Checking Spurious emissions without a SA

[kb3nqd]

I have been watching an interesting topic over at the Amateur-repairs mailing list that I thought some of the more "seasoned" individuals here might have some interesting answers to.

From  "Greg W:-)" <onegammyleg@yahoo.com.au>
Date  2007/09/05 Wed AM 04:21:26 CDT
To  Amateur-repairs@yahoogroups.com
Subject  [Amateur-repairs] No spectrum analyzer

Hi all.

What did the old tube guys do before the popularity of spectrum
analyzers to check for off frequency or spurious emissions. ?

I am pretty sure there was a meter of some sort that was used..

If so , is there any ways to recreate this device.?

I would buy a spectrum analyzer if I were to be needing it often , but
just to check the occasional qrp rig its hardly worth the expense.

Regards

gregW:-) OH2FFY

What did we use before the advent of the Spectrum Analyzer?

[W1EUJ]

A dummy load and reciever?

Dave Goncalves
W1EUJ

[kb3nqd]

Well you could do that but what if you didn't have one?  I am guessing that back in the day not everyone had enough money to purchase a second receiver and for those that homebrewed they would have not necessarily been too keen to build a second receiver just for such a test.  An interesting suggestion that someone had during the discussion was to use a GDO in the detect mode.  I am not sure that I get that since I only recently picked up an old Heathkit GDO at a Hamfest and haven't had a chance to play with it yet.  IIRC it did not have any kind of "detect mode" switch on the side though...

[Tom WA3KLR]

As you know a neon bulb will glow it's normal pink in the presence of r.f., but I seem to recall something about the bulb's color changing to blueish in the the presence of an r.f. signal with parasitic oscillations.

In line with the grid dip meter in the detector mode, there were these little handheld McMurco Silver meters which would be the same as the grid dip meter detector mode - just a crystal set essentially - LC circuit (air var. cap.), diode and meter movement.  Plug-in coils like the grid-dip meter.  The meter was in a separate meter box with sensitivity pot.  May have been referred to as an absorption meter.

Another device along these lines was the Millen "absorption meter".  No circuitry other than the bare LC circuit.  There was a set of 4 fixed LC circuits with a calibrated hand-held frequency scale.  My old set covers 3 -140 MHz.

As a variation between the 2 mentioned above, there were units with an incandescent  pilot lamp connected in - much cheaper than a meter movement and metal work, but an indicator never the less.

Not an actual radiated power test, but the behavior and meter readings of the amplifiers as they are tuned and operated is an indication that something is not well.  Most of us have experienced this.

A young old buzzard would hold the type with a pilot lamp up to his antenna coupling network and tune through the HF bands to show that his main signal was where it was supposed to be; a good check that the multiplied signal is coming out on the right band.

Photo of a Millen 4 band set below:

[The Slab Bacon]

Manaen,
           Originaly the ham bands were layed out to be harmonically related to each other, hoping that your harmonics would not interfere with other services. If you were radiating harmonics, someone (hopefully not the FCC) would usually let you know agout it.
One would also check it our with his station receiver as well. Maintaining the proper Q of the tuned circuits, and using a PI type or link coupled tuna also acted like a low pass filter reducing the spurious outputs.

All of that is kind of a pain in the a$$ so I bought a spectrum analyser it was just easier.

Just about all of the better GDOs have a monitor function. they simply switch the osc tube to a detector (usually a diode connection) and use the coil and tuning cap as a simple TRF receiver.

The old hand held "absorbtion wavemeters" were also about the same, If you had a light bulb hooked to them, the spurious emissions werent usually enough to light the bulb.

                                            The Slab Bacon

[Tom WA3KLR]

So how were the indicator-less hand-held absorption meters used, like the ones in my photo above?

I think an old buzzard would hold the meter up to an interstage tank circuit in the transmitter while keyed and look at the grid current in the following stage and when the "meter's" resonant frequency matched energy in the tank circuit, some would be absorbed into this additional tuned circuit.  At this point, there is a dip in the grid current of the following stage.

[WU2D]

The wavemeter was what I used for my QRP rigs of the 1980's. It was a lafayette box if I recall. It is litle more than a tuned circuit with a diode detector and a sensitive meter. You wave the antenna around the tank and check for spurs and harmonics.

The lafayette was band switched and had a collapsable whip.

The picture is of an antique Eddystone Wavemeter in which the plug in coil itself is the pickup much like a grid dip meter.

Mike WU2D

[The Slab Bacon]

Sorry,
        I love old gear, but i have ABSOLUTELY no use for old test equipment. That kind of szht just makes you crazy!! That is why I body slammed my Measurements 809 signal generator 3 times, so I wouldnt be able to fix it any more!! the only piece of "vintage" test gear that I still use is my GDO.

