[Elecraft] Testing how loud it is!

[Don Brown]

Hi Tom and gang

Your testing approach seems good to me. Anyone doing the tests needs to c=
heck the calibration on the scope and the compensation on the X10 probe. =
Most scopes have a calibrator signal on the front panel with an accurate =
square wave signal at 1000 hz or so. Just make sure the scope measures it=
 correctly taking into account the X10 probe. The square wave should be n=
ice and square with no rounding or peaks at the corners and the top shoul=
d be flat (check in DC coupling mode). If the calibrator does not look go=
od adjust the compensation on the probe. It will be a little trimmer on t=
he probe or in the little box at the BNC end. Many people forget this, ge=
t screwy readings and wonder why. As you are measuring audio, any scope s=
hould be adequate. However many people are confused when the measured sig=
nals get up close to the bandwidth of the scope. A scope is speced at 3 d=
b down at its bandwidth this means it may measure 1 volt as .707 volts at=
 its bandwidth and still be within specs. The vertical amplifiers fall of=
f quickly close to the upper bandwidth limits. This is not a problem with=
 most modern scopes with 60-100 mhz bandwidth when used in the HF bands. =
But if you have a scope with 10 or 20 mHz bandwidth keep it in mind.

BTY the only part of a scope that can be calibrated to NBS is the calibra=
tor. So when you see the little cal sticker on the scope its for the cali=
brator not the scope. It is the operators responsibility to check the ver=
tical amplifier and timing against the calibrator to make sure the scope =
is accurate.

As to your last paragraph I guess I could be considered an expert as I wo=
rked for over 10 years in the metrology business. With five years at Tekt=
ronix as a Test and Measurement instructor. So if anyone has any question=
s on how to make any of the various measurements using most any type of t=
est equipment I will be glad to help. I have a well equipped lab and 2 K2=
's I could make measurements on if required. SN1808 and 2174. I also have=
 2611 under construction and 2186 adding a KPA100 in final check out.

Don Brown
KD5NDB
----- Original Message -----
From: Tom HammondN=D8SS
Sent: Monday, September 30, 2002 8:07 AM
To: Francis Belliveau; Jim Campbell; elecraft@mailman.qth.net
Subject: Re: [Elecraft] Testing how loud it is!

Good morning folks:

Sorry for the delay in replying, but I was out of town this weekend and =20
just returned home Sunday evening.

>You are correct about the variations in side-tone level for the same
>setting, I figured the last measurement would filter that out.  I guess =
we
>could find the right test point and have Ron and Tom give us a measureme=
nt
>at a setting of 128 or maybe even 200 and then we could all try to match=
 it.
>A more complete test would be to measure things at high, low, and medium
>level inputs but I wanted to keep the test simple.  I guess we need to
>decide if the problem is not enough overall gain or maybe non-linear gai=
n
>such that only the loud or weak signals are not amplified enough.

Here's what I'm planning on trying... today, if time permits... and I thi=
nk =20
this may fly... at least for the AF section.

I think that merely setting ST L to an arbitrary NUMERICAL setting may =20
still not allow us to have confidence in our measurements. BUT, for those=
 =20
of us who have o'scopes which we believe to be relatively accurately =20
calibrated, we should be able to set the output of the Quad, 8-Bit DAC =20
(U8), to a certain measurable peak-to-peak level and THEN proceed with =20
performing our output level tests for the rest of the AF chain.

I think it would probably be appropriate the begin by setting the P-P =20
output of the DAC by measuring the sidetone level at the drain of Q5 =20
(2N7000). I've not done this yet, but I'll try to provide a usable level =
a =20
bit later today. I DO think we should attempt to first use a P-P voltage =20
somewhere in the lower to mid range of the volume level control. Probably=
 =20
somewhere below a DAC count of 100, and probably closer to 50.

I recently bought a new (for me) o'scope, so now I'll finally actually ha=
ve =20
a REASON to put it to some good use. I will be using the X10 probe, just =
to =20
help to ensure that I'm not 'loading' any possibly high-impedance part of=
 =20
the AF circuit during my testing.

>The question I am looking to answer is what percentage of us have an Rx
>problem and how many have a simpler audio problem?  This would be a star=
ting
>point so that those with an audio problem do not waste time looking in t=
he
>wrong place.  I figured that finding a way for us all to inject a matchi=
ng
>signal into the antenna to trace down an Rx problem will take a while.  =
This
>seems like a place to focus efforts while the experts work out an approa=
ch
>for the more complex possibility.

Once we can/cannot vindicate the AF chain itself, THEN we can focus on th=
e =20
RF chain.

Results later today.

73,

Tom  N0SS

PS

Just as Fran has 'fessed' up, I shall as well (as I've done before). I ha=
ve =20
NO formal electronics training. Whatever small amount of 'knowledge' I ha=
ve =20
has come from several (44) years of hands-on experience, and having done =
it =20
WRONG enough times to at least have a feeling for possibly doing it =20
right(er) the next time... so I may still be going off on the wrong =20
tangent, but at least it'll be MY tangent...<G>


--- StripMime Report -- processed MIME parts ---
multipart/alternative
  text/plain (text body -- kept)
  text/html
The reason this message is shown is because the post was in HTML
or had an attachment. Attachments are not allowed.
Please post in Plain-Text only.---