[Hammarlund] HQ-110 alignment

[Kenneth G. Gordon]

On 6 Apr 2006 at 19:53, Julian Bunn wrote:

> Thanks to Ken for his comments on my web page describing IF sweeps.
> 
> Some clarifications on my setup: I do in fact use the sweep
> generator's ramp output as the X input on the scope, and it is this
> ramp that I run at 1ms. I.e. the sweep generator runs between 400kHz
> and 500kHz in 1ms. I probably need to make that clearer on the web
> page. (I use the 10second sweep temporqarily while visually checking
> that the frequency limits are where I want them.)

Ah! I did not read your write-up carefully enough. In that case, your fast 
sweep rate can cause ringing in the IF chain and give you false 
indications. In order to see a true representation of the IF passband, 
you have to use a sweep rate that is very slow. There is a formula for 
calculating that, but I don't have it at my fingertips.

> There are two types of display that are produced, and I haven't quite
> understood why yet. If you look at this picture:
> http://pcbunn.cacr.caltech.edu/jjb/hammarlund/if-fg504-4-1.jpg you
> will see the standard IF bell curve on the 7904. This is what appears
> when the Y plates are connected at the input to the demodulator tube
> i.e. pin 2 of V7, the output from the T5 IF can, and is what I believe
> you see in your setup.

Not quite, although that is close to what I get. Where I connect the 
vertical probe of my scope is at the plate (output) of the detector. Then 
what I get is a trace somewhat like yours, but "filled in". I.e., I don't get 
just an outline of the IF passband, like yours looks there, but more like 
your trace below, except only the top half. The bottom half of your trace 
below is not included.

If you think about it, what is happening is that as the SG starts on the 
left side of the scope (when the horizontal sweep is synchronized to the 
SG output) the vertical input to the scope will rise as the SG sweeps 
across the IF passband, and the indication on the scope will be a 
higher and higher vertical signal, to a peak, then lower and lower to the 
right hand edge and out of the IF passband. Thus a "filled in" IF 
passband shape on the scope screen.

But in order to accomplish this, you have to keep the sweep rate very 
low in order not to get the IF chain ringing. It is somewhat of a 
compromise between keeping the IF chain from ringing, and keeping 
the scope trace bright enough and persistent enough to see. In my 
case, thus the darkened room, with a spot-light on the receiver IF 
adjustments.

> However, when the Y plates are connected to prior IF stages, the IF
> response curve looks like this:
> http://pcbunn.cacr.caltech.edu/jjb/hammarlund/sweep2.jpg

That looks like ringing to me. What you are seeing is an oscillation in 
the IF chain, and thus an incorrect representation of it. You are getting 
a trace above and below the center line, and that isn't correct.

> 
> In the former case, the signal appears to be demodulated/rectified. In
> the latter it is not. That's my theory, anyway (I am by no means an
> expert in any of this).

What I was trying to avoid by connecting to the DETECTOR output, 
was loading the IF chain. If you connect to the output of the last IF 
stage as you do in the first photo, you cannot really properly adjust that 
stage. In all the setups I have read about, the vertical scope probe is 
connected to the output of the detector instead, as I have done.

BTW, there is an entire military book on this method of visual alignment 
for the Collins R-390 receiver. I have a copy of it here somewhere. It is 
really TOO detailed and is quite extensive. Further, it goes into the 
theory of why this is the best method of alignment. I think it is entitled 
something like "Using the Visual Method of Alignment for the R-390 
Receiver".

As I said in a previous e-mail, you sure have led the way, and shown 
everyone on the list how to properly connect the system up.

All you need is a few tweaks.

Good job! :-)

Ken Gordon W7EKB