What is a Sweep Generator?

[DMcMahon]

What is a Sweep Generator and how does it differ from a Signal/Function generator ?

Is it just a frequency sweep and is it manual or automated. If manual how does it differ from manually sweeping a signal generator output ?

David

[kellys_eye]

As the name implies, it can sweep a range of frequencies either manually (not often used) or automatically from a triggering source.

The sweep function can either be built in to the generator or, if the generator has the appropriate control input, applied from a separate external sweep generator. The driving waveform is often a simple ramp signal with variable amplitude and duration.

A frequency generator is controlled by a voltage and the applied voltage tunes the output. The tuning voltage can be linear or logarithmic to give the appropriate display on an oscilloscope which has it's x-deflection triggered at the same time the sweep voltage starts.

Most commonly IMHO they are used to plot the frequency characteristics of filters by, for example, sweeping between 400kHz and 500kHz to plot the bandpass of an IF (at 450kHz), producing the familiar curve (hump) that can also be adjusted in real time for perfect symmetry.

Indeed you can purchase (or build) dedicated sweep generators for the common IF frequencies to assist in receiver alignment.

Plotting filter characteristics (low pass, high pass, bandpass/stop etc) is also done for RF front ends and/or specific types of filters used in RF work.

[6AL5W-Martin]

It differs from the manual sweeped signal generator by the fact that it must provide a very constant signal level over the complete range, this is required to use it for filter curves.

[trh01uk]

A sweep generator has a variable frequency output, fully automatic, and is just a frequency sweep between two frequencies, which are sometimes referred to as "start" and "stop". The rate of the sweep between the two frequencies is usually adjustable.

[paulsherwin]

In Britain we tend to call them wobbulators (originally a trade name). Sweep generator is generally used in the US.

[peter_scott]

I generally think of them for displaying receiver bandwidth plots on a scope whose X input is fed from a sawtooth sweep signal provided by the sweep/wobulator in addition to its RF output.

Peter

[DMcMahon]

Thank you everybody, a lot clearer now, yes remember the name wobbulator.

Quickly looking at a few radio (domestic) service manuals I see that sweep generators are sometimes called up as required test equipment, e.g. for FM IF alignment but not always, i.e. RF signal generator only sometimes called up as a signal source.

David

[Electronpusher0]

If you fancy making one the Haigh design is very good. I have made this myself.

https://www.vintage-radio.net/forum/...d.php?t=172685

https://www.vintage-radio.net/forum/...d.php?t=186073

Peter

[PaulR]

I looked for one for ages. I bought a Cossor 1324 but never got it to work very well. Eventually I found a Philips PM5324 reasonably priced on ebay. It works really well and is great fun to use (if that is your idea of fun!). It has also found a use for my ancient oscilloscope. https://www.vintage-radio.net/forum/...d.php?t=195853

[Bazz4CQJ]

Is a sweep generator a wobbulator; I don' think it is.

A sweep generator is surely an item in itself, and could be used for a number of purposes, but a common one would be as connected to a wobbulator, or as a part of a wobbulator.

In the case of a wobbulator, the sweep generator provides a sweep voltage (increases over a period of time before resetting to the starting voltage) usually to drive a varicap diode to make it to sweep the (RF) frequency of the wobbulator.

Isn't it the case that a sweep generator has been referred to as a timebase generator?

So, I am at this very, moment building the Armstrong sweep generator and that will be a connected to a separate wobbulator (Haigh), and that to an oscilloscope to make measurement, and all three are needed to make a measurement.

It's just occurred to me that some time a go I bough a nice little function generator off ebay to do some tests on solenoid valves, using its square wave function. If that has a got saw tooth function I could probably use it with the wobbulator!

Where on Earth did the term wobbulator come from?

B

[DMcMahon]

According to Wikipedia -The term "wobbulator" is a portmanteau of wobble and oscillator.

David

[ms660]

Not forgetting the ones that obtained the sweep by electro mechanical means, which might have been where "wobble" originated.

Lawrence.

[Bazz4CQJ]

Originally Posted by ms660

Not forgetting the ones that obtained the sweep by electro mechanical means, which might have been where "wobble" originated.

Lawrence.

Yes, didn't some have electric motors with either cams or magnets creating the desired effect. Maybe a bit like badly loaded washing machine on spin, the unit would steadily "walk" along the lab bench?

