Lots of egg frying going on with nasty spikes – worse on the left than right channel – and not solved by the change of filter caps to new EBay fake Nichicon 4700uF – that aren’t even 4100uF…I’ll leave them in as the original caps are 25 years old but test fine.
Seems more like driver transistors and as most components have black leg rot, then probably noisy transistors somewhere. Not helped by the equally noisy balance pot either…
The obvious transistors to change are q420-21 I guess?
And maybe the 50V, 3.3uF electrolytics to C401-2 in the audio chain of the Voltage amp IC401.
All I can do is go with my gut feeling that this is a cap breaking down at power on and remove one at a time for a recap. I have no other logic to guide me here as I can’t even split the audio chain in half by removing one, as a collapsing DC rail could cause these noises, as well as noisy components (which I think there are too).
Please note: The unit powers on and amplifies however there are multiple issues when the unit does so. When linked to any type of speakers the left speaker connections do not work at all, meaning the only sound that can be heard is out of the right speaker. When amplifying music, the sound is completely distorted and warped when using any type of media. I have tested this with both cassette and cd player and experienced the same results. Unless repaired, the amplifier cannot be considered working as the sound output is faulty as described.
Stereo phones, but no left speaker channel. Low volume for 0dB right channel for a 40W amp. Driver stage missing?
The SVI 3102B is right before the speakers in the block diagram.
It has +/-34 and -15V but no +15V to it’s pins which is correct from the sux840 SVI3102A pinouts :
Guess it’s backtrace the signals from the SVI3102 pins back through the circuit to see where it’s lost…
A nice 40W monster from the 80s! Was working – stored for years, now not turning on after the last owner tried to recently…allegedly…
Poor old bugger! Look at the dust in this thing – no wonder it won’t turn on!
First thing – get the hoover out…!
That did not work at all – it’s that sticky, greasy shit – so acetone, contact cleaner and a brush required…and a wide open window!
I’ll check the big 6600uF, 56Vcaps first as there’s an easy underside panel removal for that – these are in good shape – 0.003 and 0.004 ohm ESR and the dual PSU DC will help reform them – they should be changed as they’re 35 years old if original:
Now clean enough to work on (fuse is OK too), it can be powered to 30V DC gently by dual rail PSU across the rectifier diodes.
This is one of those 4 legged beasts like in the SA-5L that is difficult to get croc/scope leads on the legs – had to use pliers then make really sure the insulation was over the crocs to not touch the AC in legs.
It powers to 30V with no drama, but no mains bulb.
This is a 12.6Vpp AC from the PT so should be replaceable by an LED:
Note the 145C thermal cut outs, the thermistors on the small board test unblown at 39K each:
Unfortunately, the LED didn’t sit centrally when clipped back in so not all lettering is lit well, but better than nothing…
Running off the mains shows all PT pins have the correct voltages referenced to ground (black).
The diodes are:
D505-8 all check out rectifying from 0V-(-/+7.5V)(+/-15V). The caps nearby also have expected values..
Assumptions again! There are NO LEDs on the front (my eyesight is getting BAD!) to imply channel logic switching (as on the last amp I was looking at – talk about going senile..) so there actually is NO problem with this amp at all now it seems!! That’s also letting the seller’s opinions affect your perception – he just turned it on, saw no light and assumed dead I guess? All that was required was to switch the speakers on at the front, probably!
Read the bloody schematic Steve!! The schematic shows the Voltage amp AN7062 pinouts (all correct within tenths V) AND the 16 pins for the SVI2003 itself (all correct).
So, now I have headphone and speaker outs, with an output of 45Vpp I can attach the 8 Ohm loads:
44Vpp = 22Vp = 15.556V RMS
P=VV/R = (15.556)^2/8 = 242/8 = 30.25 Watts per channel full volume.
Hmm..that’s MUCH lower than the 45W specs and with a 500mV input…??
I didn’t know the RIAA filter was in the feedback circuit of the phono op amp.
