#### FG 504 function generator odd behaviour

Colin Herbert

I have an FG 504 which is showing some odd behaviour. The triangle-wave shows asymmetry in that the negative peaks of the wave are rounded-off at any frequency and the positive ones have a similar but lesser problem at higher frequencies. This also seems to manifest itself as what looks like cut-off on the negative peaks of the sine-wave and lesser cut-off on positive peaks. While the triangle-waves are more-or-less symmetrical, the sine-waves are definitely not so, showing a nearly three-to-one voltage ratio in favour of positive peaks. There is also noticeably overshoot on the square-waves at anything above 1kHz and only a DC offset at frequencies below 80Hz.
I am totally ignorant of how this function generator produces triangle-waves, but I do understand that the square-waves and sine-waves are derived from the triangle-wave. Does anyone have any help that they can offer? I would be grateful to get this useful function-generator working properly.
Colin.

Dave Daniel

The first thing to do is measure the power rails to make sure that the voltages are in spec and clean.

DaveD

On Jul 10, 2020, at 13:04, Colin Herbert via groups.io <colingherbert=blueyonder.co.uk@groups.io> wrote:

I have an FG 504 which is showing some odd behaviour. The triangle-wave shows asymmetry in that the negative peaks of the wave are rounded-off at any frequency and the positive ones have a similar but lesser problem at higher frequencies. This also seems to manifest itself as what looks like cut-off on the negative peaks of the sine-wave and lesser cut-off on positive peaks. While the triangle-waves are more-or-less symmetrical, the sine-waves are definitely not so, showing a nearly three-to-one voltage ratio in favour of positive peaks. There is also noticeably overshoot on the square-waves at anything above 1kHz and only a DC offset at frequencies below 80Hz.
I am totally ignorant of how this function generator produces triangle-waves, but I do understand that the square-waves and sine-waves are derived from the triangle-wave. Does anyone have any help that they can offer? I would be grateful to get this useful function-generator working properly.
Colin.

Colin Herbert

Yes, Dave. I did indeed check and slightly adjust the +15 V and -15 V rails
some time ago. I also checked the +25 , -25 and +5 V supplies and had a go
at checking ripple using the differential method in the Service Manual, but
I didn't really comprehend the ripple results at the time. I should say that
this didn't cause the odd behaviour, it was already there and I was
attempting to solve it. I suppose I should have another go at looking at the
supplies, power rails and ripple again...
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Dave
Daniel
Sent: 10 July 2020 18:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] FG 504 function generator odd behaviour

The first thing to do is measure the power rails to make sure that the
voltages are in spec and clean.

DaveD

On Jul 10, 2020, at 13:04, Colin Herbert via groups.io
<colingherbert=blueyonder.co.uk@groups.io> wrote:

I have an FG 504 which is showing some odd behaviour. The triangle-wave
shows asymmetry in that the negative peaks of the wave are rounded-off at
any frequency and the positive ones have a similar but lesser problem at
higher frequencies. This also seems to manifest itself as what looks like
cut-off on the negative peaks of the sine-wave and lesser cut-off on
positive peaks. While the triangle-waves are more-or-less symmetrical, the
sine-waves are definitely not so, showing a nearly three-to-one voltage
ratio in favour of positive peaks. There is also noticeably overshoot on the
square-waves at anything above 1kHz and only a DC offset at frequencies
below 80Hz.
I am totally ignorant of how this function generator produces
triangle-waves, but I do understand that the square-waves and sine-waves are
derived from the triangle-wave. Does anyone have any help that they can
offer? I would be grateful to get this useful function-generator working
properly.
Colin.

Chuck Harris

that fed into a polarity switcher that fed into an integrator.

The output of the integrator was a linear ramp that changed
direction whenever the polarity switcher changed polarity.

To make the function generator oscillate, the output of the
integrator was fed into a Schmidt trigger that in turn fed the
polarity switcher, and made it switch whenever the ramp got
above or below the Schmidt trigger's threshold voltages.

Draw yourself a picture using blocks.

The output of the integrator is a triangle wave, and the output
of the Schmidt trigger is a square wave.

The sine wave was made by feeding the triangle wave into a soft
clipper that clipped off the triangle wave's pointy peaks giving
an approximation to a sine wave. A good bit of art went into
that soft clipper.