B

[ms660]

Yes they did, for VHF/UHF work electro-mechanical was the preferred method as in here (1st column last para.):

https://worldradiohistory.com/hd2/ID...20capacitor%22

Full article here, Electronics, November 1958, mag. page 88:

https://worldradiohistory.com/Archiv...1958-11-07.pdf

Lawrence.

[Phil__G]

Crucially it has a voltage output which ramps proportionally to the frequency, which can be fed to your scope X in place of the timebase - so a filter response can be seen on the Y (level) as X (freq) increases, which gives a textbook filter diagram on your scope screen

[Radio Wrangler]

Let's start at the beginning...

They're all signal generators. They all generate signals. The term is so wide it doesn't really help much.

Repairman's sig gen:

Around here, if you say 'Signal generator', most people will assume you mean something basic, intended primarily for aligning receivers/ There are plenty of old models by Advance, Taylor, Avo etc in circulation. Inside there is an LC oscillator and usually an audio oscillator to drive an amplitude modulator. Frequency accuracy and stability can be reasonable. Phase noise is usually not bad. Attenuator range is usually limited and amplitude accuracy is awful. FM is rare. Coverage over 30MHz is unusual, some claim higher frequencies than the actually go to, by including their harmonics. Screening can be minimal and signal leaks out of the case.

One of these is perfectly fine for aligning broadcast receivers. What it won't do is measure sensitivity.

Lab sig gen:

Getting more serious, you'll see an attenuator capable of taking the signal down into the microvolt region, and a well enough shielded case to make that attenuation real. You might get FM with this level of thing. Frequency coverage could be anywhere depending on model. The audio oscillator for internal modulation might be tuneable across the audio range, or you can use an external audio source. Frequency precision gets better. Old models may have a crystal marker system, plug in headphones and hear tweets every 100kHz up the band sort of thing. Newer models may have a frequency synthesiser setting the frequency. Some synth ones can do frequency stepping, giving a slow, stepped, but accurate sweep.

Phase noise is usually reasonable (less so for synths). Models from 9kHz to many GHz.

These things are good for measuring almost all aspects of receiver performance.

Some have a sweeper built-in, EG Marconi TF2002

Sweeper

Subset of signal generator, but a wide range in itself. THis is mostly used to mean RF signal sources used to make plots of behaviour versus frequency. Maybe gain, maybe phase, maybe the transfer function of a discriminator. THey might have rudimentary AM, many don't have much attenuation available. Quite a lot are for microwave bands for testing filters, waveguide couplers, antennae etc. Some work down at more mundane frequencies. EG the HP8901A does 100kHz to 110MHz and has AM and FM and has an attenuator which goes all the way down. It can do wide and narrow sweeps as well. If you see one, grab it!

Wobbulator

This is an old term, and unlikely to be understood by most in the industry. Ask a modern equipment supplier and there is a fair chance he'll suggest you look in your local "Adult shop"

They are a subset of sweepers, usually offering only narrow sweeps, just enough to scan an IF bandwidth of a receiver (ones for AM radio, FM radio and TV common IFs) Good for aligning stagger-tuned filters and frequency discriminators. Usually not much attenuation capability.

Function generator

File this under 'Jack of all trades and master of none'

THe heart of one of these is an opamp type integrator. It drives a comparator with hysteresis and the comparator switches the direction an input voltage drives the integrator. So the integrator bounces back and forth between the switching points of the hysteresis, making a triangular waveform. The input voltage sets the speed of the sides of the triangle, and hence the frequency. So it's a voltage to frequency converter. This allows external voltage control, sweeping etc.

The voltage to frequency relationship is linear and usually fairly accurate. For sweeping audio systems, some have an internal anti-log voltage converter to give you a logarithmic sweep which is expected for audio plots.

You aren't tied to triangles. The comparator switching gives you a squarewave. Usually people want sines, so a complex diode network is used to distort the triangle to form a sine. THe more complex, the better the approximation. Many have a second func gen used to modulate or sweep the main one.

These things are handy, but as an R-C oscillator the 'resonator' has Q below unity, so phase noise isn't the best. The shaped sine form better units isn't, bed, but cheapies aren't very good. So one of these is not a low distortion oscillator for audio distortion testing. Don't expect attenuation down to radio input levels. These are general lab bench instruments, not really targetting radio activities.

Wien Bridge oscillators

THis is a simple circuit, which needs some care over amplitude stabilisation. Bill Hewlett invented and patented the thermistor method, but Jfets etc can be used. One of these can be made to have very low distortion. THey can be cheap and still effective for audio. Don't expect to sweep, change frequency and the amplitude bounces. This is the tool for measuring amplifier distortion.

