Return-Path: X-Spam-DCC: paranoid 1002; Body=2 Fuz1=2 Fuz2=2 X-Spam-Checker-Version: SpamAssassin 3.1.3 (2006-06-01) on lipkowski.org X-Spam-Level: X-Spam-Status: No, score=-2.0 required=5.0 tests=BAYES_00,DNS_FROM_AHBL_RHSBL, HTML_30_40,HTML_MESSAGE autolearn=no version=3.1.3 Received: from post.thorcom.com (post.thorcom.com [195.171.43.25]) by paranoid.lipkowski.org (8.13.7/8.13.7) with ESMTP id u0DGEJlh011453 for ; Wed, 13 Jan 2016 17:14:19 +0100 Received: from majordom by post.thorcom.com with local (Exim 4.14) id 1aJNza-0006WY-5b for rs_out_1@blacksheep.org; Wed, 13 Jan 2016 16:09:50 +0000 Received: from [195.171.43.32] (helo=relay1.thorcom.net) by post.thorcom.com with esmtp (Exim 4.14) id 1aJNzZ-0006WP-KX for rsgb_lf_group@blacksheep.org; Wed, 13 Jan 2016 16:09:49 +0000 Received: from mail-wm0-f52.google.com ([74.125.82.52]) by relay1.thorcom.net with esmtps (TLSv1.2:ECDHE-RSA-AES128-GCM-SHA256:128) (Exim 4.86) (envelope-from ) id 1aJNyS-00086W-M7 for rsgb_lf_group@blacksheep.org; Wed, 13 Jan 2016 16:09:48 +0000 Received: by mail-wm0-f52.google.com with SMTP id f206so379486980wmf.0 for ; Wed, 13 Jan 2016 08:08:24 -0800 (PST) X-DKIM-Result: Domain=gmail.com Result=Good and Known Domain DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; bh=68kJjOdnKlS6ZLqI32Od599jbTVGLguqPX+0ZV5E0IU=; b=ltTj+zrbsJNIMC03QJy+EUCNIHIGCwmYVKgTd3bee713fOSuK4CJCMLLq8UBlOmlwR JlFe6s2RpgG768vnvooXPZWwxMu91hPr/AcwAEve2O9GYsS0ul7RdSAv4v06MYiGlB5+ gMUllAIdj+dxe7VPrVIfGfvVU5kYnARAlNe5uhDI0xpUGMSKkpILxSdC/W4gxl9tfAbz XcOQxOe+/vewLTvOUZzcRC/XiHh5VF9E1FJ/Nrtem4vLkbr5pEXyphTDhUnEPt1RpSXw SjXI6+HPNfSO5n0KR8XCnWxyiMH5SZihNbZYdU5HHTGlr7KE+KoFQk7LRXemlRFktXGZ yj0A== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:in-reply-to:references:date :message-id:subject:from:to:content-type; bh=68kJjOdnKlS6ZLqI32Od599jbTVGLguqPX+0ZV5E0IU=; b=INab4gsZppa0Z4JCb3Uoj1AivGVCOZc54WdFY9kW1hxeZK4UGmKLoSpu5Az1qscN07 sQt+Bt2S90IzyVgdHVDB8WLMkefXM2NWpFGJsiWH42Jqoc517ih5sa94CXroj97fL9l2 gvCWxuKSgS7zd9OENmJkVuoE60B03RPBJu9SR5AHG9VNlKwXcw/KbKyLhltCINVeMhEj cADejx8nt3pmBAYbj94SjLcCn7CWbK1IwQWV21+oXl8ZTj9ojnk455DpJtAiSrxqZLjZ CmFJJqPLUGBCGtI3tL35JSzA2Xb3OJoPg9KUBjeteP2oqawnFlf7MUWVNC98V8B/lJqG a+Cg== X-Gm-Message-State: ALoCoQnTMkLUwiy1RPs826N8oub1f3kzHu6yBxVJFyMbckhU3xeigVsMkYwHyyRxq6l3ThKzahrIYmdh7hWxh1D1cIdaw5WEEQ== MIME-Version: 1.0 X-Received: by 10.194.103.164 with SMTP id fx4mr128395087wjb.56.1452701298621; Wed, 13 Jan 2016 08:08:18 -0800 (PST) Received: by 10.28.150.6 with HTTP; Wed, 13 Jan 2016 08:08:18 -0800 (PST) In-Reply-To: <1523b989d9e-5764-7820@webprd-a106.mail.aol.com> References: <1523b989d9e-5764-7820@webprd-a106.mail.aol.com> Date: Wed, 13 Jan 2016 16:08:18 +0000 Message-ID: From: Andy Talbot To: rsgb_lf_group@blacksheep.org X-Scan-Signature: e51eb24639420ed2b9d8115488127c6b Subject: Re: LF: Decca 5501 Content-Type: multipart/alternative; boundary=047d7bf198e4de8a570529396060 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-Scanned-By: MIMEDefang 2.56 on 10.1.3.10 Status: O X-Status: X-Keywords: X-UID: 6301 --047d7bf198e4de8a570529396060 Content-Type: text/plain; charset=UTF-8 I've knitted together Fig4a and 4b. the driver circuitry as http://www.g4jnt.com/DownLoad/Decca5501_Driver.gif It's not at all clear, and looks a ling winded way of doing things, but I guess that is more to do with component availability in the 1970's The "on" drive via TR2 comes direct from the squarer IC1a and two gates The "off" or clamp drive takes the output of the squarer, goes through a delay (R10, C8 and inexplicably D7), another opamp to square things up again and into D1. It looks as if all that complexity is there to give some dead time between the A side on and B side on periods. That will get passed through to the output devices and prevent current spikes due to unequal switch on / off times. In which case the various zeners and clamp diodes are there to protect the driver devices themselves from high voltage transients during the dual-off period. Oh the joys of circuit design in that era when everything had to be done with basic components - instead of an EE lookup table or a DSP chip Andy G4JNT On 13 January 2016 at 15:26, wrote: > OK thanks Andy I do have the circuit diagrams but I understood that the > clamp only operated if the drive was lost. The description was not that > clear to me but a probe with the scope may clear that up. I have added > resistors across the fet gates so that if the transformer fails the fet > should turn off. I think I was lucky when the transformer failed and the > fet was not killed! The insulation in the transformer is only what is on > the litz wire and it has sleeving out of the coil. It failed as the ferrite > is earthed and it had degraded the insulation over time. It was a pig to > wind but was worth it, no spares are probably available unless from a scrap > unit. > Dave > > > > -----Original