Solid state relay blows the fuses in my speakers

[szekelyisz]

Hi everybody,

I want to turn my active studio monitors on and off with a solid state relay. I wired up everything, and it worked as it should for about a day, but then suddenly both my speakers' fuses blew. I replaced one of them, turned it on, worked fine. Then I replaced the other one, powered it up, and after a couple seconds both blew again. The SSR works fine, tested it with a light. I'm wondering if the issue has anything to do with the SSR. Before if worked fine for months connected directly to the electric network (no relay). What do you think?

Thanks!

[NiHaoMike]

Sounds like a defective SSR that's creating a DC offset under some conditions.

[Circlotron]

DC offset would only matter if the PSU it was feeding had a transformer. If SMPS then shouldn't care.

[NiHaoMike]

DC offset would only matter if the PSU it was feeding had a transformer. If SMPS then shouldn't care.

Linear power supplies are still very common in audio equipment.

[Circlotron]

Linear power supplies are still very common in audio equipment.

That's true, especially for stand alone amplifiers. But for an active speaker I imagine SMPSs are common nowadays, but not universal of course.

[Gyro]

It sounds as if you are using an SSR that switches on a mains zero-crossings. This is fine for resistive loads, but will cause a transformer in a linear supply (if it is a linear supply) to saturate on every switch on. If this happens every time you switch them on, it will stress the fuses until they ultimately fail.

You can get 'random switching' SSRs that turn on as soon as they receive the control signal, rather than waiting for the next zero crossing.

[Zero999]

A random firing SSR will improve things, but it's not guaranteed never to turn on at zero. You want to turn on at the peak voltage, or add a soft start, such as a NTC (Negative Temperature Resistor) or another timer relay which bypasses a current limiting resistor after a few hundred ms.

[szekelyisz]

Thank you guys! It is a linear PSU and the SSR has zero crossing detection. I'll get a random firing one.

[jonpaul]

Chinese SSR are junk, poor reliability and performance

Use the EU or USA brand not Chinese knockoffs

j

[Zero999]

Chinese SSR are junk, poor reliability and performance

Use the EU or USA brand not Chinese knockoffs

j

Are you sure you're responding in the correct thread? Nothing was mentioned of Chinese knockoffs here. The same thing could happen with high quality products made in Germany. Transformers are known to blow fuses, especially when turned on during the zero crossing point.  Toroidal transformer, which are more commonly found in high quality audio products are especially prone to this.

[jonpaul]

Rebonjour, The SSR market like others has been affected by shortages.

Counterfeit SSRs are rife. Most counterfeits have very poor safety and reliability and performance.

Question to OP is the model/photo/mfg of the SSR in question.

Original genuine or knockoff?

Enjoy!


Jon

[Zero999]

Rebonjour, The SSR market like others has been affected by shortages.

Counterfeit SSRs are rife. Most counterfeits have very poor safety and reliability and performance.

Question to OP is the model/photo/mfg of the SSR in question.

Original genuine or knockoff?

Enjoy!


Jon

How would a cheap knockoff SSR cause the fuse to blow in this instance?

The original poster said it is a linear PSU and the SSR has zero crossing detection. He would get the same result whether the SSR is a cheap knockoff, or a branded one from Germany.

[jonpaul]

Zero999
inductive loads can saturate with any DC bias, like a mains transformer

SSRs have SCR, Triaacs, even a perfect zero cross firings may not reliablely prevent all DC bias eg asymmetrical Vfwd.

Poor quality snubber RC or no snubber is another possibility

Jon

[Zero999]

Zero999
inductive loads can saturate with any DC bias, like a mains transformer

SSRs have SCR, Triaacs, even a perfect zero cross firings may not reliablely prevent all DC bias eg asymmetrical Vfwd.

Poor quality snubber RC or no snubber is another possibility

Jon

You're missing the point: zero crossing increases the risk of a transformer blowing the fuse!

I take your point about asymmetric forward voltage drop, but the fact the original poster was using an SSR with zero-crossing is more likely to be the cause and needs to be ruled out first.

Either way, I'd drop the SSR, use an ordinary mechanical relay and add an NTC resistor, to limit the surge.

[jonpaul]

Zero999, ACK, 100% agree.

We used double break relays, wired for arc blowout.

very likely no thermistor needed if load is under 100w

large multi KVA need a contactor....

Still think the Chinese SSR are failure prone....

Bon journée

Jon

[Doctorandus_P]

The fuses should not blow, even with an SSR.
If the fuses blow because the SSR switches at the zero crossing, there is some serious design flaw with those amplifiers.

A DC bias form the SSR's seems more likely, especially if it takes some time between turning it on and blowing the fuses.

[bdunham7]

You're missing the point: zero crossing increases the risk of a transformer blowing the fuse!

