6 metre 4 pole filter
 

Hi peeps.
Just an interesting bit of info... well to me anyway..

Doing some prep work on a VHF Triband Tranceiver.. I like to make modules I have not made or tested before....

The 6 metre front end filter was first..At 2 Mhz bandwidth, I chose 4 X Ind and 4 X caps... the inductors were suitably tapped to match the analyser.

In lashup mode.. all seemed well.. a reasonable flat ish top ...however looking at other circuits I decided to use a Double gate BF96 fet, to provide a bit of gain and AGC control.. Thus feeding the fet gate directly from the top of L4, I was getting a bit of very peaky responses, thus setting up was... difficult.. I played with various coil spacings and coupling, some were better than others, and I had no room or inclination to put screens between the coils..... so I hit on damping the "Q" of each coil.. I tried 33K at first and it was a much improved situation...finally 22K was chosen. Now I know its not perfect.. see photo..but in the final version, final tweeking should be ok.
There will be outer screening in the finished model :D

Re: 6 metre 4 pole filter
 

Just a couple of pointers.

Not wishing to appear to be bragging, but back in the 80's I built a dual band 4 / 6 mtr processor controlled all mode transceiver which was published in RadCom and won the VHF Cup for that year.

You might to take a look at how the rx and tx chains worked and how the filtering was done etc.

The PLL would be a lot simpler these days - in fact if I was to do it again I'd make it SDR based with a A/D and D/A as superhets are 'yesterdays technology'.

Looking at your filter / amplifier it appears you have four tuned circuits ahead of the FET which will likely incur significant loss and so degrade the FET noise figure - important if it's going to be a rx LNA !

Also FET's at 50MHz can be difficult to get stable especially with 'hi q' tuned circuits on inputs and outputs.

Again take a look at my design or have a good google for best practices.

The band pass filters I used employed a combination of inductive and capacitive coupling to create a nice shaped filter symetrical on both sides.

regards Tim

Re: 6 metre 4 pole filter
 

Design was in RadCom, May issue 1990 - how time flies and technology has progressed !

Re: 6 metre 4 pole filter
 

Hi All.. Apparently the RSGB ordered the American History publications to remove all Radcoms.

So.. If anyone has a pdf of the May 1990 articla as above I would be grateful... unless of course it goes against Forum policy. :D

Re: 6 metre 4 pole filter
 

Here's the circuit from Tim's article.

Question for Tim though, is there an error in that circuit or am I just reading it wrong - RL1 coil appears to be in the g1 circuit of TR14 with no obvious way to activate the relay.

Re: 6 metre 4 pole filter
 

Hi Terry,

there's at least two mistakes in that drawing.

RL 1 should be powered from the switched tx/rx 12V line and C55 should couple to TR12.

Looking at that design makes me realise how old it is with the choice of components etc.
I still have rig and it still works. I even have the 144MHz version which was published in Wireless World in 1981/2 !

I dread to think the hours spent designing and writing the code for it - but very satisfying at the time.

73 Tim

Re: 6 metre 4 pole filter
 

The collector of TR12 or the middle winding of T3 (as per T2) ?

Re: 6 metre 4 pole filter
 

the middle winding of T3 (as per T2) is where it should connect.

Re: 6 metre 4 pole filter
 

Thanks Tim, I'll make some notes on my copy, sorry for hijacking this Wendy.

Re: 6 metre 4 pole filter
 

No problem Terry.. You have helped me many times :D
Its a pity the article has no coil information.. such as turns .. diameter etc.. all Toko parts which are not available..But hey ho.. it gives me a starting point. :D

Re: 6 metre 4 pole filter
 

Check your email Wendy ;)

Re: 6 metre 4 pole filter
 

Considering this, I'm reminded of a rather good German military receiver, which used lengths of 'micro-hardline' UHF coax as meandered-quarterwave-lines in its front-end.

"Snakes" they were nicknamed.

Perhaps you could adopt this approach? To be honest it really confused the first set of bench-techs who had to deal with it.

"What's this 'schlange-schwingungskreis' thing? There's no front-end tuned-circuits! How can this work!"

