Well, back in August the car came home (link here) , and it's been languishing back in the garage awaiting re-charging of funds.
I decided to bite the bullet and get a proper whole front panel. It wasn't cheap, but the front valance was shot. I was going to sped a few hours repairing it, but it had gone round the holes in the front badly, and that was going to be difficult to get right.
Measurements are made, and spot welds drilled out.... there's some unpleasant bodgery around the headlamp surround. Thankfully the wings are sound.
"I've got a Marantz amp, intermittent low volume on the left hand channel, usually when changing between speakers A and B , and it only happens on one set of speakers..it's too good to throw out. Can you take a look?"
Yeah ... why not?
Nice looking modern amp...
... and very nicely constructed.
We're going to have to get the main board out, which is a bit of a struggle as usual, as all the sockets are connected straight to the main board. But before that, let's see if the amp will misbehave for us... Yep, one of the speaker protection relay's doesn't always make it across. They're only rated to 5 amps, which seems a little low for my liking, so some 10 amp relays are obtained, and fitted.
Here are the guilty parties. Actually only one was faulty, but I changed both to be on the safe side. A check of the bias (it was fine) and a good soak test. An easy fix.
When I returned the amp, the guy said it had always done this since new, and had been returned under warranty twice. Each time it was returned stating "no fault found". I would have thought it was obvious, once the customer's speakers and wiring were eliminated from the equation...
This arrived ... "Can you have look mate, been in the attic for years, plugged it in and it hums loudly and gets hot"...
... Yeah, why not.
I initially thought it was some sort of cheapo Japanese "nomark" amp...
Closer inspection reveals it to be made by L&G ... nope, me neither! A google search reveals it's made by Luxman, a well thought of brand in hifi circles, and this is from their L&G "more affordable" range. Apparently designed by Kimura-San of 47Labs & Gain-card fame.
It's of bog-standard 70's Japanese construction. Thankfully you can get to the print side of the board through a hole in the bottom.
A close look at the board shows there's either been some leakage of electrolyte form the caps, or there's been some sort of spillage. It certainly smells like a cat's litter-tray..
More signs of spillage.
This is a bit weird... there are two caps, mounted upside down, despite having more than enough space on the board... Here's one, there's the same on the other side of the board....
Nothing unusual about the output pair, NEC 2SD388A & 2SB541A on the left hand channel ...
nothing unusual that is, unit you look at the right hand channel....
... which is fitted with Toshiba 2SD371 & 2SB531 ! Apparently it's never been repaired before , so it must have left the factory like this! Perhaps they switched supplier and ran out halfway through making this amp!
A few cursory checks, and on with the mains, don't hold out much hope for the caps, but let's find out how bad things are. Left hand channel has a horrible 8V DC offset, and it's all humming very unpleasantly.
I think we'd better look at those caps. One by one, they are removed, tested and replaced. Some are open, some electrically leaking, many are physically leaking, and none are to tolerance, except the main split-rail smoothing caps, and those two caps mounted upside down!
After a comprehensive re-cap, things are improved, the DC off-set on the left channel is still way too high at 2V. I measure the bias of the right-hand channel at the 0.33 ohm emitter resistors. It's set for 80mA. I set the left the same, and then set the balance up for both channels. DC offset is less than 5mV for both channels.
Now the switch-gear and pots are all noisy, but really nicely made. A good clean up with some switch-cleaner and repeated operation sorts any switching issues and noisy pots are no more.
I soak test the amp for a couple of hours with some fine Yes music, re-checking the bias and offset after an hour, and it's all well. Sounds great, the phono stage is quiet, and it makes a good 50 watts into 8 ohms. Despite different transistors in each channel, I'm damned if I can tell!
I'm working on a small commercial project at the moment, which requires me to store some data into the Arduino's EEPROM. When you upload a new sketch to the Arduino, the EEPROM remains unchanged.
Now a "virgin" arduino has every address in the EEPROM set to 0xFF. Here's a quick sketch to reset the EEPROM to as new, so you can test your code on a fresh arduino.
It outputs to the serial monitor, and blinks Pin 13 on completion.
