Improving Selectivity In Vintage Receivers

[Bazz4CQJ]

Hi Andrew, great to hear from you. As you'll have seen we have been discussing the ceramic filters for a while, and I thought I checked the forum for previous work, but only came across what you had done over the Xmas break.

I wonder if I could ask you to do a simple sketch circuit of what you did?

Although you say that you've become a bit dissatisfied with it because it isn't good with CW, for someone like myself (not a CW user) it would be interesting to see.

As you'll realise, Jon is looking at using a Murata (ceramic) ladder filter in his HRO while I'm looking at the possibility of using a modified crystal filter, as used in several Hammurland receivers, but I'm sure we'd like to know more about what you did.

@Jon, yes Laurence's R106 does look pretty good, and presumably, it was a late model as it uses the 0-1mA meter rather than the original S-meter used in the earlier sets.

B

[ms660]

Originally Posted by Bazz4CQJ

yes Laurence's R106 does look pretty good, and presumably, it was a late model as it uses the 0-1mA meter rather than the original S-meter used in the earlier sets.

I think it was made during the last months of WW2, the cabinet and front panel wasn't in that condition when I got it, it was refurbished by myself as were all the coil packs which were in a poor state when I first got them.

In the background in the photo of the R106 I posted earlier is a HRO Senior which was fitted with the original S meter.

The front panel and cabinet of the R106 was refurbished using some elbow effort and some semi gloss black paint, the coil packs were refurbished using some more semi gloss black paint, considerably more elbow effort and some clear acrylic sheet.

The HRO is my favourite receiver, a simple elegant circuit and layout from the 1930's.

The HRO Senior mentioned had a plug in Xtal that you could easily take out and scrub

Lawrence.

[Radio Wrangler]

A reasonable AM filter is not good for SSB or CW.
A reasonable SSB filter is not good for AM or CW
A reasonable CW filter is not good for AM or SSB.

Once you start getting into better filters, they start to be much more mode specific, so you wind up using mode specific ones, switched for the modes you want to be able to use.

Don't be put off a filter by any reports that it's not good for modes it wasn't designed for. The modes are sufficiently different in their needs that this is actually a good sign!

David

[Jon_G4MDC]

That is certainly true. In the ideal case I would hope one could have the original full BW AM response just determined by the original IFTs - already fine for AM. A choice to use an SSB BW ceramic filter is obviously something I want.

If the crystal filter can be retained to work in conjunction with either of those then that is a bonus. Here is another reason why I quite like putting the ceramic filter on a coupling winding. It can be switched in or out as needed.

By the way I have a few other Murata ceramic filters which are quite narrow. They are not CFJ455K performance but they might still be useful. They originally came from a speech processor kit supplied by a DL amateur - make SSB, clip it, demod back to AF principle. It was sold for fitting in FT817/857/897 microphones. TBH it didn't do much for me so I didn't ever build the second one.

[Radio Wrangler]

Ceramic filters can age, and this usually menas a drift of centre frequency. So before committing to any large project., check the filters you have are still OK.

David

[Bazz4CQJ]

Would somebody like to suggest the preferred means (a sketched circuit?) for checking the ceramic filters (preferably using the sort of low-grade test kit I own)?

I did a very crude check on my HRO crystal, simply putting it in series between an old sig-gen and an old scope and did see the scope respond at 455kHz, but I suspect that a slightly more elaborate set up would have been better. I've seen one suggestion to put a 10R resistor in series with the crystal so that the crystal current can be observed, might be better?

In terms of one filter not suiting all modes, yes, but the Hammarlund switched loading of a single crystal seems to have been a good attempt at a "filter for all seasons". If the directly-wired version of the HRO crystal, devoid of any connections to short it, comes good, the Ham. filter might still prove a good option and is not a drastic re-work of the existing HRO circuit (and still be able to listen to music).

B

[Jon_G4MDC]

I found the filter used by the DL OM in one version of the speech processor that I mentioned. (Search for DF4ZS)

Below is a snippet from the Murata data book about it along with some advice about how to test them.

It shows a sig gen with series resistor to pad up the impedance to suit the filter. It doesn't terminate the generator so you will get the open circuit Voltage (EMF) out of it.

I would prefer to add a termination resistor to ground so you can have a decent length of cable between generator and filter. Of course the resistor network needs to be mounted on the filter at the far end of the cable. SG Voltage is now divided by 2 (PD if that is how it's output is calibrated - depends on manufacturer).

On the far side the RF Voltmeter can just as easily be an oscilloscope. The high input impedance is just shunted by a resistor to suit the filter in question.

Have I got it right that you wanted to test filters rather than single resonators? The Murata doc I have features a few resonators but they are intended for use in FM discriminators. As such it natters on about S curves and presents test circuits for each type of FM IC MC3357 etc. Not much help for what you want I fear.

[Jon_G4MDC]

Been thinking on adding a link winding to the HRO Mixer Anode Coil assy.

The universal law of $^d says that the tuned winding is up against the end cheek. To add an extra winding is only possible alongside the untuned Anode coil and that is not the one that I want best coupling with.

Has anyone experimented with moving the coil pies further along the former? I would like to create about a 1/8" gap where the new winding could be wound in a slot. I wonder what dope sticks the coils to the former and what might dissolve that without also attacking the former material?

[GW3OQK Andrew]

Bazz, I tried to attach a sketch but the procedure didn't complete. I'll try it on reply to private message.