Re: 6 metre 4 pole filter
 

Starts getting a bit big on 6 metres... about 1 metre long each assuming 69% velocity factor for PTFE dielectric 'semi-rigid coax'

But quarter-lambda lines are rather useful for narrowband filters. For 70MHz in a small space, helical resonators are a good possibility. You can get SAW filters covering various ranges around that frequency... see Golledge's website. But the insertion loss of SAWs can be off-putting.

Re: 6 metre 4 pole filter
 

I have used "tuned stubs" on 2m for TVI problems back in the day.. Mind you the Germanic way of lumping several words together, was utterly confusing to me when I got some tech gear from Grundig...It was IF alignment equipment for a big batch of TV's that were bought to convert to PAL "I" from "M".

Re: 6 metre 4 pole filter
 

If it helps, a typical 50MHz BPF with 4 resonators and about 5% bandwidth should have an insertion loss of maybe 2dB - 4dB depending on the quality of the components. There are several BPF types that can be used for a 5% BW and I would normally choose one that has the steepest rejection on the same side as the LO and image.

The best place for this is just after the preamp as per Tim's impressive design. If you have RF switching here, the same BPF can be used for both Rx and Tx.

It's more than 30 years since I operated on 6m but I doubt there is much advantage in having a system noise figure of much lower than about 6dB. This would include feeder loss in the overall system. It might be different for diehard DXers in a rural location with a beam but for most operators the noise levels on the band will be dominated by manmade and galactic noise.

With this in mind it might be possible to use a basic preamp made with a U310 JFET in grounded gate and this will be easier to manage in terms of stability at the expense of a slightly higher noise figure. Otherwise, some care will be needed with a dual gate mosfet as Tim warns. They can go unstable very easily and the BF981 can generate negative resistance at the drain up past 2GHz if care is not taken with the design.

I think TR10 should have a PNP symbol on Tim's schematic but it's a bit fuzzy in places so it might already be OK. It looks like an NPN symbol to me.

Re: 6 metre 4 pole filter
 

TR10 shows as NPN on the schematic - BD132.

Re: 6 metre 4 pole filter
 

I'd be careful with that Radcom schematic as it doesn't seem to be accurate but my hand drawn originals are - so not sure how it got past proof reading back in the day....

However the filtering using ferrite tuned S18 coils does work very well, the two inductively coupled coils were / in metal cans. I could post photo's of the physical implementation as I still have the radio.

Good idea about using a grounded gate fet for the front end - easy and stable providing the subsequent stages NF are not too high.

Getting the correct distribution of gain / NF and filtering for optimum performance is critical - I have an old spread sheet for just that purpose.

Adding a high gain low noise pre-amp can sometimes degrade the overall performance due to inter modulation products / gain compression.

Lot to be said for using valves !

73 Tim

Re: 6 metre 4 pole filter
 

Hi Jim...... I would be grateful for a photo of the filter inductors as the Toko catalogue does not show any Cirkit numbers and conversely I cannot find a CIRKIT catalogue.. Or Dimensions.. and number of turns.... I can work from there.. Thanks :D

Re: 6 metre 4 pole filter
 

Try this.
We mourn their passing along with 10K, MC108 etc.
I have the similar info for those parts.

They were 1.5 turns to max 8.5 turns (= max capacity of former) with 3 different slugs, 2 types of ferrite and aluminium.

Winding OD is 7.5mm, wire dia 0.66mm, slug OD 4.5mm (could be M4.5 x0.75 thread? - unconfirmed)
Former material appears to be some form of polythene.

The colour of the former followed the resistor colour code. Green 5.5T Violet 7.5T White 8.5T etc from examples I have here also confirmed in the pdf.

Re: 6 metre 4 pole filter
 

Coilcraft did an equivalent to S18. In fact they came first. A coil craft rep showered us with gorgeous sample boxes and we designed them into various things. But our buyers trying to get manufacturing quantities was a disaster. I don't know why. But then I found the Toko S18 family and we could get those. So Toko got the business.

Coilcraft are now great to deal with and have a site near cumbernald in the UK.

I think they still do their S18 type things.

I use their stuff quite a bit, and had a bit of fun at the reaction of their CEO when I told him about the fun we'd had at HP.

David