The manual will describe the FG504's specific implementation, but
I would concentrate on the soft clipper stage. It typically
has a lot of trimmers, and diodes.

-Chuck Harris

Colin Herbert via groups.io wrote:

I have an FG 504 which is showing some odd behaviour. The triangle-wave shows asymmetry in that the negative peaks of the wave are rounded-off at any frequency and the positive ones have a similar but lesser problem at higher frequencies. This also seems to manifest itself as what looks like cut-off on the negative peaks of the sine-wave and lesser cut-off on positive peaks. While the triangle-waves are more-or-less symmetrical, the sine-waves are definitely not so, showing a nearly three-to-one voltage ratio in favour of positive peaks. There is also noticeably overshoot on the square-waves at anything above 1kHz and only a DC offset at frequencies below 80Hz.
I am totally ignorant of how this function generator produces triangle-waves, but I do understand that the square-waves and sine-waves are derived from the triangle-wave. Does anyone have any help that they can offer? I would be grateful to get this useful function-generator working properly.
Colin.

Roger Evans

Colin,

If you have a problem with the triangle wave symmetry then I would begin with that before looking elsewhere. The operation is well described in the manual but I need the big screen to look at that, not the phone! If memory serves me right there are two constant current sources driving the integrating capacitor and limit comparators to define the end points. If the waveform is distorted at the integrating capacitor then you need to look at the current sources, any associated decoupling circuitry and the FET that buffers the integrator. If the triangle wave looks OK at its source then you can follow it through the switching and amplifying stages. The output stage is quite beefy to drive 50 ohm loads and I had to replace some components in mine to get it up to spec.

The push buttons on the 500 series are unreliable with age, there is a thread here about cleaning so try exercising the waveform selector buttons several times.

Regards,

Roger

Colin Herbert

Oops! I have checked the voltages and while the +25V (+24.99), +15V (+14.98)
and -15V (-14.98) are just about bang on, the -25V is sitting at -3.58V. I
really can't remember if this was the case when I checked previously, but it
is now. I did manage to short something out with the DMM probes, so that may
have given this problem, but sure as sure it needs to be fixed, but how?
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Dave
Daniel
Sent: 10 July 2020 18:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] FG 504 function generator odd behaviour

The first thing to do is measure the power rails to make sure that the
voltages are in spec and clean.

DaveD

On Jul 10, 2020, at 13:04, Colin Herbert via groups.io
<colingherbert=blueyonder.co.uk@groups.io> wrote:

I have an FG 504 which is showing some odd behaviour. The triangle-wave
shows asymmetry in that the negative peaks of the wave are rounded-off at
any frequency and the positive ones have a similar but lesser problem at
higher frequencies. This also seems to manifest itself as what looks like
cut-off on the negative peaks of the sine-wave and lesser cut-off on
positive peaks. While the triangle-waves are more-or-less symmetrical, the
sine-waves are definitely not so, showing a nearly three-to-one voltage
ratio in favour of positive peaks. There is also noticeably overshoot on the
square-waves at anything above 1kHz and only a DC offset at frequencies
below 80Hz.
I am totally ignorant of how this function generator produces
triangle-waves, but I do understand that the square-waves and sine-waves are
derived from the triangle-wave. Does anyone have any help that they can
offer? I would be grateful to get this useful function-generator working
properly.
Colin.

Harvey White

Generic steps:

turn scope off, wait 5 minutes or so, turn scope on.  See if -25 is good (something could have latched up).

with power off, check continuity of fuses or with power on, check for very low voltage drop across the fuse.  There must always be a voltage drop because the fuse acts like a resistor.

Look at the bulk supply for the -25.  Check against voltage readings on the schematic.  Go through regulator circuit to output and see what doesn't match.  Any transistor where the base-emitter voltage is the wrong polarity indicates transistor is off.  Any transistor (except a darlington) where the B-E voltage exceeds 0.7 volts is likely bad.

Typical things to fail:  pass transistor.

in the event of a short circuit, check the current sense resistor (the low value one from the pass transistor to the output) and then check the (frequently done like this) transistor which has the base/emitter junction connected across the resistor.  Too much current and the transistor turns on, stealing base drive from the pass transistor, and turning it off.  Current limiting.

In many tek supplies, the - supply is a positive supply in topology, with what would be the ground being the negative lead, and what you'd think ought to be the positive output grounded. The voltage across the pass transistor is the difference between the bulk supply and the desired voltage.

