Marconi 2305
Forum rules
Use tags for the type of equipment your topic is about. Include the "repairs" tag, too, when appropriate. If a new tag is needed, request one in the TEAdministration forum.
Use tags for the type of equipment your topic is about. Include the "repairs" tag, too, when appropriate. If a new tag is needed, request one in the TEAdministration forum.
Marconi 2305
I've carried out a repair after the longest time living with a fault.
In this case it must be around 15 years since I bought this useful modulation meter from our "favourite" government surplus ebay seller who can't test it because he doesn't have the knowledge to do so. It arrived dented in every surface because it had been shipped wrapped in a mouldy rug and some cardboard. The rug was probably an antique!
After I straightened the panels it worked perfectly, even to manufacturer's specification. There was one more problem, though, the optional audio distortion board that was fitted would not operate, the computer insisted that the option wasn't fitted (Error 06 when the option was selected).
The option board is detected by grounding one pin on a big edge connector which fits into a small backplane board and routes the signal to an octal tri-state buffer on the computer's I/O and memory board. I took all that apart and checked that the 0V was reaching the buffer many years ago, the buffer also appeared to be working properly in every other respect so I just assumed there was some non-volatile setting that wasn't documented but was capable of disabling the option and used it as-is while I looked for something better.
The alternative arrived after a few years with a nice HP 8901B but I didn't sell the 2305 because I preferred it. Fast forward to this year's Rugby rally and I bought a 2305 for £60 because it was a 2305 for £60 and maybe I could use it to figure out my existing 2305.
I had a moment to investigate this morning and during substitution of boards I narrowed the problem to the buffer chip, a 74LS244 on AC5. Replacing this brought the option board back into action. I wish I'd done that years ago.
I now need to do some capacitor replacement, the blue Philips electrolytic capacitors in the power supplies on these Marconis are almost always down to a few µF instead of 22µF. In the end I've used nothing from the "new" 2305, bar the security of having it there in case I broke something and a source of boards to swap to narrow down the fault. Every board ended up back where it started.
In this case it must be around 15 years since I bought this useful modulation meter from our "favourite" government surplus ebay seller who can't test it because he doesn't have the knowledge to do so. It arrived dented in every surface because it had been shipped wrapped in a mouldy rug and some cardboard. The rug was probably an antique!
After I straightened the panels it worked perfectly, even to manufacturer's specification. There was one more problem, though, the optional audio distortion board that was fitted would not operate, the computer insisted that the option wasn't fitted (Error 06 when the option was selected).
The option board is detected by grounding one pin on a big edge connector which fits into a small backplane board and routes the signal to an octal tri-state buffer on the computer's I/O and memory board. I took all that apart and checked that the 0V was reaching the buffer many years ago, the buffer also appeared to be working properly in every other respect so I just assumed there was some non-volatile setting that wasn't documented but was capable of disabling the option and used it as-is while I looked for something better.
The alternative arrived after a few years with a nice HP 8901B but I didn't sell the 2305 because I preferred it. Fast forward to this year's Rugby rally and I bought a 2305 for £60 because it was a 2305 for £60 and maybe I could use it to figure out my existing 2305.
I had a moment to investigate this morning and during substitution of boards I narrowed the problem to the buffer chip, a 74LS244 on AC5. Replacing this brought the option board back into action. I wish I'd done that years ago.
I now need to do some capacitor replacement, the blue Philips electrolytic capacitors in the power supplies on these Marconis are almost always down to a few µF instead of 22µF. In the end I've used nothing from the "new" 2305, bar the security of having it there in case I broke something and a source of boards to swap to narrow down the fault. Every board ended up back where it started.
Tags:
Re: Marconi 2305
Excellent stuff! All faults are simple once found...
I have no sympathy with sellers who don't pack things properly. Or those who claim (despite being in the business of selling electronic junk) that, "they don't have the knowledge or equipment to test it properly."
I have no sympathy with sellers who don't pack things properly. Or those who claim (despite being in the business of selling electronic junk) that, "they don't have the knowledge or equipment to test it properly."
Re: Marconi 2305
To play Devil's Advocate and depending on the meaning of "properly", I think that is almost inevitable for high-end equipment such as a 7.5digit multimeter. My approach is to state "can't test full spec, but look at the pictures and make up your own mind".EC8010 wrote: ↑Thu Nov 12, 2026 10:33 pm Excellent stuff! All faults are simple once found...
