need advice aligning a vaccum tube spectrum analyzer
 

Hello, everyone.

It looks like I have a new, exciting, rather daunting project ahead of me.
I'd like some advice
from anyone interested in providing it.

I have now adopted into my household (and home lab) a Lavoie LA-18M spectrum
analyzer unit,
circa 1950's, completely tubes plus a few microwave crystal diodes. I
also have a manual.

The manual is missing the overall schematic diagram of the entire unit but
has schematic diagrams
of all the major sections, such as the IF strips, and it also has general
troubleshooting procedure.

Given its age, it is in outstanding condition, because everything about it
works. Given the fact that
I bought it for only 50 bucks, as a working antique it is an absolute
treasure. However, I am concerned
about the weak signal sensitivity.

For calibration equipment, I have a late year 100 Mhz analog scope, a
Wavetek 3001 signal generator,
and a digital multimeter, if that helps.

With the signal generator I was able to determine that the minimal
discernable signal at 10 Mhz was
-34dBm. The manual on the Lavoie indicates that the sensitivity at that
frequency should be 40 microvolt,
so that seems to tell me that age has affected the sensitivity.

There are several IF strips on the unit, and the manual tells me which strip
needs to be aligned first, but
it does not go into any detail about how to align each of the IF strips for
best performance... it merely
says "replace IF strips as necessary".

If I had a time machine I could probably follow the manual's suggestions,
but short of that I am out of
luck there. I also understand that the alignment procedure for AM IF
strips is differerent from FM IF
strips, and for IF strips used by spectrum analyzers it could be different
still.

So, does anyone have any advice for me on how I should peak each vaccum tube
IF strip. I will supply
specific information on each, and will email scans of the schematics for
each, to those who want to
consider the problem.

Thanks,

The Eternal Squire

I don't have any knowledge of your particular unit, but in general the
shape of the IF filters is very important to spectrum analyzer
operation. Unless you use a good sweep generator and calibrated
detector, or better yet a network analyzer, along with good information
about what the proper IF response should be, you're almost certain to
ruin the response by tweaking. You also won't recover anything like 40
dB, which is what it's lacking, by fiddling with the alignment. If I
were you, I'd first make sure that I measured the MDS properly. If the
unit has a selectable bandwidth (or dispersion, as I recall it was
called), be sure it's at the correct setting for the specified MDS,
since the bandwidth will directly impact the MDS. (This won't, however,
account for any but a small part of the total shortage.) With about 40
dB deficiency in MDS, you should be looking for something that's broken,
not just out of alignment. I'd suspect the diodes first, since they're
fragile and vulnerable. Next I'd check the tubes. You should be able to
trace a moderate amplitude signal through the analyzer with your scope,
and find replacement parts on eBay.

Roy Lewallen, W7EL

The Eternal Squire wrote:
Hello, everyone.

It looks like I have a new, exciting, rather daunting project ahead of me.
I'd like some advice
from anyone interested in providing it.

I have now adopted into my household (and home lab) a Lavoie LA-18M spectrum
analyzer unit,
circa 1950's, completely tubes plus a few microwave crystal diodes. I
also have a manual.

The manual is missing the overall schematic diagram of the entire unit but
has schematic diagrams
of all the major sections, such as the IF strips, and it also has general
troubleshooting procedure.

Given its age, it is in outstanding condition, because everything about it
works. Given the fact that
I bought it for only 50 bucks, as a working antique it is an absolute
treasure. However, I am concerned
about the weak signal sensitivity.

For calibration equipment, I have a late year 100 Mhz analog scope, a
Wavetek 3001 signal generator,
and a digital multimeter, if that helps.

With the signal generator I was able to determine that the minimal
discernable signal at 10 Mhz was
-34dBm. The manual on the Lavoie indicates that the sensitivity at that
frequency should be 40 microvolt,
so that seems to tell me that age has affected the sensitivity.

There are several IF strips on the unit, and the manual tells me which strip
needs to be aligned first, but
it does not go into any detail about how to align each of the IF strips for
best performance... it merely
says "replace IF strips as necessary".

If I had a time machine I could probably follow the manual's suggestions,
but short of that I am out of
luck there. I also understand that the alignment procedure for AM IF
strips is differerent from FM IF
strips, and for IF strips used by spectrum analyzers it could be different
still.

So, does anyone have any advice for me on how I should peak each vaccum tube
IF strip. I will supply
specific information on each, and will email scans of the schematics for
each, to those who want to
consider the problem.