Usual idiocy by me making assumptions so fault finding when I didn’t need to, but good experience for logical voltage checking, non slipping probe practice and familiarity with the different pin voltages of the different SVI models and diodes – and new components – the Voltage Amp AN7062N and the 145C thermistors measuring 39k Ohms at room temp.
A good amp for a beginner to fault find as all components are accessible by a sharp probe and the schematic is detailed with all voltages at all point for this:
The LED light now adds a custom aspect to the front. Blue always looks good! Would have been nice to have channel select LEDs on this model. The -20dB cut switch with Loudness on is really effective in reducing noise from a badly tuned S-5L!
The rectifier ripple is 130mVpp on the +/- 35V DC rails:
This will need filter caps replaced I think, despite good readings if they’re original at 35 years old as there is a nasty mains hum sometimes at power up that can shut the amp down. I’ll order them now. This may be a cap temporarily leaking noise onto the rails if it’s breaking down at powerup?
Don’t think it’s a diode breaking down else I think that would likely be a permanent fail?
Volume pot is really itchy and scratchy; heat sink runs very hot; one channel missing (left); the volume LEDs flicker to higher values on pot increase but don’t stay lit for the pot position and return to unlit; other channel too quiet.
So for the experience, I cleaned up the old messy, white, SVI2129 and power transistor thermal grease and applied new copper coloured paste. There’s an insulating silica heat pad behind the regulator – the old paste was only on the heat sink side, not the transistor to pad side so not sure if this was factory work, so unsure if the SVI2129 is the original or not.
My logic behind the heat sink being hot, is that if the old stuff is too insulating, then the heat will rise in the SVI2129 and regulator making them too hot, which will eventually heat up the heat sink as the the rate of cooling is insufficient, so the default temperature becomes above normal.
After doing this, the heat sink did run cooler, but not definitive yet until full signal is restored.
The main board needs removal for access to the dirty/damaged volume pot..
…and what a pig to get to! Had to dismantle the whole case and so many screws to get it out! Pain to wick as well, nearly lost a track.
Now cleaned, it has made a big difference already – really easy to turn and the almost dead channel is a bit more audible and the good channel now shows continuous LEDs. The trimmers on the LED board were also measured and made equal at 6.2k Ohms almost exactly back where they were paint locked before.
After a lot of wasted time checking components that were fine, I found it was the pot again!
This volume pot caused me no end of grief after testing OK out the chassis – it had me chasing down other components for the missing channel when it was the pot being intermittent bad contact even after replacing and cleaning! It took a second removal, clean, check and some solder heat damaged tracks to get this thing working properly and re-installed despite the bad tracks. Only then did all else work correctly.
The chase down was good experience though, going through the signal chain at convenient points though some components were hard to find with my eyesight.
The odd thing that caught me out too was the speaker AB system – when I had signals at the SVI2129 I got no speaker sound! The connections are separate grounds so you can have 2 or 4 speakers without the usual switches you find on amps so the single system connections aren’t next to each other like usual!:
There is no difference between speaker loaded and unloaded output, using 8 ohm dummy loads to test with a 500mV input signal (I can’t get a clear sig gen/scope signal at much less than this level with the cheap sig gen), showing 40Vpp at full volume just shy of clipping:
Gain at max volume (0.5V in) before clipping = 40Vpp/o.5Vpp = 80
P=IV = 40Vpp = 20Vp/root2 RMS = 14V RMS
14Vp x 14Vp / 8 ohm = 24.5 W per channel
About right for the higher input signal of 500mV over nominal 150mV line input 25W per channel at minimal harmonic distortion:
A solid little amp – well built as expected from Philips. I don’t like the horrid big heat sink sticking out the back though, and it can get quite hot – be aware what it’s resting against.
The SVIs are all the same in most amps so should all sound the same power for power output.
The minimus 26s make the difference – they have a really nice imaging quality and are pretty bassy for their size with the Loudness button on!
Beware soldering under board horizontally, as solder can run down larger component legs – like volume pots! – and cause similar problems to the ones you are trying to solve by cleaning!
So the carbon crap on the board was the problem with this nice little, great sounding unit.
