Harvard H407 CB RX audio problem

[Damo666]

In addition to other things, I recently acquired a Harvard 407 home base CB which uses the PCMA002F board - same as the Fidelity 2001, Amstrad 900, Harvard 402, etc'..

Transmit is fine, however, on Receive, strong signals are somewhat muffled & distorted as if they're off frequency or slight AM is involved. Weak signals sound fine, although not perfect, & are fully legible.

When I first turn the rig on, Receive audio on strong signals is fine for around 10 minutes, but after that, things start to degrade.

Does anybody know where I should start looking for the fault(s)?


Here is the schematic - http://www.cbtricks.com/radios/amstrad/cb901/index.htm

[Biggles]

The first thing I would try is adjusting the frequency of a signal gen slightly above and below the correct channel frequency while putting in a high level of RF (-47dBm/1mV) to see if the signal improves at all off frequency. You would need a good signal generator for this though. Normally high signal strength distortion is down to discriminator problems. Assuming it hasn't been got at, the quadrature tank coil may be a little out of adjustment, if the IF chip uses one. If it uses a PLL discriminator chip for audio demod there may be a trim pot to get it back on tune. Take care not to adjust the discriminator quadrature coil too far in as I have seen cases of the core damaging the windings. Common on Pye 496 etc.

[SiriusHardware]

Well - before we go in that deep I think we need to ask Damo which of the following items of tools and test gear he has, or has access to, in no particular order of importance. We know (from previous postings) that Damo does have a meter, so that isn't in the list. We need to tailor our suggestions based on things we know he can actually try.

So, Damo, which of these do you have?

-Desoldering equipment (hand desolder pump, solder wick, electric desolder tool)

-Freezer spray

-Audio signal generator

-RF signal generator capable of working at (Ideally) 27Mhz but certainly at 10.7Mhz and 455Khz.

-Frequency meter

-Oscilloscope

-Signal Injector

-Signal Tracer or simple audio amplifer with volume control and speaker

-Crystal earphone / earpiece

-A known working radio with the same chassis as this one, for comparative measurements, borrowing parts from (temporarily) and 'signal jumping' (I'll explain that one later).

I would caution against adjusting anything just now unless there is good reason to presume that the set has been maladjusted, and instead assume that the symptoms are those of a faulty, but otherwise correctly tuned radio which was working fine before the fault developed.

[Damo666]

It definitely sounds like this probkem isn't going to be quite as simple to fix as the last!

Equipment wise I'm limited to; A soldering iron, decent desoldering wick, a DMM & analogue meter, Handheld frequency counter only covering 50 Mhz - 2.4Ghz, Crystal Earpieces, and a good inventory of components.

I have a little LM386 amplifier on route from eBay.

I think I'll be able to temporarily borrow a Fidelity 2001 which uses the same board, and if not, I'm sure the guy who owns it can convey voltage measurements at various points. I've known the owner of the Fidelity for many years on & off.

In terms of a signal source, I only have CB's to hand, but I do have a dummy load & attenuator to feed the signal into to act as a weak source.

As far as I'm aware - the rig itself hasn't been got at.

[SiriusHardware]

Ok, I thought you probably wouldn't have a purpose-made signal generator. It's good that you do have a dummy load, I forgot to put that on the list.

There are various possibilities here, but let's start with some simple stuff. I know Jeremy loves this chassis, so I'm sure he'll have some good suggestions.

Turn the radio on and leave it on for long enough to let the fault develop / come on. Then,

1) Measure the receiver supply voltage on Q19 emitter. It should be about 8.6V.

2) If (and only if) this radio has a CB / PA switch, plug an extension speaker into the PA socket (not the 'EXT SPKR' socket), switch the radio from 'CB' to 'PA', key the microphone and either talk into the microphone or play music into it with the microphone held next to the speaker of a small audio source, such as a transistor radio, set to a fairly quiet volume. The audio coming out of the extension speaker (the one plugged into the 'PA' socket) should sound clear and undistorted. (This is to check out whether the audio power amplifier is working OK). If the audio turns out to be distorted in PA mode as well, then you would be looking at a faulty LA4422 IC, or one of the components associated with it.

