How much DC on your mains ?

[emeritus]

In the 1970's, before mum had central heating, she used to do her sewing in a small room. A single 1kW bar of her Belling electric fire got it too hot, so I wired a silicon diode across the switch of one of the elements to provide a 500W setting, which was just right. I wasn't too bothered about the resulting 2A of DC as I knew that contemporary TVs used half wave rectification, so though t it should have been alright.

[Herald1360]

Originally Posted by Lucien Nunes

.

A historical note about egregious dumping of DC into the mains: One of the first attempts in the UK at a fully electronic theatre-lighting dimmer was Strand's 1949 thyratron-based unit that came to be known as 'Woody's Electronic' after J.T.Wood, its designer. Probably to sidestep a patent by Century in the USA who were working towards a similar goal, Wood started out by using three thyratrons per channel, one on each phase, to provide controlled 3-pulse rectification feeding variable voltage DC to the lamps. The entire current used by the installation was DC and the sum of the three line currents, totalling some hundreds of amps of DC, returned via the neutral! Not surprisingly, one of the first installations to be conmissioned (possibly Reykjavik Opera House) resulted in the neutral of the theatre's main supply cable burning out. It had other problems including thyratrons running away, so the company soon returned to electromechanical dimmers pending the arrival of thyristors.

They did a range of saturable reactor control systems as well in the '50s and early '60s before thyristors took over. Oddly enough, the only type of theatrical dimmer system where I ever got my hands on the inner workings.

[Craig Sawyers]

Originally Posted by Station X

No posts have been deleted from this thread.

I must be cracking up. Perhaps I typed it, then went somewhere else without sending it!

Craig

[Radio Wrangler]

The DC on AC is exactly like the issue we have with single-ended audio output transformers. They must be gapped enough and large enough in cross sectional area to avoid saturation, and still have enough primary inductance to cover the wanted frequency range. The same happens at substation transformers. THEY acted as the DC return path for the rectifiers of every telly in their area. It was a significant design concern.

For an alternator running with a three-phase resistive load, the torque the generator takes from whatever is spinning it comes out as constant and smooth. Forces on the stator windings vary through the cycle, but transfer from phase to phase to keep the total constant.

Now add some non-linear loads. Rectifiers peak charging, discharge lights igniting only on the tops of the cycles and any thyristor variable on-time switching.

You now get non-sinusoidal currents - worse! ones with ruddy great jumps in them. Back at the generating station, these hit the alternators making mechanical shock loads which can shake alternators apart, fatigue off blades in turbines and create mayhem. The nature of rectifiers with capacitive loads is that all the mini-shocks cluster at the peaks of the waveforms at stable positions in the rotation of the rotors. Shafts can get bent.

The power industry really doesn't like non-resistive loads.

THat's why requirements for power factor correction came in for larger appliances. It also allows you to get more power before fuses pop.

My TIG welder rectifies mains (no reservoir) and feeds an inverter to charge the reservoir bank. tHis inverter is modulated to make the current drawn sinusoidal following the mains voltage. In this way it fakes a resistive appearance to the mains network. THe reservoirs feed another inverter that converts down to the welding voltage, provides safety isolation and regulates the welding current. Enough inversion yet? Ha! no way! that then feeds a third inverter Most of the time it's switched fully one way or the other to give me DC of whichever polarity I choose. Good for steels. For aluminium, the third inverter comes to life and gives me AC. I can have it switching for square wave, and I can vary the frequency mark-space ratio to vary between heating and cleaning effects. But it's noisy and as it's a posh machine I can select sinewave and it pulse-width modulates the third inverter to make the sine. It's less powerful, but easier on the ears. Oh and I can still vary the frequency and mess with mark-space distortion. S when I'm sticking aluminium antenna bits together, It's nice to know I'm using triply-inverted electricity!

What the regulations missed was that they decided that small appliances didn't count, they couldn't afford PFC. However, they hadn't a clue how much tiddly little devices would increase in number, and they all sync up their current demand pulses to the mains waveform.

What proportion of generated power now goes into electronic loads, now that it includes every light bulb, charger and gizmo?

All the smartness in the metering world isn't going to affect the hammering the alternators get.

David

[Craig Sawyers]

Originally Posted by Craig Sawyers

I must be cracking up. Perhaps I typed it, then went somewhere else without sending it!