The front face was a pig to get back on – you have to remove the analogue meter to get it to fit properly!?
The digital dial light is a 6.2V AC circuit like on valve amps – why? Just for a wheat grain bulb type? Seems a waste of a 6.3V winding on a transformer, unless it is used elsewhere and rectified, but without a circuit diagram, I can’t say..I don’t see why this dial section needs a light either, with the digital display?
It has an alignment/quartz lock issue on FM as it can’t scan and lock, and I don’t have the kit to align it.
The LW and AM bands scan and lock ok, but there’s no stations worth having anyway.
The dual PNP and NPN power amps are well heat sunk and really kick out the bass with the Bass Boost on for a small unit.
P= IV = V^2/R = 7.1 x 7.1 / 8 ohm = 50 / 8 = 6.25 Watts per channel
This is at volume 7.2 for 0.5V 1kHz input.
I didn’t get the power transistor types as they were covered by the heat sinks and I can’t find any info for this unit at all!
It is now not necessary to remove the main board to check the caps (underside) with the Peak ESR meter now I have a dual rail supply, as I can connect to the rectifier with +/- 30V DC and ramp up slowly watching the amps flow for any shorts OR NO current flow for an open circuit somewhere (probably a constant voltage transistor soon after the rectifier). Shorts on other boards can be proved by disconnecting relevant ribbon/push connectors if fed from the main board.
Low 4 ohm values on secondaries on the Peak ESR can be deceptive! It does NOT necessarily mean the windings are short, as I have since learned on the SUX840 as these values increased once disconnected from the main board (that amp needs more investigation – no voltage on the voltage regulators though they tested OK when removed).
When testing power output with the cheap 100W chinese wire wound resistors – be careful! They get REALLY hot if left at full volume for too long (seconds) and they don’t cool well AT ALL once disconnected and are a burn risk.
Now I have a working dual rails PSU, I tested the main board to +/- 30V DC by connecting the dual PSU in series to the rectifier diodes, slowly bringing the voltage up in 5V (+/-10V) steps with the slave channel display in current mode. No current is drawn by the board at all, so there is no short on this board that could have caused the PT secondaries to short out.
But, no current at all isn’t good either..?
The only way I can test for any power amplification is to direct inject signals into pins 11 and 13 and measure AC output at pins 4 and 1 :
Pins 2 and 3 are the +/- 36V power rails resp.
Nothing definitive by injection as I can’t change the volume as it’s a rotary controller on this model, so logic levels, but by turning up the sig gen input from 0-20V I get gain from pins 1 and 4 though behaviour is alternately asymmetrical:
DUH! Dick! I forgot I removed the regulators to check them and forgot to solder them back in – just clicked them back..that’s why there’s no current flow!
Once soldered back the main board powered and drew about 4mA – same as the SA5L. Positive 30V is red on the left of the diodes, negative 30V right, black ground is on the heatsink. (Don’t always assume the heatsink or chassis IS ground though – ground may be floating on some equipment)
All front panel lights lit so there is a logic issue on the front board but probably needs the additional windings voltage from the PT..which maybe blown still unless the problem is on a front board?
I removed the front panel to find the flouroscan screen bulb blown black top left – won’t be finding a replacement for THAT on Ebay..:
The schematic says this provides 24V on some pins but it’s external feeds from the main board are 14V max on any pins:
“Assumption is the Mother of all Fuck Ups”! again…
Right tools for the job as usual – should have waited until the dual PSU arrived, but had to start somewhere. It would have saved more work and done a better job if I’d waited for all the needed kit like solder wick to do the de-soldering properly as I lifted a track but got away with the soldered connection – lucky!
The only fix is a new futaba flouroscan, then maybe a PT…doubt it. They look cool when working – shame..
Removal of it was good practice for solder wick technique at least.
This PSU from Taiwan has a Master side voltage max out to 45V as soon as you turn the current sense control up and seems to be an issue for others on the Net. This unit seems an exact clone of the Tektronix PS280:
Suggestions for checking the control pots makes sense first as any bad pot here could switch the logic to the max out put.