Check also to see what the reference voltage is for the -25 volt supply.  It must be exact, otherwise the supply will be off.

Harvey

On 7/11/2020 1:56 PM, Colin Herbert via groups.io wrote:
Oops! I have checked the voltages and while the +25V (+24.99), +15V (+14.98)
and -15V (-14.98) are just about bang on, the -25V is sitting at -3.58V. I
really can't remember if this was the case when I checked previously, but it
is now. I did manage to short something out with the DMM probes, so that may
have given this problem, but sure as sure it needs to be fixed, but how?
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Dave
Daniel
Sent: 10 July 2020 18:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] FG 504 function generator odd behaviour

The first thing to do is measure the power rails to make sure that the
voltages are in spec and clean.

DaveD

On Jul 10, 2020, at 13:04, Colin Herbert via groups.io
<colingherbert=blueyonder.co.uk@groups.io> wrote:
I have an FG 504 which is showing some odd behaviour. The triangle-wave
shows asymmetry in that the negative peaks of the wave are rounded-off at
any frequency and the positive ones have a similar but lesser problem at
higher frequencies. This also seems to manifest itself as what looks like
cut-off on the negative peaks of the sine-wave and lesser cut-off on
positive peaks. While the triangle-waves are more-or-less symmetrical, the
sine-waves are definitely not so, showing a nearly three-to-one voltage
ratio in favour of positive peaks. There is also noticeably overshoot on the
square-waves at anything above 1kHz and only a DC offset at frequencies
below 80Hz.
I am totally ignorant of how this function generator produces
triangle-waves, but I do understand that the square-waves and sine-waves are
derived from the triangle-wave. Does anyone have any help that they can
offer? I would be grateful to get this useful function-generator working
properly.
Colin.

Andy Warner

On the FG504, the -25V rail is referenced off the +25V one.
Check for bad tantalums on the -25V rail and also independently check the
pass transistor used by that rail, it is in the chassis, not the module.

I have to work on the sweep section of my FG504, which looks like it will
be a right pain. Anyone got any tips for working on the middle board in
this unit ?

On Sat, Jul 11, 2020 at 13:09 Harvey White <madyn@...> wrote:

Generic steps:

turn scope off, wait 5 minutes or so, turn scope on. See if -25 is good
(something could have latched up).

with power off, check continuity of fuses or with power on, check for
very low voltage drop across the fuse. There must always be a voltage
drop because the fuse acts like a resistor.

Look at the bulk supply for the -25. Check against voltage readings on
the schematic. Go through regulator circuit to output and see what
doesn't match. Any transistor where the base-emitter voltage is the
wrong polarity indicates transistor is off. Any transistor (except a
darlington) where the B-E voltage exceeds 0.7 volts is likely bad.

Typical things to fail: pass transistor.

in the event of a short circuit, check the current sense resistor (the
low value one from the pass transistor to the output) and then check the
(frequently done like this) transistor which has the base/emitter
junction connected across the resistor. Too much current and the
transistor turns on, stealing base drive from the pass transistor, and
turning it off. Current limiting.

In many tek supplies, the - supply is a positive supply in topology,
with what would be the ground being the negative lead, and what you'd
think ought to be the positive output grounded. The voltage across the
pass transistor is the difference between the bulk supply and the
desired voltage.

Check also to see what the reference voltage is for the -25 volt
supply. It must be exact, otherwise the supply will be off.

Harvey

On 7/11/2020 1:56 PM, Colin Herbert via groups.io wrote:
Oops! I have checked the voltages and while the +25V (+24.99), +15V
(+14.98)
and -15V (-14.98) are just about bang on, the -25V is sitting at -3.58V.
I
really can't remember if this was the case when I checked previously,
but it
is now. I did manage to short something out with the DMM probes, so that
may
have given this problem, but sure as sure it needs to be fixed, but how?
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Dave
Daniel
Sent: 10 July 2020 18:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] FG 504 function generator odd behaviour

The first thing to do is measure the power rails to make sure that the
voltages are in spec and clean.

