Menahem, thank you for all the information.

Nope, no scope for measuring, it's true that i'm not really well equipped for these kinds of projects. All i have is a low-cost digital multimeter.

Dillen:

Try putting your multimeter in AC-volts mode and measure across the large

filter capacitor and let us know the result. Not the best way to measure ripple but it will
give us an idea of the ripple level.

Measured in same conditions (power & FM on, volume at zero) it reads 68.5

Dillen:

There's an adjustment for the 24V supply on the main board, did

you measure/adjust this after repairing the power supply ?

I did, some, but as i never got it right everywhere, i just left it at a closest acceptable across the board, not very different from the original value.

Answered to your PM, thanks.

Sebastian

68.5 - is that volts or millivolts ?
If volts, the filter capacitor is not doing anything and the rectifier not the best either.
If millivolts, no problem at all.

Martin

Erm, uh... that would be volts i'd say. Never done AC measuring before (nobody slap me please). I measure it with the multimeter at the range of 200V~ that, let me see if i get this right, measures at millivolts precision (hence the .5). So a reading of 068.5. When measured at the range of 600V~ (volts precision), it's a round 069. So yeah, must be volts. Oh dear... I don't like the sound of your answer, Martin.

I do have a bunch of new generic caps somewhere in my spares box, of which a couple of Hermei 1000uF 35V. Should i swap and test with one of them ?

Sebastian

Yes, that's bad news but mainly for the large filter cap.
You can disconnect the old large capacitor and connect a new.
Make sure to get the polarity right.
The BM900 uses positive ground. Use the old cap for polarity reference and DON'T assume red
leads being the positive.
If it does not fit into the mounting bracket, you can use two cable ties crossed
around the cap and bracket.

I strongly suggest you do not power up the radio again before replacing the filter cap.
There's a good chance that the original zener and transistors died from overvoltage
and the new components will do the same.

Martin

Nope, seems it wasn't that after all, i get the same reading with a new cap (and the polarity is good). Could it be something before the cap, even if the cap itself was good ?

Ok, so there is, what, the rectifier also that was suspected, right ? So i rechecked it once more, and this time i took notes of it all (since i already did it before, but now it's impossible for me to say how, to know if i was right or wrong, but more likely wrong, but it's always nice to have proof). Notes are good. Makes a note on making notes.

And to me, the rectifier, now, i don't know. But i'm done guessing, and will leave the analysis to those who are qualified. So ; i desoldered the four leads (so it was completely "out of circuit"), and tested it in two ways, with ohm meter, and a diode test. Between each two points the measure was taken 2 times, +/- and -/+, so a total of 8 measures per test.

With diode test i got an average of 300 in one direction, and 1 (infinity) in the other, each time.

With ohm test i was absolutely out of the simple ohm level (it read 1 immediately, both ways), should that matter. With the multimeter switched to 200k, it averaged to about 17K in one direction, and to around 50K in the other, each time.

So, i have not come this far to not make yet another mistake, and so i venture and say it could be finally the rectifier ?

Sebastian

The rectifier readings sounds about right.
The huge ripple can be caused by the filter capacitor but it could also be a very heavy load on the power supply.
However, that would probably see some power supply components heat up pretty badly and as I understand it, they don't.
Still, a ripple that high would have the radio hum very much (if running at all) so I must admit to having doubts about that reading.
Try measuring AC voltage across cap 514.

Martin

Yeah, no, there's no heating in particular. To tell you, i didn't really notice anything unusual before this ripple measuring started.

Here are the latest measures :

- cap 514 : AC = 56.4V / DC = 25.8V

- power cap (1000uF) : AC = 65V / DC = 29.7V

And here's a photo of the method used to measure the value, maybe you could confirm if it's done correctly (and if i read it correctly) ?

Sebastian

Sebastian

Something is really wrong here.

Whatever you measure electrically, you can't have 56VAC with 25.8VDC - this situation would be catastrophic.

Depending on the circuit, it can be 56VAC with 26mVDC (millivolts), OR 56mVAC with 26VDC.

Can you get a better multimeter to make sure that you get the right readings?

As much as I would like to help you, and Martin is already doing a fine job, it's impossible if we can't get correct readings.

Menahem

yachadm:

Can you get a better multimeter to make sure that you get the right readings?

Oh, so it could be the multimeter, yeah i hadn't really thought of that. It's true that it's an el cheapo and has had a few years of service.

This said, i couldn't really justify to get a top-of-the-line one, but i'll have a look on alternatives today (el less cheapos).

Thanks Menahem.