Beat Frequency Oscillator

Two RF oscillators are mixed together to make an audio frequency product which can be tuned over many octaves. Sweeping one RF oscillator a little will sweep the output across the full audio band, with no Wien Bridge wild amplitude bounces.

Once upon a time Bruel & Kjaer equipment was the standard for audio testing. It used a beat frequency oscillator linked by a 'speedometer cable' to a strip chart plotter. Logarithmic frequency scale. They did and still do gorgeous microphones for speaker testing, headphone testing etc. Look in seventies hifi reviews and you'll see their strip plots reproduced everywhere.

Oddities

The above is full of sweeping generalisations (sorry!) but should be a decent guide to the big picture (and I got my favourite wobbulator joke in)

The HP8903 is a low distortion oscillator and distortion/millivolt/SINAD meter. It uses a state-variable filter under microprocessor control to be run into instability, but kept oscillating at a controlled level. This can be swept and can do swept amplitude and distortion measurements across and beyond the audio frequency range. It's old hat now. Later instruments amount to super duper soundcards and DSP software.

David

[G6Tanuki]

Yes, a sweep generator is essentially a voltage-controlled-oscillator that varies its frequency between defined limits and is handy for aligning IF tuned-circuits or transmitter-driver stages.

Used in conjunction with an oscilloscope it can give you a great insight into the 'shape factor' of filters; the big problem being that to properly visualise, understand and optimize the shape-factor of a filter you need to put the output into a *logarithmic* detector, so you can see what is happening 60dB down a few KHz away from the filter's peak.

Good logarithmic detectors with a 60dB dynamic range are a rather fun area to investigate. Expect to pay a lot for a good one.

[ms660]

Here's a photo of a sweep motor and capacitor from one, other methods were used including capacitor plates attached to a moving coil unit that vibrated at mains frequency:

https://richardsears.files.wordpress...-capacitor.jpg

Lawrence.

[Bazz4CQJ]

This all reminds me the project undertaken by my late, great friend G8DSW. Back around 1970, he hastily built a 8MHz valve VFO on a biscuit tin... and coupled it to his 2m rig. The x18 enhancement of the drift meant that results were 'not good', and we all joked that instead of "Calling CQ, and listening 146 down" he need to call "Calling CQ, and drifting 146 down". He only used it very briefly but provided much amusement. That too was "some kind" of sweep generator.

Nice picture Lawrence!

B

[Vintage Engr]

Picking up on the reference to mechanical wobbulators, reminds me of my early days in domestic TV. I worked for Rediffusion, who had a specially designed Wobbulator manufactured by Samwell & Hutton, of Little Ilford lane Ilford Essex. It was constructed mostly out of ex-WD parts.

This was the model 41B/R, which covered not only the normal RF/IF frequencies, but also Rediffusions S.F. frequencies for their wired systems.
The 41B/R incorporated a mechanically driven variable capacitor as the sweep element. On switch-on, one had a 50/50 chance that the motor, and thus the sweep would run in the right direction! The motor switch was therefore separate from the main on/off switch. It was a very satisfying wob to use, with built-in xtal markers, and its own built-in CRT display. It was a two man lift! I wouldn't mind one now, but it would probably cause a
disharmony at home..

Grundig Produced their own swept frequency oscillator, known as a 'Wobblesender". This was entirely electronic. There was also the Polyskop. I still own an early Cossor Band I/III sweep generator, which uses the TB sawtooth out to drive it.

Of course there are sweep-frequency oscillators for many other purposes other than what we might consider normal I.F. alignment.

In the broadcast industry, I used Rhode & Shwartz & also Leader video SFO's for alignment of various video equipment such as equaliser, VTR's etc.

Lastly, there are also SFO's used for audio engineering, for instance my Lindos unit can do automated audio-frequency sweeps for checking response & bandwidth on audio equipment .

A fun note to end on.

Our local antique market has had a Jason Kit (Remember them, - F.M.tuners?) Wobbulator for sale hidden amongst old radios & telephones. It sported a notice "Buy yourself a Wobbulator, every home needs one". There was actually a fairly accurate explanatory notice with the unit, informing the would-be- purchaser of its' function. Sitting next to it was a Jason FMT 1.
Both items were there for over a year, but I think by now they must have gone.

David.