Apologies for lack of comments, but it's straightforward enough...
#include<EEPROM.h>//EEPROM interfaceintaddr=0;floatlen;intpcComplete;voidsetup(){Serial.begin(9600);Serial.println("EEPROM Eraser");Serial.print("Size of EEPROM =");len=EEPROM.length();Serial.print(len);Serial.println(" bytes");delay(1000);pinMode(13,OUTPUT);}voidloop(){EEPROM.write(addr,255);Serial.print(addr);Serial.print(" ");pcComplete=(addr/len)*100;Serial.print(pcComplete);Serial.println("%");addr++;if(addr==len+1){addr=0;Serial.println("Complete");blinkPin();}}voidblinkPin(){while(1){digitalWrite(13,HIGH);delay(1000);digitalWrite(13,LOW);delay(1000);}}
Ugh ... soundbars. No. Not for me. A decent amp and speakers, why not? But the thought of an amp and speakers placed in a tiny enclosure to put under your telly, to improve the (usually) woefully inadequate speakers in a flat screen ... anyway ... rant over.
This is the Panasonic SC-HTB10 sound bar. It belongs to a colleague.
"It just clicks. Can you have a look?"
Yeah , why not...
Dis-assembly involves removal of the myriad of screws from the back, there are 4 hidden under rubber bungs on the corners. The back fits really tightly, I needed to gently pry it free using a "spudger" (one of those cheap plastic tools you get for taking mobile phones apart).
This is the only time I've ever had to take a photo of a piece of kit in panorama mode!
You'll need to disconnect the sub-woofer which is in the back...
Neat tuning port!
Unplug the other two speaker plugs (blue and white, just by the green circuit board)
You should now be able to lift the electronics clear from the case.
Now, I'll just draw your attention to the disclaimer up there on the right of the webpage. Yep, that one. I'm doing that, because this one caught me out a bit. Under the pressed steel shield is the power supply (actually two supplies). Because of the symptoms, my guess was it was failing to start up. I carefully removed the shield and removed the supply.
Now, we can see there are two switched mode supplies. There's a small switched mode transformer above the date code, and a larger one below.
The small power supply supplies the micro-processor and control electronics. It's always on, so the remote is kept powered. This appeared to be functioning correctly.
The larger power supply, and the components in the area marked "HOT" are used to supply the amplifiers, This is where the fault lay. Despite the fact that this supply had been disconnected for many minutes, the large electrolytic still held an impressive ~300V DC. Whilst I avoided getting a shock, I was surprised to find this charge still remained. Having safely discharged the cap with a watty 10K resistor, I started a few checks. See how dis-coloured the board is around the switched mode controller and chopper IC 5701? Thing's have been running HOT! (The HOT marking on the board is to warn you that this part of the power supply is connected to live mains, and is not isolated! Not the temperature of the electronics!) Poor old C5726 had suffered as a result. It's ESR had risen to about 25 ohms... far too much for this capacitor to be in any health. A replacement was fitted and restored operation!
Float back to the 60's (man..) and Selmer was producing some really good "Truvoice" amps. Quite expensive though... then in the early 60's (around 1963, to match their Futurama guitars) they released some budget amps.
This has wandered in to the workshop...
It's a series two amp, dating from 1965. Schematic can be found here. It's a simple enough circuit, two ECC83's and EZ80 rectifier and an EL84 output valve.
Cosmetically, it's in really good nick...
There's a little storage compartment in the back for the mains lead and tremolo foot switch.
This one's not so good electrically. Switching it on, it just hums loudly.
The amp is constructed on a small PCB, mounted on the rear of the front panel. The output transformer is on the left of the chassis, and the mains transformer on the bottom of the cabinet. It has a three core mains lead, and is nicely earthed, unlike the WEM copycat!
Removal of the chassis for service requires removal of the four screws attaching the front panel.
Investigation into the hum shows the main HT smoother cap to be totally unservicable. Once again the dreadnaught lets me down, you can't win 'em all....!
To preserve the look of this lovely amp, we'll re-stuff the original 32+32uF cap with modern equivalents...