Sebastian

Even the cheapest multimeters are generally quite useful. Those readings are so odd that it makes me as well think your multimeter's AC range is faulty or there is something really strange going on in the measurement that makes it show potatoes.

Try measuring some known pure AC signal, like the voltage over the dial light bulbs (should show 6-7 volts).

As even the smallest AC voltage range in your multimeter is 0-200V, it will never be too accurate at measuring small voltages... I don't think it is very useful to try and measure the ripple voltages with it.

tournedos:

Try measuring some known pure AC signal, like the voltage over the dial light bulbs (should show 6-7 volts).

Well, the curious thing is, i just tried and it's a steady 6VAC, which on the other hand should prove that it measures just fine ?

What on earth is going on here ?

Sebastian

I actually managed to find the schematic of your multimeter on the friendly interwebs. Seems to me that it simply isn't designed to measure an AC voltage riding on a DC bias, as is the case when you try to measure the ripple voltage on the supply lines. It has half-wave rectification on the AC range and probably shows the DC bias + the AC voltage, multiplied by some fuzzy factor that depends on the AC waveform. Not a big wonder, since proper AC measurement is a complex task and very few, if any, cheaper multimeters do it right.

So, just don't trust it - or if you wish to experiment, put a small capacitor (preferably non-electrolytic) in series with one of the probes when doing the ripple measurement. It should isolate the DC component out of the measurement and maybe you'll see a more sensible reading.

Hey, thanks Mika for your efforts

I actually had the manual of the meter downloaded, after having read something of the sort while googleing yesterday about AC measuring. But i wasn't able to decipher this in the manual, it said being able to read it. So you saw it directly on the schematics ? Wow.

Ok. I read about that method of a small cap in series, but what would one call a small enough cap ? I just tried with one 0.1uF Wima polyester cap, and then one tantalum 330pF, and it read 00.0 with both. It kinda jumps a few millivolts when i move the leads to verify contact, and then zeroes in immediately. So i guess in principle the trick works. But i would've preferred a "small something" reading rather than nothing. How to be sure, now ? Is there anything i could measure with it that should be sure to have ripple ? Note : it gave 00.0 with both the new cap, and the old big one that i resoldered back on to test.

Lord. If anybody can recommend the cheapest good hobby multimeter on eBay, i would've needed a new one sooner or later anyway.

Sebastian

I waited 15 years until I bought my first Fluke, and then slapped myself for waiting so long. I thought that I could get by with Japanese (now read Chinese) knock-offs.

The Fluke just removed all doubt when measuring.

My second Fluke is a model 187, now 5 years old, bought 2nd-hand on ebay, for a fraction of the new price.

It's the best investment I ever made. I wouldn't even look at anything else. And you can always get it serviced, if required.

Just check that the model you choose can do capacitance up to at least 10,000uF as well as the regular AC/DC/Ohm/Diode functions.

Keep your Mastech for doing transistor measurements.

Another excellent tool is the EVB ESR meter (made in Portugal) for checking capacitors in-circuit.

A good (used) Fluke DMM and an ESR meter is a very comprehensive toolbox for troubleshooting electronics.

Menahem

I actually just opened up my 900M (please don't praise me for that effort, I was going to do it anyway ) and took a similar measurement across the main filter cap - I'm not at my "lab", so I only have an el cheapo multimeter here, not unlike the one you have.

Guess what, it shows 77.5 V on the 200V AC scale, and there's definitely nothing wrong with the power supply in this Beomaster.

When measuring through a series cap, the jump of the reading when connecting is normal as the cap is being charged to the DC voltage. I guess the zero reading is about right - as I recall, the accuracy of your multimeter on the 200V AC range was given as 1.something% of the reading, +/- 10 digits! So if there was 2 volts of real ripple, the meter might show anything between 0 and 4V and still be within specs.

I'm afraid I can't recommend any particular meter... Fluke or Meterman (currently Amprobe) should be safe bets, although quite expensive for sporadic use. Don't get rid of your current one anyway, it has the transistor test function which is sometimes very useful, and not too common in even more expensive meters.

I was away for a couple of hours and you guys solved the mystery meanwhile. 

I agree on the multimeter and instrument suggestions above, I use Fluke meters myself,
but I take it you are not planning on setting up a complete and stylish repairshop
but rather deal with the things the "kitchen-table" way. Correct me if I'm wrong,
in which case, as Mika suggests, the cheaper multimeters will do fine in most cases.
That's also my experience and most will easily tell you if something is wrong or right,
no need to go full steam ahead at this point.
There's also some meters that don't like measuring 100Hz AC or AC that is
not a nice sine wave, even more expensive ones.
A scope may be the better investment if it comes to that.