DaveD

On Jul 10, 2020, at 13:04, Colin Herbert via groups.io
<colingherbert=blueyonder.co.uk@groups.io> wrote:
I have an FG 504 which is showing some odd behaviour. The triangle-wave
shows asymmetry in that the negative peaks of the wave are rounded-off at
any frequency and the positive ones have a similar but lesser problem at
higher frequencies. This also seems to manifest itself as what looks like
cut-off on the negative peaks of the sine-wave and lesser cut-off on
positive peaks. While the triangle-waves are more-or-less symmetrical,
the
sine-waves are definitely not so, showing a nearly three-to-one voltage
ratio in favour of positive peaks. There is also noticeably overshoot on
the
square-waves at anything above 1kHz and only a DC offset at frequencies
below 80Hz.
I am totally ignorant of how this function generator produces
triangle-waves, but I do understand that the square-waves and sine-waves
are
derived from the triangle-wave. Does anyone have any help that they can
offer? I would be grateful to get this useful function-generator working
properly.
Colin.

--
Andy

Dave Daniel

Ir you end up having to replace any of the TO-3 pass transistors in the chassis, be aware that many (all?) new ones may have a different pinout than the originals.

Just something to be aware of.

DaveD

On Jul 11, 2020, at 17:22, Andy Warner <andyw@...> wrote:

On the FG504, the -25V rail is referenced off the +25V one.
Check for bad tantalums on the -25V rail and also independently check the
pass transistor used by that rail, it is in the chassis, not the module.

I have to work on the sweep section of my FG504, which looks like it will
be a right pain. Anyone got any tips for working on the middle board in
this unit ?

On Sat, Jul 11, 2020 at 13:09 Harvey White <madyn@...> wrote:

Generic steps:

turn scope off, wait 5 minutes or so, turn scope on. See if -25 is good
(something could have latched up).

with power off, check continuity of fuses or with power on, check for
very low voltage drop across the fuse. There must always be a voltage
drop because the fuse acts like a resistor.

Look at the bulk supply for the -25. Check against voltage readings on
the schematic. Go through regulator circuit to output and see what
doesn't match. Any transistor where the base-emitter voltage is the
wrong polarity indicates transistor is off. Any transistor (except a
darlington) where the B-E voltage exceeds 0.7 volts is likely bad.

Typical things to fail: pass transistor.

in the event of a short circuit, check the current sense resistor (the
low value one from the pass transistor to the output) and then check the
(frequently done like this) transistor which has the base/emitter
junction connected across the resistor. Too much current and the
transistor turns on, stealing base drive from the pass transistor, and
turning it off. Current limiting.

In many tek supplies, the - supply is a positive supply in topology,
with what would be the ground being the negative lead, and what you'd
think ought to be the positive output grounded. The voltage across the
pass transistor is the difference between the bulk supply and the
desired voltage.

Check also to see what the reference voltage is for the -25 volt
supply. It must be exact, otherwise the supply will be off.

Harvey

On 7/11/2020 1:56 PM, Colin Herbert via groups.io wrote:
Oops! I have checked the voltages and while the +25V (+24.99), +15V
(+14.98)
and -15V (-14.98) are just about bang on, the -25V is sitting at -3.58V.
I
really can't remember if this was the case when I checked previously,
but it
is now. I did manage to short something out with the DMM probes, so that
may
have given this problem, but sure as sure it needs to be fixed, but how?
Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Dave
Daniel
Sent: 10 July 2020 18:13
To: TekScopes@groups.io
Subject: Re: [TekScopes] FG 504 function generator odd behaviour

The first thing to do is measure the power rails to make sure that the
voltages are in spec and clean.

DaveD

On Jul 10, 2020, at 13:04, Colin Herbert via groups.io
<colingherbert=blueyonder.co.uk@groups.io> wrote:
I have an FG 504 which is showing some odd behaviour. The triangle-wave
shows asymmetry in that the negative peaks of the wave are rounded-off at
any frequency and the positive ones have a similar but lesser problem at
higher frequencies. This also seems to manifest itself as what looks like
cut-off on the negative peaks of the sine-wave and lesser cut-off on
positive peaks. While the triangle-waves are more-or-less symmetrical,
the
sine-waves are definitely not so, showing a nearly three-to-one voltage
ratio in favour of positive peaks. There is also noticeably overshoot on
the
square-waves at anything above 1kHz and only a DC offset at frequencies
below 80Hz.
I am totally ignorant of how this function generator produces
triangle-waves, but I do understand that the square-waves and sine-waves
are
derived from the triangle-wave. Does anyone have any help that they can
offer? I would be grateful to get this useful function-generator working
properly.
Colin.

--
Andy