There is so much really good test gear now being surplussed out that it just doesnt pay to even think about putting up with some of that old crap. In the olden days you had what you had. But nowadays anyone who can afford a nice shiny ricebox and a big leanyour can afford good test equipment. Believe me, it pays for itself in the long run in saved time and aggrevation.

A really good scope can be had used for peanuts these days. An MFJ antenna analyser is a poor mans network analyser and they are well under $300. a good used Simpson 260 multimeter can be had anywhere around $20-25 at most hamfests. Really nice DMMs that do everything from VOM functions to also measuring L & C as well as frequency can be had in the $100 range (this booger is a whole lab)
nice digital PLL signal generators are now in the $200-250 range. Not to mention what other nice goodies can be found from many of the slurpuss dealers at hamfests.

there is so much really good test gear out there that it just doesnt pay to mess with junk!! Old test gear can be fun to play with for nostalgia reasons, But it would be really foolish to rely on it for accurate results!! Sit that junk on the shelf for nostalgia and get good test gear. You will never be sorry.

                                           the Slab Bacon
 

[WA1GFZ]

Test gear is just another addiction once you have a couple good things.

[John K5PRO]

In the days of old, International Crystal or somebody made a little neon lamp probe that ran off 120 VAC - it was set up to easily light in the presence of weak RF. You would stick this thing inside an amplifier around the output network, not too close to arc over, and if it lit purplish, it was a parasite at VHF, and if it was orange, it was fundamental RF.

In big industrial RF oscillators (for dielectric heating), I used to have a neon on the end of a long plastic stick, that I could poke inside big systems, and see the distribution of RF voltage visually. If it went purple, was supposed to be a parasite happening.

Crude, yes. I have a 7L14 now in a 7603 mainframe as a S/A. Still works great - digital storage.

[KB2WIG]

  "  I love old gear, but i have ABSOLUTELY no use for old test equipment. "

The fats in the fire now.......  I love my RM45a as it will take HV all day; my 453 gets used on the lower voltage stuff' caus I use factory probes.....   I'm still looking for the spectrum analyzer plugins for the 'rm45...............  klc

[WQ9E]

The Tek 1L10 and 1L20 plug-ins still do the job OK.  The 1L10 really makes setup of phasing exciters easy.  I acquired a 7L13 last year and have used it a few times in my 7623A and 7854 mainframes and it is a bit easier to setup than the 1L series but it is rare to find a working 7L series at a reasonable price and with a bit of care the older 1L series are very useful.  I calibrated my first Tektronix scope when I was 12 and many years later I still like the big 500 series mainframes.  I have a Type 555 dual beam with its power supply sitting on a scopemobile near  my Desk KW for monitoring and I leave a Type CA and 1L10 plug-in setup most of the time.  For true BA usage, you can use the Tek 1L series SA plug ins as a receiver by taking audio from the record out plug.  If you want more audio, on most of the mainframes you can run the record out to one of your vertical plug-in inputs and you will have sufficient audio at the scope vertical out connector to drive a small speaker.  You can easily copy AM on 3885 with a fairly small antenna. 

Rodger WQ9E

[The Slab Bacon]

 The fats in the fire now.......  I love my RM45a as it will take HV all day; my 453 gets used on the lower voltage stuff' caus I use factory probes.....   I'm still looking for the spectrum analyzer plugins for the 'rm45...............  klc

Kevin,
I used to have a TEK RM45, that was a really great scope!! It was just huge and heavy!! that is not what what I mean when I say old test gear.  But it is definately a BA

                                                     the Slab Bacon

[KB2WIG]

    "  It was just huge and heavy!!  "

It was huge and not so heavy in my 20's; its a lot heavier now!!   klc

[WU2D]

Here is an example of old vs new.

WU2D Mike

[WA1GFZ]

At work we have a R&S ESIB 40 that goes from SA to receiver at the push of a button........and costs more than my new QTH.

[WD8BIL]

Even the basic HP 8500 series SAs can hit yas for $8K+ Frank.
Our 8591EM is great though.

[kb3nqd]

I am not sure that I agree about the old test gear.  If all old test gear was crap than how did anyone ever fix their rigs back in the day?  If a Vintage piece of test gear is properly restored I am guessing it would still be a valuable addition to someone's test bench.  If the piece has not been thoroughly restored or if it was poorly engineered in the first place than you very well could be in for the heartache that only "a drifty piece of junk" can bring you.

