FT-480R woes

[SiriusHardware]

Found this info in another thread for a possible bespoke crystal supplier:

voyjako(at)gmail.com or vincentvoy(at)hotmail.co.uk

Vincent Jakomin IQ Electronic design.

He may be the individual alluded to in post #19. (Joachim / Jakomin?)

[Terry_VK5TM]

Last reference I can find was in 2018, hopefully still going.

I-Q Electronic Design
69 Angus Close
Chessington
Surrey KT9 2BN

Tel: 0181-391-0545

[ex seismic]

That is the chap I was thinking of.

Gordon

[G0HZU_JMR]

I think to begin to reverse engineer the crystal parameters you would need to know the inductance of L09.

From this I think the Cm and Lm of the crystal can be approximated by doing a basic circuit analysis. The missing info is the inductance of L09.
Presumably it is a small fixed inductor. Is this intact and can it be measured without damaging it?

[Jon_G4MDC]

Thanks to David RW for some interesting clarification about overtone VXOs.
Very pertinent to the thread and it explains much that I had wondered about for a long time.

Maybe of some further relevance, I once worked on some 29MHz FSK transmitters which put two identical crystals in parallel, supposedly to increase the pulling range. I can't be sure whether they were overtone or fundamental. I always supposed they were 3rd overtone but now I begin to doubt it.

It definitely seemed that a pair would really pull further than one.

[murphyv310]

Originally Posted by Terry_VK5TM

Last reference I can find was in 2018, hopefully still going.

I-Q Electronic Design
69 Angus Close
Chessington
Surrey KT9 2BN

Tel: 0181-391-0545

Brilliant.
Thank you.
I'll call after the holidays.

[G0HZU_JMR]

I tried doing a negative resistance analysis of the circuit and I think L09 could be something like 700nH. This would make sense if the aim was to starve the circuit of negative resistance up at the 5th overtone with the RL network at the base of the 2SC1674L. The circuit is also starved of negative resistance at the fundamental by the 680nH and 47pF in the emitter of the 2SC1674L BJT.

To give the tired crystal one last chance you could try boosting the negative resistance at the desired third overtone by putting another 1k resistor across R27. This would increase the transconductance of the BJT.

Reducing the feedback cap C48 from 22pF to 15pF or even 12pF would also boost the negative resistance. If you are lucky and the crystal ESR has crept up since you took it apart this might be enough to restore oscillation.


These don't look to be regular overtone crystals. I don't think you can just phone up a crystal company and ask for 63.9151MHz 3rd overtone crystal. Even if they offer you a hundred attempts at getting the centre frequency correct this is only part of the answer. This is because you also need to know Lm and Cm because these are tuneable crystal oscillator circuits rather than fixed frequency and Lm and Cm dictate the ease of pulling. I've had a go at calculating Lm and Cm based on the overall circuit and the desired pulling range but I'm not sure I'd want to spend money ordering a crystal based on this. I think it could take several attempts to get it right in terms of centre frequency and correct pulling range.

[murphyv310]

Hi.
I took the crystal out and put it in another circuit and also a counter that has provision for testing crystals. The result was nothing. I decided to reopen the crystal and I discovered it was in two parts, no doubt caused by heat. So that idea is dead in the water.
I'll leave it now unless the chap further up the thread can help or I land another radio or PLL board. I did think at one point of using one radio for mobile use but I also have to get a PA as well as this failed in the first radio, so time now I think to give up.
Thanks to everyone for the very useful info, its been very informative and a good learning curve.

Perhaps time for the moderating team to close the thread.
Again thanks to all.

[Radio Wrangler]

Well, you still have one FT480 with presumably a functional crystal, and you have a second one for spares. That isn't a bad position to be in. If you came across a third FT480 you'd have the quandary of wanting to get all three going. But a dead and very difficult to make specialised crystal says that one set is going to be a scrapper.... If you come across an FT780, I wouldn't be surprised if the synthesiser had a similar open-loop interpolator in it, possibly on the same frequency.

I'll leave this thread open in the hope that someone discovers it and has a PLL board to donate.

I had a lot of fun with that FT480 and it would be good to know it was on the air again. That brown guck adhesive they used to combat vibration really is evil.

David

[G0HZU_JMR]

One thing to bear in mind... if you look at the diode switching and work out the max/min change in capacitance at the 63.9151MHz crystal it only amounts to about a 1pF change so this crystal doesn't need to be pulled much (5Hz steps, 50Hz total?). It seems to be providing the 10Hz steps for the radio. It appears to be a highly pullable crystal that is only being tickled by a tiny change in capacitance to get the required tuning range.