I take your point about asymmetric forward voltage drop, but the fact the original poster was using an SSR with zero-crossing is more likely to be the cause and needs to be ruled out first.

Either way, I'd drop the SSR, use an ordinary mechanical relay and add an NTC resistor, to limit the surge.

In the OP it is stated that the problem occurred after a day of use, and that the fuse was subsequently observed to blow 'after several seconds'.  That's not a turn-on issue.  He also stated that the amplifiers had been working normally before the SSR was installed.  I'm going to guess that a different SSR will solve the issue, but not because it is not zero-crossing. 

[Zero999]

You're missing the point: zero crossing increases the risk of a transformer blowing the fuse!

I take your point about asymmetric forward voltage drop, but the fact the original poster was using an SSR with zero-crossing is more likely to be the cause and needs to be ruled out first.

Either way, I'd drop the SSR, use an ordinary mechanical relay and add an NTC resistor, to limit the surge.

In the OP it is stated that the problem occurred after a day of use, and that the fuse was subsequently observed to blow 'after several seconds'.  That's not a turn-on issue.  He also stated that the amplifiers had been working normally before the SSR was installed.  I'm going to guess that a different SSR will solve the issue, but not because it is not zero-crossing.

Fuses can wear out over time.

The original poster didn't say whether it's happening on turn on, or not.

It could be the zero crossing, the SSR's asymmetry, or both. It is easy to jump to conclusions and I admit I've been guilty of that plenty of times.

[jonpaul]

We can save time on speculation if the OP can just provide more details of the SSR used, and amplifier.

Model #s, a diagram and photos will be useful

Jon

[tooki]

Are they fast-blow fuses? If so, is this a situation where using slow-blow fuses would be a viable solution? (I don’t have enough background in magnetics to understand the mechanics of the transformer saturation problem, hence the question.)

[CaptDon]

Someone enlighten me on why zero crossing turn-on blows fuses with transformers as the load? We always used zero-crossing SSR's in industry to turn on inductive loads. Connecting a transformer during the peak voltage point is sure to cause a huge inrush fuse blowing event! Our safest SSR devices for transformer and inductive loads had an internal bridge rectifier and SCR as opposed to the triac models. The bridge/SCR units were less likely to cause D.C. offset as compared to triac models. For some of the high current (more than 10 amps) devices we created our own bridge/SCR SSR packages. With a really terrible power factor sometimes the turn-off events were weird. I suppose with an inductive load a zero crossing SSR could be bad in a dimming/power regulation application? Our usage simply called for full-on or full-off which is what I supposed is the original posters application.

[Gyro]

Someone enlighten me on why zero crossing turn-on blows fuses with transformers as the load?

Transformers usually rely on some collapsing residual magnetisation from the previous half-cycle to prevent their cores from saturating at mains peak. If you power them on at a mains voltage zero crossing, they will tend to saturate on the first peak, causing a high current surge, limited only by the primary DC resistance. The size of the surge depends on the size of the transformer, how close the transformer is to saturating in normal operation (quality / 'economy'), and type (torroidal tend to run closer to saturation).

Connecting a transformer during the peak voltage point is sure to cause a huge inrush fuse blowing event!

Actually, no. A transformer is inductive, so current lags voltage by quite a margin. Current will rise relatively slowly on a fast voltage step, and the core won't have time to saturate before the voltage has dropped away.

At one point I worked in traffic control systems. The halogen lamp head transformers would audibly 'bong' and cause structural resonance when switched at zero crossing. They also suffered high failure rate. Shifting the triac switching point away from the zero crossing cured both.

[CaptDon]

Interesting, the current rise time with high peak voltage turn-on relationship. I work very often with a particular 1Khz switch speed SMPS and since the SMPS transformer is never allowed to go anywhere near saturation (It has a lot of built in safety margin to unusually high line voltages and long PWM pulse widths at low line) we see a very nice linear current ramp up with a nice clean turnoff. The voltages on the drains of the mosfets on the other hand look horrible!!! The O.E.M. has released a few snubber mods!!! Excessively high ring voltage peaks were killing the mosfets. There was a lot of capacitor energy storage ahead of the transformer, and the transformer had very low primary D.C. resistance. The mosfets left town in a spectacular way when they shorted!!

[Circlotron]

For a transformer that is already running in a steady state, a half cycle of voltage will cause the magnetism in the core to go from fully south to full north. If the transformer is switched on at the voltage zero crossing the core magnetism starts at zero instead of fully south and so by the time the voltage gets to the 90 degree peak the magnetism has already reached fully north and for the remaining 90 degrees of voltage it cannot increase further and so the core saturates and the current goes off the planet.

[Zero999]

In an ideal inductor, the startup current is double if the waveform starts at zero, compared to what it would be if it starts at the peak voltage.