help w/ start-up current-inrush limiter for tube amp

[cyclecamper]

Help. I need some advice, or maybe just an exact parts spec advice.

I need a current-inrush limiter to protect the power transformer while the power supply caps fill at turn-on. The tube amp is rated at 180 watts output. B+ is 500V. This big heavy CLC-filtered bass guitar head now has so much capacitance in its big last-stage filter caps and the even bigger cap in its damper for that filter that it has a lot of current inrush at turn-on, and then it continues to fill the damper caps thru a resistor for a while. The bridge rectifier module I have can probably take the punishment if I put it on a heat-sink, but turn-on is too dangerous for the old stock power transformer that wasn't designed for the abuse of my tinkering.

So I figured I would add a thermistor in series, between the bridge and the filter ladder.

Then I started looking at the thermistor devices.

You can use a PTC thermistor to protect the transformer just before it blows up...the PTC allows high initial current while cool, and gets hot as excessive current flows and increases resistance to limit current. A PTC thermistor acts kind of like a fuse. Its temp kind of models the transformer temp,and limits as it gets hot. It kind of ruins your transformer's regulation, but only just before it would blow up.

It's more common to use an NTC thermistor as a more conventional initial-current inrush-limiting device. When it starts out cold it has high resistance and limits the inrush current right from the start, unlike the PTC! As it gets warm it starts letting more current thru. By the time the power supply caps are full it's warm and stays warm during use. I guess this is the more normal use. What I worry about is the time it takes to change temp. Done carefully, this should not interfere with playing loud, and a momentary loss of power of any duration should not be a problem when it comes back on either, if you set the cap bleeders to drain at about the same rate as the caps fill thru a cool thermistor and as the thermistor cools? Sounds like it could involve some experimental testing, and then require adding or subtracting thermal mass, heat-sink disspation surface area, or insulation around the thermistor, as well as consideration of the environmental temp.

Some people put an NTC thermistor on the power transformer PRIMARY, in series with the wall supply. That seems a bit removed from the cause of the problem, and assumes the transformer IS working transferring excessive current; saturation might prevent protection.

I see some mfgrs say you can put NTC thermistors in series, up to a few hundred volts. But I didn't find any that say I can use them on a 500-volt line like my B+. Thinking about how they'd split the load in various states (assuming they don't act in exact unison) makes my head hurt.

Then I started thinking about getting it completely OUT of the circuit after turn-on. Probably not necessary if properly specified, but worth discussing.

I was thinking that I could have the thermistor connected via the normally-closed contacts of a relay, and convert the standby switch to a momentary toggle with another latching relay. That way, if power is ever momentarily interrupted the amp reverts to standby mode and has the cold thermistor in series. I could even devise some find of failsafe-lockout so that the standby momentary spring-loaded toggle (think of it like a button) wouldn't work unless the caps were filled enough to reach a reasonable B+ voltage. But then why even use a thermistor? If I'm going to short across it in normal use I'd might as well use a resistor with more consistent behavior.

Or I could make my amp turn on like my gas clothes dryer, where I have to hold down a button or lever until a light comes on, then I can let go and it stays on.

Surely this is old territory others have gone over before. So please clue me in. What makes the most sense?

So far I'm in favor of making 'ON' mode (as opposed to standby) only connect the B+ supply to the load as the latched-on mode of a relay that's normally in 'standby' mode. And 'standby' mode would always have a resistor or some kind of thermistor limiting current. But 'on' mode would take it out of the circuit.

If I wanted the amp switches to look and operate more traditionally, I could make the "standby/on" switch a conventional toggle, and get a make-then-break relay so that when the switch is 'on' it would connect the latch power but disconnect its direct power to the relay. So it would look and work like a conventional standby switch, except if power is momentarily interrupted the user has to turn it back to standby then one again.

There must be a better way of making the standby completely automatic, and have is come on only when B+ is reasonable and the heaters are hot.

