Piezo Filters etc

[radiozero]

Hi. Can someone please tell me the type of filter that's fiited in my TRIO JR-500S. The second filter in the radio is described as a mechanical filter centred on 455 Khz. I attach the drawing of the filter & it's selectivity curve. You will notice that it has a piezo transducer at input and output with a "Coupler" joining them. Questions: Is this truly a "mechanical filter? Is it a ceramic filter? Is it a crystal filter? It's a very small thing, the same size as IF transformers you find in pocket portable radios of the 60's etc.

I'm hoping I can source the manufacturers specs in a TOKO catalogue, but I won't hold my breath on that. Manufactured around 1967. Thank you.

[G6Tanuki]

I guess the definition of a 'mechanical filter' can be a bit open.

In my mind the term applies to the classic Collins sesign where the incoming signal is fed to a coil-and-magnet transducer then fed through a series of coupled resonant discs before being converted back to IF by another coil-and-magnet.

The piezoelectric/ceramic type is still 'mechanical' in that the IF signal is piezoelectrically converted to a mechanical wave which propagates through a calculated mechanical path to the piezo element at the output whre it gets converted back to an electrical signal.

Definition-wise, you could argue that a classic crystal-filter [single-crystal with phasing-control, or the more modern KVG "Ladder" types] are both really mechanical filters since they depend on the mechanical resonance of lumps of Quartz.

[kevinaston1]

Ambit used to distribute these, look on the American Radio History website. Here is a link to part 1 of their catalogues.

https://worldradiohistory.com/UK/Mis...-Catalog-1.pdf

[Radio Wrangler]

The classic mechanical filters were a range created by Collins Radio. They used a group of resonator discs mounted in a row on a common axis. Discs were coupled together with wires running along the outside of the gapped cylinder the discs made up. At each end there was an electromagnetic transducer.

This structure was a mechanically-realised coupled resonator structure and could be used to make many poles and to synthesise the classic filter shapes. Collins made some unusual centre frequencies, but 455 kHz was by far the most common. I've come across them down to 45kHz. They were very expensive in their day, but not as expensive as a comparable crystal filter.

This sort of small, precision instrument was rather attractive to Japanese industry and the new kid on the block was Kokusai, making similar filters to the Collins ones, but at a lower price. Kokusai used a stack of metal discs as resonators with coupling wires, but they used piezo-elextric transducers at the ends.

I remember reading reviews reckoning that the Kokusai filters had advantages, the Collins magnetic transducer was more non-linear at the levels encountered and suffered intermodulation effects.

Over the years we've learned that the Kokusai filters and some of the Collins have died due to chemical breakdown of the foam used to support their active structures.

These are the filters that everyone called 'mechanical' Note that the transducers and the coupled resonator structure are quite separate.

Along came some other filters that were mechanical in terms of their filtering mechanism. They were made out of a piezo-electric ceramic like lead-zirconium titanate. In these parts the piezo material had the density and elasticity that allowed self resonant structures to be made. No need for transducers to be added, the stuff itself was a transducer, electrode pairs were added at the ends. The most common of these filters are at 10.7MHz and found in most FM tuners and radios. They are really just 2-pole resonators in each package. Two packages are often concatenated to give a 4-pole response. Also, around this time, the radio makers wanted something for 'AM' radios where a 455-ish kHz IF was largely implicit in the band frequencies. Piezo ceramic worked here, too, but manufacture as a number of slabs was more expensive than the 10.7MHz ones. Murata and later Toko were active in this market. Brush-Clevite had been a fore-runner in the US. Various names were bandied about, usually 'mechanical' or 'ceramic' though all were mechanical and piezo-electric. Your filter in the JR-500S will be one of these.

A little later another mechanical filter style came along. Two pairs of electrodes could be formed on the surfaces of a single slice of quartz. Each facing electrode pair constituted a quartz resonator, and there was acoustic coupling between the two resonators. Quartz is piezo electric, so the transducer is implicit, and the unit operates as a 2-pole resonator filter, the resonances being mechanical, of distorting material. More complex filters were made of concatenated 2-pole resonator elements, with shunt capacitors regulating the electrical coupling from element to element. These get called "Monolithic crystal filters" and largely dominated amateur radio equipment for a few decades. These rarely were used much below 9MHz. 10.7 was their happy hunting ground.