Thanks,

The Eternal Squire

I don't have any knowledge of your particular unit, but in general the
shape of the IF filters is very important to spectrum analyzer
operation. Unless you use a good sweep generator and calibrated
detector, or better yet a network analyzer, along with good information
about what the proper IF response should be, you're almost certain to
ruin the response by tweaking. You also won't recover anything like 40
dB, which is what it's lacking, by fiddling with the alignment. If I
were you, I'd first make sure that I measured the MDS properly. If the
unit has a selectable bandwidth (or dispersion, as I recall it was
called), be sure it's at the correct setting for the specified MDS,
since the bandwidth will directly impact the MDS. (This won't, however,
account for any but a small part of the total shortage.) With about 40
dB deficiency in MDS, you should be looking for something that's broken,
not just out of alignment. I'd suspect the diodes first, since they're
fragile and vulnerable. Next I'd check the tubes. You should be able to
trace a moderate amplitude signal through the analyzer with your scope,
and find replacement parts on eBay.

Roy Lewallen, W7EL

The Eternal Squire wrote:
Hello, everyone.

It looks like I have a new, exciting, rather daunting project ahead of me.
I'd like some advice
from anyone interested in providing it.

I have now adopted into my household (and home lab) a Lavoie LA-18M spectrum
analyzer unit,
circa 1950's, completely tubes plus a few microwave crystal diodes. I
also have a manual.

The manual is missing the overall schematic diagram of the entire unit but
has schematic diagrams
of all the major sections, such as the IF strips, and it also has general
troubleshooting procedure.

Given its age, it is in outstanding condition, because everything about it
works. Given the fact that
I bought it for only 50 bucks, as a working antique it is an absolute
treasure. However, I am concerned
about the weak signal sensitivity.

For calibration equipment, I have a late year 100 Mhz analog scope, a
Wavetek 3001 signal generator,
and a digital multimeter, if that helps.

With the signal generator I was able to determine that the minimal
discernable signal at 10 Mhz was
-34dBm. The manual on the Lavoie indicates that the sensitivity at that
frequency should be 40 microvolt,
so that seems to tell me that age has affected the sensitivity.

There are several IF strips on the unit, and the manual tells me which strip
needs to be aligned first, but
it does not go into any detail about how to align each of the IF strips for
best performance... it merely
says "replace IF strips as necessary".

If I had a time machine I could probably follow the manual's suggestions,
but short of that I am out of
luck there. I also understand that the alignment procedure for AM IF
strips is differerent from FM IF
strips, and for IF strips used by spectrum analyzers it could be different
still.

So, does anyone have any advice for me on how I should peak each vaccum tube
IF strip. I will supply
specific information on each, and will email scans of the schematics for
each, to those who want to
consider the problem.

Thanks,

The Eternal Squire

Hi,


circa 1950's, completely tubes plus a few microwave crystal diodes.


Just a suggestion but the commonest cause of low sensitivity in spectrum
analysers, that I've experienced, is the front-end mixer diodes. I wonder if
you have checked them out?


Cheers - Joe, G3LLV

Hi,


circa 1950's, completely tubes plus a few microwave crystal diodes.


Just a suggestion but the commonest cause of low sensitivity in spectrum
analysers, that I've experienced, is the front-end mixer diodes. I wonder if
you have checked them out?


Cheers - Joe, G3LLV

Roy Lewallen wrote:
With about 40 dB deficiency in MDS, you should be looking for something
that's broken, not just out of alignment.

The worst thing you can ever do is start to re-align something that's
actually broken. It won't make anything better, and can easily make
things worse.

Electronic equipment is quite unlikely to wander so far out of alignment
that it completely stops working. So if does stop (and -40dB pretty much
qualifies as "stopped") you know that something has broken.

In that case you need to look for it, find it, and then either fix it or
change it. There is no other way.

There are many different ways to find a faulty component - but *none* of
them begins with: "Before I do anything else, I'm going to twiddle every
adjustment in sight."


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek

Roy Lewallen wrote:
With about 40 dB deficiency in MDS, you should be looking for something
that's broken, not just out of alignment.

The worst thing you can ever do is start to re-align something that's
actually broken. It won't make anything better, and can easily make
things worse.

Electronic equipment is quite unlikely to wander so far out of alignment
that it completely stops working. So if does stop (and -40dB pretty much
qualifies as "stopped") you know that something has broken.

In that case you need to look for it, find it, and then either fix it or
change it. There is no other way.

There are many different ways to find a faulty component - but *none* of
them begins with: "Before I do anything else, I'm going to twiddle every
adjustment in sight."


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek