Ceramic Capacitors - Where can I use them?

[G8HQP Dave]

Some small ceramic caps have a colour code (or was it 'had' a colour code?). A black band (like a hat) across the top often means NP0/C0G. Colours mean higher permittivity and possibly higher nonlinearity. I am not sure whether the colours were standardised, and I am not sure whether modern caps still use colours, as most of my stock is NOS stuff. Anyway, these will almost certainly have too low a voltage rating for much use in valve circuits.

[bobskie]

Well i have 4 more caps to replace now, the radio was sounding good in testing but then i lifted up the speaker (it was lying flat on the table) it is motorboating? (I think that's what it's called) very badly, the vibrations must have been absorbed by the table and were unnoticeable. I am fairly confident it wasn't doing it this badly before. I still have 4 more to go though so hopefully that will fix it, could this be caused by the ceramics though? Also i must add it makes this noise even with the volume turned down.

[paulsherwin]

It is not being caused by ceramic caps.

The lengthy posts by Dave and others are about specialist caps used by designers of professional equipment and aren't relevant to mending old radios.

The chances are you have made a wiring error.

[Biggles]

Could be IF instability caused by poor decoupling. May be capacitor related.

[bobskie]

I will have a check over my wiring, interestingly it seems to occur after a while, i can hear a clear audio signal for a good few minutes before the hum builds and you can hear it in the background. I still haven't replaced those four though but I doubt it has anything to do with them.

[Radio Wrangler]

I wouldn't say NP0/C0G and X7R ceramics are special in any way. They are the mainstream of what is being made now. NP0 and G0G are great for tuning applications, X7R are best kept for decoupling.

Think of them as simply modern codes for what used to be termed 'ordinary' and 'high-K' ceramic capacitors.

The very high density jobs for replacing small electrolytics aren't commonplace yet, but they're coming, and once their price drops below what they can replace, the switchover will be fast.

David

[TrevorG3VLF]

Motor boating is a very low frequency oscillation and will be caused by low frequency feedback. Check the electrolytic capacitors.

[G0HZU_JMR]

I don't have any experience of fitting ceramic caps to radios like this Bush model. However, I would be a bit cautious if I didn't know the specs of the capacitor. If I assume the voltage rating is adequate then there's the problem of temperature stability. Some cheapo dielectric materials can shift >70% in value if in a hot environment. Looking at the Bush schematic there are several caps used as filter (tone) caps. I assume you would want these caps to be fairly stable over temperature as the radio warms up?

I doubt distortion would be an issue but then again I've not got the experience of working with various ceramic disk caps rated for high voltage.

Hifi nuts would fall on their sword if they had to endure -50dBc harmonic distortion (unless nobody told them it was there) and I'd expect that a typical modern ceramic cap used in a filter (tone control?) would need quite a change in signal voltage to produce distortion at -50dBc. Several volts pkpk? This contribution to overall system distortion would be completely irrelevant in a vintage radio anyway. There will be ceramic cap types that will be worse than this but as always, it's best to consult a datasheet or maybe do a few tests to see how the cap performs if distortion is a concern?

Some classes of modern SMD ceramic cap suffer quite severe capacitance derating if you run them up towards their rated voltage. This is a hidden trap for modern designs where the trend is to move towards tiny little 0402 sized caps as much as possible to save PCB space. Generally, the smaller the package for a given capacitance and voltage rating and class rating the worse the capacitance derating is as you head towards 60% of the voltage rating. For example, 0402 caps will be much worse than 0805 even for the same capacitance and voltage rating and class (eg X7R = class 2). I don't know how much this reads across to big old school high voltage ceramic disk caps but it is a trap for the unwary in modern design for sure.

[G0HZU_JMR]

Looking at the schematic of the Bush radio the final valve V4 looks to have quite a high anode voltage. How much voltage swing would there be here at AF at highish listening levels? I'd imagine that the caps used here would need to be spec'd carefully to maintain a fairly steady capacitance over the full swing range. If the caps can drift with temperature it will affect the tone as well. I have no experience of this radio and I don't know what class of ceramic capacitor is being tried there or its voltage rating so maybe someone can advise what is needed here?

[peter_sol]

If you have ceramic caps of suitable value and voltage, just use them, they will be ok.
There is too much speculation as to the do's and don'ts of such things.

[Radio Wrangler]

The tone correction capacitor across the output transformer of a single-ended class-A stage is normally rated for several hundred volts because of the transients that can be generated by the stored energy in the primary. Whatever dielectric is used, it will need an equal or better voltage rating.

David

[paulsherwin]

1kV is normal in that position. It is usually a waxie. I have a bag of 1.6kV 0.01uF polyesters which I use as replacements, though the original can often be left in place as some leakage will have little or no effect.

[Radio Wrangler]

Just watch out for the odd set with the tone correction capacitor wired anode-to-ground rather than anode-to-HT (EG AR88) failure of this part will do a lot of damage, while failure of one across the transformer just makes the set go quiet until it's replaced.

David

[G0HZU_JMR]

I guessed it might be rated at 500V to 1kV but wasn't sure. If there is a lot of voltage swing there with respect to the voltage rating then I guess some classes of ceramic cap will be prone to distortion. But probably still not relevant.

But if you know the capacitance vs bias voltage characteristic of the chosen capacitor and you know the signal level and dc offset it is possible to simulate the distortion the cap would introduce in a circuit. This would only be a crude ballpark analysis but modern simulators can accept a lookup table model for a capacitor where you give the capacitance vs voltage as a table or 'map'. This would be used in a simulator to predict the distortion. I've done this type of analysis quite a few times over the years for diode analysis and non linear resistor and capacitor analysis. Such a model is very crude on its own but often very revealing and the simulation is really simple to set up and quick to run. On the simulator I use here it runs fluidly almost like real time on real test gear showing IMD and harmonic distortion levels.

[G0HZU_JMR]

I dug out an old Murata designers kit for GRM series ceramic caps and in this very large kit of caps there are class 1, 2 and 3 ceramic caps available. Sadly, these are only low voltage caps (50V typical) but I tested them with 4Vpkpk test waveforms.

I made up a simple test jig and tested three 100nF ceramic caps in a simple RC lowpass filter with the caps biased at 5V and with a two tone test waveform with 4Vpkpk max at the envelope peak at the capacitor itself.

The first cap was a class 1 ceramic cap COG 100nV 25V and this was essentially distortion free as you can see in the image below. There are some tiny IMD terms but these are in the blocking caps used in the jig. So this cap is very linear as expected.

The second cap was a class 2 ceramic cap X7R 100nF 50V and this gave moderate distortion as you can see below.

The third cap was a class 3 ceramic cap Y5V 100nF 50V and this gave quite noticeable distortion as you can see below. The distortion is this cap was highly dependent on the bias voltage and this is because of the capacitance vs voltage profile of this cap.

It would be interesting to see how a typical high voltage ceramic disk cap would perform with large AF waveforms. Obviously the class of the cap will be important here as well as its voltage rating. At a guess there won't be an issue when it comes to using one in a typical vintage radio as the AGC and the AF stage and the speaker would be the main causes of distortion along with propagation effects as well?

[peter_sol]

The motorboating may be the smoothing caps even if you don't have much hum