Low studio voltage[/audio quality,equipment failure,etcetc

[epp]

yea we put all our electricity in 2000 year old holes

[jrakarl]

Originally Posted by Mixing Suite

Great advise!!! 9 time out of 10 this is the culprit with older wiring. This would definitely be my next move.

I had a dodgy 'fuse' on the front of my house, which was built in the late 50's. We spent a good amount of cash last month replacing the connection to the house plus wiring up to the electricity meter and then to the fuse box.

I also use an isolating transformer (massive terroidal) which supplies clean power. I've had it under my mixing desk without any worries. Based on some previous posts, should I be relocating this?

Sent from my GT-I9300T using Tapatalk

[cinealta]

Originally Posted by drBill

What about Orange County? Developed even later than LA, and most towns there have everything underground.

Cost is an issue as well. OC is rich, many "planned" communities. Cost of excavating is exorbitant compared to stringing on a pole (and no fiber optic communication companies paying for it). LA population growth exploded (pre-WWII). No planning. No money. Haphazard jury-rigging.

[Geoff_T]

Originally Posted by cinealta

Hey Geoff, you have to remember that cities like modern London have undergrounds because they were built on thousands of years of previous civilization (Roman garrison Londinium established c.43 AD). Whereas, a town like Los Angeles has only been around since about 1770 when the Spanish came, so no previous underground cities. Thus, we have to put most everything in the air.

Hi

With respect, the Romans have very little to do with anything modern day except the paths of their road system and the baths at Bath.

I was born in a victorian era house in Watford. All my memory of that town, and those surrounding, all the services except telephone were underground. Transformers were on the ground in buildings or fenced off enclosures.

Electricity was invented well over 100 years ago. Los Angeles has expanded from desert and orange grooves over that same period. They had plenty of time to incorporate the electricity underground with the water, gas and sewerage systems laid in the ground.

Other cities and countries did it....

And we now have major chunks of the city covered with power poles and lattices of high tension wire.

[drBill]

Originally Posted by cinealta

Cost is an issue as well. OC is rich, many "planned" communities. Cost of excavating is exorbitant compared to stringing on a pole (and no fiber optic communication companies paying for it). LA population growth exploded (pre-WWII). No planning. No money. Haphazard jury-rigging.

Agreed. But I don't think when a city was/is built has anything to do with whether or not lines are strung above or below. It's all about planning and foresight....

[Bill Way]

I've used one of the power "regenerators" from PS Audio, and it makes gear sound about as good as it can. They are geared to the audiophile market, so the stuff's expensive, but the build quality is outstanding, and I've heard nothing but good about their customer service.

WW

[juniorhifikit]

Slightly off-topic, but a long while ago a friend of mine was graduating university as a civil engineer, with not very good marks. He said the easiest place for him to get a job with his low marks was in California.

I've found not only the electrical to be a bit shoddy in California, but the roads, roadsigns and general planning/civil engineering way below the rest of the country. The Three Stooges comes to mind...

[Geoff_T]

Originally Posted by juniorhifikit

Slightly off-topic, but a long while ago a friend of mine was graduating university as a civil engineer, with not very good marks. He said the easiest place for him to get a job with his low marks was in California.

I've found not only the electrical to be a bit shoddy in California, but the roads, roadsigns and general planning/civil engineering way below the rest of the country. The Three Stooges comes to mind...

Hi

Also off topic, I totally agree...

The roads are so bad with pot holes that I had to have the suspension of my car re-tightened as the bolts that held the front wishbones to the sub frame had shook loose from the constant hammering. When they resurface roads, manhole covers are not raised and reset so they are 1" to 2" below street level.

Also, roadworks have the arrow sign redirecting traffic against the last vehicle in the road works causing jams rather than cones and warnings 50 yards back.

The whole infra structure was set up by morons because, in a normal world, the water consumption, sewage disposal and electrical supplies are calculated as each development is made. There should be no reason for voltage variations and drop outs if planned correctly.