It’s a nicely built unit and the front control board comes away easily after removing the pots knobs and the nut/washer.
The fault on mine was the 10k master voltage (and slave side value) pot had cracked from being loose, so pushed back which clean broke the bakelite board tracks, leaving the pot open circuit. This accounts for the behaviour as soon as the current sense (1k pots) is turned up so the green sense LED lights, the logic senses the voltage pot at maximum so switches the all 3 relays on in quick succession to give a maxed out voltage of 45V on the output. The odd thing is that working the slave side only goes to 30V maximum in Independent Mode, so it seems the master side should also – so not reach 45V but the max 30V voltage also. Maybe there’s a clue in the schematic..
As the transistors (Q706-8?) on the main transformer board (which had the black scorch marks) fail with the Peak checker (as do the diodes) and don’t give definitive readings with diode mode meter except from the bases to collector/emitter (the diodes check ok too) they had to be desoldered using the Pro kit at work, as the most likely damaged part from an arc/burn. But the 3 tested ok…? Hmm..what now?
I soldered them back in using the typically inaccurate new 878D Chinese crap solder station with pointless temperature display that arrived today (220C – general soldering temp? Haha..set to 350C to get even a possibility of soldering a connection – and my free solder flux pen in a box of free goodies from a kind, retired old tech made my day! Classic: “..I’m getting too old and stupid..”). Know the feeling…:
All the capacitors in circuit check out with the Peak ESR meter, and the resistors in circuit give sensible readings so I’m not convinced I’ve found the problem was just black carbon shorting out power to this board as I can’t find a duff component…I want to check this board with the DC PSU first before I return it to the chassis and attach mains.
Before I do that I have to dismantle the entire horribly built thing to get to the main board for the rectifier and filter caps (they’re good with the Peak ESR).
Apart from a slightly broken resistor body I can’t see anything obvious on the main board either.
Another Ebay “does nothing, no power” casualty. (This is always a risky buy due to mains transformer loss – but mostly you get relatively simple faults in my limited experience – at best a surge blown fuse with not other damage, obviously.. )
Spotlessly clean, hardly used, no fuse blown, but I suspected the power transformer at first tests as one connection is open circuit but this could be a 110V connection point that this amp does not offer as it’s UK 240V model only, but need the manual to check (Yes, it is unused).
Pins 5-6; 4.4 ohm (This should not connect for the UK model)
Pins 6-7; 4.4 ohm
Pins 6-8; 10.2 ohm
Pins 7-8; 15.3 ohm
Looks like a secondary is gone…and here’s the proof:
The primary above shows 700Vpp (UK mains 240V RMS = 480V RMS pp = 480 x 1.414 = 679Vpp but showing more than 700Vpp here!) on x10 probes and 100V/div:
Only a couple of secondaries have a tiny 400mVpp signal at best, the rest are flat 0V. The white of the ribbon is pin 1 left below and marked on the main board. All others are the same.
Seems to have taken out the two FETS Q501-2 also as they are short from Drain to Gate – but I need to research FET checking in circuit.
Replacement transformer will be expensive… these are a pain too as they have to be desoldered or cut from the board unlike the SU 600.
The power amp SVI 3102 A seems original by the old off colour heat paste.
Op amp not inserted fully – what happened to factory QA?
Guess the only fix for this is another broken unit with working transformer.
I can only dismantle this, remove suspect parts and check them out of circuit for something to do. I can cut the PT ribbon to the board as it’s soldered on anyway. Solder station should arrive this week.
The rectifier diodes failed on the Peak except 1, but because of black oxide on the legs – when scraped off it still caused a problem, but on connecting to the underside of the board, all passed as 2.57V @ 2.4mA diodes.
Beware bad connections to the component under test!!
The legs of lots of components have black oxide type crap on them.
The 3300uF 50V filter caps test good at:
3005uF ESR = 0.04 ohm
3094uF ESR = 0.0 ohm!
I removed the 2 x D1265 NPN power transistors and the PNP B941 and they tested OK – no other obvious fails that could have shorted the PT secondaries or been damaged when the PT shorted..?