FT101ZD local oscillator crystal problem

[tomwalter]

I'm currently re-awakening a Yaesu FT101ZD Mk3 from many years slumber on my shack shelf. It appears to be working well and I had a very successful QSO yesterday on 40m. Everything was working fine (apart from one dial lamp) until I came back to it after it had cooled down and turning back on resulted in random numbers on the 40m display. All other bands behaved normally and on touching the 40m crystal on the premix local unit (board PB-2153) it sprung into action and changed to the correct display with signals coming through strongly. I've resoldered the crystal and checked it on my GDO and it appeared to be fine. Once activated the crystal behaves ok for the rest of the radio session. This morning however it had reverted to its crazy display antics on 40m upon switching on. Touching the crystal case was enough to get it going again. Any idea what's going on here? Is it a lazy crystal? Can I rejuvenate it? I've heard of baking in the oven etc but don't have any experience of cooking crystals! By the way, has anyone any comments on the accuracy of Yaesu manuals? In my Mk3 manual p42 on Premix Local Unit Alignment, it refers to Pin 1 on the board plugin connector for measuring RF volts; it isn't Pin 1 - that's earth, its Pin 3! They say the RF volts are 300mV; mine are all 10-20mV and all work fine (except the 40m band). I've got a good Boonton RF millivoltmeter with correct probes. That's quite a discrepancy. It doesn't give me much confidence in my amateur struggles! Has anyone a spare 21.4875 Mhz crystal?
Hoping someone can help. PS. Some pics attached.
Tom G4KWL

[frsimen]

It might not be the crystal that is at fault. Before doing anything drastic to the crystal, check the resistors and decoupling capacitors in the area. A low gain transistor could also cause similar symptoms, so consider replacing the 40m band 2SC372. The 2SC372 has a fairly low collector to base capacitance but something like a BC547 or a 2N3904 ought to be a suitable replacement.

Paula

[G0HZU_JMR]

It might also be worth looking at the setting of the adjustable inductor T03 on the output of the oscillator. This RLC network (100pF, 560R and variable inductor T03) is there to make sure the BJT only produces a sharp/focussed spike of negative resistance (between collector and base) at frequencies in the region of 21.5MHz. Without this RLC network at the collector, the BJT would produce negative resistance here across a very broad bandwidth from LF through into VHF. This would be a bad thing because the crystal could then operate on the wrong overtone mode (if it is an overtone crystal). The RLC network should be designed such that the negative resistance can only be significant in the region of 21.5MHz.

There might not be enough negative resistance to offset the ESR of the crystal at 21.4875MHz if T03 is not at the optimal setting. So, the oscillator may become a bit fussy about starting up.

Obviously, only use a plastic tool to adjust T03. It will be very easy to crack the core of T03 when using a hard metal screwdriver.

[g4wim_tim]

I had this problem on a unit I restored some years ago - over time the crystal became 'lazy' so I increased the gain in the transistor osc stage ever so slightly.
I posted a youtube video about it.
Search for G4WIM on youtube and you should find it along with the fix - assuming yours is the same problem.

[G0HZU_JMR]

Hi Tim
That does look to be the same problem, and it's even on the same band. Your mods will hopefully work for Tom's radio too.

Reducing the 560R resistor at the collector will definitely produce more negative resistance and this will help start up the oscillator. If the ESR of the crystal is slowly going up and up over time, then there will come a point where the existing circuit won't be able to produce enough negative resistance to offset the increased resistive losses in the ageing crystal.

There might also be something else ageing in the oscillator circuit as others have suggested but it does look like it will be possible to get the oscillator running reliably without having to buy another crystal in the near future.

[Cruisin Marine]

The first thing is to ascertain is whether it is oscillating at all. A basic diode probe will indicate that. A second receiver tuned to that 25 MHz frequency will help of course.
Beyond that a freq. counter, scope of adequate bandwidth and speccy analyser will show a much clearer picture of course.

[G0HZU_JMR]

Yes, it could even be jumping across to oscillate on another overtone mode or even somewhere else. The little TinySA would probably be a cheap and useful tool here to remotely sniff the circuit to look to see if it jumps across to somewhere else. A diode probe would be useful to see if it is still oscillating out of band as you say.

I have some very accurate VNA derived 2 port models of quite a few common BJTs and if I select the 2N3904 at 5V and 2mA then the optimal value for the 560R resistor in the collector is about 600 ohms if the aim is to minimise negative resistance at the (undesired) fifth overtone at 35.81 MHz. However, reducing it to 390R gives a useful increase in the negative resistance at 21.4875 MHz and only increases the negative resistance at 35.81 MHz by about 12 ohms.

This assumes this 21.4875 MHz crystal is a third overtone type. It might be a fundamental type although some of the crystals on that board will definitely be overtone types.

[tomwalter]

Thanks everyone - all suggestions are much appreciated. I have adjusted T03 to bring RF volts out to the average of the other bands, about 15mV, but without much effect. Jeremy, you are quite right, Tim's problem is exactly what I have experienced. Your video was really helpful Tim. I'll try to up the gain as you did. Thanks Paula, I'll try the resistors first though. I've got to find the right resistors now - my ¼ watt collection is only marked up with tiny bands and my colour vision is very dodgy! At least I have a good de-soldering tool! I'll get back and report results. Thanks all - really fabulous support!! Regards, Tom.
PS 40m crystal, online auction house asks twice the money for this one compared with rest of crystals. Interesting.

[m0cemdave]

I have a spare 40m crystal from a scrap board (early version PB1965, but the xtal is the same). I can send it to you for the cost of postage. With any luck, it will be better than the one you have.
PM me if interested.

[tomwalter]

Thanks Dave - pm sent
Tom

[tomwalter]

Upping the gain in the 40m band oscillator did the trick. Crystal now behaving well from cold. A spare one might still be useful for the future though, just in case. Many thanks Tim for your spot on diagnosis and solution. Also thanks Paula - you were quite right. Also many thanks Dave for your kind offer - much appreciated! This forum is simply brilliant!!
Best wishes,
Tom

[g4wim_tim]

Hi Tom

glad you got it sorted and the youtube video was relevant to your problem.

Regards Tim

[m0cemdave]

Your spare should arrive in the next day or so. Given that your existing xtal is known to be dodgy, you should definitely keep it as a backup!

[tomwalter]

You are an absolute star Dave!
Very many thanks!!
Best wishes,
Tom