Creeping back on topic, this is one of the main reasons for drop outs and incorrect voltages... You would not design a product with insufficient headroom in the power supply to handle the load... nor should a city have insufficient current to supply the load without voltage drops.

I guess it's 50 years in the UK then coming here and being blown away by the obvious inadequacies...

[nosebleedaudio]

I didn't read all the posts so: I would look to see how many if any were on the same power transformer, then see if you could check the voltage there and compare with yours ...
Old dirty connections are 100% worth checking...

[Fleaman]

Originally Posted by Geoff_T

Hi

Also off topic, I totally agree...

The roads are so bad with pot holes that I had to have the suspension of my car re-tightened as the bolts that held the front wishbones to the sub frame had shook loose from the constant hammering.

It's way worse in NYC...

[Geoff_T]

Hi

Like a third world poverty driven country... Obviously expenditure spent elsewhere.

In CA when Arnie was Governor they dug a hole for him to be photo op'd refilling it.....

Posted from my iPhone 4S

[TheSoundSteward]

Just a couple of comments on Line Voltage as it affects Studio Gear.

Assuming that the power supplies in the gear are linear supplies with transformers, rectifiers, filter capacitors and voltage regulators, a good designer will design the power supply for at least +/-10% line voltage variation over nominal and will make sure that everything works fine both when the power supply is lightly loaded (maybe even in standby or bypass mode) or maximum load (All channels going, every LED lit up, meters at maximum, etc.

This is harder than it may first appear, and based on the many pieces of audio gear that I have tested at light load, high line and heavy load, low line conditions which are the two extremes -- there are many pieces of gear that fall short. At high line conditions, the rectifiers, the filters and the regulators must handle higher than normal voltages so the voltage ratings on those parts must be able to handle the highest expected voltage to prevent damage. The higher raw voltage may cause the regulators to dissipate more heat than normal and they may even go into thermal shut-down mode.

At low line conditions with the heaviest load, the filter capacitors will have maximum ripple due to heavy loading and minimum average voltage due to the low line. If the lower ripple troughs drop below the minimum input/output differential voltage of the regulator, then 120 Hz ripple will show up at the output of the regulator.

I have seen commercial audio gear that had regulators falling out of regulation at 113 Vac. (BTW it was a mic preamp). I designed a new power transformer for that one, because the original had wrong secondary voltages.

The performance of the gear will suffer in varying degrees depending on how the circuitry handles ripple and voltage variations from the regulators. Some circuits reject ripple handily, but other gear really freaks out. In general, it is best to make sure that all the power supplies in your unit have enough AC line Voltage so they stay working correctly at all times..

[Geoff_T]

Originally Posted by TheSoundSteward

Just a couple of comments on Line Voltage as it affects Studio Gear.

Assuming that the power supplies in the gear are linear supplies with transformers, rectifiers, filter capacitors and voltage regulators, a good designer will design the power supply for at least +/-10% line voltage variation over nominal and will make sure that everything works fine both when the power supply is lightly loaded (maybe even in standby or bypass mode) or maximum load (All channels going, every LED lit up, meters at maximum, etc.

This is harder than it may first appear, and based on the many pieces of audio gear that I have tested at light load, high line and heavy load, low line conditions which are the two extremes -- there are many pieces of gear that fall short. At high line conditions, the rectifiers, the filters and the regulators must handle higher than normal voltages so the voltage ratings on those parts must be able to handle the highest expected voltage to prevent damage. The higher raw voltage may cause the regulators to dissipate more heat than normal and they may even go into thermal shut-down mode.

At low line conditions with the heaviest load, the filter capacitors will have maximum ripple due to heavy loading and minimum average voltage due to the low line. If the lower ripple troughs drop below the minimum input/output differential voltage of the regulator, then 120 Hz ripple will show up at the output of the regulator.

I have seen commercial audio gear that had regulators falling out of regulation at 113 Vac. (BTW it was a mic preamp). I designed a new power transformer for that one, because the original had wrong secondary voltages.