I have no sympathy with sellers who don't pack things properly. Or those who claim (despite being in the business of selling electronic junk) that, "they don't have the knowledge or equipment to test it properly."
But I know what you mean.
Personally I reserve my ire for the "it works because it turns on" brigade. Ditto yoootooobers who produce "unboxing vids".
Re: Marconi 2305
Agreed on all your points. I have three times broken my resolution not to watch amateur videos. Each time, I have regretted it; they were dire, very s-l-o-w, poor camera work, and zero information content. I can only imagine what a unboxing video must be like. I am still scarred from watching Mr EEVblog's video about the HPAK 34470A where he kept mindlessly repeating the instrument's name and took forever to get inside and find that, yes, it does use an LTZ1000 as a reference (he failed to mention that it operates it at too high a temperature).
I am about to venture into the workshop (9C) and make some swarf that will lead to a (much warmer) voltage reference temperature compensation experiment.
I am about to venture into the workshop (9C) and make some swarf that will lead to a (much warmer) voltage reference temperature compensation experiment.
Re: Marconi 2305
Yes, such vids are usually full of ums and ahs.
I like to point out a major change since we were kids. Back then information was hard to obtain (not many books and magazines) and often slow to obtain (weeks for inter-ibrary loans). Consequently we spent a lot of time gleaning everything we could from those sources - and doing that was a key skill to be learned. Nowadays we are deluged by a firehose of dross, and the key skill is the polar opposite: quickly deciding what to ignore.
Now I can speed read at maybe 2000wpm to speedily determine whether an article is worth my time. Can't do that with a <60wpm video. Guess where I don't spend my time. Dave Jones' vids are better than most, since he has the courtesy to give a table of contents allowing skipping.
One perverse potential benefit of the recent spate of AI videos is that maybe youngsters will start to realise that.
I like to point out a major change since we were kids. Back then information was hard to obtain (not many books and magazines) and often slow to obtain (weeks for inter-ibrary loans). Consequently we spent a lot of time gleaning everything we could from those sources - and doing that was a key skill to be learned. Nowadays we are deluged by a firehose of dross, and the key skill is the polar opposite: quickly deciding what to ignore.
Now I can speed read at maybe 2000wpm to speedily determine whether an article is worth my time. Can't do that with a <60wpm video. Guess where I don't spend my time. Dave Jones' vids are better than most, since he has the courtesy to give a table of contents allowing skipping.
One perverse potential benefit of the recent spate of AI videos is that maybe youngsters will start to realise that.
Re: Marconi 2305
Even worse, they often use an automated subtitle program that faithfully writes those ums and ahs down and they don't edit it.
Perzackly, you can flick through and speed read a paper document before casting it aside or deciding it needs photocopying and read properly.
I was wrong about the temperature in the workshop. It was 8C. It's amazing how incentivised you can be to work quickly when it's cold. Not my finest work, to be sure, but good enough. I now have two pieces of clear Perspex, one fits in the bottom of the die-cast box, the other in the lid. Both have their corners filed to match the reverse fillets of the box. The box needs to be oil-tight, so I can't drill holes to bolt the Perspex in place, and I can't use double-sided sticky tape to secure it because the white oil would do something unpleasant to it. I can only make everything a reasonable fit so that it doesn't rattle around very much and can't move such that it would short anything. Just need to drill and tap some 1/4" UNEF holes and then I'll be ready to solder stuff up in the (much warmer) lab.
Perzackly, you can flick through and speed read a paper document before casting it aside or deciding it needs photocopying and read properly.
I was wrong about the temperature in the workshop. It was 8C. It's amazing how incentivised you can be to work quickly when it's cold. Not my finest work, to be sure, but good enough. I now have two pieces of clear Perspex, one fits in the bottom of the die-cast box, the other in the lid. Both have their corners filed to match the reverse fillets of the box. The box needs to be oil-tight, so I can't drill holes to bolt the Perspex in place, and I can't use double-sided sticky tape to secure it because the white oil would do something unpleasant to it. I can only make everything a reasonable fit so that it doesn't rattle around very much and can't move such that it would short anything. Just need to drill and tap some 1/4" UNEF holes and then I'll be ready to solder stuff up in the (much warmer) lab.