Then there's problems switching high-voltage DC using relays of reasonable dimensions. If I have extra contacts and want to use more than one set, would I be better off putting the contacts in series (to divide the voltage) or parallel (to divide the current)? The problem is they never actuate at the exact same time, and for voltage you need speed and distance of breaking and certain contact materials, and putting more contacts in series won't really help much, yet I hear it recommended all the time. I also hear that it's bad to momentarily interrupt the power...though I don't understand why; the filter ladder makes sure B+ is OK first and anything else can never exceed it, and my dampers make sure the power filter will not over-shoot.

What about the use of an NTC thermistor in a hot tube amp? Would it only work correctly the first time you turn the amp on? If you turn it off for just a minute or two and then back on, the amp is still hot; if the bleeders depleted the caps much you'd want the thermistor cool. So would I want smaller bleeders and to mount the thermistor near the outside air?

What do you think is the best way to limit in-rush at start-up? I'd like the amp to have a somewhat conventional Fender appearance and operation.

If I use an NTC thermistor, can you recommend one to try first?

I guess I could forego the relays etc. and try to get the inrush and the bleed and the cooling rate of the thermistor to match or at least to be safe. I know I'm not the first newbie to over-do the power supply for a guitar amp, I'm just OK with what it will do to the sound because I'm going for a clean bass amp, not some saggy swampy bluesy thing.

 

[cyclecamper]

Then my next question is why the mfgrs don't put more fuses in an amp, especially where they do no harm. For instance, why isn't the filament supply fused? Any problems there would make the power transformer just as useless as a problem in the B+ circuit, but a filament short might not even blow the main primary-side fuse on a transformer this big.

So where should I solder in additinal fuses, and where shouldn't I?

 

[cyclecamper]

I'm finding more resources on the web now for specifying an TNC Thermistor for my application.

 

[nigelwright7557]

Manufacturers are under pressure to save money so leave out fuses where possible.

A thermistor is the cheapest solution.

 

[Enzo]

A thermistor in the mains works because the inrush current is large, relatively speaking, then settles down while the NTC heats up. Your thermistor, however you arange it, must heat up to work. How much current do you expect to have flowing out your B+ after the first couple seconds of power up? Enough to warm up a thermistor? If idle current is enough to heat your thermistor, then every note you play will make it hotter, reducing teh B+. SOunds like more sag than anyone would want to me.

I for one am not worried about your transformer from starting surge. The transformer is just wire. The amount of heating it might see for 1 or 2 seconds won;t warm it up enough to harm anything. Inrush is a danger to rectifiers, especially tube ones, and conceivably caps. And a large inrush can arc your power switch contacts and burn them. But not the transformer.


Inrush limiters are intended as slow start devices, not active safety valves. They slow down the first charge up a little then get the heck out of the way the rest of the day. And that is why they are generally in the mains lead. Your inrush lasts only a second or two. That is not enough time for a bridge rectifier to overheat. It might exceed its surge current, but that is not the same as the package overheating. If you burn out on of the diodes in a bridge, it was because the diode iutself vaporized, but because the thing had no heat sink.

Have you looked at surge currents for rectifiers? The lowly 1N4007, a 1A rectifier has a 30A surge rating.

Many many many amps have NTC inrush limiters, the time they take to warm up is minimal.. They don;t take ten minutes, they take a very few seconds.

Why don't they fill the amp with fuses? In some places they do, as required by law. In the USA all those secondary fuses are not required, so they don;t go to the added expense. One problem with them is that they cannot guarantee anyone will install proper fuses once they own the amp anyway. Ther is also the problem of nuisance blows. A fuse opens, then an amp owner has to disassemble his amp to change that fuse. The factory doesn;t want anything to get in the way of a positive impression on an owner.

180 watts is not so huge an amp. It is way short of an SVT and even shorter of a Peavey Classic 400. Those are 300 watt and 400 watt tube amps respectively. 180 watts is just a tiny bit larger than a 160 watt Peavey Mace. Go look at what those designs did. The Mace has no inrush limiting, the Classic 400 uses a triac as a power switch, and those triacs have a surge rating of something like 250A. The original SVT had no inrush limiting, the modern reissue, the SVT-CL has a simple 3A inrush device.