But recap the poor thing and let's see what happens then.

Martin

Magnificent.

yachadm:

A good (used) Fluke DMM and an ESR meter is a very comprehensive toolbox for troubleshooting electronics.

Menahem

Thanks Menahem, but i saw the Fluke's while browsing for meters on eBay just now. I could never afford one, not for the use i would have for it. Not today anyway. If i would really use it at least every week, i'm sure that it could justify itself, but as of now it's just once or twice every 6 months. I just couldn't. However i saw that you recommended an UNI-T UT70A in another thread, and that i could do.

For the ESR meter, i've seen it and sure it would be nice too, but it'll be for another day also. I'm in shock already finding myself repairing a 45+ years old B&O radio, you must leave me a few minutes to adjust before convincing me into believing that i'll be able to do it daily and need pro equipment  But i've written the references down for the future, so one day...

tournedos:

I actually just opened up my 900M (please don't praise me for that effort, I was going to do it anyway ) and took a similar measurement across the main filter cap - I'm not at my "lab", so I only have an el cheapo multimeter here, not unlike the one you have.

Guess what, it shows 77.5 V on the 200V AC scale, and there's definitely nothing wrong with the power supply in this Beomaster.

When measuring through a series cap, the jump of the reading when connecting is normal as the cap is being charged to the DC voltage. I guess the zero reading is about right - as I recall, the accuracy of your multimeter on the 200V AC range was given as 1.something% of the reading, +/- 10 digits! So if there was 2 volts of real ripple, the meter might show anything between 0 and 4V and still be within specs.

Yeah the Mastech is rated at 1.2% for AC. This is great news, and really the only confirmation needed then. It was the multimeter. Whew. Thanks so much for the trouble, even if you meant to service your 900M anyway. That was very nice of you.

So i guess it's back on track, then.

Martin has, in his infinite kindness, proposed to send me a recap kit free of charge, in exchange of some detailed informations about my specific subversion of the 900K, which apparently isn't the most common kind (well, uncommon enough for him to not have encountered one).

So i'd like to take profit of this unusually calm moment on my "restoration" project, where everything seems once again going well (before something else will come up), to repeat my sincere thanks to everybody around here, and in particular those who have helped me directly. And of course in first place Martin, who really has shown a patience out of this world.

This said, i won't say more, because it's still not finished

Cheers guys.

Sebastian

sebastian:

Martin has, in his infinite kindness, proposed to send me a recap kit free of charge, in exchange of some detailed informations about my specific subversion of the 900K, which apparently isn't the most common kind (well, uncommon enough for him to not have encountered one).

Both of you!

To tell the truth, I didn't know about this kind of problem in AC measurements myself before this, as I've always used a scope to check for ripple... So this has been a highly educational thread all over

Yes Mika, I'll second that - I also learnt something new here. Actually on the Fluke 187, it's so stable, that if I see fluctuations on the AC or DC scale (decimal points as appropriate), I'll suspect ripple, and connect my scope to confirm it. If the Fluke is rock stable, I know there's no ripple.

Those Uni-trend UT models are no longer suitable for our type of audio work. They changed their models, and I couldn't find anything over 100uF, let alone the minimum 10,000uF I'd want.

Perhaps check new Amprobe, as Mika suggested. But I can tell you that plenty used Fluke 8x or 11x go for well under $100 on ebay. You don't need a 187 like mine.

Menahem

The next chapter ;

i thought to report on how it went, so that maybe one day it can help someone else, in any way, as my good Karma payback.

So, got the recap kit (thanks Martin !), nicely documented, and attacked the job.

Luckily this is old circuitry we're talking about, because my soldering equipment isn't what you would call the most advanced kind (nor in the best condition, to top it all). But it went well. Well, i'd say so, and i won't show you any too-close-up photos so that you could tell me the contrary  

Turns out that this was indeed a version of the 900K that was still partly undocumented in Martin's information vault (just how unlikely is that ?!). So photos were taken. Notes were written. PM's exchanged. And here's a résumé of all that. A recapping of a Beomaster 900K, subversion S 290, export model with a six input voltages selector, and whose wooden cabinet has a stamped date of 8 JAN 1965.

This is an extremely image-heavy post, so my apologies to those who it might derange, but personally i would have loved to find one like it when i first encountered my radio, so here goes. The first few photos are for identification, of the untouched unit before the recapping. You'll notice (well, i just did) that the volume and radio dial buttons are reversed (the volume button has the black point) for the moment, during the project. It goes pretty much in the following order ; the original unit, the recap kit, and then the recapped unit.