First of all cut around the bottom of the can. I use a sharp craft knife, scoring round until the blade is able to pierce the aluminium can. Wear some rubber gloves.
Next, warm the can up with a hot air gun in a well ventilated area, and get a large screw, and screw it into the capacitor. (This is a concrete screw, they're really good for putting up shelves! No wall plug required!)
Grab the screw with a pair of pliers and gently pull to extract the insides...
Now make up two modern 33uF (450V in this case) capacitors, and solder them to the base. The can is the negative for both caps in this instance (it usually is, but best to check!)
If you can't get the leads through the original rivets, drill them out (or drill holes though them) so you can pass the leads through.
A lap of PVC tape to prevent any nasty shorts catching us out....
And get the whole thing together in a vice so we can re-attach the can to the base...
Here I've used a little Alusol (aluminium solder) to tack the can to the base, so there's an electrical connection there. It's a bit of a pig to do, and you'll definitely need some fume extraction, that Alusol gives off some really nasty fumes.
Finish off with some hot melt glue... and tidy up with a craft knife.
Once this was replaced, I was rewarded with a hum-free amplifier... but that wasn't quite the end of the story. The tremelo didn't work (Why, oh why, am I always plagued with tremolo circuits?!?!)
A few checks round showed a 100K anode load resistor to one half of the ECC83 (V2, the tremelo low frequency oscillator) was very high in value, having risen to ~4 Megohm. Replacing this restored tremelo operation ... but I couldn't switch it off ! The wire to the foot switch proved to have a break in it somewhere.... replacement sorted that. A quick squirt of de-oxit sorted any noisy pots out, and that completed the repair.
A little while back, a colleague bought this radio to me...
"Can you look at my late father's radio? I've found it in the attic. I think it stopped working in 1964. I'd love to hear it going again"
... and, as regular readers will know, I love a challenge, and hate seeing this stuff go to landfill....
It's a Ferranti 146, manufactured from around 1946. Electrically, it's almost identical to the 145, and the circuit can be found here.
The case has a nasty crack across the bottom, and it's missing it's back and two knobs. The mains wiring is lethal, and there's a lot of rubber insulated wire that the rubber has broken down, and cracked, so that will require replacement, but it's complete....
Here's a really bad picture of the underside of the chassis... There's plenty of horrible wax caps, one "traditional" electrolytic, and, (you can *just* make them out on the right hand side of the picture), some electrolytic caps, in wax covered cardboard boxes! , a quick check shows these to be really badly leaky. No matter how long the ole' dreadnaught is left connected to these, they're never going to be any good. Replacement is the only course of action.
Warming the boxes up gently with a hot air gun (not too hot!) allows the wax to soften, and the old capacitor to be removed from the box. It's best to allow plenty of ventilation when doing this. I'd hate to think what the chemicals given off will do to you! Rubber gloves are another sensible precaution.
One the old capacitor is removed it's simple enough to solder a new capacitor to the wires, and slide it back inside the box. This preserves the "look" of the set.
It's a run-of-the-mill job to replace the other capacitors with vintage looking modern equivalents, changing any rotten rubber wiring as we go...
It's obvious that there have been a few repairs made in the radio's long past. There's a "Radiospares" capacitor in there that dates from the late 50's. There's also the anode resistor to the audio preamp, which has been replaced, soldered at one end... and "wire-wrapped" with a bit of fuse wire at the other! A quick check and resoldering sorts that!
The chassis top is cleaned up a little, it's just a bit grimy and dusty. It's not rusted or pitted, so a clean is all that's required...
Checking for obvious short circuits first, mains is applied via the variac... ... and almost nothing ....
Tuning up and down the medium waveband produces one very weak local station. (Note: This set has no ferrite or internal aerial, the green wire you can see connected to a socket on the chassis is a long wire, draped around the workshop) Checks in the RF stages, and testing the valves shows nothing is amiss. Alignment is performed as shown in the service sheet, and... bingo! Dozens of stations on Medium wave, a couple on long wave and a few on Shortwave....