Re: Marconi 2305
If the subtitles are visible in the automated transcript (which is better than nothing), then they offer a useful hint 
For your perspex insert, I'll note it isn't too difficult to heat and bend perspex into a crude 5 sided box shape. Perspex cement could seal gaps and/or simply make a box, and optionally hold spacers in place.
For your perspex insert, I'll note it isn't too difficult to heat and bend perspex into a crude 5 sided box shape. Perspex cement could seal gaps and/or simply make a box, and optionally hold spacers in place.
Re: Marconi 2305
Indeed, Perspex can be folded under heat. It's not too difficult, but it is tricky to get the right degree of softening without bubbles appearing within the Perspex. I made a fan in the summer of '76, heating the Perspex under the kitchen grill and twisting it.
Re: Marconi 2305
I spent some time yesterday poking around in the newer "donor" 2305 to try to work out why it's occasionally displaying too much deviation, the symptom seems to be noise in the form of 50Hz pulses. This noise is coming from the -15V rail which was falling outside the dropout voltage of the LM337T regulator, the reason was a single weak diode in the bridge rectifier pack, which tested 100mV higher VF on the multimeter's diode test. I replaced the bridge rectifier and the voltage regulator and everything looked fine, just over 20V on the smoothing capacitor (manual says it should be 21V, but either is plenty). I left it to warm up and two things happened, firstly the LM337T goes into a state where the adjustment resistor seems to do very little and secondly the in/out drop-out goes bad. I wasn't looking at the input voltage to the regulator at the time and there's not completely straightforward way to measure the current so I can't tell for sure whether the load is drawing too much current or there's a problem with the mains transformer's secondary. The other voltages all seem to be unaffected. It is a mystery for now but I think if I measure the secondary current and voltage for the winding that drives the -15V supply I should be able to figure out where the problem lies.
Now I'm wondering whether the rectifier was a red herring, the problem *is* intermittent but I would expect a shorted turn to have consequences for all the voltages and that doesn't seem to be the case. Maybe I should run the next tests with an AC power monitor, too.
Now I'm wondering whether the rectifier was a red herring, the problem *is* intermittent but I would expect a shorted turn to have consequences for all the voltages and that doesn't seem to be the case. Maybe I should run the next tests with an AC power monitor, too.
Re: Marconi 2305
If you measure the ripple voltage, knowing the reservoir capacitance, you can make an estimate of load current. That should enable you to pin the problem to one side or other of the rectifier.
Re: Marconi 2305
How easy is it to separate the parts of this problem? For instance, testing the transformer on its own with a suitable load, and supplying the rest from something suitable such as a bench supply or a variac. A DSO makes it easier to catch intermittent problems.
Re: Marconi 2305
Thanks, that got me out of the dread of measuring the exact current and moving forward with diagnostics… this shows the "fault" condition does increase current to around 1.2A but now the instrument has revealed a third state it can get into where the -15V current drops slightly below what it should be (350mA instead of 450mA), readings stop but weirdly self-cal still passes, IF signal weakens and demodulation drops to very low amplitude noise - presumably because the IF is too weak to demodulate but maybe not. At least that seems to have made it clear that the fault is downstream and not the transformer.
Re: Marconi 2305
After the latest "everything stops" error I decided that I should take advantage of the fact that each PCBA has its own power connector and the two screened boxes and front panel each have their own power feed from the PSU board. First I isolated the fault to the lower RF box, then to the AB9 board, which seems to be drawing most of the current in the whole instrument. This is quite an achievement when you consider that it's only an audio filter with op-amps and exotic analogue switches. Loads of those blue Philips electrolytic capacitors, too. The switches are not common, so I expect that's what's broken!Zenith wrote: ↑Sat Nov 14, 2026 3:11 pm How easy is it to separate the parts of this problem? For instance, testing the transformer on its own with a suitable load, and supplying the rest from something suitable such as a bench supply or a variac. A DSO makes it easier to catch intermittent problems.
Re: Marconi 2305
Ha ha! The AB9 board's output oscillates at 234kHz while powered up on my bench, there are TL071s, a 741, a NE5534N and LF357Ns on here. Poking my finger in doesn't seem to have much effect. I will pull an electrolytic capacitor or two to see what their capacitance is.
