SOUND ON SOUND

[James Perrett]

I would say that the 6800 uF smoothing capacitor is most likely duff. Even if your 'scope trace was to scale, around 1.5V of peak-to-peak ripple is excessive.

That was my first thought too.

Me too. It was probably fairly marginal when new and getting worse with age.

The diode data sheet is at
https://www.mouser.co.uk/datasheet/2/11 ... -70454.pdf

which shows they are 6 amp diodes.
The parallel caps are probably 100000pF (100nF or 0.1uF) as the first two numbers are the value and the last number refers to the number of zeros after the first two numbers and this is usually measured in pF.

[ef37a]

There's a good explanation of simple linear unregulated power supplies here, complete with formula to calculate the ripple:

http://www.zen22142.zen.co.uk/Design/dcpsu.htm

As you'll see from the expected waveforms in that article, your supply appears to have very little smoothing going on. So it may well be that the big electrolytic has dried out and/or gone open circuit, or there's a dry solder joint in its connections.

It shouldn't be too difficult to replace that C with a good quality new one, and you could even increase the value if you want. Just make sure you get the right voltage rating and fit it the right way around.

Indeed and I notice the incumbent is a 105C component. Due to its proximity to the transformer probably best to keep to the same rating.
I would go for 10,000 or, as said two 4700mfd. Don't go raving mad with the value increase though, the transformer won't like it.

Dave.

[wireman]

I'll get a replacement capacitor and see how it goes. Bear in mind that two are not likely to fit if they are a similar size to the existing one.
I did look at RS a while back thinking that I was going to open up the supply and see a bulging capacitor and was completely amazed at the low lifetimes of some of these electrolytics (as bad as a light bulb). So I think that is another reason to be sure of the temperature rating as I assume operating well below it dramatically increases the lifetime.

When I first saw the scope trace I assumed the dips were due to a faulty diode but realise the dips occur at the wrong frequency and I should have paid attention to the offset so both diode paths must be providing a voltage.
One of the ceramics has a black spot but as long as those fail open I presume this is not an issue, one of my meters can measure capacitance but probably not up to the capacity of the electrolytic.

It will be an interesting learning exercise even if I do end up just getting another supply as a replacement/backup.

[jjlonbass]

wireman - Are your 10 ohm and 20 ohm resistors in series i.e. you're applying a 30 ohm load to the power supply?

John

No, they are made up from multiple resistors in a metal box that I made up a while ago.
10 ohm at 500W
20 ohm at 100W

Not that you put in anything like that power for long as the box heats up.
The box has 4mm sockets.

Ah, I see, I should have paid more attention to your original post.

In that case, you can expect about 1.1V peak-to-peak ripple with the 20 ohm load and 2.2V with the 10 ohm. Is this what you are actually seeing? To be certain we need to know the Volts / division setting for the 'scope channel showing the ripple and whether you are using a x1 or x10 probe on that channel.

Note that to take the full 4A specified assuming that the output will remain at 15V, you'd need a 3.75 ohm load resistor and the ripple would be of the order of 5.88V peak-to-peak.

These ripple figures may sound poor, but your equipment will most likely have internal linear regulators for small signal parts which will remove the ripple. If the power amps are of the traditional class B or class AB type, these are very good at rejecting ripple if correctly designed. There may be a switch-mode DC to DC converter for the power amps or the amps may be class D types, in any case the ripple should not be audible.

John

[wireman]

The diode data sheet is at
https://www.mouser.co.uk/datasheet/2/11 ... -70454.pdf

which shows they are 6 amp diodes.
The parallel caps are probably 100000pF (100nF or 0.1uF) as the first two numbers are the value and the last number refers to the number of zeros after the first two numbers and this is usually measured in pF.

Thanks, useful info.

[Trevor Johnson]

one of my meters can measure capacitance but probably not up to the capacity of the electrolytic.

Don't forget to discharge it first!