A basic mod is to connect 3.3pf from 2nd IFT to V6 grid.
Connect resonator with trimmer in parallel from grid to ground.
Connect 2M2 from the AGC to the grid.
It will make some BW improvement but reduce gain a bit.
73
Andrew

[Bazz4CQJ]

@ jon the only ceramic components I have are the one or two element ones, so I guess those are resonators rather than true filters.

Re moving the coils, I think the former is polystyrene and that does not have much chemical resistance to solvents other than perhaps IPA. I would be be reluctant to start re-arranging 80 year old Litz wire. I suppose you could wind on the outside, but then the problem would be re-using the existing screening can.

@andrew I'll PM you with my email and perhaps you could send the sketch to me and I'll find away of getting on the forum. Thanks.

B

[Jon_G4MDC]

I'm feeling stupid. Lawrence and Bazz's figures have shown the two coils are identical.

Why don't I just swap the tuning capacitors over to the other one and put the new winding alongside it? I think will try this soon. Much safer than trying to move the coils.

Do you have any pictures of your ceramics please Bazz? I'm a bit curious to see what you call the two element ones. Are they three connection terminals or four?

[Radio Wrangler]

"Two-pole resonators" There are two resonant structures, with a controlled amount of mechanical coupling between them. There are two electrodes for each resonator, but often one on each is taken to a common ground pin, making a 3-terminal package. This saves the cost of a whole pin! but it limits ultimate attenuation. Accountancy shoots, and scores again!

These structures are done also at 10.7MHz as the common FM IF filter elements.

You can get narrower band ones done on quartz.

A double-tuned IFT is another 2-pole coupled resonator device. Mathmatically they are all similar.

DAvid

[GW3OQK Andrew]

Hello Bazz, I've sent email with attachments. One is of mods I made to bandspread the coil packs which was really successful.

Is anyone here thinking to retune the IF to use a 455KHz ceramic filter?
73
Andrew

[Jon_G4MDC]

Hello Andrew, welcome back. Yes as per the title. My own interest is an HRO at the moment but of course there are many other vintage sets which were designed for AM, provided with a BFO for CW and perhaps a narrow crystal filter on the posh ones. Of course once upon a time Q multipliers were being used to get around not having a crystal filter.

Ceramic resonators and filters have come along since most of these were designed and the interest is how might they be used to make improvements to suit modes such as SSB. Q multipliers didn't help much there.

I have looked through my Murata ceramic filter info and for 455kHz I can find 4 to 11 elements. At 10.7 (10.8) MHz there are certainly ceramic duals. There have been past suggestions of using single 455kHz resonators which of course is possible but might involve carefully selecting parts.

I want to use one of the ready made ceramic filters intended for SSB. My present interest it how best to match that to the HRO Mixer/IF stages in the least invasive way. I would like to retain the old AM & CW bandwidths too but at least to try the SSB filter out to see how worthwhile it is.

SSB reception on the old sets is almost always the station you want to hear accompanied by upside down monkey chatter coming from the other side of the BFO.

73

[Bazz4CQJ]

Ok, this is the circuit sent to me from Andrew; I think the original was over 1MB so was over the limit for attachments.

The comments he made were, "Note the top of the IFT going to V6 grid has been cut and grounded. The bottom wire under the chassis is taken from the AGC line and goes to the filter".

The idea of putting the circuit between the 1st and 2nd if is interesting.

I think Jon and myself would both like to have HRO's which cope better with SSB than they do in their original config. but quite what we each want, and how we might get there are still a bit hazy. As mentioned before, the thread was not intended to be specific to HRO's and might equally interest anyone with any one of a number of old receivers which lack selectivity control.

As I have already taken the HRO crystal filter out, I could think of copying Andrews circuit but maybe putting in a stage earlier. Were I go from here needs thinking about!

B

[Bazz4CQJ]

Pics of some of my resonators; some are plainly a single element and none have more that 2 elements.

Lurking in my junk box is the transistorised wideband IF stage of an old Pye Vanguard; it has 4 stages of IF amplification and then the the detector and it occurs to me that maybe that could be fashioned in to some means of comparing different ceramic resonators. I don't have data sheets for all the ones I have. It would be interesting to find a point to insert a resonator on the board and the read mV from the detector as I move the sig-gen from ~450-460kHz; should pick out the good from the bad??

B

[joebog1]

Bazz, there were also some available that took the place of a traditional transistor IF transformer.
They have five pins and used to be coloured red. They are something less than 1/2" high and wide, and about 1/4" thick.

Joe

[FrankB]

The main solution here was to take an ARC5 and convert it to a "Q5er". Plenty of articles on this in QST, CQ, and ARRL hand books of the post Korean War era.

[G3PIJpeter]

Originally Posted by FrankB

The main solution here was to take an ARC5 and convert it to a "Q5er". Plenty of articles on this in QST, CQ, and ARRL hand books of the post Korean War era.

IF response curve attached - See QST for January 1948. I use a standard BC453 'Q-Fiver' with one of my HROs and another (modified) with a BC348-C. Both give a huge improvement subjectively. Just don't ask about shape factors and all that.
Peter G3PIJ

[Bazz4CQJ]

I've been pursuing this area of interest by getting a wobbulator and sawtooth generator constructed, and those are now near to completion.

With those available, I was wondering about how best to start to learn a little about how the ceramic filters work in-circuit (and how wobbulators work ). I've extracted an old Pye Vanguard IF board from the junk box and think it might make a useful testbed (at least for transistor circuits).

The attached figure shows the IF stage and is marked up to show some positions where I thought filters could be installed. I think the AGC should be disabled and the winding that feeds MR2 used to take an RF output from the board.

Any thoughts?

B