Return-Path: Received: from post.thorcom.com (post.thorcom.com [195.171.43.25]) by mtain-mg02.r1000.mx.aol.com (Internet Inbound) with ESMTP id B3F5B38000086; Mon, 26 Dec 2011 12:00:32 -0500 (EST) Received: from majordom by post.thorcom.com with local (Exim 4.14) id 1RfDtF-0001di-8D for rs_out_1@blacksheep.org; Mon, 26 Dec 2011 16:59:09 +0000 Received: from [195.171.43.32] (helo=relay1.thorcom.net) by post.thorcom.com with esmtp (Exim 4.14) id 1RfDtE-0001dZ-6D for rsgb_lf_group@blacksheep.org; Mon, 26 Dec 2011 16:59:08 +0000 Received: from relay2.uni-heidelberg.de ([129.206.210.211]) by relay1.thorcom.net with esmtp (Exim 4.63) (envelope-from ) id 1RfDtC-0000GI-Vp for rsgb_lf_group@blacksheep.org; Mon, 26 Dec 2011 16:59:08 +0000 Received: from freitag.iup.uni-heidelberg.de (freitag.iup.uni-heidelberg.de [129.206.29.204]) by relay2.uni-heidelberg.de (8.13.8/8.13.8) with ESMTP id pBQGx3Q4028042 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO) for ; Mon, 26 Dec 2011 17:59:03 +0100 Received: from [129.206.22.206] (pc206.iup.uni-heidelberg.de [129.206.22.206]) by freitag.iup.uni-heidelberg.de (8.12.11.20060308/8.11.2) with ESMTP id pBQGx3Xi022196 for ; Mon, 26 Dec 2011 17:59:03 +0100 Message-ID: <4EF8A765.6000609@iup.uni-heidelberg.de> Date: Mon, 26 Dec 2011 17:57:09 +0100 From: =?ISO-8859-1?Q?Stefan_Sch=E4fer?= User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; de; rv:1.9.1.8) Gecko/20100227 Thunderbird/3.0.3 MIME-Version: 1.0 To: rsgb_lf_group@blacksheep.org References: <4EF3689B.1030902@iup.uni-heidelberg.de> <805AC05E834C4F7DA019C36A04013479@JimPC> <4EF4B307.2010609@iup.uni-heidelberg.de> <4EF4D6C4.6010407@iup.uni-heidelberg.de> In-Reply-To: X-Spam-Score: 1.4 (+) X-Spam-Report: autolearn=disabled,HTML_MESSAGE=0.001,RATWARE_GECKO_BUILD=1.426 Subject: Re: LF: Re: Automatic Variometer adjustment at DK7FC Content-Type: multipart/alternative; boundary="------------090500070800060308090405" X-Spam-Checker-Version: SpamAssassin 2.63 (2004-01-11) on post.thorcom.com X-Spam-Level: X-Spam-Status: No, hits=0.0 required=5.0 tests=HTML_MESSAGE autolearn=no version=2.63 X-SA-Exim-Scanned: Yes Sender: owner-rsgb_lf_group@blacksheep.org Precedence: bulk Reply-To: rsgb_lf_group@blacksheep.org X-Listname: rsgb_lf_group X-SA-Exim-Rcpt-To: rs_out_1@blacksheep.org X-SA-Exim-Scanned: No; SAEximRunCond expanded to false x-aol-global-disposition: G X-AOL-SCOLL-SCORE: 0:2:495872544:93952408 X-AOL-SCOLL-URL_COUNT: 0 x-aol-sid: 3039ac1d60ca4ef8a82f5e71 X-AOL-IP: 195.171.43.25 X-AOL-SPF: domain : blacksheep.org SPF : none This is a multi-part message in MIME format. --------------090500070800060308090405 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by relay2.uni-heidelberg.de id pBQGx3Q4028042 Hello Andy, Thanks for the discussion. Always good to see different ways and solution= s. Yes, that's another way to reduce the 4V problem in my case. But maybe=20 it is even a bit more critical. Depending on the mechanical load /=20 torque and temperature of the motor it seems that sometimes even 4V is=20 not enough to start the motor running. Once it runs, then 4V is OK. So i=20 need more voltage to _safely_ start the motor. Yes yes, this can be=20 compensated by inserting another C in series to the feedback R and=20 switching another R in parallel to that R :-) I replaced the 33 Ohm resistor by a series configuration of 5 pairs of=20 anti-parallel switched UF4007 diodes that were available, giving a 6V=20 voltage drop. It was the easiest way to optimise the circuit here. This=20 safely starts the motor now and allows a reasonable adjustment time now. I'm QSYing from 136.172 to 137.7. and start to TX on 137.7, say in=20 