If your filter is dead, then I'd go looking for anyone with any of the MuRata 455kHz types left in stock. They were made in wide AM bandwidths, SSB bandwidths and somewhat rarer narrower ones for CW.

Ambit were big stockists. Bonex (BEC nowadays) were pretty much the last of the UK stockists of Toko types. They might be worth a phone call.

David

[Bazz4CQJ]

I think they were generally referred to as Ceramic filters (see attached). There were a handful of different companies each making a range of components so you need a type number. They seemed to fall out of favour, but it was possible to find NOS components in the fairly recent past.

B

[radiozero]

Hi. Thanks for the interesting and informative responses. So, mechanical filter covers a whole lot of filter products. And the filters in my JR-500S would perhaps more often be called ceramic filters. That's my understanding. I've been trying to establish the actual components, but not easy since no markings on the 10mm square cans. The only thing I was able see on both filter cans was "RCL PATENT No 542912 etal". I think that refers to the can. I attach a picture of the two filters. One can is a matching transformer, the other the filter. 1967-ish year of manufacture, not sure as the RX serial looks to be 840293. I'm interested in MF2, the 455 Khz filter, which is at left in the picture. MF1 is the first IF filter, that passes 8.9 - 9.5 Mhz. MF2, I would call a SSB filiter. Certainly used as such, with fixed local carrier / BFO at 453.5 Khz.

[Bazz4CQJ]

I, for one, would very much like to see the schematic of the Trio if you have one. Could you confirm that the reason you want to know about the filter is that think it has failed, so is the Rx dead or just working poorly?

I haven't bought any ceramic filters for some time, but I certainly did acquire a number of different ones on ebay. I have a realistic DX-160 which uses them and I also considered attempting to use them to build a better filter for my HRO, for which the original filter is CW only.

B

[radiozero]

Hi. The filter is working OK. What I am trying to do, just for the fun of it, and the fact I'll increase my knowledge, is to fully understand the operation of all the vintage receivers I have. I mean a fairly deep techinical understanding. I've been into ham radio since the 1970's but I've never studied radio or electronics in general above an elementary level. I should have been an electronic engineeer, but that never panned out. At the moment I'm curious about the filter arrangement in my JR-500S. Ideally I'd find out the filters specs, but I think that's problematical. I do wonder if I'll find the filter models in old magazine adverts. Or, I could measure, and discover the input and output impedances. I have all the Practical Wireless magazines for the entire 1960's (save about half a dozen copies). Perhaps TOKO advertised their filters in that mag. I must cycle thru them, see what gives.

[Bazz4CQJ]

Attached is a Toko datasheet which contains info on some of their products. I would guess that even if yours is not exactly the same, it will be fairly close.

As for careers, I became interested in radio/electronics as a kid, and it's OK as a hobby, now and then, but not sure about doing it as a day job .

B

[Radio Wrangler]

Piezo-ceramic and quartz filters have a few problems inherent in them. They also have a number of overtone modes as well as some unpredictable spurious modes. To combat this, it's usual to have some L-C based filtering. It doesn't need to be as sharp as needed to shape the passband, but it does need to be narrow enough to block signals at the spurious responses of the mechanical aspects.

On top of this, the piezo stuff often isn't at any preferred impedance regime, so the associated tuned circuit can be arranged to do impedance transformation to make the overall composite filter more house-trained.

Toko were the kings of Japanese IF transformers and had a crazily huge range.

Now, flip back to the photo in post 6. One of those two coils in its can looks quite normal, but the otheris extended on one side, so the hole for the tuning slug looks off-centre.

Toko did a range of IFTs with built-in ceramic filters. The sort of 2-pole resonator filter that Bazz fas attached, but mounted inside the enlarged can.

Guess what they look like?

