigger level seems to vary erratically which is also
seen on the internal test points.

The symptom is clearly a malfunction of the high speed circuit, plus, I
can see erratic drift of the DC values both on the meter and with the
scope. The drift is small, but the DC values are also wrong on the -5
volt half of this circuit. I just can't pin any of this down since I'm
not clear on how the circuit works. The parts I understand will all
affect each other if any of them malfunction.

TP5 should be as negative as TP7 is positive; i derived that from the
necessary biasing for that DFG, and the schematic voltages noted confirm
that.
Resistor and voltage values from one to the other in the DFG will be
different, but more-or-less on a ratio basis, they should track.

Once you get the triangle symmetrical, then it should be fairly easy to
find a bad resistor or diode in the DFG (am guessing this as being a
secondary problem).
Maybe one of the stared capacitors in the sine shaper amplifier is bad,
but i doubt it; if there is a problem there HF-wise, it would be due to
a transistor (Q2 or Q3).

I'm not worried about the DFG actually, I'm confident the problem is in
the high speed circuit. I just can't figure out which part is bad.
Someone suggested the caps often fail, at least the electrolytics
anyway. There is one cap that looks like an electrolytic, but the
schematic says it is 0.1 uF and the parts list says it is a 1.0 uF part.
I'm starting to loose confidence in HP docs. I may have to shotgun
it... or just use it on the lower ranges only.
 
J

josephkk

Jan 1, 1970
0
Yes, that is exactly the problem. The triangle is not symmetrical. The
high speed circuit is not about clipping the triangle into a sine wave,
it "adjusts" the circuit that detects the triangle has reached +-5 volts
and reverses the direction of the waveform. The -5 volt detect seems to
work ok on the lower ranges. On the 10k and 100k ranges the high speed
circuit kicks in to modify the operation of the +-5 volt detect circuit
so that it triggers just early enough to give a proper +-5 volt triangle
in spite of whatever delays it has. This is when it malfunctions and
looks like it is not going far enough negative, which is confirmed on
the scope. The trigger level seems to vary erratically which is also
seen on the internal test points.

The symptom is clearly a malfunction of the high speed circuit, plus, I
can see erratic drift of the DC values both on the meter and with the
scope. The drift is small, but the DC values are also wrong on the -5
volt half of this circuit. I just can't pin any of this down since I'm
not clear on how the circuit works. The parts I understand will all
affect each other if any of them malfunction.



I'm not worried about the DFG actually, I'm confident the problem is in
the high speed circuit. I just can't figure out which part is bad.
Someone suggested the caps often fail, at least the electrolytics
anyway. There is one cap that looks like an electrolytic, but the
schematic says it is 0.1 uF and the parts list says it is a 1.0 uF part.
I'm starting to loose confidence in HP docs. I may have to shotgun
it... or just use it on the lower ranges only.

May i presume that you are using this document for the schematic?:

http://cp.literature.agilent.com/litweb/pdf/03310-90003.pdf

So that we can all be on the same pages.

?-)
 
T

tm

Jan 1, 1970
0
Yes, that is exactly the problem. The triangle is not symmetrical. The
high speed circuit is not about clipping the triangle into a sine wave,
it "adjusts" the circuit that detects the triangle has reached +-5 volts
and reverses the direction of the waveform. The -5 volt detect seems to
work ok on the lower ranges. On the 10k and 100k ranges the high speed
circuit kicks in to modify the operation of the +-5 volt detect circuit
so that it triggers just early enough to give a proper +-5 volt triangle
in spite of whatever delays it has. This is when it malfunctions and
looks like it is not going far enough negative, which is confirmed on
the scope. The trigger level seems to vary erratically which is also
seen on the internal test points.

The symptom is clearly a malfunction of the high speed circuit, plus, I
can see erratic drift of the DC values both on the meter and with the
scope. The drift is small, but the DC values are also wrong on the -5
volt half of this circuit. I just can't pin any of this down since I'm
not clear on how the circuit works. The parts I understand will all
affect each other if any of them malfunction.



I'm not worried about the DFG actually, I'm confident the problem is in
the high speed circuit. I just can't figure out which part is bad.
Someone suggested the caps often fail, at least the electrolytics
anyway. There is one cap that looks like an electrolytic, but the
schematic says it is 0.1 uF and the parts list says it is a 1.0 uF part.
I'm starting to loose confidence in HP docs. I may have to shotgun
it... or just use it on the lower ranges only.

May i presume that you are using this document for the schematic?:

http://cp.literature.agilent.com/litweb/pdf/03310-90003.pdf

So that we can all be on the same pages.

