Harvard H407 CB RX audio problem

[SiriusHardware]

Real life sometimes gets in the way here as well.

The traditional way to go hunting for a fault with freezer spray is to use it on the circuit when it is ON, and the fault is present, to see if cooling a suspected component instantly fixes the fault condition while you are observing it.

However, this applies strictly to low voltage circuits, or strictly to the low voltage side of any unit with mixed voltages inside.

Your 407 has an integral mains PSU but there shouldn't be any mains voltages on the main radio PCB which is the same one used in 12V mobile radios, so it's OK to use freezer on the radio PCB when it is powered. Just keep the spray well away from any mains voltage in the unit.

I tried emailing the original author of the redrawn diagram to let him know we'd noticed a few errors but unfortunately the email address given is defunct.

[Damo666]

Sirius,

I managed to check the voltage on Q19 Emitter & it's at 8.33V at all times, so only 0.27V under the figure indicated on the redrawn schematic (8.6V).

I left the rig on for about 15 min's until audio became somewhat distorted, albeit the incoming signals weren't strong enough to cause major distortion - but a quick blast of the freezer spray on the LA1230 seemed to improve the audio.

I'll wait until the local regulars with strong signals are on later when audio goes bad & blast the LA1230 with some more freezer spray.

At this stage, I'm guessing that the LA1230 is the prime suspect - but I was under the impression that semiconductors either worked or didn't; IE an "all or nothing" affair?

Could a defective 100uF Capacitor connected to pin 7 of the LA1230 cause this? I did spray this component but I heard no noticeable improvement.

[SiriusHardware]

Semiconductors, including linear ICs like the LA1230 are quite capable of failing "partly" and often do. You said you have a good stock of capacitors so by all means try changing the electrolytics closely associated with U3, starting with the one associated with pin 7. It will cost you nothing to try and may even fix the fault. If that doesn't fix it then I personally would change the LA1230 - this is where it would have been handy to have a working example to borrow U3 from to prove things one way or the other before ordering a replacement.

[Damo666]

Thanks for clearing that up on the Semiconductors confusion.

I'll replace some Capacitors in the area, specifically that 100uF one - and if that doesn't fix it, I'll get that LA1231 from Knight's as there's one or two other things I'd like to buy from his shop.

[SiriusHardware]

Originally Posted by Damo666

Thanks for clearing that up on the Semiconductors confusion.

If you look at the data sheet for a linear IC it will typically show you a block diagram and even something like an internal circuit diagram, from which you can see that a typical IC is a box full of interwired parts, any one of which can go wrong. What happens when they do can vary greatly depending on exactly which part goes bad, and in what way. It might, as you expected, kill the operation of the IC entirely, or it may stop just one part of its function working, or may degrade the performance of the IC without stopping it from working altogether.

When it comes to ordering another IC, you need to be reasonably happy in your own mind that that is probably the problem, so I'm assuming you did carry out the 'strong signal' test you said you were going to and found that there was reason to believe that U3 or something around it was causing your fault.

If so, and if changing those capacitors doesn't change anything then, with the replacement IC only costing £4 I think it's worth gambling that much on a replacement. If the replacement IC cost £25, that would be very different and I would be looking to explore every other possible avenue first.

I mentioned back in the 'All about CB' thread that I had clubbed together with a friend to buy a few CB retro spares from a company over in Ireland - it's a pity we didn't know someone might need an LA1230 at the time as I noticed they did list them (LA1230, not LA1231) as available. Unfortunately, as said then, there is a minimum order value of £20 from that company so it wouldn't be economic just to order an LA1230 from them. I haven't looked into it but I'm assuming that the 1231 is an improved version of the 1230 - maybe the LA1230 proved to be quite unreliable?

[Damo666]

Ever since doing the voltage checks, I've kept the lid off the rig & the problem seems to have virtually disappeared. I kept the rig on for a good 30 min's last night, went for a bath, & when I returned there was barely, if any degradation in audio with strong signals. There is good ventilation in this room, though, so I can only guess it's keeping the offending part of the circuit cool and preventing the problem.