The performance of the gear will suffer in varying degrees depending on how the circuitry handles ripple and voltage variations from the regulators. Some circuits reject ripple handily, but other gear really freaks out. In general, it is best to make sure that all the power supplies in your unit have enough AC line Voltage so they stay working correctly at all times..

Hi Steve

It's not hard to design a unit that accommodates line voltage drops, it often the manufacturer skimps on the transformer to save costs.

If you took the basic formula that the dc produced is 1.414 x the ac volts (I don't have a square root two symbol on my keyboard) then a 24 secondary would give a dc (after rectification and smoothing) of 33.94 volts.

A 24v regulator generally needs a minimum of 4 volts headroom between the dc input and the dc output so it will dropout of regulation at 28v dc... = hum city.

If the ac input drops 10%, the dc follows to 30.5 volts and you still have a smidgen of headroom.

I am reminded of one (I won't name) OEM that used an 18v secondary transformer, that gave 25.45 volts pre regulator and was then surprised at how noisy the power supply was on load....

Give a bit more from the transformer specs and there should not be a major issue... or go to switching supplies that go down to around 80 volts.

[JohnRoberts]

Perhaps a newbie design mistake. When I worked at Peavey we had to design products to work all around the world, and there are many countries and regions with much worse power variation than in the US. Another odd problem I recall was in Oz where they would sometime encounter extremely high mains voltage, when sparsely populated areas tried to compensate for line losses over too long power runs, by running really hot at the start, to still get usable voltage at the very end of the line.

Prudent experienced designers already understand this and deal with it, some new guys may not have figured it out yet.

We now have better tools to work with thanks to LDO (low drop out regulators) and switchers. A newer LDO regulator may even rescue a marginal design. If it drops into the same PCB layout since standard 3 terminal regulators needed a volt or two to work.

JR

[Richard Crowley]

Yes, and what a good little sheep you Californians are. If we can attribute sufficient intelligence to the politicians, that is their plan. To choke off all the conventional sources of power to force everyone to shingle their homes and workplaces with Chinese solar cells, and put up windmills (at least until the bird lobby discovers them). At least that is one promise Obama delivered on. Doubling gas prices.

[Geoff_T]

Originally Posted by rcrowley

Yes, and what a good little sheep you Californians are. If we can attribute sufficient intelligence to the politicians, that is their plan. To choke off all the conventional sources of power to force everyone to shingle their homes and workplaces with Chinese solar cells, and put up windmills (at least until the bird lobby discovers them). At least that is one promise Obama delivered on. Doubling gas prices.

Hi

Count me out with the sheep bit... You can see my views on the loopy admin in my recent posts

It's the morons who keep voting the same guys in, irrespective of the damage it does...

We are about 48th in USA state tables for schooling but have the highest paid teachers, lots of other statistics and a state billions in the red because of overspending and poor budgeting ...

I don't see a fix either

Posted from my iPhone

[Fleaman]

Oh no.....politics

do

not

go

there

[TheSoundSteward]

Originally Posted by Geoff_T

It's not hard to design a unit that accommodates line voltage drops, it often the manufacturer skimps on the transformer to save costs.

If you took the basic formula that the dc produced is 1.414 x the ac volts (I don't have a square root two symbol on my keyboard) then a 24 secondary would give a dc (after rectification and smoothing) of 33.94 volts.

A 24v regulator generally needs a minimum of 4 volts headroom between the dc input and the dc output so it will dropout of regulation at 28v dc... = hum city.

If the ac input drops 10%, the dc follows to 30.5 volts and you still have a smidgen of headroom.

I am reminded of one (I won't name) OEM that used an 18v secondary transformer, that gave 25.45 volts pre regulator and was then surprised at how noisy the power supply was on load....

Give a bit more from the transformer specs and there should not be a major issue... or go to switching supplies that go down to around 80 volts.

Although a really good power transformer (toroid or conventional laminated) can have limited performance due to cost constraints, it seems to me that the main reason for linear power supplies not properly handling high and low mains voltages properly is lack of taking into consideration all of the things one must take into account when designing a transformer/rectifier/filter/regulator system for an audio power supply.