?-)

+++++++++++++++++++++++++++++++++++


http://ko4bb.com/Manuals/HP_Agilent/HP_3310AB_OS.pdf


This is a better copy.
 
R

Robert Baer

Jan 1, 1970
0
tm said:
May i presume that you are using this document for the schematic?:

http://cp.literature.agilent.com/litweb/pdf/03310-90003.pdf

So that we can all be on the same pages.

?-)

+++++++++++++++++++++++++++++++++++


http://ko4bb.com/Manuals/HP_Agilent/HP_3310AB_OS.pdf


This is a better copy.
Yes, better, but still a lot missing.
At least, see page 76 there for a complete DFG schematic, and pages
79 & 80 for a complete A3 (order is srdawkcab).

HF problems aside, fix the triangle FIRST and make it as perfect as
possible.
 
R

rickman

Jan 1, 1970
0
Yes, better, but still a lot missing.
At least, see page 76 there for a complete DFG schematic, and pages 79 &
80 for a complete A3 (order is srdawkcab).

HF problems aside, fix the triangle FIRST and make it as perfect as
possible.

The HF circuit is the *cause* of the triangle problems. The triangle is
just a constant current into a cap. When it reaches +5 volts the detect
circuit switches the current source to ramp down to -5 volts, then the
cycle repeats. The HF circuit modifies the threshold of the +-5 volt
detect circuit which is part of forming the triangle wave. The problem
is the -5 volt detect is happening too soon and the triangle is not
going far enough negative. The *only* happens when switched to the 10k
or 100k ranges. The only difference in the ranges is that the HF
circuit is kicked in (and the value of the integrating cap, but they are
different between the 10k and 100k ranges).

Just to be clear, the triangle still looks just like a triangle. The
symptoms are:

1) doesn't go negative enough,
2) faster than it should be,
3) frequency is erratic,
4) sine wave is truncated on the negative half

To me, this clearly points to the HF circuit. The wave shaper only
affects the sine wave and won't change the frequency. The sine wave is
affected because the triangle doesn't go negative enough.
 
R

rickman

Jan 1, 1970
0
Good site, thats.. Put that in my Book marks.

Jamie

Yeah, I thought it was pretty impressive too. BTW, are you British?
That sounds like a British expression to me... "Good site, that". Or
maybe an Aussie?
 
R

rickman

Jan 1, 1970
0
My grand father was British born (Holmes), but I was born in Maine where
every one is related. THe tree does not branch out too far!!!!!!!!!!

Some one told me the other day they don't bother to do crime
investigations with DNA in Maine any more, the results are always
inconclusive.

LOL!
 
J

Jamie

Jan 1, 1970
0
rickman said:
That schematic has missing info. I found a better copy at KO4BB.com and
edited the pages a bit to make it easier to view.

HP_3310AB_OS_1971_edited.pdf
Good site, thats.. Put that in my Book marks.

Jamie
 
J

Jamie

Jan 1, 1970
0
rickman said:
Yeah, I thought it was pretty impressive too. BTW, are you British?
That sounds like a British expression to me... "Good site, that". Or
maybe an Aussie?

My grand father was British born (Holmes), but I was born in Maine
where every one is related. THe tree does not branch out too far!!!!!!!!!!

Some one told me the other day they don't bother to do crime
investigations with DNA in Maine any more, the results are always
inconclusive.

Jamie
 
R

Robert Baer

Jan 1, 1970
0
rickman said:
The HF circuit is the *cause* of the triangle problems. The triangle is
just a constant current into a cap. When it reaches +5 volts the detect
circuit switches the current source to ramp down to -5 volts, then the
cycle repeats. The HF circuit modifies the threshold of the +-5 volt
detect circuit which is part of forming the triangle wave. The problem
is the -5 volt detect is happening too soon and the triangle is not
going far enough negative. The *only* happens when switched to the 10k
or 100k ranges. The only difference in the ranges is that the HF circuit
is kicked in (and the value of the integrating cap, but they are
different between the 10k and 100k ranges).

Just to be clear, the triangle still looks just like a triangle. The
symptoms are:

1) doesn't go negative enough,
2) faster than it should be,
3) frequency is erratic,
4) sine wave is truncated on the negative half

To me, this clearly points to the HF circuit. The wave shaper only
affects the sine wave and won't change the frequency. The sine wave is
affected because the triangle doesn't go negative enough.
Well, then work on the negative threshold sense circuit as well as
its "compensation": threshold first at low frequencies for symmetry and
then the comp.
 
Top