DFCW-4 without resonating the antenna before. This is done automatically=20 now in the first DFCW element which appears a bit weaker then at some=20 receivers. So it takes just 4 seconds. I'm often running the system from=20 remote since not beeing in Heidelberg since 4 days :-) It works :-) Haven't done a phase measurement of the actual phase yet (with an=20 oscilloscope) but i estimate the reactance offset is < than +- j 1 Ohm. 73, Stefan/DK7FC Am 24.12.2011 21:37, schrieb Andy Talbot: > In that case - you could try a non-linear amplifier, with a high gain > until the magnitude of the output voltage is around 4V then reducing. > Makie the opamp feedback R of two elements, then shunt one of these > with a pair of back to back zeners of about 4V rating. At low Vout, > the two resistors operate in series giving a high gain. As 4V output > is reached, the diodes start to conduct, removing the contribution > from the R they are shunting, and leaving the remaining series R for > reduced gain. The speed of the gain change can be modified with > another R in series with the diode arm > > Bit hit and miss, perhaps, but it ought to be possible to compensate a > non-linear load by a non-linear drive network. > > Andy > www.g4jnt.com > > > > 2011/12/23 Stefan Sch=E4fer: > =20 >> Hi Andy, >> >> Thanks for the hint. >> >> I intended to use a hard drive since the motor does not rotate when su= pplied >> with< 4 V DC. So it would always stop before the resonance is reached= . That >> problem may be reduced by increasing the gain of the input stage but t= hen i >> expect oscillation problems and small currents running through the mot= or all >> the time... >> >> 73, Stefan >> >> Am 23.12.2011 20:24, schrieb Andy Talbot: >> >> =20 >>> Stefan.. >>> >>> An idea. You have implemented a bang-bang driver with a dead zone. >>> Switched hard on in either direction, or stopped close to the middle. >>> >>> Change the circuit so you linearly amplify the output from the phase >>> detector, and make youself a bidirectional bridge driver, using say a= n >>> op-amp and NPN-PNP pair - as in audio amplifiers. Or it may even be >>> possible to use an audio amplifier chip if the gain can be brought >>> down low enough.. >>> >>> Now, the motor will automatically slow as the phase error is reduced, >>> and you 'shouldn't' have to include any dead zone. >>> You may have to compensate the loop to prevent overshoot or high >>> frequency instabilities if there is too much mechanical lag in the >>> system or inertia, but that is something that can be done pure with C= R >>> networks. >>> >>> Andy >>> www.g4jnt.com >>> >>> >>> 2011/12/23 Stefan Sch=E4fer: >>> >>> =20 >>>> Jim, >>>> >>>> Another thought: >>>> >>>> Am 23.12.2011 12:17, schrieb James Moritz: >>>> >>>> =20 >>>>> [...] I guess the inertia of the motor and gearbox will produce som= e >>>>> hysteresis, moving the variometer to a very slightly over-adjusted >>>>> position >>>>> after the motor drive is removed, which will also help to prevent t= he >>>>> system >>>>> continuously searching around the correct tuning point.[...] >>>>> >>>>> =20 >>>> >>>> This effect can be compensated by choosing the dead band treshold va= lue. >>>> Then the motor gets stopped a bit before the resonance point and wil= l >>>> exactly land on it. As good as having an infinitesimal small dead ba= nd >>>> ;-) >>>> Anyway we are talking about phase angles which are very very small, = much >>>> below that what could become critical for a PA. >>>> >>>> 73, Stefan/DK7FC >>>> >>>> >>>> =20 >>> >>> =20 >> >> =20 > =20 --------------090500070800060308090405 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Hello Andy,