Worth noting that the early versions of this set (and it's earlier brother, the 145) use a 6V6G output valve. This set is fitted (from new) with an EL33B. There's some global negative feedback added in to tame it's output.
After a while of testing, it suddenly crackled and went silent... everything appeared to still be working except there was no audio. There was nothing present on the anode cap of the EBC33. Wiggling the wire temporarily restored operation. The wire is screened, so a new one was fabricated from RG59 co-ax and fitted.
Next thing to sort out is the dial-cord or string. This is in a mess. The tension spring is rusted and has lost most of it's springy-ness. At some stage the string has broken, and been shortened, which has meant it's strung around the shaft which connects to the tuning knob the wrong way round. It slips because of this. New drive-cord is obtained from eBay here, and a suitable spring found in the junk box. Re-stringing is not an easy task, and involves a lot of cursing, but we got there! It now functions smoothly. The string connects the knob to the variable capacitor, and via 4 pulleys to the dial pointer.
Some thin super-glue is poured into the crack in the case from the inside and allowed to harden. As the crack is along the bottom edge, it's out of sight.
The case is cleaned up using some Servisol "Foam Cleanser 30" (other foam cleaners are available, although this stuff is good!) and 69 years worth of grime from coal-fires, fags and lord only knows what runs off in a brown slurry. I thought the case was a dark brown Bakelite, but it's jet black! Quite the handsome set!
Now to fabricate a back. I obtain a piece of hardboard (the type that's shiny on both sides) and laid the case of the set on it. I draw round the outside, and work out I need to make it about ~6mm smaller so it fits inside the recess of the set. Now some measuring to get the various holes in sensible places, and cut it out with a craft knife. I screw the back to a piece of stout board to stop it warping, and paint it with two coats of 50-50 PVA/water mix to prime it, and once this is dry, I lightly spray the back with some left over black paint. The set is transformer isolated, so the chassis is not connected to mains, so it's safe to operate without knobs until something suitable turns up.
Just downloaded Arduino 1.6 to find some of libraries, notably TIME don't compile properly, and throw errors. I used the library manager to update them, and it's still bad.
The trick here is to delete the offending library from your ..../arduino/libraries directory and re-load it using the library manager.
Now, a few years back, I restored a 1951 Bush TV22 for my friend Colin.
I made a video at the time...
The Bush TV22 is a very common set, even now. They seem to crop up at regular intervals on eBay. It's a 9" TRF set, and had various modifications and different designs during it's long production run.
Now this set was used regularly until last year, when Colin rang to say it was acting up a bit, and could I take a look... sure.
And so it arrived. Poor line linearity and decreasing width as the set warmed up, over a period of 30 mins or so. Not good symptoms for a vintage television, as it's usually signs that the line output transformer is on it's way out, and good secondhand transformers are difficult to find, and new ones unobtainable.
That proved to be the cause with this set. Thankfully the Bush transformer is quite simple, and there's a marvelous chap called Ed, who can rewind the things. It's unusual to find a duff high voltage EHT winding in the Bush set, which is a blessing, as rewinding this part is not possible, So it's the primary and lower voltage secondaries that fail. Off to Ed it goes....
... Within a few days it returns, nicely rewound, and I reinstall it into the set .... the line stage warms up, and whistles strongly (a good sign). There's no glow from the EHT rectifier valve, and no EHT. I check the line stage from one end to the other. Nothing amiss... ???
I remove the transformer again, and perform a "ring" test. This is a basic test to ensure the transformer has no shorted turns. If it was the overwind, there's a few options open to me to create some EHT to drive the final anode of the CRT without using the overwind.
The ring test involves connecting the transformer to the square wave calibrator on your oscilloscope, at the end which is usually fed from the anode of the line output valve, and connecting the scope probe to the eht wiinding.
The ring test allows the transformer to oscillate. The more oscillations, the better the transformer.
That all looks great...
A duff transformer looks like this.. See the "ringing" stops well before the next cycle...