Message----- > From: Andy Talbot > To: rsgb_lf_group > Sent: Wed, 13 Jan 2016 10:45 > Subject: Re: LF: Decca 5501 > > I've just looked again, and there's more to to the drive circuitry than > that. It is single ended drive for current into the inductor, as I stated, > but there is another IRF520 used as an active clamp (the circuit diagram > was drawn a bit ambiguously, so didn't spot the arrangement > straightaway). So when these two FETs are driven alternately there will be > a proper equal amplitude symmetrical drive waveform generated. > > It is definitely used as an energy storage inductor though, hence a gapped > core is essential. > > 'jnt > > On 13 January 2016 at 10:23, wrote: > >> Andy thanks for that I am looking to find a scrap amplifier module to >> play around with with a view to rebuild the PCB and drive circuits. Looking >> on a scope the switching is very symmetrical but that may be only luck. I >> have used pushpull fets to drive another amp I designed and it worked well >> using a rewound driver transformer from a pc power supply. The complete >> setup is now on the QRP labs website if you want to have a look, look under >> GW8GLO. >> 73 Dave >> >> >> >> -----Original Message----- >> From: Andy Talbot >> To: rsgb_lf_group >> Sent: Wed, 13 Jan 2016 9:51 >> Subject: Re: LF: Decca 5501 >> >> Bit late now, but I found the circuit of the Deca transmitter on G3XDV's >> website. >> http://homepage.ntlworld.com/mike.dennison/index/lf/decca/transmitter.htm >> . >> (Look at Fig 4, 'right' ) >> >> The driver circuit is just a single IFR520 FET switching into the >> transformer, so there is a net DC component - and of course it has to be a >> gapped core. In fact I'm a bit surprised at that circuitry, considering it >> has to switch the FETs alternately on each half of the output waveform, a >> single ended, flyback drive like that cannot possibly be symmetrical. >> Especially as the only reverse clamp appears to be a 68V zener.. But no >> doubt it is '*good enough'* In my 137kHz Tx design - very loosely >> based on the Decca output stage, I used a ferrite transformer (ungapped) >> and AC coupling from the driver IC into that. >> >> Andy G4JNT >> >> >> On 9 January 2016 at 09:26, wrote: >> >>> Thanks for the reply the transformer failed because of a small cut in >>> the rubber sleeving where it entered the ferrite. As the ferrite is earthed >>> the insulation on the litz was not good enough. I have a diagram and used >>> that to help rewind the new transformer. There does not appear to be any DC >>> component flowing in any winding according to the diagram. This unit has >>> possibly been exposed to damp conditions as two of the input transformers >>> were also OC primary and were rebuilt. One set of gate windings are running >>> at the OP from the bridge and this is the one that failed. I am thinking of >>> rebuilding the whole drive board to improve the isolation. I have already >>> added voltage regulators to supply the ICs, this makes the 27v current a >>> lot lower! >>> 73 Dave >>> >>> >>> >>> -----Original Message----- >>> From: Andy Talbot >>> To: rsgb_lf_group >>> Sent: Fri, 8 Jan 2016 18:53 >>> Subject: Re: LF: Decca 5501 >>> >>> The fact it is a GAPPED core is of more significance than that of the >>> type of ferrite. In a transformer application - as that appears to do - >>> it is surprising that a gapped one is used. I do not have the circuit of >>> the Decca transmitter to hand, but can only assume the driver has DC >>> passing through the winding, hence the need to introduce an air gap to >>> prevent ferrite saturation. >>> >>> In which case, the fact your unit overheated suggests your gap was too >>> small and the Vt product (volt-seconds is what saturates ferrite, not >>> current). BUT if that was the was the original unit, you need to ask the >>> question why it should have failed before trying to replace it. Those >>> transmitter units were very reliable under rated things, so components >>> failing need investigating. >>> >>> Andy G4JNT >>> >>> On 8 January 2016 at 11:09, wrote: >>> >>>> >>>> >>>> >>>> Does anybody know what ferrite grade is used on the fet driver >>>> transformer. I have had one burn out and have rebuilt it but would like to >>>> redesign it with better isolation between windings. Have tried various >>>> cores but cannot get the same clean waveform. The original is a gapped core >>>> of unknown material ferrite. Does anybody have a scrap amp unit that I can >>>> use for testing. The finished mods have given 1 kw at 53 volts, can be >>>> driven up to 75 volts but dummy load cannot take it!!! >>>> 73 Dave GW8GLO >>>> >>> >>> >> > --047d7bf198e4de8a570529396060 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
I've knitted together Fig4a=C2=A0 and =C2=A04b. t= he driver circuitry =C2=A0as http://www.g4jnt.com/DownLoad/Decca5501_Driver.gif

