Merlin's standby - reduced heater voltages

[Ears]

I like Merlin's Standby arrangement, simple and seemingly effective. See www.valvewizard.co.uk

However I came across this exchange (in of all places Ultimate guitar.com - a place usually full of complete drivel) [ http://www.ultimate-guitar.com/forum/sh ... ?t=1294336 ]
In it a reply to a post takes exception to reduced heater voltages..
discuss .
Richard.

First poster...

Is the way that Merlin talks about it really all that much worse than just having a normal standby switch. Cathode poisoning can occur as long as there is a heater voltage and no anode voltage on the tube. So in that case, wouldn't cutting the heater voltage in half lessen the effect.

Merlin's switch has to be used in conjuncture with a standby switch in the B+ line.

Maybe I am missing something here, that you could explain more to me.

Reply...

The absolute worst thing you can do is run the cathodes at a reduced temperature and apply DC to the anode. That's exactly what happens the way he does his circuit.

Do you see the 47k resistor across the standby switch? This turns the PS into a current source, rather that a voltage source. As long as the cathodes are fully heated, this is not a real problem. The output tubes attempt to conduct, but the resistor causes most of the voltage to be dropped across it, because of the current flowing through it. Life is pretty good.

But when you reduce the temperature of the cathodes, everything changes. The tubes won't conduct as normal. So they won't allow as much current to flow. Now we don't have the idle current that was causing most of the voltage to be dropped across the resistor. So the plate voltage rises. At temperatures slightly below those required for thermionic emission, bad, Bad, BAD physical and chemical processes occur. Ions will bombard the cathode. The cathode is neither cold enough to keep them from embedding in the first surface, nor hot enough to keep them mobile.

I'd rather have the cathodes completely cold and full DC on the anodes, or fully heated with no voltage on the anodes, instead of this condition. If this was the only standby switch possible, you'd be far ahead to not use one at all. Just switch off the mains when not using the amp. Coming through this condition for a very short time as the cathodes come up to temperature is not all that bad. But staying in this condition for any length of time is. "Shocking" the heaters with the power-on surge, isn't a wonderful thing to do. But I'd gladly do that rather than do this.


TBH, I've never tried this in an amp. But from what I know from working with thermionic sources, this is not a good place to be. I predict problems by doing this. But if anyone wants to try it and find out for themselves, by all means do it.

[FYL]

I wouldn't worry: we're dealing with small tubes and guitar amps, not big transmitting bottles and transmitters where cathode poisoning can be a real problem.

Anyway, Merlin's approach allows for a reduced B+ and heater voltage, thus the tubes are still conducting somewhat.

[Merlinb]

I like Merlin's Standby arrangement, simple and seemingly effective. See www.valvewizard.co.uk

However I came across this exchange (in of all places Ultimate guitar.com - a place usually full of complete drivel) [ http://www.ultimate-guitar.com/forum/sh ... ?t=1294336 ]
In it a reply to a post takes exception to reduced heater voltages..
discuss .

The poster's argument would be reasonable if the anode voltage were kept at a normal working level or thereabouts. However, my intention is that you choose a resistor in parallel with the standby which is large enough that the voltages are brought down to some very low level during standby, so that even though the cathodes are not very hot, the tubes still operate well below saturation, so the "bad bad processes" (whatever they are) cannot occur. 47k to 100k is generally enough to drop the B+ down to maybe 10% of its normal value with the heaters at half power. Perhaps I should update my site.

[ampdoc1]

This guy seems to be as knowledgeable as Merlin. Here is his take on standby switches:

[Firestorm]

All reasonable except for the bit about electrons "bumping into the cathode coating" causing cathode poisoning. This is what electrons do. After all, they are bumping into the coating from the other side (from the metal part of the cathode). Is that bad too?

[Structo]

I find all of this confusing.
On one hand you have guys that say the standby switch is not needed for audio amps and were really designed for radio transmitter tubes.

Then you have the guys that say you must have the first reservoir cap before the standby switch so it charges and you don't get the large in rush current when switched to operate.

Then you have the guys that say no, the standby switch must be on the AC side of the rectifier.

Then some that say a 100K resistor across the standby switch allows the caps to slowly charge while the filaments heat the tubes so that the in rush isn't as bad.
I think this is what Merlin recommends.

On my D'lite I changed it so the B+1 filter is before the standby switch so it gets charged first.
I thought that was something that the majority of this type of amp said was best?

