Use of Class X2 Capacitors as Droppers.

['LIVEWIRE?']

The link on your most recent post works, Chris, but the one on your post # 58 still doesn't, yet I can't see any difference between the two!!!

[Philips210]

Originally Posted by 'LIVEWIRE?'

The link on your most recent post works, Chris, but the one on your post # 58 still doesn't, yet I can't see any difference between the two!!!

Hi.

Although the two links look identical, placing the mouse over each link reveals the problem with the first link. In Firefox you can see the full web address appear in the bottom left of the screen. The first link is missing quite a few characters.

Regards
Symon.

[julie_m]

The forum software alters the displayed version of links, inserting three dots to replace characters from the middle in order to make the link displayable even on small monitors; while leaving the machine-readable version intact.

If you highlight the human-readable version and middle-click, you will get an invalid link. You have to either right-click the link and choose "copy link location"; or middle-click it to open it in a new tab (thus proving it works), highlight it in the address bar, then tab back and middle-click in the editing area to insert the highlit text.

[Herald1360]

That explains it. I just highlighted the visible link in another post then ctrl C copy followed by ctrl V paste in the new post.

[TrevorG3VLF]

Originally Posted by Leon Campion

A poor man's true RMS meter:

Obtain two pilot lamps from the same source, rated to light below full brightness at the required RMS heater current. For a set with 100mA heaters, I would use a 150mA lamp.

Place one in the set's heater chain and connect the other to a DC power supply and supply it with the required current - 100mA in this case.

Either subjectively or (better) use a film camera exposure meter to judge when the two lamps are at equal brightness, adjusting the dropper capacitance to achieve this. The heater RMS current in the set then has the required value.

It's a remarkably effective method. You can interchange the lamps and take an average value to mitigate against the effects of mismatch, but within the allowable tolerance of heater current, it's good enough.

Leon.

VMARS are having a discussion on calibrating an RF ammeter, this can be done using DC or low frequency AC. They work by measuring the temperature of a heated wire.

This device would seem to be ideal to measure the effective current in a valve heater string.

[Radio Wrangler]

Thermocouple ammeters are getting a bit thin on the ground. They are rather fragile and haven't been made for a long time.

More easily obtained is a "True RMS" multimeter. Their AC current ranges use the same rectifier/processor so the current ranges are true RMS also.

But even with a capacitive dropper in the circuit, causing the current to be out of phase with the applied mains, the current and voltage on each valve heater will still be in phase with each other, and will be sinusoidal unless the mains is distorted. So an ordinary AC meter with the usual peak/RMS fiddle factor intended for sinewaves should be OK.

David

[TrevorG3VLF]

A simple current or voltmeter will do if the capacitive dropper is driving the heaters only.
If the dropper supplies the HT as well, there will be a DC contribution in the heaters. I have used 5spice to look at the effect of varying HT curent but the version I have will not calculate power. Perhaps LTspice or TINA will do it.

[G8HQP Dave]

With low pass and high pass filters it might be possible to separately measure the AC and DC components of the heater current, and so calculate the total power. However, given that the capacitor, by design, is dropping some voltage there may also be a contribution from the 'hole' left by the rectifier charging pulses. Only a true RMS meter will cope with that.

[TrevorG3VLF]

I have done a transient analysis, using 5spice.
I used 3uF driving 600R, a diode and three different HT loadings, 1k, 3k, 5k with a 47uF bypass capacitor.

As can be seen the positive side of the heater waveform is clipped.
I am surprised that the negative side appears to be unaffected even though there must be a DC component.

As said previously, I cannot get the heater dissipation.

[kalee20]

Yes, there's a DC component. The wave shape of the negative half cycle looks unaffected (as expected), but there is a negative bias.

This again is as expected - the HT rectifier takes current in one direction, so the heater current must be offset in the other direction to maintain zero average current through the capacitor.

[G6Tanuki]

To equalise the effect of the rectified-HT-component is why quite a few half-wave power rectifiers designed for AC/DC radios had a tapped heater - you connected the dial-lamp bulbs across the tap to provide an extra path for the 60-odd milliamps of HT current.

I'm also told that back in the 50s and 60s in some rural areas fed by small substations, during TV-viewing hours there was often a noticeable DC component on the mains because of the asymmetric effect of half-wave rectification.

[dinkydi]

Hi Trevor,

Thank you for your instructive transient analysis.

If you have a spare moment, for comparison it would be very informative to have a companion analysis of Robert Casey's preferred configuration. Casey's configuration uses a smaller capacitor, let's choose 2uF, but otherwise keep the other component values as you have chosen.

For Casey's circuit the capacitor is in parallel with "a diode and three different HT loadings" including the bypass capacitor, and that combination in turn drives the 600R.

I imagine that the new transient waveform becomes magnified/extended in the place where it was previously clipped and this is offset by a lower amplitude of the fundamental.

Peter

[Herald1360]

I reckon you'd get near enough to the power by treating the two halves of the waveform separately. The bottom half will simply be half the power of the corresponding "complete" sine wave. The top can be treated either as half a sinewave of lower amplitude or perhaps as half a simple square wave, depending on the degree of clipping. Maybe someone more analytically inclined could work out the error doing it that way?

[TrevorG3VLF]

Hi Peter,

Yes I can do what you want but am not too clear on the circuit. Could you post a sketch?

I have also done a simulation using a 600 ohm resistor insted of the input capacitor, simulating a resistive dropper cable.

I have even managed to find out how to convert the strange output file to jpeg!

Trevor

[dinkydi]

Hi Trevor,

Sounds like you're cooking with gas!

Here is a sketch of Casey's dropper configuration.

Peter

[Radio Wrangler]

Yes, the HT rectifier current certainly mens that a true RMS meter is essential, and this would also be true with resistive droppers.

With the spice simulation, it would be interesting to also plot the rectifier current. This issue gets worse with larger smoothers. The voltage drops less through the dead part of the cycle, and then the rectifier turns on a bit later on the rising edge of the mains, and less time is left to re-charge the smoother back to the peak. Consequently the current through the rectifier becomes a shorter pulse, of greater current.

I use LT Spice, which is a freebie and was originally created and tweaked for power supplies so it handles fast transients and slow control loops together very well. I use it for GHz microwave work as well. There are internet user groups and lots of guidance on importing 'foreign' libraries.

David

[dinkydi]

Yes, the HT rectifier current certainly means that a true RMS meter is essential, and this would also be true with resistive droppers.

But this would not be the case if separate droppers (capacitive or resistive) were employed for the HT and filament string, as has been suggested.

Peter

[G8HQP Dave]

If you use a capacitive dropper for the HT alone then you need something, either another diode or big resistor, to deal with the impossibility of passing DC through a capacitor. A diode turns it into, roughly, a voltage doubler. A resistor simply replicates the heater chain problem, but simplified (because you don't care too much about the resistor current) and wasteful (the resistor heat is just thrown away instead of heating cathodes).

You can't pass DC through a capacitor!

[Radio Wrangler]

If you went for seperate droppers, as Dave says, the HT one couldn't be a capacitor. You don't want a voltage doubler so that means a resistive dropper, and if you're doing a resistor in the HT path, why not put it in the DC side of things between a reservoir and a smoother. OK the reservoir has to handle full rectified mains, but the dropper and smoother give you a good bit less hum on the final DC. The capacitor dropper for the heaters saves you more energy than the HT dropper handles, so you've got most of the savings.

But I really don't like live chassis. Give me a proper transformer any day!

David

['LIVEWIRE?']

David(and others) I am still thinking of going down the Class X2 Capacitor route for a combined Heater/HT dropper for reasons stated in my other posts, but I'm a little puzzled as to why a capacitor can't(shouldn't?)be used to drop the HT. It will be in the Mains side of the rectifier circuit so will not be in a DC path (or am I missing something here?) I have worked over 40 years at repairing radios, etc., but admittedly have never before converted a 120v radio using capacitor droppers.