[TMC] SBE-3 & SBE-6 on AM

Sat Mar 1 10:49:46 EST 2014

Hi Duncan,

AM was already becoming outdated when these exciters were produced, and 
TMC regarded themselves as sideband specialists and considered AM 
largely irrelevant.   So there was no effort to provide high quality AM 
in the SBE and SBG series exciters. Although you can insert a carrier 
and two sidebands, the resulting DSB AM is poor due to the disturbed 
phase relationships between the two sidebands, and between the sidebands 
and the reinserted carrier.  This causes severe distortion when received 
on an AM receiver with conventional envelope detection.  As you 
observed, no amount of sideband and carrier level adjustment corrects 
the problem.  I have several SBE and SBG series exciters, all of which 
use the same crystal filters, and they all exhibit this problem. The 
problem mainly stems from group delay variations between the two 
sideband filters.  As you noticed, TMC provided a carrier phase adjust 
in the later versions of these exciters, and this does help 
considerably, but it's only a compromise.  It does nothing to compensate 
for the filter problems, and the correct phase relationship can never be 
obtained between the carrier and both sidebands at all modulating 
frequencies.

You shouldn't see that severe modulation asymmetry you describe, though, 
which indicates more serious problems.  Make sure the sideband filters 
are in good shape.  I've found these old crystal lattice filters (made 
for TMC by Bulova Electronics), have often aged considerably, and are 
well out of their original specs.  They often suffer from severe 
passband ripple and poor opposite sideband rejection.  I'd say 75% of 
the SBE/SBG series exciters I've worked on had filter response problems 
of varying severity. You may have a badly aged and mismatched pair of 
filters in your exciter.  Fortunately, they can sometimes be repaired 
and realigned, provided there are no dead crystals.  I unsoldered the 
filter cans, and built up a test fixture to allow repair/realignment on 
a network analyzer, and was able to get the majority back in spec.  Most 
problems seem to be from aged crystals, and the occasional defective 
capacitor.  If the crystals haven't aged too far, the filter may be 
realigned with the internal piston trimmer capacitors.  Also make sure 
the carrier notch filter hasn't failed or aged off freq and encroached 
on one of the sidebands.  You probably would have noticed poor carrier 
suppression on SSB if this were the case.

Filter alignment is tricky, and requires a vector or scalar analyzer 
with synthesized swept LO to do it right.  A vector analyzer is really 
the way to go, as you can observe group delay and amplitude response 
together, which makes alignment much more convenient and gives a better 
overall response.  The object is to get the optimum combination of 
passband frequency response and group delay, with good unwanted sideband 
rejection.  It should be possible to get passband ripple down to within 
about 1.5 dB, and unwanted sideband rejection of 55-60 dB at 500 cps to 
1kc into the opposite sideband.  Rejection may degrade slightly above 1 
kc, but should never drop below 50 dB.  I should point out that all my 
exciters have the wideband, 7.5 kc BW sideband filters, not the 3 kc 
filters.  I've never repaired one of the 3 kc jobs.  They  will have 
much worse group delay at around 3 kc than the 7.5 kc types, of course.  
The 3 kc BW, 250 kc CF filters used in the SBE-3 are completely 
different from those in the SBG-2, BTW.

If you're really serious about using this transmitter on AM, I would 
recommend installing a carrier phase adjust network similar to that used 
on the SBE-9 exciter.  This will definitely improve AM, but even with 
in-spec sideband filters, it won't be perfect. I guess it depends on how 
good you want it, but the bottom line is that two independently 
generated sidebands with a reinserted carrier is just not a real good 
way to make AM.  These were among the best sideband exciters of their 
time, but AM was an afterthought, at best.  TMC did it right in their 
later solid state MMX series exciters, which have a separate DSB 
full-carrier amplitude modulator, separate from the ISB modulator.  
These produce near broadcast-quality AM.

Regarding the hum problem, if it's not a power supply problem, it may be 
a corroded chassis ground return in the audio section, as Richard K. 
suggested.  I've seen this before, as has Mark KI0PF. It may be possible 
to re-tighten ground lug screws with a nutdriver.

Best of luck, and let us all know how things develop...

73,

John K9WT

On 2/28/2014 10:10 PM, Duncan Brown wrote:
> The AWA Museum has a SBT-1KV transmitter with a SBE-6 (similar to the 
> SBE-3) exciter.  I tried to set it up for AM for the AWA QSO Party a 
> few weeks ago and the AM signal looked very distorted. I pulled the 
> SBE-6 and brought it home to work on, but have not been able to find 
> the problem.
>
> Basically, in AM, I can get 100% positive modulation, but only about 
> 50% negative modulation.  Played around with audio levels and carrier 
> insertion, changed tubes, used external oscillators and realigned 
> everything, but to no avail. I wonder if others have had the same 
> problem???
>
> The problem seems to be in the 250Khz mixer/IF area.  The only 
> conclusion that I have come to is that perhaps there is something 
> wrong with the way the 250Khz is reinserted with the sidebands. The 
> SBE-3 manual never talks about 100% carrier AM (only 10-50%).
>
> In the later SBE-8,9,10, there is an added 90 degree phase shift 
> circuit (that is not in the SBE-3 or -6) between the 250KHz oscillator 
> and the carrier insertion point (V126-1).  The SBE-8,9,10 manual says 
> that this is to provide a "true AM signal".  Implying that previous 
> designs did _not_ provide a "true AM signal" ???   (It is well within 
> the realm of possibilities that the SBE-3 was originally designed 
> mainly for SSB and that full AM (by turning up the Carrier Insert 
> control all the way) was just thrown in as an "extra".  Then later on 
> (after the SBE-6 was in production) some customer wanted to use the 
> SBE-x for real AM and the TMC engineers realized that they had to 
> correct the phase of the carrier inset, so it was put into the 
> SBE-8,9,10 version.)
>
> Has anyone with a SBE-3 or -6 had this problem??
>
>
> Another problem I found with the SBE-6 is a lot of 60 hz hum that 
> seems to be getting into the signal in the 250Khz section.  It is 
> about -40 to -50dbc on the output, but audible.  I spent a lot of time 
> trying to track down the hum source thinking it might be connected to 
> the distortion, but never found a smoking gun. Probably due to a bad 
> ground somewhere and the fact that they run the tube heater returns 
> through the chassis.  I tightened up all the tube socket screws (they 
> provide all the ground points for each tube) but didn't make much 
> difference.  Is this a common problem?? (I'm thinking about running 
> separate wires for the filament returns in this section.)
>
> Thanks for any help/comments,
>
>
> Duncan Brown, K2OEQ
>
> Antique Wireless Association Museum Asst. Curator, Commercial Equipment
> (also Chief TTY operator & repairman)
> http://www.antiquewireless.org/
>
>
>
> ---
> This email is free from viruses and malware because avast! Antivirus 
> protection is active.
> http://www.avast.com
> ______________________________________________________________
> TMC mailing list
> Home: http://mailman.qth.net/mailman/listinfo/tmc
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:TMC at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
>