So those boards have an IFT with integral ceramic filter, and a second IFT that gets connected to the other end of the ceramic filter. This helps with the LC selectivity reducing the ceramic spurs, and does the impedance transformation for that end of the ceramic filter.

A really foxy designer would have the two ceramic filters made slightly different so that poles did not line up, and he could have a Chebyshev response with more, but smaller passband ripples. A really foxy accountant would want the parts all the same so he could run off with the extra profit, and the customer would just have to live with the bigger ripples.

So, although these filters are, strictly speaking, mechanical, they're usually called ceramic filters and the mechanical word tends to get saved for the Collins and Kokusai parts (much more expensive) so calling them mechanical isn't actually wrong, but it is the marketing department trying it on a bit.

David

[Bazz4CQJ]

It's interesting that the Toko datasheet covers all the bases by calling them "ceramic mechanical" filters.

My impression is that they were inexpensive, and though they didn't compete with more sophisticated components, they were value for money and certainly worth having in budget equipment?

B

[Radio Wrangler]

They worked well, and their real cost to manufacturers were pennies. Good parts that filled a real need, they enabled cheap receivers to have quite good selectivity.

There are very few FM tuners without ceramic filters in their IF.

I've read of ageing effects in ceramic filters, mainly frequency drift of FM tuner types so matched pairs don't track in ageing and responses get ruined.

David

[radiozero]

Hi. OK, what I'm getting is that the actual ceramic filter inside the can will look a bit like the filters in the pdf posted in post #9. What bugs me about that though, is that if you look at the pictures I posted in post #6, you see that both cans seem to have tuneable slugs. I'm a bit confused. One of the cans in the pair is for matching and I guess maybe contains a simple resonant circuit. Actually a transformer.

[G6Tanuki]

Yes the tunable slug part is to match the impedance/stray-capacitances of the circuit used to drive the ceramic-part.

"Classic" Collins-type mechanical-filters also needed their source-and sink-circuits to be matched by some form of adjustable tuned-circuit.

[this circuit also served to isolate the mechanical transducer from any DC component which would have created an undesirable additional magnetic bias to the transducers].

[Radio Wrangler]

The ceramic filter has two ports; an input and an output. Both need matching, and the matching circuits need to be independent of each other and screened from each other or else unwanted signals off of the wanted channel can leak past the filter by stray field coupling.

So one can carries one matching transformer with a resonating capacitor and a tuning slug. The other can is pretty much the same, but in a larger can for the ceramic filter to fit inside with the transformer.

DC can also create fixed stress in piezo-electric parts. Most people see them as capacitive and think they are self-blocking of DC. As far as external circuitry is concerned, they are, but once you think about internal stresses and the gamut of signals, then you provide DC blocks to keep them centred-up.

David

[radiozero]

The larger can connects to the anode circuit of a valve. So I'm assuming the transformer (the smaller can -apparantly) connects to valve grids. I attach a schematic. In theory, the set up (i.e. the two cans that constitute MF1 & MF2) should be matched at input and output.

[radiozero]

I've not noticed it before, but in post #1, the filter drawing does indicate two matching transformers. One is explicitly shown for the input. For the output the drawing says "To 800 Ohm matching transformer". That's your two matching transformers right there. If I've not made an error, the input side goes to a valve anode circuit, whilst the output side goes to a valve control grid input circuit. Of course, the input must pass the DC needed to get DC to the anode of the valve.

[Bazz4CQJ]

I spent a little time search Google to see if I could find any info on other valve receivers using ceramic filters but didn't manage to find anything. I guess one technology was just coming in as the other went out.

If you find anything more on this subject, please do post the info.

B

[radiozero]

In the 1970 issue of the ARRL handbook, you see nothing about ceramic filters. If I find out anything, particularly about the specific filters in the JR-500S, I'll pipe up. Until then I don't think I have any more comments. Trio 9R-59DS might use ceramic filters IIRC.

[radiozero]

Just one more thing. Finding information on the filter itself, is more likely forthcoming from those who collect TRIO radios. More so if they are Japanese. I've not found yet a TRIO collectors group / club.