But Merlin says that is not a good idea because it can "pop" when switched because of the DC and:

When the switch is thrown one end of the choke is left completely unconnected to anything. Consequently the choke will develop a massive fly-back voltage which could cause arcing in the choke or switch. This could be allayed by adding a protection diode in parallel with the choke (shown faint) but is still not a great design choice. MB

Not sure I understand that.
The choke is connected in series with the other filter caps.

The D'lite like other Dumble amps, has two 220K balance resistors across the first two 100uF caps in series.
Doesn't that also allow them to charge a bit slower and more evenly?

So right now on my amp I have the Standby after the first B+1 filters and also have a .01uF ceramic disc across the Standby switch.
The hot side of the switch has the choke input and the CT of the OT connected then the other side of the choke goes to the rest of the filter string.

Is that wrong or what!
Doesn't this cap allow ripple across it when switched to Standby thereby slowly charging the other filter caps until the switch is thrown?

[Merlinb]

I find all of this confusing.
On one hand you have guys that say the standby switch is not needed for audio amps and were really designed for radio transmitter tubes.

The valves don't need need a standby. But the caps may do (in old amps, or Vox, who still use cheap underrated caps)

Then you have the guys that say you must have the first reservoir cap before the standby switch so it charges and you don't get the large in rush current when switched to operate.

Then you have the guys that say no, the standby switch must be on the AC side of the rectifier.

Now we're talking about the rectifier valve. It seems to be the case that you should not use a standby that allows the recto to warm up without any load current, because they have a tendency to flashover if they are hot and then you throw the switch on a cold reservoir cap.

Then some that say a 100K resistor across the standby switch allows the caps to slowly charge while the filaments heat the tubes so that the in rush isn't as bad.
I think this is what Merlin recommends.

Yes, anything which slows down the rate at which things can draw current can usually be relied upon to improve their life.

On my D'lite I changed it so the B+1 filter is before the standby switch so it gets charged first.
I thought that was something that the majority of this type of amp said was best?

But Merlin says that is not a good idea because it can "pop" when switched because of the DC and:

When the switch is thrown one end of the choke is left completely unconnected to anything. Consequently the choke will develop a massive fly-back voltage which could cause arcing in the choke or switch. This could be allayed by adding a protection diode in parallel with the choke (shown faint) but is still not a great design choice. MB

Not sure I understand that.
The choke is connected in series with the other filter caps.

It only applies if the switch is connected in such a way that with the switch open, one end of the choke is connected to nothing, not even a smoothing cap.

The D'lite like other Dumble amps, has two 220K balance resistors across the first two 100uF caps in series.
Doesn't that also allow them to charge a bit slower and more evenly?

No, they don't have much effect on the inrush to the reservoir.

So right now on my amp I have the Standby after the first B+1 filters and also have a .01uF ceramic disc across the Standby switch.
The hot side of the switch has the choke input and the CT of the OT connected then the other side of the choke goes to the rest of the filter string.

Is that wrong or what!
Doesn't this cap allow ripple across it when switched to Standby thereby slowly charging the other filter caps until the switch is thrown?

No, the cap is just an anti-pop / anti-arcing measure, only DC can charge the smoothing caps. Your method sounds fine.

[Structo]

Thanks Merlin.

A little clearer now.

The amp in question uses SS rectifiers.

[FYL]

Yes, anything which slows down the rate at which things can draw current can usually be relied upon to improve their life.

A PTC is one of the best solutions, provided that you don't switch back on the amp while it's still hot.

[Structo]

[Ears]

I'd say PTC means +ve temperature coefficient thermistor.

Both (negative) NTC and PTC, depending on topology, can be used for surge protection.

[Merlinb]

PTC = positive temperature coefficient. He means a thermistor, (except they're actually negative temp' co')
They're somewhat overrated IMO, because they're not widely available in values useful for a valve amp, and they can't protect against rapid on-off-on switching.

[Ears]

You're up late Merlin

[Structo]

PTC = positive temperature coefficient. He means a thermistor, (except they're actually negative temp' co')
They're somewhat overrated IMO, because they're not widely available in values useful for a valve amp, and they can't protect against rapid on-off-on switching.

Is that the device that is sometimes inserted into a power transformer that trips if the transformer gets too hot?

[FYL]

Is that the device that is sometimes inserted into a power transformer that trips if the transformer gets too hot?

No, a PTC is a thermisistor with a positive temperature coefficient: it's cold resistance is high, hot is low. The PTC heats as current is drawn thru it then stabilizes.

Modern PT's integrate a thermal cut-out, either self resetting (good) or fuse-like (silly).