That redrawn diagram is a very nice job, but it has the Audio PA IC number as 'LC4422' and the IC number as IC1, whereas my 'original' Amstrad diagram has the audio PA IC as U4, LA4422.

[G0HZU_JMR]

I know Jeremy loves this chassis, so I'm sure he'll have some good suggestions.

I'm not sure love is the right word, but this chassis was a decent allrounder in my opinion. It wasn't the best at anything but it was the one I used the most. It wasn't quite as harsh sounding on receive as the 134 chassis, the squelch was a bit better than the 134 chassis, it had a fairly decent s meter circuit and the overload performance was above average. The receiver had a bit too much gain in my opinion and it was a very sensitive receiver as well. I don't think a 27MHz CB needed to be this sensitive when measured on test gear.

The fault could be caused by quite a few things. It could be a leaky cap somewhere in the AF stages or it could be the AF chip or the LA1230 FM chip or one of the components around it.

I agree that the CB/PA test would be a good one if there is one on a homebase and if the CB/PA health check of the AF stage passes and the rx voltage test passes I'd check the voltage stability on pin 7 of the LA1230 FM chip (chip U3) and the health of the 100uF electrolytic cap here. Also check the voltage on pin 11 of this chip to see if it stays stable over the 10 minutes.

[SiriusHardware]

Originally Posted by G0HZU_JMR

...or the LA1230 FM chip or one of the components around it.

That was the next thing, but if the chip, how best to prove that without changing the chip? I'm thinking that some freezer spray might be a worthwhile addition to the toolkit, given the apparently thermal or time-sensitive nature of the fault. Freezer spray might also pinpoint an ailing capacitor.

Knights Electrocom list the LA1231 which they offer as a (presumably drop-in) replacement for the LA1230, for about £4. Possibly less than the price of a can of freezer spray? They also have the Audio PA for about the same price.

I don't know about you but I'm inclined to feel that this is NOT a case of the receiver frequency (as in the synth) drifting off frequency because any such fault ought to affect TX as well as RX. With no frequency meter available for the range of frequencies in question, that would be hard to verify in any case. An HF receiver equipped with SSB mode so you could tune for zero-beat (in other words, any typical amateur HF receiver or transceiver) could be used to monitor the VCO frequency to see if it is changing, I suppose.

[G0HZU_JMR]

Yes, it could be a tricky one to track down especially if there is no test gear on hand. Freezer spray can be very useful and could be worth a try.

One solid piece of advice I can offer with this chassis is to be VERY careful if you reach the stage where you might want to adjust one of the IF transformers. The ones in the IF strip with the big slot inside the can are very brittle and they also tend to be gummy/sticky and it is really easy to break the ferrite cup/slug into two pieces as you turn the trimtool to try and get the cup/slug to start rotating.

I know this because I did this myself on my Fidelity 2001 way back in about 1982. I broke one of the ferrite adjusters! Amazingly, I was able to get a replacement IF transformer via Fidelity UK. I assume someone got one sent over from Japan? I contacted them via their contact details and I think these details were in the manual or maybe on the box. But either way, they sent me a replacement IF transformer in the post and I took great care in fitting it and adjusting it. It was heart in mouth time in case I broke it again! But if you take care and use the right trimtool I found that these can be adjusted OK and I went on to align loads of these radios over the years

[SiriusHardware]

With no 'proper' RF signal source (Rock steady: Constantly modulated: Adjustable in fine steps from zero up to very weak) I would put trying adjustments at the very back of the queue. It works OK for the first ten minutes, so the adjustment is correct during that time.

Once any adjustments are turned away from what must be presumed, for the time being, to be their factory settings, it will be difficult to put them back exactly where they should be without a decent signal source and Damo doesn't have one, so this should be a last resort.

Some companies were pretty good for spares including RF inductors and cans: Tandy were good like that as well.