Thanks for the discussion. Always good to see different ways and solutions.

Yes, that's another way to reduce the 4V problem in my case. But maybe it is even a bit more critical. Depending on the mechanical load / torque and temperature of the motor it seems that sometimes even 4V is not enough to start the motor running. Once it runs, then 4V is OK. So i need more voltage to safely start the motor. Yes yes, this can be compensated by inserting another C in series to the feedback R and switching another R in parallel to that R :-)

I replaced the 33 Ohm resistor by a series configuration of 5 pairs of anti-parallel switched UF4007 diodes that were available, giving a 6V voltage drop. It was the easiest way to optimise the circuit here. This safely starts the motor now and allows a reasonable adjustment time now.

I'm QSYing from 136.172 to 137.7. and start to TX on 137.7, say in DFCW-4 without resonating the antenna before. This is done automatically now in the first DFCW element which appears a bit weaker then at some receivers. So it takes just 4 seconds. I'm often running the system from remote since not beeing in Heidelberg since 4 days :-) It works :-)

Haven't done a phase measurement of the actual phase yet (with an oscilloscope) but i estimate the reactance offset is < than +- j 1 Ohm.

73, Stefan/DK7FC


Am 24.12.2011 21:37, schrieb Andy Talbot: 
In that case - you could try a non-linear amplifier, with a high gain
until the magnitude of the output voltage is around 4V then reducing.
Makie the opamp feedback R of two elements, then shunt one of these
with a pair of back to back zeners of about 4V rating.  At low Vout,
the two resistors operate in series giving a high gain.   As 4V output
is reached, the diodes start to conduct, removing the contribution
from the R they are shunting, and leaving the remaining series R for
reduced gain.  The speed of the gain change can be modified with
another R in series with the diode arm

Bit hit and miss, perhaps, but it ought to be possible to compensate a
non-linear load  by a non-linear drive network.

Andy
www.g4jnt.com



2011/12/23 Stefan Schäfer <Stefan.Schaefer@iup.uni-heidelberg.de>:
  
Hi Andy,

Thanks for the hint.

I intended to use a hard drive since the motor does not rotate when supplied
with < 4 V DC. So it would always stop before the resonance is reached. That
problem may be reduced by increasing the gain of the input stage but then i
expect oscillation problems and small currents running through the motor all
the time...

73, Stefan

Am 23.12.2011 20:24, schrieb Andy Talbot:

    
Stefan..

An idea.  You have implemented a bang-bang driver with a dead zone.
Switched hard on in either direction,  or stopped close to the middle.

Change the circuit so you linearly amplify the output from the phase
detector, and make youself a bidirectional bridge driver, using say an
op-amp and NPN-PNP pair - as in audio amplifiers.  Or it may even be
possible to use an audio amplifier chip if the gain can be brought
down low enough..

Now, the motor will automatically slow as the phase error is reduced,
and you 'shouldn't' have to include any dead zone.
You may have to compensate the loop to prevent overshoot or high
frequency instabilities if there is too much mechanical lag in the
system or inertia, but that is something that can be done pure with CR
networks.

Andy
www.g4jnt.com


2011/12/23 Stefan Schäfer<Stefan.Schaefer@iup.uni-heidelberg.de>:

      
Jim,

Another thought:

Am 23.12.2011 12:17, schrieb James Moritz:

        
[...] I guess the inertia of the motor and gearbox will produce some
hysteresis, moving the variometer to a very slightly over-adjusted
position
after the motor drive is removed, which will also help to prevent the
system
continuously searching around the correct tuning point.[...]

          

This effect can be compensated by choosing the dead band treshold value.
Then the motor gets stopped a bit before the resonance point and will
exactly land on it. As good as having an infinitesimal small dead band
;-)
Anyway we are talking about phase angles which are very very small, much
below that what could become critical for a PA.

73, Stefan/DK7FC
--------------090500070800060308090405--