Re: Marconi 2305
Or be quick and dirty: tack solder a decent capacitor in parallel with one or more suspect decouplers.
If they are blue Philips electrolytics, the extra lead inductance probably isn't that important.
Re: Marconi 2305
Decoupling capacitors across op amps can cause that sort of problem. They are usually ceramic, low voltage and seldom cause problems - but......synx508 wrote: ↑Sat Nov 14, 2026 5:16 pm Ha ha! The AB9 board's output oscillates at 234kHz while powered up on my bench, there are TL071s, a 741, a NE5534N and LF357Ns on here. Poking my finger in doesn't seem to have much effect. I will pull an electrolytic capacitor or two to see what their capacitance is.
Could be a dry joint.
Re: Marconi 2305
I've replaced the two Philips blue caps nearest the NE5534N on the input, this is configured for 10dB gain, it has a compensation capacitor too. I might take that out to check the value before doing more of these electrolytics. This board has "bent over wire" soldering, which I don't enjoy. So far no change, bridging a capacitor at the output end of the board had an unusual effect so I might do those next.
Re: Marconi 2305
I've replaced six electrolytic capacitors, the old ones were measuring around 10µF and around 6-10Ω ESR, but this hasn't helped enormously. The strange supply-sapping state still exists and the "working fine" state has weird bursts of noise on it that upset the peak readings. This is *different* from before but not in a great way!
Re: Marconi 2305
Presuming there are no hints in the service manual and schematic, thing become more, ahem, empiracle.
Investigate PSU lines with a scope to see where the oscillation is strongest. Ditto with a wire coil acting as a probe antenna.
Poke components with you finger and a pen, to see if capacitance and movement indicate anything.
If you have a soldering station then you could use a narrow air jet to heat up a few components at a time. Changes in behaviour might localise the problem.
Investigate PSU lines with a scope to see where the oscillation is strongest. Ditto with a wire coil acting as a probe antenna.
Poke components with you finger and a pen, to see if capacitance and movement indicate anything.
If you have a soldering station then you could use a narrow air jet to heat up a few components at a time. Changes in behaviour might localise the problem.
Re: Marconi 2305
... and if you don't have a camera, perhaps a single pixel or single digit "camera"?
Re: Marconi 2305
A thermal camera is on my wish list, Mercedes man was telling me about how useful his new one is so I need to keep up!
Though in this instance it would not help as whatever was drawing the current has failed to the "not drawing the current" state, for now. The fault is definitely plural, the LF board still oscillates with the capacitors and NE5534N changed, the removed part tests fine. I think the circuit doesn't tolerate an open circuit input. It's smothered in 22µF capacitors that are really around half that with a high ESR. It doesn't really make sense to me as there can only be a few nH or mΩ between these capacitors and it seems unlikely that their performance would ever have been wagging the Smith chart t'other way with their relatively high ESR but maybe that's the point, the whole thing is critically damped with a distributed bad (even when new) capacitor carpet. HP would've chucked a couple of 47µF where power enters the board then spread out some 22 to 100nF near each op-amp but maybe the ground return from each op-amp (double layer board, with 0V gnd on top) would've circulated some currents that caused feedback on the high gain settings, hence the distributed big, crappy capacitor. I dunno. Now I feel like changing 6 of these and not all of them might be detrimental to stability, but that seems unchanged.
Whatever has drawn too much current then expired has affected system gain, it has lost about 30dB of sensitivity, to the point that the AM modulation mode no longer acknowledges a signal is present and FM needs more than 8dBm to register a noisy signal. The power supply is no longer overloading, so that's good. Must be some sort of amplifier stage that's common to AM and FM. There are a bunch of BFY90s with copious test points. I will read the manual a bit more…
Though in this instance it would not help as whatever was drawing the current has failed to the "not drawing the current" state, for now. The fault is definitely plural, the LF board still oscillates with the capacitors and NE5534N changed, the removed part tests fine. I think the circuit doesn't tolerate an open circuit input. It's smothered in 22µF capacitors that are really around half that with a high ESR. It doesn't really make sense to me as there can only be a few nH or mΩ between these capacitors and it seems unlikely that their performance would ever have been wagging the Smith chart t'other way with their relatively high ESR but maybe that's the point, the whole thing is critically damped with a distributed bad (even when new) capacitor carpet. HP would've chucked a couple of 47µF where power enters the board then spread out some 22 to 100nF near each op-amp but maybe the ground return from each op-amp (double layer board, with 0V gnd on top) would've circulated some currents that caused feedback on the high gain settings, hence the distributed big, crappy capacitor. I dunno. Now I feel like changing 6 of these and not all of them might be detrimental to stability, but that seems unchanged.
