re:Re: LF: Re: 136kHz WSPR experiment

["Michael Sapp" <wa3tts@verizon.net>]



 Roger:  Yes I find having the IPO "selected" is required which I believe means having the preamp "off" effectively, so we are in agreement.  It is somewhat  surprising but not entirely so how the FT-817 hears well at LF/MF with a good LPF filter in line.

 

I may give your 137 kHz E-probe a go with a high-Q inductor made with FT50-61 material. It would be curious to see the response versus a lower Q 77 or 75 high permeability core....somehow a half mH air wound version would not seem to fit well in the box ~:)

 

 

Thanks for the postings,  73, Mike wa3tts

 

 

----- Original Message -----

From: Roger Lapthorn

To: rsgb_lf_group@blacksheep.org

Sent: Thursday, January 25, 2018 3:48 PM

Subject: Re: LF: Re: 136kHz WSPR experiment

Thanks Mike. I seem to remember that I got surprisingly good results with my EFP (schematic shown a few days ago) feeding into the FT817 on 136kHz, This has not be tried since before my 2013 stroke. If I am correct the EFP is in a drawer somewhere! I shall have to try to find it. I have a lot of junk.

Next season I shall try the EFP on 136kHz RX.

BTW, I find sensitivity on 136kHz with the FT817 is better with IPO selected. Maybe this relates to your findings?

73s

Roger G3XBM

On 25 January 2018 at 20:31, Michael Sapp <wa3tts@verizon.net> wrote:

Roger: I tried to post a reply to the group prior but it did apparently did not make it past the admin monitor or it exceeded message size limit.

 

I found the key to getting the FT-817 to receive well on LF and MF is to use a low pass filter with good in-band return loss characteristics between the antenna (e-probe, EWE, etc.) and the

FT-817.  If you have ever tried listening to an AM broadcast station with the FT-817 and it's VHF rubber duck antenna, you know it is stone-cold deaf that way.  But if you use an antenna with a reasonable 50 ohm match at lower HF, the FT-817 hears AM band stations (530 ~ 1600 kHz) very well. The same principle applies to LF/MF in the 100kHz to 500 kHz range.  The FT-817 receiver front end needs to see something close to 50 ohms to receive signals efficiently.  But it also needs most of the AM band energy rejected so it will not overload.  Using an LPF with a good pass band match seems to solve the problem when using an e-probe antenna at LF/MF with the FT-817.  Also do not use the preamp in the FT-817. Turn off the AGC and NB.

 

There is a VE7BPO NDB filter with build information online that is easy to construct with commonly available parts if you do not have an LF/MF LPF filter available. I'm sure if you try it or another good LF/MF LPF with your FT-817 you will see an immediate order-of-magnitude improvement in receive performance at LF/MF.  I use a variation of the VE7BPO filter for 630m and a 9th order 300 kHz Butterworth LPF for LF.  The Butterworth LPFs have very good pass-band return loss and the FT-817 responds very well to having the filter between the antenna and the radio at LF.

 

For even better performance I run a broadband LF/MF 2N3904 preamp after my LPF and either a 3dB pad or a 50 ohm universal diplexer for the frequency of interest after the preamp and into the FT-817.  I built the 2N3904 preamp last night to try it out based on your recent FT-817 LF postings...

 

Last night I was able to decode WH2XXP in AZ from PA with the FT-817 at some 3000 km distant 55 times with the FT-817 on the 2200m band. I was just able to occasionally decode WH2XXP with only the LPF and no post filter preamp.  With the 2N3904 preamp into a 3 dB pad and then into the FT-817 I was able to obtain a best overnight SNR of -18 on WH2XXP at 3000 km distant. I did need to use the attenuator in the FT-817 to adjust the rx level.

 

Quick sketch of preamp and FT-817 rx layout...

 

https://www.dropbox.com/s/tmszk33xuyjjtnc/2N3904P.JPG?dl=0

 

Pix of filter-preamp setup but with 137 kHz diplexer on preamp  output instead of 3 dB pad...

 

https://www.dropbox.com/s/tms4gruu2x8azz4/flpredp.jpg?dl=0

 

73 Mike wa3tts

Virus-free. www.avg.com