I suppose as long as an alternative 63.9151MHz crystal can be encouraged to oscillate in the right place the accuracy of the 10Hz steps probably doesn't matter much. It is possible to define Lm and Cm for this crystal based on the mid settings of the trimmer caps and the required pulling range. This circuit will be much more forgiving of spec errors for the crystal especially if changes to the padder caps C46 and C47 are allowed. I'm happy to show how to calculate the crystal parameters but it would really need to be reviewed by a few people. Errors and typos can easily creep in.

I'm not sure how much it costs to get a custom crystal cut like this but I'd imagine it will be several tens of pounds. So mistakes will be costly if the padder caps can't compensate for this.

By contrast, the 64.595MHz oscillator is the one to leave alone for sure. It has to pull 5kHz if I'm reading the PLL block diagram correctly. This crystal would be much harder to clone successfully as this circuit isn't as flexible/tolerable when it comes to crystal specs. I'd leave this crystal well alone as long as it is working.

[murphyv310]

Thanks both David and Jeremy.
When it comes to crystals and their operation etc it's one field of electronics I'm not massively clued up on so I'll hand it too the experts.
The radio I'm using is the one from David with the first radios PLL board fitted, all is well on FM but it's a little wobbly on SSB, I'll need to check the settings on the PLL again as it did drift initially when I removed the gunk glue from that one. The bad areas of the board have been protected with a high quality and expensive protectant that stops corrosion and is an insulator, lanolin based.
So yes the first one will be used for spares.
There are three radios in the series FT-480R, FT-680R & FT-780R, 2 meter, 6 meter & 70 cm. All are very similar but have differing crystals and the 70cm version is more complex, very rare beasts here in the UK.

[murphyv310]

Hi.
Some further progress.
I received a message from Vincent Jakomin regarding the crystal. In the message it was quite clear that he was well aware of the crystal that I am looking for. The price of the crystal is cheaper than I have been quoted for by a considerable margin. ETA is between 8 & 10 Weeks.
I'll be ordering this later on today or tomorrow. It's well worth the gamble and I also have a suitable fix for the PA stage.
Believe it or not the output from the PA Driver transistor is between 150mw & 200mw @50 ohms, and I quickly connected the in to out of the PA (PA disconnected of course) and was able to make a 12 mile QSO on FM & a 20 mile QSO on SSB prior to the PLL problems...... all good QRP fun!!!!!!

[Radio Wrangler]

Originally Posted by murphyv310

The price of the crystal is cheaper than I have been quoted for by a considerable margin. ETA is between 8 & 10 Weeks.

That's good.

There is a reason for that ETA. It's built into the entire quartz industry and there isn't any way round it.

Once quartz has been ground and polished to the wanted frequency it gets metal vapour-deposited under vacuum onto its faces.

The grinding process is rather traumatic seen on the crystal lattice molecular scale. Bits of the lattice are ripped away. This leaves a surface of torn bonds and some atoms a bit floaty and unsupported. Over time bonds reattach to wherever they can and flaws that were created in the lattice by tearing partially heal.

This changes the frequency of the crystal.

It could go up after manufacture, it could go down. You have to live with it.

It does an asymptotic approach to a long-term final value. As a rule of thumb, a crystal does some amount of ageing in its first day. It does the same amount of ageing but takes the next 10 days after that, and in the next 100 days, the amount of ageing is the same again.

So the rate of ageing decelerates, thank heavens.

If someone who knows how to specify crystals comes along and specs a crystal as much as crystal manufacturers prefer (They really don't like you spec'ing motional parameters - they don't have good control of those, but they're happy with frequency specs) they will put an ageing spec on it... something like ageing by less than xx ppm over yy years.

How does the crystal manufacturer handle this?

Simples, the make the crystal and do a good enough finish polish and pick a holder suited to the ageing spec, then they just keep it long enough for the worst of the ageing to have happened before the ship it!

Consequently, for typical processing etc, they will keep your crystal for several weeks. Hence the ETA. If anyone promises faster service, think of this and go elsewhere.

For good ageing spec crystals, the ageing slope can be accelerated by temperature cycling to get the worst over with more quickly.

Soldered holders are not good. Tiny balls of condensed solder and flux float around inside leading to both steady ageing and occasionally sudden jumps in frequency. Not good.

Cold welded cases lack the crap inside and you can reckon on ageing typically five times better.

Glass cases are better still.