Craig

Definitely cracking up - it was actually a post I made to a DC on mains thread on Golborne

I have just tried two autoranging DMMS on the mains, AC volts, both about 233V. However on DC volts, one said -0.3V with the leads either way round, the other +2mV leads both ways. I guess I was hoping a bit too much from them, neither blew up.

[PJL]

The very low resistance of toroid primaries makes it believable that it could cause the core to saturate with only a small DC mains component. The capacitor approach may sort out DC but a screened isolation transformer would also help clean it up.

Do top end hi-fi amps still use mains transformers?

[The Philpott]

Question: Is this DC bias/distortion phenomenon on the mains what causes iron ballast fluorescent light fittings to suddenly change their tune?
Dave

[Craig Sawyers]

Originally Posted by PJL

Do top end hi-fi amps still use mains transformers?

A mixture. Class D can use either a switcher (either integrated with the amp, or separate) or conventional mains transformer.

Likewise class B or AB - a mixture of both.

One of the most expensive (mono) units, long obsolete - the Halcro - used two switchers one after the other. The mains facing one was power factor corrected, and the second one derived the rails that were used by the "conventional" amp. That was a FET output with both feedback and feedforward to give sub-ppm distortion.

https://www.stereophile.com/content/...plifier-page-2

Everything in the bottom half is power supply.

Craig

[Valvepower]

Hi,

Interesting thread.

We experience noise problems with larger toroidal transformers and as part of the QA process for toroidal transformers we apply a simulated “DC” mains supply to the transformers to check their susceptibility to DC. The gadget we use is a 100-150 watt old school filament light bulb in series with a 1N4007 diode, which is switched across the mains supply and we listen for excessive noise coming from the transformer.

We have found the ‘DC’ susceptibility of the transformer lies in choice of core material and the magnetising current. We have also noticed the quality of the bonding of the transformer core has an effect on the noise.

Regards
Terry

[trobbins]

That's interesting that you QA for mechanical noise. I guess that form of QA naturally happens for individual testing of large parts with a human operator nearby, but could be easily missed with mass production automated testing.

Trying to measure the DC component on mains appears to me to be an indirect method of trying to measure the more important characteristic of AC primary current waveform and its asymmetry and peaky-ness that is the driving aspect of the toroidal noise. Measuring current waveform is safer, easier and has no filtering type influences applied to it.

[Craig Sawyers]

Originally Posted by trobbins

Trying to measure the DC component on mains appears to me to be an indirect method of trying to measure the more important characteristic of AC primary current waveform and its asymmetry and peaky-ness that is the driving aspect of the toroidal noise. Measuring current waveform is safer, easier and has no filtering type influences applied to it.

That is fine if you have a current probe, which most people do not.

Craig

[Diabolical Artificer]

The old thread didn't come up in a search I did, odd, so it's not a new phenomenon.

I've no issue here on our supply AFAIK but will check. We live in a small village so are maybe less afflicted.

I find this subject hard to get my head around, as I tend to think of mains electricity ( and DC to a point being uni-directional, I mean a schematic goes from left to right, IP and an OP, that is unless you stop to think in terms of current. This multi dimensional aspect of electronics is tricky, at least for me ( and probably other beginners).

As far as I understand it then, devices that use halfwave rectification and thyristors/traics cause a DC component to be fed back into the mains. So, does this DC offset just sit there or is peroidical, IE only present when a large amount of folk are drying their hair or using drills?

The comment about thyratron dimmers made me think of the old teletype machines once common in use that used a thyratron PSU, was this an issue?

Andy.

[GrimJosef]

The DC can be time-varying. I went through a phase of servicing a few pieces of US-manufactured high-end audio kit. Their toroidal mains transformers had 2 x 115V primaries, wired in parallel for over there and in series for over here. I suspect they were designed with 60Hz in mind though, so they'd be under more stress in a 50Hz country anyway. And of course our mains voltage is hardly ever as low as 230V. All the same, the transformers were reasonably quiet during the daytime. But from early evening onwards they started to buzz. I live quite close to a large supermarket which starts its in-store bakery ovens around that time. Coincidence maybe ?