I'll be changing the Capacitors in close proximity to the LA1230 later today, and I'll plonk the lid back on without screws and await the issue returning. Then immediately hit the IC with the spray.

Re' the LA1230, I've noticed quite a few places sell them cheap. I think the cheapest source for one was £1.95, but most of the places are asking £5 or thereabouts for delivery - so I hear you on the "group buy" thing.

[SiriusHardware]

I know for a fact that I have the sad remains of a Mustang 1000 or 2000 (so, same chassis as your 407) in the loft, it came to me as a non-working hulk minus casings and probably a whole lot of other parts, and because I was always more of a fan of classic Cybernets, I never gave a thought to fixing it (never liked the looks of the Mustangs either, to be honest).

The problem is that I would (a) have to find it and (b) repair it enough to get the receiver working in order to find out whether the LA1230 is working - that's if it is even still in there, otherwise I would have offered to send it down to you as a parts donor - but if you think you can get an LA1230 for ~£2 then surely that has to be worth a try. Where would those ones be coming from? China?

[Damo666]

Not sure where I found the IC for a couple of quid as I can't find it now, but with the postage being about £5 It's no special offer.

I've put things on hold with this rig just for the time being, but I've put "LA1230" into my eBay followed searches - so when one is listed at a decent price I'll buy it.

I'm gonna change the caps in it tomorrow, so you never know, it might fix the issue.

[Biggles]

From my experience I would go for the obvious things like electrolytic capacitors and maybe any tuning components if the discriminator uses them. I have changed chips on more than one occasion confident that the problem will be solved and have been disappointed when it wasn't! The freezer spray/hair drier combination is a handy way of tracking down thermal faults.
Alan.

[Damo666]

For what it's worth, I've used a thermal laser thermometer on various components in the discriminator stage but nothing really increases in temperature.
I've used freezer spray on some of the Caps but it didn't make a difference, so I'll replace a few of the electrolytic's tomorrow & hope for the best.
The freezer spray cost me £15 and if the IC is at fault it'll be about another £7 all in, so it's turned out quite expensive - but I suppose the spray will last ages & be useful if I endure any similar problems in the future.

[G0HZU_JMR]

The LA1230 should show a slight rise in temperature according to the datasheet. The chip is powered from the PSU voltage (so 12-13V typical) and the datasheet says 20-23mA typical current.

So that means just over 250mW Pdiss. So I'd expect it to run warm enough that you could feel warmth with a light touch from a finger tip.

Note also that these radios won't sound as soft/natural as a typical Uniden radio on FM so you will have to accept some degree of harshness and distortion. But it shouldn't get worse as the radio warms up.

[SiriusHardware]

Wow, £15 is at the very high end for a can of freezer. That must have been from somewhere like CPC or RS? Sorry you ended up paying that much, but as you said, it is a generally useful tool to have.

I'm still not sure that you are sure that cooling the discriminator area makes a lot of difference. I wish we could be sure that the problem is not actually overall frequency drift, but since you don't have access to a frequency meter which covers the required range, I haven't tried to pursue that so far.

If you can lay hands on a synthesised HF receiver which can cover 17 to 18 MHz in very fine frequency steps and has SSB, let us know.

[Damo666]

I'm sorry about the late reply. I logged in earlier today & was meant to respond, but something got in the way.

I bought the Servisol freezer spray from eBay, and I bought the largest I could find. 400mL I think it is. It's a pretty hefty sized can. The price seemed about average compared to other listings. There were other brands
that were somewhat cheaper, but there wasn't much in it.

I still haven't got around to swapping out a Capacitor or two, so I apologise about that, and it often seems like there's not enough hours in the day - but I'll see to it sometime this weekend.

With the discriminator, I think I'm going to resort to using an hairdryer in the area suspected - because leaving the case off the rig seems to be preventing the problem. It's anyone's guess when a strong signal is due as well, so this doesn't help things as it means sitting around until one comes in, if at all.