A few emails swap between myself and Ed. It appears the transformer for this set was very unusual in it's construction, and one Ed had seen before. Some of the windings appeared to have been wound "out of phase" with usual convention. Could this be the cause of the issue? I experimented by removing the EHT overwind, and slotting it back onto the transformer upside down. It worked after a fashion, but the EHT was still very low. Ed asked me to return the transformer to him for further investigation.
A few days elapsed, and the second rewind arrived back. Whilst the transformer was now functioning, the EHT was still low at around 4KV (it should be ~7.5KV), and there was terrible line fold-over which I couldn't get rid of...
To take the above picture, I needed more than the available 4KV, so I had previously purchased this little device from eBay...
Described as a DC 3V to 7KV 7000V Boost Step-up Power Module High-voltage Converter Generator, a useful addition to the armory of test gear. I wouldn't recommend shorting it out like the picture! Anyway, this provided some EHT to show the picture above.
Then lady luck smiled on me... and this set arrived as a gift. Apparently found on a tip! There's a small crack to the bakelite case, but it's almost invisible. I don't actually own one of these sets, so it was most welcome.
Naked, and filthy!
Line output stage cover removed ... and dusty!
First things first, and hook up the Leader CRT-910A tube analyser to see if it's worth pursuing. If the CRT is bad, it won't be worth doing anything with.
The needle slowly swings into action ... and rises.... and rises ... 95%! I find this hard to believe, perhaps the analyser has developed a fault.... but checks on another tube show it's telling the truth!
So, I hatch a plan. Do a quick restore on the gifted set, in an attempt to prove the line output transformer. If it's OK, swap the transformer over to prove that the rewound transformer still has issues, or not, as the case may be ....
Now this model was later, and has a different receiver (or RF deck) than the one we are repairing, which is a pity, as I was just hoping to fix the top chassis, which contains the power supply, line and frame stages, and just swap out the working receiver from Colin's set... not to be.
So, our tube (CRT) is good, so first, remove the mains smoothing capacitor, and give it a few hours on the ole' Dreadnaught capacitor reformer...
It takes a while, but comes up well...
Next up is to re-cap the frame stage. This is located on tag-strip on one side of the chassis. It's essential to do this before applying power, to avoid damaging the frame blocking oscillator transformer (Ed rewinds these too!)
Then to re-cap the underside of the main chassis. Anything with wax insulation will be faulty (if not now, within hours of first use) and electrically leaky, so they are evicted on sight. There are two located under the tag-strip, which needs careful removal.
Whilst I'm doing this, I notice the line output transformer is different in appearance to the original in Colin's set. Thankfully this is simply a later revision, and the two are electrically identical.
A nicely dated capacitor shows the set was made late 1952.
Tag-strip with line output components re-capped. (There's one on the sound transformer to change yet! Can you spot it?)
Then onto the receiver or RF deck...
Half-way there!
And, once finished, on with some mains!
EHT rises, and eventually settles on ~8KV
And, after some alignment issues with the front end of the RF Deck... pictures :)
Pictures were taken in a mirror...
The set isn't finished by a long-way, as there's much rubber insulated wiring that has degraded and become brittle, but it's proved a point.
So the known good transformer is removed from the set, and an adaptor bracket is made, so we can mount it into Colin's set, as the two transformers are cosmetically different.
And Colin's set produces a good picture!
I check the resistances on all the windings, and between the windings on both transformers. They are identical... I measure the inductances on all the windings, both transformers measure the same. I have another word with Ed, and he asks me to send him both the known good transformer and the re-wound one. Ed finds the only difference between the transformers is the inter-winding capacitance. After a few test windings, Ed succeeds in making two transformers that measure identically.... and, sure enough I fit the freshly re-wound transformer back into Colin's set, and, with bated breath, switch on!
Results!
Back in it's cabinet, and undergoing a soak test for a few hours!
Big thanks to Ed for his incredible diagnostic and re-winding skills, and thanks to Colin, for donating this lovely 20" Sony Widescreen..
Apparently there's a 16" too ... that would be nice ;)
Well, back in August the car came home (link here) , and it's been languishing back in the garage awaiting re-charging of funds.
... took some cussing, and help from the wife....