Cheers,

GJ

[trobbins]

LEM make a wide range of current transducers, and due to smps applications, some models are quite plentiful and cheap and even on ebay (eg LTS6). When used with a soundcard and spectrum software they are awesome for audio span. I recently used an LA25 and scope to check out some transformer mains in-rush levels, with the LA25 in a mains AC pass-through box.

Just trying to make the point that there is a disconnect between a toroid transformer making a hum, and trying to measure some level of DCV on the mains.

Gone are the days when you needed to own a TEK A6302/AM503 wand/plug-in combo.

[Craig Sawyers]

Originally Posted by GrimJosef

T Their toroidal mains transformers had 2 x 115V primaries, wired in parallel for over there and in series for over here. I suspect they were designed with 60Hz in mind though, so they'd be under more stress in a 50Hz country anyway. And of course our mains voltage is hardly ever as low as 230V. All the same, the transformers were reasonably quiet during the daytime. But from early evening onwards they started to buzz. I live quite close to a large supermarket which starts its in-store bakery ovens around that time. Coincidence maybe ?

Cheers,

GJ

Should have mentioned that - the 60Hz/50Hz issue is quite a serious one. The core properties need to be relevant to the lowest frequency used. The two pieces of US kit that were part of my system at some point were a Krell KSA100 class A amp, and an Audio Research 100W valved amp. Both had transformer hum as a result of being 60Hz designs, quite apart from DC on mains issues.

In fact the Audio Research sat on the bottom shelf of my rack, on the top of which was my record deck. I was plagued by a hum - but only when the stylus was on the record. Turned out it was mains transformer vibration shaking the rack. I had to site it remotely from the rack for that reason.

Craig

[eddie_ce]

My understanding of the 18th Edition of BS7671 and its amendment 1 is that the situation could get worse with the increase of EV (Electric Vehicle) charging points. One of the current requirements is for automatic disconnection of the supply if a DC current of >6mA is detected. This will not necessarily be achieved with a standard RCD type A.

Other members more involved in this technology may be able to add further information.

[Lucien Nunes]

The comment about thyratron dimmers made me think of the old teletype machines once common in use that used a thyratron PSU, was this an issue?

Nothing wrong with thyratron control itself, just that the Strand dimmer only used the positive half cycle from each of the three phases. A full bridge using six thyratrons would have been fine although extravagant; a more conventional solution would have been a back-to-back pair running from one phase, as universally adopted with thyristors. Any phase-angle control can distort the mains waveform and increase the neutral current due to the presence of harmonics, but at least it should do so symmetrically.

Where multiple thermionic or vapour devices are used as a rectifier or control bridge, there is an inbuilt tendency to favour common-cathode arrangements. Solid state, however, is ambidextrous - take a pair of diodes with a common cathode, another pair with a common anode, connect their AC terminals together and you have a bridge rectifier. The common-cathode pair is easy to do with valves too, but the common-anode pair is a pain. Unless they are indirectly heated with cathodes isolated for the full peak-peak voltage, each needs a separately isolated heater supply (hence three for a bridge, one at HT+ potential and one attached to each end of the HT secondary.) If they are controlled devices, the same is true for bias supplies, so it all gets a bit complicated.

Arc rectifers are easy to build with as many anodes as you like (6 was common, 12 was available) but only the one cathode. Each anode draws from the transformer secondary a pulsating DC current just like the Strand 'electronic' dimmer, however, the transformer is integral to the rectifier plant and wound specially to accommodate the DC in its secondary. Not having a transformer, the Strand dimmer passed the DC directly into the mains.

[hannahs radios]

Having reread this topic again reminds me of something I noticed that happened a few years ago. I have a large 25 amp 13.8 volts PSU using a conventional transformer. I've had it since it was new it hums quietly, except once or twice I noticed that it was rattling a bit and at the same time there was a noise on my radio like a thermostat opening. As soon as it had opened the physical noise from the PSU stopped. I opened it up and checked inside, there arw no thermostats in the PSU in fact the only switch is the mains on/off there's no electronics in the primary side.

Does anyone have any thoughts what might have caused it to rattle like that?

[Mullard_Nut_74]

Just a question.
Could the DC on the mains be caused by someones poorly installed Solar aray and/or Battery storage unit?

If the answer is yes then DC on the mains will be a growing issue as more and more people get solar and Tesla style wall packs.