[Folderol]

Funny things diodes. They actually have three current ratings. There's the average one, 6A in this case, then the seldom quoted peak one, 5-10 times as much, and finally the startup surge one prolly around 500A. Also each diode is only conducting for half the time so that effectively doubles the rating (well, not quite double).

With caps, as well as value, voltage and temperature, look for ripple current rating. The higher it is the cooler the cap will run, and the longer it will last. It is related the ESR (effective series resistance), which correspondingly wants to be a low as possible. All of this lot has to be balanced with price of course!

[Folderol]

one of my meters can measure capacitance but probably not up to the capacity of the electrolytic.

Don't forget to discharge it first!

Via a resistor!
NEVER discharge a cap by sticking a screwdriver across it! That can do considerable damage - and not just to the cap itself :(

[ef37a]

Modern capacitors are deal smaller than those of even just a decade ago so you might get two 4700 in there? Even if not another 1000uF would help and at 22V they are pretty wee.

I do get hissed at some of the thoughtless designs* around"! With just a moment's thought that PCB could have been put closer to the transformer and the cap fitted on the diode side and given at least some shielding from traff heat. The bit of fibre board looks like an afterthought. If you have a bit of gash 1mm ally you could back that up as a heat shield but I would guess a new cap(s) will last another ten years anyway.

*I had a charger for a mobility scooter in a few years ago. An 80mm fan at one end blew air FROM the power transformer (around 200va) OVER the rectifier/regulator PCB. Needless to say, PCB burned up!

Dave.

[wireman]

Have ordered a few capacitors to try but in the meantime I discovered something quite interesting.

There is a circuit analysis tool at this site: https://www.falstad.com/circuit/

You can build a circuit and choose a component to measure in an 'oscilloscope'.

So I built a bridge rectifier (before realising this is a pre-defined one you can pick).

With a little experimentation I found that I can get a similar waveform to the ones I posted in this thread by adding a resistor (around 2 ohms) in series with the smoothing capacitor. This must approximate directly to the resistance (specifically the ESR) of the capacitor getting really high.

[James Perrett]

With a little experimentation I found that I can get a similar waveform to the ones I posted in this thread by adding a resistor (around 2 ohms) in series with the smoothing capacitor. This must approximate directly to the resistance (specifically the ESR) of the capacitor getting really high.

Just checked the ESR of a few high value capacitors on my cheap Chinese component tester and all gave values of either 0.01 or 0 ohms. Not sure how accurate it is but 2 ohms is certainly far too high.

[ef37a]

I would not get diverted about "ESR" in capacitors?

Only of interest to beardy, tweaky audiophools who think audio goes through the power supply! Maybe designers of switch mode power supplies?

Dave.

[James Perrett]

The ESR reading on my cheap Chinese tester seems to give a good indication of whether a capacitor needs replacing or not. Or at least that holds true for all the dodgy capacitors in the Fostex G24 that I'm working on - one of the few pieces of gear where a wholesale capacitor replacement is justified it seems.

[ef37a]

The ESR reading on my cheap Chinese tester seems to give a good indication of whether a capacitor needs replacing or not. Or at least that holds true for all the dodgy capacitors in the Fostex G24 that I'm working on - one of the few pieces of gear where a wholesale capacitor replacement is justified it seems.

Possibly James but I would trust a harmonic distortion test far more. If the capacitors have been chosen with low LF distortion in mind they will show an increase in THD at say 40Hz ref 1kHz long before they restrict the LF response.

Dave.

[Folderol]

The ESR reading on my cheap Chinese tester seems to give a good indication of whether a capacitor needs replacing or not. Or at least that holds true for all the dodgy capacitors in the Fostex G24 that I'm working on - one of the few pieces of gear where a wholesale capacitor replacement is justified it seems.

Possibly James but I would trust a harmonic distortion test far more. If the capacitors have been chosen with low LF distortion in mind they will show an increase in THD at say 40Hz ref 1kHz long before they restrict the LF response.

Dave.

I think you're mixing up ESR with EMC Dave.
ESR is Effective Series Resistance