Front

Back

Power unit ; 1 cap to change

Tuner ; 1 cap to change

Bottom side ; in my model version there are 12 caps to change (+ the two 800uF output caps that are fixed to the back wall)

Top side (on topmost there are the amp board on the left, and the stereo decoder on the right)

Amp board top ; in my model version there are 17 caps to change

Stereo decoder top ; in my model version there are 6 caps to change

Stereo decoder bottom

The recap kit from Martin

Bottom after recapping

Amp board after recapping

Power unit after recapping

Here are all the original caps from the main board, with their respective designations on the schematics that i have. After discussion with Martin, i've come to learn that these designations may well vary between model versions and/or schematics, so it just might not be the same ones for you. But this is the complete set on my specific model/version.

And here's the same thing for my amp board, to be considered with the same grain of salt.

And here's something i discovered only while taking the photos, even though i had manipulated the board frequently during days before (i've bent the cap slightly to make a point, i hadn't noticed it at all before). A severed + leg on a cap on the stereo decoder board. Just for the curiosity factor. It'll be fixed by the recapping of the stereo decoder board, coming in the next chapter.

So there, for now. What remains is the recapping of the stereo decoder, and the single one on top of the tuner box, both to be done in the coming days. I've also completely redone the wooden cabinet and given new luster on the gray speaker grilles with some polishing. It's starting to look good

Sebastian

Good work, and good pictures! Thanks for taking the time to post them.

Seems like your Beomaster has had a pretty good life. Mine have surface rust on the metal parts of the connectors, screws and on the chassis, probably from years in the attic or something. Especially the oldest which is a S290 as well, only 1390 serial numbers away from yours

Thanks Mika,

yes, for it's age it's pretty pristine, just old but no structural deterioration, not one bit of rust. Same for the speaker elements, not the smallest hole. Except for the fuse holders that i changed before anything else, it was just a question of lubricating and fixing the component problems. It would indeed look like it has spent it's life indoors in the dry, like it sure deserves.

It's pretty amazing actually that you own another that is as close a cousin of my one  Here's to them  

Sebastian

Well, I don't see any substandard work - looks like you did a good job!

Check those carbon resistors which are in the Power supply circuit, especially those whose function is to drop voltage - you'll probably find a couple out of spec.

Did you do a voltage check across the power cap, as before?

I'd be interested to see the difference in measurement results.

You're very lucky that the structure is in such good condition without corrosion.

Enjoy it for many years to come!

Menahem

Nice job, I agree.
Worth checking the 1,5Kohm resistor that sits right next to the cap with the severed pin on the stereo decoder.
It looks chipped but may still be OK.

Martin

yachadm:

Check those carbon resistors which are in the Power supply circuit, especially those whose function is to drop voltage - you'll probably find a couple out of spec.

Did you do a voltage check across the power cap, as before?

I'd be interested to see the difference in measurement results.

Well, in fact they all seem to be just a bit above the supposed values, or close enough, so nothing to report really.

As for the power cap, no real change in values, i just noticed that depending on the input voltage it's more or less close to specs (normal, i suppose). So at 220V it's 34,5V, and at 240V it's 32V. Given that 34V is aimed at, i guess that operating it at 220V it is, then.

Dillen:

Worth checking the 1,5Kohm resistor that sits right next to the cap with the severed pin on the stereo decoder.

It looks chipped but may still be OK.

It measures 1.55K, so seems ok as well. Looks like these dinosaurs of resistors do resist to time too.

So far, after the recapping and a few measures at same points around the power supply, things seem just a tad more balanced. I managed to adjust to a spot where i get around 24V/34V where needed, but the same problem than before remains at below voltages. This is to say that i have 34,6V on the power cap, it goes to 24V & 24,2V after the 2N555, from there it goes to the AC128 (24,2V / 24,4V / 34,6V), and it's after that, at AC126 that it goes a bit weird, and i don't know what it means ; i get 24,4V / 8,6V / 8,7V where i should have 24,4V / 9V / 9,4V ?

To the same effect, the amp board gets exactly 24V at arrival, but for example at the first input transistors i measure 7,9V / 8V / 16,5V where it should be 6,8V / 6,9V / 18,5V. And so it goes, everything below 24V is a little off.

Any idea what is to be looked for ?

[EDIT] I also get a somewhat high "bzzzzz" from the transformer which was already there from the beginning, but maybe just a bit louder now. [/EDIT]

Sebastian