[GreenKeys] ST6 questions

[Bill Henry]

Hi Eugene and greenkeys guys,

Regarding the removing of turns to tune an 88 mHy toroid, we found:
         at 2125 or 2295, the frequency increases 3 Hz for every turn removed.
         at 2975 Hz, the frequency increases 5 Hz for every turn removed.

Be careful when removing turns - take off "almost enough" and re-check!

To lower the frequency, adding 100 pF in parallel (silvered mica, 
polystyrene, or mylar - NOT disc ceramic):
         at 2125 or 2295 Hz, the frequency lowers by 2 Hz
         at 2975 Hz, the frequency lowers by 5 Hz.

When measuring the "center frequency" of these filters, do NOT try to find 
the peak.   The filters are too broad.  Rather, look at the tuning meter 
and look for side frequencies equally down on each side of the peak and 
computer the center.  For example, if the tuning meter peaks at 0.7, 
measure the frequencies above and below resonance when the meter reads 
0.6.  Add them, divide by two, and that's your true center frequency.

I have noticed that over the years the tuned circuit frequency of an 88 mHy 
toroid filter seems to move up.  I have no idea why but it has been 
consistent.  Maybe the toroids "age"?  I have never caught an orange drop 
or silvered mica capacitor aging but I suppose it's possible.

In regard to absolute calibration, what others have said is true.  On 
receive, only the shift is critical and the filters are broad enough that 
anything from 160 to 200 Hz M-S difference works well for "170 shift".  We 
always tried to set set them "dead-on" but as I said something seemed to 
drift with time.

HOWEVER, if you are going to use a transceiver, it is very important that 
your transmit tones exactly match your demodulator receive 
filters.  Otherwise, you won't be sending and receiving on the same radio 
frequency and your QSO will probably "walk up and down the band" as the 
other guy tries to stay tuned to your frequency.

In regard to L-C passive filters vs active filters, LC filters are a lot 
harder to design and to tune.  But, their dynamic range is very very 
good.  It takes a LOT of signal to overload an LC filter - such as in 
saturating the toroid or arcing-over the capacitor!  Conversely, active 
op-amp filters are a lot easier to tune but also have a very limited 
dynamic range.  Desiging good active filters is an art.  Some circuits work 
a lot better than others.  If given a choice ALWAYS choose an op-amp 
circuit that uses more op-amps - like 2 op-amp sections rather than 
1!  Also, there is a LOT to be said for power op-amps from +12 and -12V 
rather than trying to make do with one that runs off of 5V 
only.   Actually, a vacuum tube (valve?) running on 100V  is VERY hard to 
beat for dynamic range (no doubt a "secret" of the TTL-II)!  If you 
over-drive an op-amp filter, it becomes a clipper or limiter, probably 
capturing on the unwanted signal.  Forget it's filtering 
capability.  Digital filters can be designed to produce virtually any 
band-pass you want - BUT - they also suffer from dynamic range problems 
only more so when overloaded.  If you over-drive the A/D converter, it's 
all over - forget the signal or interference.

The ST-6000 design was triggered when we couldn't get enough 88 mHy toroids 
for ST-6's - and the labor to tune them was eating us alive.  As with all 
new designs, we kept adding features and the ST-6000 evolved into quite a 
bit more of a demodulator than just an active filter ST-6.  There were 
quite a few "active filter ST-6's" written up in the RTTY Journal in the 
early 70's - see the CD ROMs.

In regard to the transistors on the autostart board, to the best of my 
recall, we used either an MPS3394 (plastic TO-92) or 2N697 (Metal can TO-5) 
at Q401.  We always used the 2N697 at Q404 and 405 and a MPS3703 (plastic) 
or 2N2905 (metal) at Q402, 403, and 406.

That's what occurs to me this morning!

73,
Bill Henry, K9GWT


At 10:31 PM 7/16/2006 +0000, Eugene Hertz wrote:
>Hello all,
>
>I've been using an ST6 from ebay for several years now.  It has always 
>done good service but the motor control anti-space circuit never worked 
>which was ok by me.  I never had a chance to really look into it in any detail.
>
>Recently I acquired 3 more ST6s on ebay and now have been able to do some 
>comparative anatomy between them all. Very interesting findings. One has 
>obviously been built from a kit, 2 seem to be factory and the fourth is 
>unknown.
>
>So I took the best of the 3 new ones and began comparing to my regular ST6 
>and I started looking at the frequency response. Here's what I got using 
>my R&S Polarad SPN audio generator. ST6 #1 is one of the new ones, ST6#2 
>is my regular use TU.  Note that my regular unit does not have 425 shift.
>
>ST6 #1  (new one)
>MARK    SPACE   SHFT
>2150    2330    180
>2150    2580    430
>2150    3010    860
>
>ST6 #2  (regular use)
>2160    2325    165
>2140    2975    835
>
>So, how important is it to have the frequencies right? Or is it more 
>important to have the total shift right. In the case of the shift for ST6 
>#1, they are pretty good varying 5-10 hz from expected. #2 is 5-15hz off a 
>little worse.
>
>Should I try to get the shifts more accurate to 5hz each?
>
>My receivers only have BFO, no product detectors (YET) so I seem to always 
>be able to tune the receiver or BFO to produce the right tones. However, I 
>hope soon to own a product detector (outboard). How will this affect the 
>need to have the tones accurate vs the shift?
>
>Second batch of questions.
>
>Seems out of a total of 4 ST6's only 1 auto start/anti-space board 
>works.  Can anyone tell me if this is a common problem? In one case, there 
>is a transistor that seems to be in backwards compared the one only board 
>that works. Is it possible this never worked? This board looks like it 
>came from the factory.  Could there have been a substitute transistor used 
>that either had the EBC in a different order?
>
>Thanks all!
>Eugene
>
>
>
>
>
>
>
>
>
>_______________________________________________
>GreenKeys mailing list
>GreenKeys at mailman.qth.net
>http://mailman.qth.net/mailman/listinfo/greenkeys