Whatever has drawn too much current then expired has affected system gain, it has lost about 30dB of sensitivity, to the point that the AM modulation mode no longer acknowledges a signal is present and FM needs more than 8dBm to register a noisy signal. The power supply is no longer overloading, so that's good. Must be some sort of amplifier stage that's common to AM and FM. There are a bunch of BFY90s with copious test points. I will read the manual a bit more…
Re: Marconi 2305
Loss of signal appears to be entirely down to the front panel N-type socket being not especially attached to its RG178 or whatever Teflon® mini coax it uses. It was torqued down so tightly that the nut exploded before loosening and may not be in a suitable condition to be refitted so now I have to find my box of N-type stuff, where I'm almost sure that I won't find anything suitable.
Re: Marconi 2305
I'm not sure I follow what's going on with this. Perhaps if you explain the symptoms and steps it may make things easier for you.
There was what appeared to be a PSU fault. Was that fixed or traced to something other than the PSU?
There was a fault with an audio board producing oscillations at 234kHz when powered on a bench PSU outside the instrument. It seemed like the kind of instability you can find with op amp circuits. Was that fixed? Did it do that inside the instrument? Some caps were changed.
There's a problem with the N-type connector on the front. Does this have any bearing on the other faults you've observed?
Have you tried board swaps with the good Marconi 2305? How about voltage and waveform comparisons?
There was what appeared to be a PSU fault. Was that fixed or traced to something other than the PSU?
There was a fault with an audio board producing oscillations at 234kHz when powered on a bench PSU outside the instrument. It seemed like the kind of instability you can find with op amp circuits. Was that fixed? Did it do that inside the instrument? Some caps were changed.
There's a problem with the N-type connector on the front. Does this have any bearing on the other faults you've observed?
Have you tried board swaps with the good Marconi 2305? How about voltage and waveform comparisons?
Re: Marconi 2305
I'm as confused as you, the power supply *was* in terrible shape to start with, most of the capacitors were seriously degraded including one big one that had lost all its capacitance. I started with this, then saw the low voltage on -15V, changed a weak bridge rectifier and a perfectly fine voltage regulator and discovered that something was still dragging the -15V current too high. The power supply's three outputs were disconnected one at a time and the output feeding the lower RF box was the only one to relieve the -15V loading problem, so then I removed power from boards in that box until I found the one that made the difference. Now, it's possible that I didn't find the cause of that but I did find that this board self-oscillated with no input and all its decoupling capacitors were seriously degraded. I changed the NE5534, as this was the input op-amp, but even after these changes it still self-oscillated on the bench with no input. Now it's possible that it was self-oscillating when back in the system, but I didn't see any evidence of that - if it had been going off at 200kHz that would still slip through the 300kHz LPF option, so should've been visible.
After refitting it, though, the only problem seemed to be an intermittent and worsening loss of RF gain which was entirely the fault of a fractured coax centre wire inside the N-type socket on the front panel. I've repaired that, reassembled everything and the remaining issue was one of the filters not working - which I've not mentioned yet, but it was just a bit of internal coax wiring touching a select-on-test polystyrene capacitor in the filter bank and shorting it to ground, I must've snagged it during reassembly.
It all seems to be working now but I've a feeling I haven't seen the last of the "high current" fault, which was the worst sort of intermittent.
After refitting it, though, the only problem seemed to be an intermittent and worsening loss of RF gain which was entirely the fault of a fractured coax centre wire inside the N-type socket on the front panel. I've repaired that, reassembled everything and the remaining issue was one of the filters not working - which I've not mentioned yet, but it was just a bit of internal coax wiring touching a select-on-test polystyrene capacitor in the filter bank and shorting it to ground, I must've snagged it during reassembly.
It all seems to be working now but I've a feeling I haven't seen the last of the "high current" fault, which was the worst sort of intermittent.