[Elecraft] Field Day rig experience

Thu Jun 14 18:38:14 EDT 2018

Hi Guys.

For what it's worth, I find it interesting how the term "Roofing filter" has changed a bit over time and with different 'ownership'. 

Personally, I first came across the term in around 1966 as a junior design engineer working on the Redifon R550 series of HF receivers. I understood then that the term "roof" referred to the "top of the house" filter used to provide the first measure of protection against adjacent unwanted signals.

These and other similar HF receivers used an up-conversion architecture, and the R550/551 employed a first IF at 38 MHz with the local oscillator running 38 to 68 MHz.  The bandwidth of this filter, which followed the first mixer, was around 15 KHz as I recall. The second IF was at 1.4 MHz (or 1.6 MHz in other similar designs) and featured a number of selectable crystal filters typically providing close-in band-widths from around 200Hz to 12 KHz.  Employing a first IF above 30 MHz shifts the first image into the VHF spectrum and allows the use of a 30 MHz low pass filter in the front end, with sub-octave band pass filters to provide a measure of front-end selectivity.  We would have loved to provide close-in selectivity at the first IF frequency and so avoid a down-conversion to the second IF, but achieving the required passband /stopband characteristics just was (is) not possible at 38 MHz.  However, decent close in selectivity (passband and stopband) can be provided with crystal filters at around 9 MHz or thereabouts, and many of the earlier purely analogue designs of amateur equipment took advantage of this, including TenTec.  I do not personally view this particular application as a roofing filter as is not protecting further stages of selectivity.t it All now ancient history...things have moved on a bit since then! Can anyone trace the term further back in time?

However, It seems to me that the term "roofing filter" still makes perfectly good sense in the context of the K3 design, with the selectable crystal filters providing the maximum possible (mode dependent) selectivity protection in front of the final IF, even if that is now implemented using DSP techniques! 

Terry
G3VFO
-----Original Message-----
From: elecraft-bounces at mailman.qth.net <elecraft-bounces at mailman.qth.net> On Behalf Of WILLIE BABER
Sent: 14 June 2018 16:02
To: elecraft at mailman.qth.net; Wes Stewart <wes_n7ws at triconet.org>
Subject: Re: [Elecraft] Field Day rig experience

Hello Wes,

I took a look.  Both designs are using the idea of "roofing filter" to refer to up-conversion radios similar to the use of up-conversion 3khz filters as roofing filters in Icom radios.

"Roofing filter" (a mode specific filter after the first mixer including narrow cw filters) only makes sense in the context  of the history of superhet design and in particular the use of one broad 15 khz first I-F (so that all modes may pass through it) typical of all Japanese radios until recently.  Calling a 45 mhz filter at the first I-F a "roofing filter" as noted in the info you sent entirely misses the point of what roofing filter means.  Or, to put it another way, all Ten-Tec radios had roofing filters in them (and were ssb and cw only) well before the term roofing filter was coined!  Which is why an Omni C will out perform any wide (15 khz) first I-F Japanese radio, even those built well after the 1980 vintage Omni C.

Unless mode specific up-conversion crystal filters can be made and as narrow as 200 hz (this is possible with down-conversion) then "roofing filter" and up conversion doesn't make sense historically or in reality.  

Actually, Icom says that did it with 1.2khz filter at 64 mhz in the Icom 7851, though I'm not convinced the filter is that narrow, and 1.2khz is far from the 200hz filter that my K3 has in it (however, the placement of this filter is why the 7851 is among the best radios in Sherwood's chart, on cw).

It is possible to make very narrow and precise crystal filters as narrow as the 200 hz inexpensively, and this is the point of having multiple roofing filters at the first I-F.  So, this is the origin of the term roofing filter---in comparison to the barn-door up conversion first I-F.

73, Will, wj9b

CWops #1085
CWA Advisor levels II and III
http://cwops.org/

--------------------------------------------
On Wed, 6/13/18, Wes Stewart <wes_n7ws at triconet.org> wrote:

 Subject: Re: [Elecraft] Field Day rig experience
 To: elecraft at mailman.qth.net
 Date: Wednesday, June 13, 2018, 3:08 PM

 Certainly not to disparage the
 K3(S) architecture (I have two of them) there is  nothing inherently wrong with an up-conversion  receiver, if modern hardware is used.

 See: https://martein.home.xs4all.nl/pa3ake/hmode/g3sbi_intro.html

 and my friend Cornell's,
 Star-10 transceiver. 
 https://pdfs.semanticscholar.org/eb33/5c12858779a653d9b9b93ca20120aebb7616.pdf

 Wes  N7WS

   On 6/13/2018 11:38 AM, WILLIE BABER
 wrote:
 > Robert is talking about the
 crystal filters, also known as roofing filters now-days,  that are typically placed after the first mixer (I  mistakenly typed "ahead" but I meant  "after" as Robert notes), though there is a post  amp and NB before these filters in K2 and K3.
 >
 > The idea is that a
 crystal filter right after the first mixer gives high  dynamic range because high selectivity comes before the  receiver has developed stages of gain that otherwise could  cause blocking or IMD, especially when selectivity is  postponed to the second mixer while ignoring gain  distribution in prior stages of the receiver.  This basic  idea was popularized in Solid State Design for the Radio  Amateur, and it was applied to Ten-Tec radios for decades  (at a 9 mhz I-F).
 >
 >
 Roofing filter gets defined in relationship to Japanese  radios that had up conversion 15 khz filters at the first  I-F, and generally lower dynamic range as a result, (but you  got all modes, general coverage, and optional crystal  filters at the second I-F).
 >
 > Good for everyone radios.... but with  lower dynamic range and phase noise from the early  synthesizers.  This is why Ten-Tec radios were so popular  among contesters, especially Omni V and VI (modified with a  narrow cw filter at the first I-F).
 >
 > 73, Will, wj9b
 >

 ______________________________________________________________
 Elecraft mailing list
 Home: http://mailman.qth.net/mailman/listinfo/elecraft
 Help: http://mailman.qth.net/mmfaq.htm
 Post: mailto:Elecraft at mailman.qth.net

 This list hosted by: http://www.qsl.net  Please help support this email list: http://www.qsl.net/donate.html  Message delivered to wlbaber at bellsouth.net ______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:Elecraft at mailman.qth.net

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html Message delivered to terry.hart at btinternet.com

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus