Over-voltage reduction - tips?

[Al (astral highway)]

I've just bought a new transformer (epxensive, aren't they?) with 230 pri and 350-0-350 secondary.

When I wired it into the chassis and before wiring it to the double rectifier and + filter circuit, I tested my work and checked out the voltage across the double HT windings. I found 415-0-415, using a precision voltmeter. Now, the manufacturer's spec for the double rectifier diode says 350 max across each half of the valve, so this is a high overvoltage and will shorten the life of this old tube. The other windings to the heaters are 5.0 and 6.3 and read correctly under load.

You'll tell me, of course, that under the correct load conditions, this voltage will drop somewhat, but it won't be nearly enough. What would you do? Of course I could put balanced R's into the anodes such that when the HT+ circuit is under a typical current conditions, the voltage drops to below 350 at these anodes. But is there a neater or better way?

The circuit under question is an old military signal generator with six valves and a cold-cathode cathode voltage regulator.

[Station X]

I'd be inclined to check what the actual voltage is with the circuit working. You don't say what the current consumption is, but obviously the higher it is the more voltage will be dropped in the resistance of the windings. You could use a variac to bring up the voltage slowly if need be. There's no need to worry about in rush current into the smoothing capacitors, as the voltage will build up slowly as the rectifier valve heats up.

If you need to drop more voltage, then resistors in the anode feeds should be OK so long as they're suitably rated power wise.

[Al (astral highway)]

Thanks Graham - pretty much confirming that you'd probably do the same as me, which is reassuring.

And of course, if I do that, the two anode dropper resistors will have to be exactly matched. I don't want a voltage inequality across the two windings.

[Skywave]

Originally Posted by astral highway

I've just bought a new transformer with 230 pri and 350-0-350 secondary . . .

Please excuse me if this sounds daft / obvious - but since this is a new transformer, do you not have, or able to obtain, the manufacturer's data on it - i.e. "350-0-350" at x mA? This will help you estimate the regulation / volt-drop on load.

Al / Skywave

[Al (astral highway)]

Ah, that is far from daft. Indeed, Al, the secondary HT is rated @57 mA, but this is a maximum rating. I didn't think it would run so high at zero load.

[PJL]

Duncans Amps has a free download for a PSU modelling tool that works well...

[Herald1360]

Another little point- nominal mains here is 240V, so you've already got the 350 up to about 365 before you start doing anything.

With a rating of only 57mA on the HT secondary, it will be quite resistive and may drop a lot more on load than you expect.

Otherwise- proceed as advised.....

Cheers

Chris

[kalee20]

Astral Highway - you probably don't want equal value resistors in series with the anodes.

Measure the off-load voltages accurately for the half-secondaries - are they really both 415V off load? They should be equal. This will confirm the half-secondaries have equal turns (which is as it should be).

Now, measure the DC resistances. You'll probably find they are anything but equal. The inner secondary, wound first, will have a shorter turn length, hence lower resistance.

In order to make the rectifier valve work its anodes equally hard, you want the total series resistance to be the same for each half secondary. Adding external resistances gives you the opportunity to do this. You add more resistance to one anode than the other, to bring them up to the same level. And what that level is, you'll have to fiind out in practice, to drop the HT to the desired level.

Incidentally, the regulation of your transformer does seem a little bit high at 415V off load, 365V (corrected for 240V not 230V primary voltage) full load, is 12% drop on load. But, adding your series resistors, although worsening the regulation further, will give the rectifier an easier time by reducing the peak current (I'm assuming a capacitor input filter). The peak inverse voltage for each anode is a bit difficult to pin down by calculation, as it is the DC output voltage plus the peak input voltage. The peak input voltage is 415V x root 2, because there's no resistive voltage drop when the anode is negative; the DC output voltage is whatever it is. If the valve is rated for 350-0-350V input then yes, you will be exceeding its ratings, but with this type of over-run, it's likely to either survive or flash-over - reduced life is unlikely. So you may want to add fuses in series with each anode, to guard against flashover.

Best of luck!

[Al (astral highway)]

Hi and thanks for this detailed post.

Yes, the half secondaries really are 415 off-load, by the same meter that determined 5.0 and 6.3 for the heater windings under load. They are exactly equal, 415.

I was suprised about the 415 reading. Last year, I investigated the mains voltage stability here and found it ranged up to about 8 volts above 230 at various demand conditions, day and night (but never below).

Incidentally, under what conditions is flashover 'fatal' for electron tubes? I destroyed a Svetalana 572B with just one flashover a few years ago.

[HG MICKE]

Before you run up the unit check the rectifier spec, one thing that is quoted is [ Rt ] this is the minimum input resistance for each anode. The value is made up of the resistance of the secondary winding plus a fixed anode resistor , For instance the value of Rt for each anode of an EZ80 is 300 ohms. If the value of resistance is to low it will result in internal arcing of the rectifier.

Colin

[kalee20]

HG Micke's dead right, it would be worth checking this. I'd almost assumed that adding external series resistance would bring the total resistance to well above the minimum, hence my comment about giving the rectifier an easier life. But, check the data anyway. It's good practice to never assume anything!!

Astral, if you have flashover in a rectifier valve, it's probably because the valve has had it anyway. Internal gas, or a flake of cathode material breaking away as a result of a big blow or jar at some point in the valve's history, could account for this. But if you think about it, you'll put a near short-circuit across the transformer secondary: one anode will be conducting normally to cathode, and the other will have flashed over to cathode. Result: one burned-out transformer! So, because the voltages are fairly high for your unit, flashover becomes more likely and it's worth adding fuses. And a single fuse in the secondary CT, or the rectifier cathode, will not protect against the incident I've described - it has to be two separate fuses!

[Al (astral highway)]

Originally Posted by kalee20

HG Micke's dead right, it would be worth checking this. I'd almost assumed that adding external series resistance would bring the total resistance to well above the minimum, hence my comment about giving the rectifier an easier life. But, check the data anyway. It's good practice to never assume anything!!

Astral, if you have flashover in a rectifier valve, it's probably because the valve has had it anyway. Internal gas, or a flake of cathode material breaking away as a result of a big blow or jar at some point in the valve's history, could account for this. But if you think about it, you'll put a near short-circuit across the transformer secondary: one anode will be conducting normally to cathode, and the other will have flashed over to cathode. Result: one burned-out transformer! So, because the voltages are fairly high for your unit, flashover becomes more likely and it's worth adding fuses. And a single fuse in the secondary CT, or the rectifier cathode, will not protect against the incident I've described - it has to be two separate fuses!

Thanks Colin and Kalee20 -- these are both thoughtful and useful responses.

[Al (astral highway)]

Originally Posted by HG MICKE

Before you run up the unit check the rectifier spec, one thing that is quoted is [ Rt ] this is the minimum input resistance for each anode. The value is made up of the resistance of the secondary winding plus a fixed anode resistor , For instance the value of Rt for each anode of an EZ80 is 300 ohms. If the value of resistance is to low it will result in internal arcing of the rectifier.

Colin

Just checked! Rt is 50R for each anode, but someone seems to have 'modified' the circuit and Rt is simply the transformer's internal resistance (which I haven't yet checked.) So thanks for pointing this out! I can now correct the situation when I rebuild the power supply.