[G0HZU_JMR]

I would put trying adjustments at the very back of the queue.

Agreed. But I thought it was worth putting in an early warning about the very brittle slugs. Anyone could have popped up and advised him to try a quick diagnostic tweak here and there. These IF cans look very easy to tweak with the big inviting driver slot in the top but they are very fragile and definitely a trap for the unwary

[Damo666]

Thanks for the great input guys.

I think there's a PA option on the rig, but I won't know until I get to see it later today.

I'll report back with answers to all the suggestions given as soon as I can.

Should it be a faulty Electrolytic Cap, I have a load of those - and I also have a couple of sets of trim tools should the need to use them at a later date arise.

[Techman]

I think freezer spray will very quickly diagnose the fault with this one

[SiriusHardware]

Somewhat surprisingly for a Homebase, it DOES have a CB/PA switch.

While of limited use in most home base setups, it does give us a nice diagnostic method of testing the audio power amplifier independently from the receiver, so we can eliminate (or implicate) an entire section of the unit quite easily. That's helpful.

Here's an image of the 407 (image #1) for anyone who might be curious to know what we are talking about.

A lot of the radios released on the UK market re-used tooling which had previously been used for models targeting the USA or European markets, as with the Cobra 21X (UK) which was descended from the Cobra 19X (other markets).

While I was looking for an image of the Harvard 407 to see what was on the front panel, I also came across its ancestor, or one of them: The Royce 619 (image #2).

I also found the 407 user manual here: Unfortunately it doesn't seem to include the specific diagram for the set, which is quite unusual.

https://www.manualslib.com/manual/87...vard-H407.html

[Techman]

I've just realised that I think this is the same homebase radio that someone in the local city to me bought recently from another local without hearing it on the air first. I gather that it's in mint condition and in its original box, but without its original microphone. The first thing he did when he first connected it up was to plug his Ham-Master base mic into it and it sounded absolutely dreadful. He was only 'readable' with the gain on the mic turned right down to nearly off. It was much better when he plugged the non-original hand mic into it, but it was a big let down from the radio that he'd been using before. He bought this radio on its looks and condition, rather than its functional ability. He eventually stopped using it after we'd all complained about the way it sounded. As I said to him, it sounded like an Amstrad 901 on a very bad day. I did suggest to him that someone had perhaps 'wound' something up inside to make it sound 'louder' sometime in its previous life.

I perhaps shouldn't say all of the above seeing as Damo now has one of these radios, but I was never a fan of the Amstrad type sets on transmit - strangely, the receive was actually quite good on them. The homebase version that you have certainly looks quite smart and I can see why the person that I know, bought his example for its looks.

[Damo666]

I quite like the Amstrad boarded rigs, particularly the Harvard 402 whose S meter glows red on TX. The RX sensitivity is a cut above the rest from what I remember compared to a good few of the various sets I had back in the day.

Sirius; I managed to crack open the Harvard 020 for you today, and the telescopic antenna measures exactly 74CM from the threaded base to the tip.

Re' the Harvard 407, I didn't get around to opening it up & pursuing any voltage checks, but I did manage to test the PA function with an external speaker at initial switch on, and also ~10 min's later when the RX audio degrades. The PA audio reproduction is perfect at both switch on & later with no noticeable difference to my good hearing. It was extremely clear in both cases, with volume set from minimum to full pelt.

Pardon my very limited knowledge as usual, but I'm guessing this narrows the suspect area down to the discriminator stage?

I'll get around to doing some voltage checks tomorrow, and I've also ordered some freezer spray, so that should be with me in the next day or two.

One last thing; When the RX audio degrades after around ten minutes, I turned the rig off for about 5 minutes, then when I turned it back on again the RX audio reverted to poor audio after about 3 or 4 min's - so I'm thinking it's an heat issue around some area - hence me ordering some freezer spray.

Also, the S meter is incredibly generous on this thing, showing S8 with no signal whereas my Pama GX-19 shows about S3 - but, I do recall these PCMA002F rigs being this way from when I owned them many years ago. Either way, it's easy enough to amend this.