It's not at all clear, and looks a ling winded wa= y of doing things, but I guess that is more to do with component availabili= ty in the 1970's

The "on" drive via = TR2 comes direct from the squarer IC1a and two gates
The "of= f" or clamp drive takes the output of the squarer, goes through a dela= y (R10, C8 and inexplicably D7), another opamp to square things up again an= d into D1.

It looks as if all that complexity is t= here to give some dead time between the=C2=A0A side on and B side on period= s. That will get passed through to the output devices and prevent current s= pikes due to unequal switch on / off times.=C2=A0=C2=A0=C2=A0In which case = the various zeners and clamp diodes are there to protect the driver devices= themselves from high voltage transients during the dual-off period.
<= div>
Oh the joys of circuit design in that era when everythin= g had to be done with basic components - instead of an EE lookup table=C2= =A0or a DSP chip

Andy=C2=A0 G4JNT

On 13 January 2016 at 15= :26, <gw8glo@aol.co.uk> wrote:
OK thanks Andy I do= have the circuit diagrams but I understood that the clamp only operated if= the drive was lost. The description was not that clear to me but a probe w= ith the scope may clear that up. I have added resistors across the fet gate= s=C2=A0 so that if the transformer fails the fet should turn off. I think I= was lucky when the transformer failed and the fet was not=C2=A0 killed! Th= e insulation in the transformer is only what is on the litz wire and it has= sleeving out of the coil. It failed as the ferrite is earthed and it had d= egraded the insulation over time. It was a pig to wind but was worth it, no= spares are probably available unless from a scrap unit.
Dave



-----= Original Message-----
From: Andy Talbot <andy.g4jnt@gmail.com>
To: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Sent: Wed, 13 Jan 2016 10:45
Subject: Re: LF: Decca 5501

I've just looked again, and there's more to=C2=A0to the drive = circuitry than that.=C2=A0 It is single ended drive for current into the in= ductor, as I stated, but there is another IRF520 used as an active clamp=C2= =A0(the circuit diagram was drawn a bit ambiguously, so didn't spot the= arrangement straightaway).=C2=A0=C2=A0So when these two FETs are driven al= ternately there will be a proper equal amplitude symmetrical drive waveform= generated.

