[R-390] Re: Spurious Resonances

Sat Oct 27 12:36:14 EDT 2007

Hi

Any filter is going to respond to it's termination impedances. Both  
the input and output impedance are significant.

The grid of the RF stage looks like > 100 meg ohms resistive at RF.  
It's got a bit of capacitance, but not a whole lot.

If you look at the preselector like a single tuned circuit, it's going  
to have Q. The Q is going to be determined by:

1) The resistive portion of the source resistance driving the  
preselector, relative to the reactances.
2) The Q of the coils themselves
3) The resistive portion of the load resistance, relative to the  
reactances.

The "relative to the reactances" part gets a bit crazy because you  
normally have part of the resistances in series form and the other  
part of the resistances in parallel form. There is a great thing  
called the "Q formula" that lets you convert them all back and forth.  
It was a staple in EE texts up through about 1954.

Vacuum tubes have really low Gm compared to just about anything else.  
To get a stage to have gain, you need very high impedances. The normal  
way to make stuff work was to run impedances way higher than we do  
today. You can easily fake yourself out going from solid state to  
tube. Loading is a much bigger deal when checking tube circuits.

A measurement you can make on a solid state amp with very little  
interaction, is much more difficult to make on a tube circuit. An easy  
way to see this is using a DVM on an R390. The numbers don't come out  
right. The input impedance of the DVM is not as high as that of a  
VTVM. The DVM is fine on solid state, it's not up to the task with  
tubes.

By going to the plate for your measurement, the load resistance is no  
longer interacting with the preselector at all. The source resistance  
is still there, but the antenna has source resistance as well. Once  
you have the load impedance out of the way you can be reasonably sure  
that what you are measuring is "real".

Since you are after second order distortion, every db you knock down  
the input counts as two db at the output. If you were talking about  
third order (intermod) distortion, each db of attenuation would count  
three db. A 30 db selectivity in the preselector will take second  
order distortion down 60 db and third order down 90 db. At low levels,  
a tube amp should have second order distortion down > 30 db with no  
preselection at all.

If the response you see is at exactly twice the input frequency, it's  
second order distortion. The conversion scheme in a 390 does not  
produce other spurs that follow a 2:1 ratio like that. You can write a  
book about all the possible spurs that can occur (and people have).  
The math is pretty simple. It's been mentioned before in this thread.  
Take the two sine waves, add them or subtract them. Take that and  
raise it to an integer power. Powers of 0,1,2,3,4, and 5 are commonly  
used. If you have a mixer you have three signals rather than just two.  
The same "raise it to a power" is used. You can include am, fm, and pm  
in the signal equations. AM, and FM modulation transfer differently  
through each spur generating process. AM is transfered by all of them  
to the extent that the amplitude of the output will follow he  
amplitude of the input through the db/db ratio for that response.

So much fun .....

Bob

On Oct 27, 2007, at 11:39 AM, Tony Casorso wrote:

> Hi Bob,
>
> I'm not sure what you are getting at. I used a 10x scope probe  
> connected to the Grid of the RF stage. I want to measure the  
> response of the preselector. I do not want to swamp anything. The  
> two pictures show a sharp response from the preselector in one case,  
> and a not so sharp response in the other case. The only difference  
> is the impedance of the signal source connected to the antenna.
>
> Tony
> ----- Original Message ----- From: "Bob Camp" <ham at cq.nu>
> To: "Tony Casorso" <canthony15 at msn.com>
> Cc: <r-390 at mailman.qth.net>
> Sent: Saturday, October 27, 2007 9:04 AM
> Subject: Re: [R-390] Re: Spurious Resonances
>
>
>> Hi
>>
>> What are you picking off the output with?
>>
>> Picking off at the plate would be a *much* better idea if you are   
>> trying to measure the input circuit. You will have far less impact   
>> from the probe that way. If you want to suppress the effect of the   
>> output tank, swamp it with a low probe impedance.
>>
>> Bob
>>
>>
>> On Oct 27, 2007, at 10:35 AM, Tony Casorso wrote:
>>
>>> How lucky can I get? I posted a set of images this morning that  
>>> had  a "spurious response" in it. It turns out it was cause by  
>>> the  settings on my sweeper (DC Offset was cranked over to one  
>>> side for  some reason). The mailer wus unable to post it for some  
>>> reason so I  get to try again with good pictures.
>>>
>>> Here are links to couple of pictures. I am sweeping the  
>>> preselector  in my R-390a. The left edge of the sweep is 1.5MHz.  
>>> The right edge  is 4.5MHz. The center is approximately 3MHz. The  
>>> receiver is tuned  to 3.00MHz and the power is off. I am measuring  
>>> the response at the  grid of the RF stage.
>>>
>>> The broader response picture is with the 25 Ohm (gen with 50 Ohm  
>>> rterminator) sweep gen directly connected to the balanced input.  
>>> The sharper response is with a 2.2K resistor in series with one  
>>> side of  the balanced input.
>>>
>>> It looks like I was right that low source impedance broadens the  
>>> frontend response.
>>>
>>> http://i25.photobucket.com/albums/c90/tonysradios/IMG_0825.jpg
>>>
>>> http://i25.photobucket.com/albums/c90/tonysradios/IMG_0826.jpg
>>>
>>> Tony
>>>
>>> ----- Original Message ----- From: "Tony Casorso" <canthony15 at msn.com 
>>> >
>>> To: <r-390 at mailman.qth.net>
>>> Sent: Saturday, October 27, 2007 8:09 AM
>>> Subject: [R-390] Re: Spurious Resonances
>>>
>>>
>>> I should havethought about it for a few more minutes.
>>>
>>> Maybe the sweeper has a strong second harmonic that is visible  
>>> near  1.5.
>>>
>>> Tony
>>> _____________________________________________________________
>>> R-390 mailing list
>>> Home: http://mailman.qth.net/mailman/listinfo/r-390
>>> Help: http://mailman.qth.net/faq.htm
>>> Post: mailto:R-390 at mailman.qth.net
>>> Unsubscribe: http://mailman.qth.net/mailman/options/r-390
>>> _____________________________________________________________
>>> R-390 mailing list
>>> Home: http://mailman.qth.net/mailman/listinfo/r-390
>>> Help: http://mailman.qth.net/faq.htm
>>> Post: mailto:R-390 at mailman.qth.net
>>> Unsubscribe: http://mailman.qth.net/mailman/options/r-390
>>>
>