[SiriusHardware]

Damo, thanks for measuring the aerial on the 020 - I've been trying to find out that exact info for decades, so that is appreciated.

You seem to have proved that the audio PA stage on the 407 is OK, so, you're right, the next area to look at is the discriminator IC U3 (LA1230) and its associated components.

If you had a scope we'd ask you to look at the audio coming out of the LA1230 on pin 6 when the fault condition is present - you don't have a scope, but you can try connecting your crystal earpiece between pin 6 and 0V or between C95 positive and 0V and using it to listen to the audio at that point to see if it is distorted or clear when the fault condition is present. If your LM386 amp arrives, you can try using that to listen to the audio at those points as well.

As you've ordered some freezer spray, we'll give that time to arrive and see how you get on with it.

The most obvious target to hit with the spray will be the LA1230 IC, but if it doesn't seem to be that you could try cooling a few of the capacitors immediately around that IC, such as C86, C89 and C95.

Re: The meter, I saw something recently - maybe Jeremy said it, but I can't find it - that said that radios with this chassis which have an analogue S-meter need one, sometimes more diodes in series with the meter output signal to make the meter indicate zero when there is no signal. In the diagram you pointed us to, there is one diode (D13) shown. I've seen a suggestion somewhere that there sometimes needs to be another diode in series with that one to make the meter read sensibly.

[Damo666]

No worries at all on the 020 antenna. In fact, I was just browsing eBay for a possible replacement and it looks like this might be the perfect candidate albeit a bit longer. Just don't extend it fully - http://www.ebay.co.uk/itm/2222935224...%3AMEBIDX%3AIT

Back to the H407, I think I'll hang fire until the freezer spray arrives and that way I can kill two birds with 1 stone; Freeze various components, check the audio, and voltages at various points as requested.

In the meantime, as before, a huge thanks to all again for your educated good input & patience with me on this. It's really very much appreciated.

[Damo666]

The freezer spray arrived yesterday, but unfortunately I didn't have time to report back then.

I took readings as requested, and here's what I found; LA1230 pin 7 at switch on, voltage = 5.89V, after 2 min's = 5.77V, 6 min's = 5.70V, 10 min's = 5.61V, 15 min's = 5.60V.

LA1230 pin 11 at switch on = 12.66V, 2 min's = 12.68V, 6 min's = 12.68V, 10 min's = 12.68V, 15 min's = 12.68V - so the voltage is stable on this pin.

After 15 min's, I turned the rig off & sprayed freezer spray on the LA1230 , turned back on & the voltage on pin 7 was 6.13V.

[SiriusHardware]

OK, can we just ask you to clarify, if you...

-Turn the radio on and allow the fault to appear
-Leave the unit on, with the fault present
-Spray the LA1230 and its associated components

...does that make the RX audio distortion go away?

I don't know if you've used freezer spray in the past, if not, you need to aim a series of short bursts at the component you are trying to check until it is visibly covered in frost.

Thanks for the voltage reading on the supply voltage to the LA1230 (Pin 11) which appears consistent. The slow decline on the voltage on pin 7 could be significant. Could you also do the same for the voltage on Q19 emitter?

[Damo666]

Hi Sirius,

I failed to notice you asking for the voltage on Q19 Emitter previously, so please accept my apologies about me omitting this result. I'll get some readings later today on that part.

I was doing all the aforementioned tests with no antenna connected unfortunately, and when I was about to do the next freezer spray test with an Antenna connected & at a poor audio stage, my baby Son was coming up to bed so I had to vacate the room. Again, I'll get onto this tomorrow as well.

I was comparing the redrawn schematic of this rig with the original & it looks like there are a good few errors, so I'm going to base the location of Q19 from the actual image of the PCMA002F board layout - the yellow image.

By the way, I found that the output Voltage of the internal PSU was set too low at 12.68V, so I turned it up to 13.2V via the preset for what it's worth.