It is definitely used as an energy storage inductor though, hence a ga= pped core is essential.=C2=A0

'jnt=C2=A0

On 13 January 2016 at 10:23, = <gw8glo@aol.co.uk<= /a>> wrote:
Andy thanks for that I am looking to find a scrap amplifier module t= o play around with with a view to rebuild the PCB and drive circuits. Looki= ng on a scope the switching is very symmetrical but that may be only luck. = I have used pushpull fets to drive another amp I designed and it worked wel= l using a rewound driver transformer from a pc power supply. The complete s= etup is now on the QRP labs website if you want to have a look, look under = GW8GLO.
73 Dave



-----= Original Message-----
From: Andy Talbot <andy.g4jnt@gmail.com>
To: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Sent: Wed, 13 Jan 2016 9:51
Subject: Re: LF: Decca 5501

Bit late now, but I found the circuit of the Deca transmitter on G3XDV= 's website. http://homepage.ntlworld.com/m= ike.dennison/index/lf/decca/transmitter.htm.
(Look at Fig 4, 'right' )

The driver circuit is just a single IFR520 FET switching into the tran= sformer, so there is a net DC component - and of course it has to be a gapp= ed core.=C2=A0 In fact I'm a bit surprised at that circuitry, consideri= ng it has to switch the FETs alternately on each half of the output wavefor= m, a single ended, flyback drive like that cannot possibly be symmetrical.= =C2=A0 Especially as the only reverse clamp appears to be a 68V zener..=C2= =A0 But no doubt it is 'good enough'=C2=A0=C2=A0=C2=A0 In = my 137kHz Tx design - very loosely based on the Decca output stage, I used = a ferrite transformer=C2=A0 (ungapped)=C2=A0 and AC coupling from the drive= r IC into that.

Andy=C2=A0 G4JNT


On 9 January 2016 at 09:26, &= lt;gw8glo@aol.co.uk> wrote:
Thanks for the reply the transformer failed because of a small cut i= n the rubber sleeving where it entered the ferrite. As the ferrite is earth= ed the insulation on the litz was not good enough. I have a diagram and use= d that to help rewind the new transformer. There does not appear to be any = DC component flowing in any winding according to the diagram. This unit has= possibly been exposed to damp conditions as two of the input transformers = were also OC primary and were rebuilt. One set of gate windings are running= at the OP from the bridge and this is the one that failed. I am thinking o= f rebuilding the whole drive board to improve the isolation. I have already= added voltage regulators to supply the ICs, this makes the 27v current a l= ot lower!
73 Dave



-----= Original Message-----
From: Andy Talbot <andy.g4jnt@gmail.com>
To: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Sent: Fri, 8 Jan 2016 18:53
Subject: Re: LF: Decca 5501

The fact it is a GAPPED core is of more significance than = that of the type of ferrite. =C2=A0 In a transformer application - as that = appears to do - it is surprising that a gapped one is used. =C2=A0 I do not= have the circuit of the Decca transmitter to hand, but can only assume the= driver has DC passing through the winding, hence the need to introduce an = air gap to prevent ferrite saturation.

In which case, the fact your unit overheated suggests your gap was too= small and the Vt product (volt-seconds is what saturates ferrite, not curr= ent). =C2=A0 BUT if that was the was the original unit, you need to ask the= question why it should have failed before trying to replace it. =C2=A0 =C2= =A0Those transmitter units were very reliable under rated things, so compon= ents failing need investigating.

Andy =C2=A0G4JNT

On 8 January 2016 at 11:09, &= lt;gw8glo@aol.co.uk> wrote:



Does anybody know what ferrite grade is used on the fet driver = transformer. I have had one burn out and have rebuilt it but would like to = redesign it with better isolation between windings. Have tried various core= s but cannot get the same clean waveform. The original is a gapped core of = unknown material ferrite. Does anybody have a scrap amp unit that I can use= for testing. The finished mods have given 1 kw at 53 volts, can be driven = up to 75 volts but dummy load cannot take it!!!
73 Dave GW8GLO




--047d7bf198e4de8a570529396060--