I've got a receiver that covers the frequencies you mention, but there's no SSB option & the minimum step is 1Khz, so that's surely a no go.

[SiriusHardware]

Unfortunately that's not quite up to the job. It was just an idea, an alternative way of checking that the VCO frequency wasn't drifting over time.

To heat a specific suspect component for the purposes of thermal fault finding, just hold the flat face of the tip of a soldering iron 1-2mm or so away from the component in question for a while.

If you are going to carry on collecting radios which may or may not turn out to have minor faults, keep an eye out for another old frequency meter which can work down to as low as 9Mhz (so you can use it to check / adjust crystal oscillators commonly found in CB radios) and obviously at least up to 30Mhz at the upper end. It also needs to have a resolution of at least 10Hz, ideally 1Hz, at 27MHz so a frequency of 27.000Mhz would be displayed by the meter as 27.000.00 or 27.000.000.

[G0HZU_JMR]

I managed to find my old Fidelity 2001FM which uses the same PCB inside and I did some tests on frequency stability whilst the footie was on. I used a signal/spectrum analyser on a very narrow span to watch the warmup drift of the reference crystal inside the radio. I used a near field probe to sniff the signal so I didn't have to open the case of the radio and I could watch it drift as it warmed up over about 40minutes on receive only. Hopefully this will give typical figures for the drift of the reference crystal.

It started out at 10.240478MHz from cold and eventually drifted to 10.240486MHz. I can remember setting this radio on frequency a few years ago so I expected it to be close and this result was impressive indeed. I think this crystal typically needs to be set to be within +/-50Hz and this radio is still well in spec here.

I also noted that the ref frequency shifted down slightly when on Tx. It drops about 35-40Hz but that's no big deal. A 40Hz error at the reference crystal would mean less than 150Hz error up at 27MHz. Easily good enough for NBFM use.

[G0HZU_JMR]

I also had a look at the FM chip with a thermal camera after leaving it for >30 minutes with the cover on. See below for the image. The hotspot marker in the centre of the LA1230 chip is 59.6degC and the background temperature inside the radio is about 30degC. So this chip does run quite warm, the centre of the chip is approx. 30 degrees hotter than the cooler parts of the radio on receive.

[Biggles]

That's surprising. After all, it's just a low level signal processing chip. I wouldn't even consider that normal for an audio amp chip, at least when under no signal conditions.

[SiriusHardware]

If the IC runs warm in normal operation that makes it a little bit more likely to fail eventually. Like Biggles, I can't imagine what it must be doing which would need so much current.

However, I have an old Ferrograph tuner which uses the CA3089, of which the LA1230 is apparently an advanced derivative, and I can remember that used to run warm to the touch as well.

Jeremy, good work on the drift analysis, but could you also check what happens to the voltage on the LA1230 on pin 7, to see if the voltage slowly declines over time the way it does on Damo's?

I haven't been able to lay hands on the datasheet for the LA1230 but I do have the one for the LA1231, and it appears to be pin-for-pin the same. It does beg the question, why introduce the LA1231 when the LA1230 already exists?

Some sort of problem with the LA1230?

[Damo666]

I'm also surprised the LA1230 runs so hot in normal operation, and as I noticed no major temperature rise on mine, perhaps mine is defective.

I went through my component drawers looking for a 100uF Cap' last night to start the ball rolling with replacements, but unfortunately I don't have that value. I have nearly all other values though. Sods law, isn't it?

Due to a severe lack of test equipment, I'm just going to bite the bullet & order a LA1230/1231 and the necessary Capacitors I don't have around the IC that are used in this rig.

Can anyone suggest a decent entry level frequency counter to look out for, and that way I can create a "followed search" on eBay & be instantly notified if something is listed?

Many thanks for everybody's continued support, and I'm sorry that this is such a long winded process.

[Techman]

I honestly think you'll verify if it's that chip by leaving the covers on until the fault occurs and then remove the covers and very quickly spray it with freezer to see if the fault instantly clears. I wouldn't risk wasting money on a chip just yet.