The starting place for finding the peak of the AC waveform is indeed 1.414 x the RMS secondary voltage. In the case of a nominal 24Vac secondary that is 33.94V. One must, however subtract from that peak value, the forward voltage drop of the rectifiers. Assuming a full-wave bridge circuit, there are two diode drops. Approximately 3V total for Silicon Carbide Schottky rectifiers (best for audio), 2V for either standard recovery (not recommended) or Fast, soft-recovery silicon diodes (much better), and about 1 volt total for Regular Schottky rectifiers. The specified nominal secondary voltage must be adjusted to accommodate the type of rectifiers being used for optimal performance.

At the filter capacitor(s), the top of the ripple waveform is the peak ac voltage less the rectifier drops referenced above. The capacity (uF) of the filter cap(s) will determine the peak-to-peak ripple with the heaviest loading causing the most droop of voltage from the peak.

Most standard regulators actually only need about 2.5 volts Input/output differential to maintain regulation, but that 2.5V is from the lowest DC voltage at the bottom of the ripple waveform compared to the regulated output voltage. The increase in ripple between light load and heavy load can cause the ripple to increase so that the bottom is too low. Low drop-out regulators can have less than 1V I/O differential required. If heavy loads cause too much ripple, then the filter caps need more uF. This is another parameter that must be worked out carefully.

When the line voltage varies +/-10% the secondary will also change the same percentage. One must choose the voltage on the secondary to be high enough to maintain regulation at the lowest expected ac line. One must also choose the Voltage rating of the caps to safely handle the maximum possible voltage at high ac line and light load. This voltage may be surprisingly high due to transformer regulation discussed below.

Another factor that must be considered is the regulation of the transformer -- not to be confused with the DC voltage regulators. The power transformer has DC resistance in its primary and in its secondary which prevents the peak voltage from coming up as high as the theoretical turns ratio. Power transformers are always rated with the secondary voltage specified @ xxx Amps. This means that the open circuit voltage on the secondary is fudged up so that it ends up being the specified secondary voltage only when the secondary is loaded down to the specified load.

If our theoretical 24 Vac secondary transformer discussed above has 20% regulation, the unloaded secondary would actually put out 28.8V not 24Vac.
Let's say that the transformer secondary is 24Vac @ 1Amp. When loaded with a resistive load that draws 1 amp, the secondary voltage would load down to the specified 24Vac. Thus, when transformer regulation is taken into account, keeping a power supply happy can involve much wider voltage swings than those caused by the drop in mains voltage alone. If the load current varies in this gear (due to relays & LEDs and LED Metering etc) this will cause an even wider secondary voltage variation than that caused by the low ac line by itself.

Cheaper transformers may have smaller cores and more turns of smaller diameter wire -- This results in poor regulation because the DCR of the windings is higher than a transformer made with fewer turns of fatter wire on a bigger core. The lower the DCR of the windings, the better the regulation and the less variation in secondary voltage due to loading. The down-side is that a tranformer with better regulation will be physically larger than one with poor regulation. (It will also most likely cost more).

For what it is worth, always try to specify a 50Hz/60Hz transformer over a 60 Hz only transformer. The 50 Hertz rating requires more turns and more core and therefore less radiated stray field than a 60Hz only unit -- all other things being equal. You may also want to purchase a transformer with a higher current rating than is necessary in order to get much better transformer regulation.

I will still maintain that getting all these parameters right is not a trivial task, but the rewards of getting it right will be a piece of gear that will be reliable and work as advertized over the entire range of specified line variations.

[Geoff_T]

Hi

Thanks for the more detailed explanation that I avoided for brevity's sake.

Besides the capabilities of the power supply the elephant in the room is the anticipated load to be applied, a load whose initial inrush current can be far higher than the normal operating current.

The Neve 10** family generally take around 110mA but the first instant of turn on, the current can be many times that, enough to trip out the over current sensing of the regulator.