[WA1GFZ]

Yupper Bud.
We have a broken 8566 or is it a 69 that I have offered to take home. I told my boss I would fix it and let them use it anytime. A good HP does the job. I'm still driving a HP141 at home.

[w3jn]

Manaen, I'm firmly with Slab in this one.  I think one reason  we find a lot of messed up boatanchors is that hams really had no decent test equipment back in the day.  Besides I really object to having to fix my test equipment, and then not really be able to trust it as far as calibration is concerned.

Good quality (Tek, HP, WaveTek) test equipment isn't too expensive and it's readily available as opposed to the 60s where hams were still struggling with WWII surplus junk.  A used HP608, which is probably what National and Halilcrafters used in their labs,  was prolly around a kilobuck.  And even those are prone to drift.

You can score a good 100 MHz scope at just about any hamfest for under a c-note.  Synthesize sig gens are around for little more than that.  ANd I scored a Tek 496 S/A on eBay for about $500.

IIRC you have one of those old mil sig gens which is great for aligning IFs and doing sensitivity checks.  A great piece of gear.  But those old Eico and Heathkit sig gens need to be dumpsterized except as collector's curiosities, IMHO.

[kb3nqd]

Actually I am rebuilding that signal generator since it drifts more than I would like it to. 

All I was getting at was that I don't think that old automatically = junk on the test equipment front necessarily.  I like new equipment as much as the next guy but shoot we even used old test equipment to help send folks to the moon.  The KSC Apollo 3B3 is just one example of tube technology used in an Apollo mission. 

I do own a fair amount of brand new gear and relatively new used gear.  I will freely admit to being a bit overly frugal with the hobby but I wouldn't object to spending up to $500 on a piece of test gear like say for instance a swept sig gen (I am still trying to get my Hammarlund aligned using an ordinary sig gen) which would speed up the alignment process for certain receivers.

[The Slab Bacon]

Mad dog,
               FWIW that sig gen that you had over here a few weeks ago is always gona drift some. It is the nature of that beast. I had just had enough of stuff like that. there are many nice PLL type sig gens out there at very reasonable prices.

Look for Motorola / Wavetek R-1010s amd R-1100s they are getting pretty reasonable and work very well once you repair them. they are PLL type and very accurate. Accurate enough to set frequencies in UHF gear. they can be found reasonably because they are known for crapping out. Replace all of the first generation tantalun electrolytics in them and they go forever. I am also seeing earlier Cushman CE-3, and CE-5 service monitors for very reasonable prices at local hamfests, many under $200!!

Believe me, you are no more frugal than I am, but sometimes you have to consider your time as money!!

                                                                           the Slab Bacon

[Ian VK3KRI]

While good test gear is great, I belive there is something to be said for experience in using simple test gear to repair and align eqipment.  The more basic the gear , the more you you need to find out und understand what is going on.

Thats not to say I think that we should be using wavemeters instead of spec-ans , but at the very least sweeping a wave meter to find parasitics gives you a much greater appreciation of what a great tool the spec-an is.
And you won't be stuck when your not surrounded by all the neat toys, and you just have a a multi-meter and your wits to work with...
                                                              Ian VK3KRI 
 

[Ed/KB1HYS]

kinda like
"endeavoring to construct a mnemonic memory circuit
using stone knives and bearskins."

Kidding aside, you can learn a LOT from the OT equipment.

Build a crystal radio, build a regen (one tube blooper) from there you can build a simple crystal controled transmitter, then a VFO, etc... Build some test equipment too.   
Heck a GDO is an oscillator with a meter in the circuit.  There is (was) a fellow marketing a kit built Spectrum analyzer. uses your oscilliscope and build the rest. IIRC it was less than $200. You will learn loads and it's fun to build your own stuff. (especially when it works)

That said, I wouldn't want to try and setup a Class E rig or complex reciever with out some modern acurate test gear.

[k4kyv]

If you are  looking for spurious signals, a good general coverage receiver with a functioning S-meter works almost as well as a spectrum analyser.  You don't even need a 2nd receiver if you have a good one as your main station receiver.  Just disconnect the antenna from the receiver, and maybe short out the antenna terminals, to reduce the signal level from your transmitter to about S-9 or a little more.  Then while transmitting, tune across the band and compare the strengths of any spurious signals you hear.  One thing to watch out for is images, but otherwise the receiver should work fine.

A SA is nothing but a receiver with a scope display attached at the output, whose frequency scans a segment of spectrum at the same rate as the horizontal trace on the scope.  The SA is not immune from image responses, either.

Spurious signals strong enough to be clearly audible over the air may still be too weak to give an indicatin on a wavemeter or GDO.