Super posh crystals go in valve envelopes and may even have a getter. Great ageing, and the vacuum improves Q a lot. BUT BEWARE OF LOW HEAT DISSIPATION CAPABILITY.

I've been designing critical spec crystal oscillators for ober 40 years, and I've made many of the possible mistakes. When I wrote the oscillators/synthesisers chapter for the ARRL handbook, I included a fair bit on what makes good crystal oscillators. It's amazing how much isn't general knowledge, and how many fallacies there are floating around. Quartz crystals are really amazing, but you have to take some precautions to keep them that way.

David

[murphyv310]

Hi
Well I have now two working FT-480R radios.
I obtained another rather tatty one on the outside but in immaculate condition inside with no tarnishing on any of the plating and absolutely no sign of corrosion anywhere. The PLL board was untouched and the cores were still covered with wax.
I discovered it was a "C" model with the USA 2 meter frequencies. The front panel was rather rough looking so I swapped out the fronts including the PLL control panel. This has now given me a good working radio and the normal UK 2 meter band. The scrap set requires a PA Module and crystal. On reassembly of the donor set the zink based threaded portion of the mode switch sheared. I did manage to repair it though with JB Weld Epoxy so it'll be fully hardened tomorrow. If I decide to sort it with a crystal I'll consider using this as a QRP radio, I did try this with only 150 mw from the driver transistor and was able to have a 12 mile QSO on FM and on SSB was received at least double the distance, not sure though as I'll bother yet.
So at least what I set out to do in the first place has worked

[murphyv310]

Hi.
The repair on the mode switch was successful.
So I think I'll still get the crystal. A change of the driver transistor to a different type that has higher FT, HFE & collector current will double the 200mw drive at least. So it might be a little bit of fun on QRP.
Here are some pictures of the radios which all now have very nicely black painted cases.

[G6Tanuki]

That sounds like a sensibly-pragmatic approach to dealing with the failed and made-from-unobtainium Mitsubishi 'thick-film' PA modules.

You could easily build a little linear-PA module using a "Cable-TV" transistor [2N4427?] to lift the RF to a couple of Watts - then feed it into one of the linear-amps that were popular in the 70s/80s for boosting the output of the likes of the FT290.

Here, I've got a "Mirage B23A" which I bought for £10 at a local hamfest; it works in Class-AB1 [so is good for both FM and AM/SSB] and has enough forward-bias so it draws 400mA without RF drive; fed with 1.5 Watts from my Yaesu-Standard VX150 handheld it churns out a reliable 30 Watts of RF without getting unhappily-warm and run into a collinear I can access repeaters 80 miles away.

[murphyv310]

Hi G6.
The 2N4427 is exactly the transistor I've ordered, the radio still transmits even though the PLL isn't locked but with careful adjustment it's pretty well spot on, on the full and medium steps. Having QSOs on 2 meters into a slim Jim on 2 meters with a good few miles is very rewarding and opening up the local as well by using just 200 milliwatts.
Thank you for the encouragement.
A happy new year to you.

[Radio Wrangler]

If you carefully saw the plastic off of the Mitsubishi module, you'll likely find the alumina thick film substrate cracked. You can fix these with tiny wire bridges. Cracked resistors, scrape them away carefully and solder an SMT one over the top.

These Japanese sets have a three layer sandwich all attached together.... ceramic substrate on copper heat spreader bolted to aluminium or pot-metal heatsink. All with different coefficients of expansion, so the damned things bend like bananas if they get hot. Designed to fail. No temp warning, no temp cutout. Just a bill.

My FT480, I think is original. I've never had to go into that area. Wish I could say the same about the TM-V7E which had a new module fairly early on, then a repair to that when it went after the things became unobtainium.

David

[SiriusHardware]

A question: I love the idea of DIY repairs to the unobtainable PA modules and have seen this suggested before elsewhere on this forum. I'm curious to know what the composition of the ceramic substrate mentioned above is likely to be.

A lot of RF power components back in the 70s / 80s used Beryllium Oxide as the insulating ceramic base or layer - how likely is it that these Mitsubishi modules contain Beryllium Oxide?

Although perfectly safe in its ceramic slab format it is reportedly extremely dangerous / injurious to health in the form of dust, as it might be when sawn or filed.

[Radio Wrangler]

The thick film substrate was Alumina in the ones I've worked on. The power transistor and driver mount in holes through the alumina. Some of them may be sitting on a tny chunk of BeO so be careful around them. The sircuit wants grounded emitter, but the processing naturally puts collector to substrate.

David