So a rack of eight such modules should take less than an amp... The 880mA... But you wouldn't fire that up with a two amp supply without tripping the over current, so I would recommend a five amp supply to feed such a eight module rack... With proportions less for fewer modules.

As the power supply forms part of the audio path, there can be no skimping with either voltage or current headroom.

Posted from my iPhone

[TheSoundSteward]

Originally Posted by Geoff_T

Hi

Thanks for the more detailed explanation that I avoided for brevity's sake.

Besides the capabilities of the power supply the elephant in the room is the anticipated load to be applied, a load whose initial inrush current can be far higher than the normal operating current.

The Neve 10** family generally take around 110mA but the first instant of turn on, the current can be many times that, enough to trip out the over current sensing of the regulator.

So a rack of eight such modules should take less than an amp... The 880mA... But you wouldn't fire that up with a two amp supply without tripping the over current, so I would recommend a five amp supply to feed such a eight module rack... With proportions less for fewer modules.

As the power supply forms part of the audio path, there can be no skimping with either voltage or current headroom.

Posted from my iPhone

The inability of any power supply to come up due to inrush current demand from the load would typically be a result of the regulated power supply using foldback vs straight current limiting. With straight current limiting such as that used by three-terminal regulators like the LT-1086, LT-1085, LM-317, there should be no problem starting all 8 Neve modules in the scenario you describe. The regulators just limit the inrush current to about 2 Amps until the caps and circuitry in the Neve modules charge up. If the current overload should persist, then the the regulators would shut down due to overheating (thermal limiting).

When attempting to power 8 modules with a typical open-frame linear OEM power supply such as a Power One, Condor, International Power, etc, you may run into a startup problem unless the power supply has been modified for audio use by someone who really understands them. These units use the LM723 IC as the heart of the power supply. The LM723 is an amazing unit. It is actually kind of a kit of parts or building blocks from which one can make almost an infinite variety of Power Supply topologies. The LM723 contains a temperature-compensated 7.15V buried zener reference that is quite stable and has relatively low noise. With proper filtering, it can have VERY low noise, so that the reference amp in the LM723 actually dominates the noise of the power supply.

The LM723 can be (and usually is) configured to use foldback current limiting in these supplies. Foldback current limiting causes the output current to be reduced to a value lower than the rated output current in the presence of a sustained current overload. Straight current limiting just goes to some maximum output current and stays there until the overload goes away. The current is reduced in the foldback circuit by turning the output voltage down. This prevents the regulator from ever starting up in the presence of a heavy overload. The foldback current limiting is used instead of straight current limiting in these supplies in order to prevent overheating of the pass transistors in the supply in the event of a sustained short circuit -- like the tripping of a crowbar overvoltage protection circuit.

As supplied from the factory, these generic supplies don't have a lot of filtering in the zener reference in order to make them turn on quickly. When used to power audio gear, a slow start up is OK if not preferred, so a larger zener filter cap used with the proper series resistor not only lowers the zener noise, but it also slows down the turn on. The slower turn-on allows the caps in the Neve modules to power up slowly and the foldback current limit doesn't get tripped. As long as the current limiting doesn't fold back, the supply voltage will come up reliably.

Another problem with these supplies is that the LM723 reference amplifer is almost always overcompensated by as much as a factor of 30. I have seen supplies that are so slugged down that the amplifier couldn't even follow the 120Hz ripple waveform correctly and the amplifier would ring in response to the impulse stimulus of the residual noise. Reducing the compensation to the proper value makes a huge improvement in the regulator's transient response.

I would always recommend a power supply that can deliver 2 or 3 times the required operating current. Properly configured, a 2 or 3 amp supply should be plenty to power a rack of 8 Neve modules. A 5 amp stock unit may be required to get a stock supply to turn on, but you just have to tweak the 2 or 3 Amp supply a little to get it to turn on reliably. When I modify these popular OEM supplies for audio use, I typically end up with 20dB lower noise (-90 dBu is typical) than factory and much better transient response.