Puppeteer

Author Topic: Why ST heads?  (Read 7595 times)

MachineNLectricMan

  • Newbie
  • *
  • Posts: 25
    • View Profile
Why ST heads?
« on: September 13, 2017, 09:13:35 AM »
Is there any reason other than price that folks are running brush type alternator heads from CGG? I am seeing numerous issues reported with things like waveform issues and aluminum windings on CGG ST heads. Brush technology is so old and mature! It was perfected during the Tesla era! It would take a team of drunk monkeys to mess up in winding and assembling one wrong. Where are they importing these from? I've seen that the 3 phase heads don't seem to have these issues. Is that true, or is it that very few use the three phase heads?

Theoretically aluminum windings would work fine if engineered and installed correctly. Unfortunately, the Chinese and Indians are not capable of installing them with the special requirements needed for aluminum in my opinion. Some of the cheap wire enamel used by these folks actually slowly sublimates over time during use. The heat rise in normal operation vaporizes this stuff very slowly. Years ago, I have noticed that on some cheap store-bought fans, most likely Chinese, the windings actually become progressively shiny during use, resembling clean copper, then they fail. Upon close examination, it is discovered that the enamel has all but disappeared. Is the same stuff is being used on imported alternator heads?

38ac

  • Hero Member
  • *****
  • Posts: 2327
    • View Profile
Re: Why ST heads?
« Reply #1 on: September 13, 2017, 01:45:42 PM »
My opinion is that a person who has the knowledge and will to make science out of his generator project would be foolish to start that project with a caveman style generator head???  But for me, an electrical know nothing that wouldn't know how to hook up a O-scope if somebody gave me one and even if I did I would not know what I was looking at on the screen the old school simpleness of the ST type suits my simplistic electrical mind perfectly!  I am not even running AVRs on any of my 3 set ups that use ST heads, just the harmonic regulation. When set for 60hz the voltage is well within the range of OK for my needs. Had two issues over the years, bad rectifier which took me 5 minutes to figure out and replace and a family of mice decided to build their home in one.  I have spares ST heads in crates and even with that cost I am still way less in cost than a hi-end head.  My experience with ST heads has been great, other's as you said,  not so good.  Nothing in my house realizes it is being powered by a substandard power supply, when the grid goes down the 1115 and 15KW ST power us same as when grid is up, TVs ,computers, satellite receiver, everything, it all runs just as on grid.
I am not trying to say my way is the only way, just answering your question as to why an ST. I admire the set ups I see guys put together on here but they are waaaay above my ability to understand, let alone build and maintain one.
Collector and horder of about anything diesel

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: Why ST heads?
« Reply #2 on: September 13, 2017, 03:43:57 PM »
Price and simplicity is the only reason to use an ST head. The waveform quality varies wildly between units; some have spikes and ratcheting of output wave bad enough to cause heating and buzzing of small timer motors and/or interfere with some electronics. This is due to variations of the stator plate stack and windings. As I reported before, I've had three stators on my bench in the last year and all three were different in length. There is also such variation in the castings and machining as to cause clearance issues between rotor and stator, which greatly affects harmonic ouput.

My original ST-3 continues to provide power of marginal (ratcheting and spikes plus harmonic hump) waveform (THD measured at 18%)  reliably with a new bridge diode and SK bearings.  It runs my well pump and washing machine reliably.

I found that the crated backup unit I had much more problems and will not be usable without AVR from mains as the harmonic is too weak to use as excitation source. It's case casting was cracked and the stator and rotor had interference problems. 

There are no other affordable double bearing 1800 rpm generator heads in 3-5KW size.  I have no qualms about brushes and slip rings; they are simple and maintenance is simple.   Gary has recommended the Stamford clones and perhaps their 7KW monster is the best way to go unless you are willing to spin a 2 pole head.

The variation of the ST heads is so great that it's much like playing Rajkot roulette.  The 3 phase STs are notable in that they do not have the same waveform issues as the single phase units.






MachineNLectricMan

  • Newbie
  • *
  • Posts: 25
    • View Profile
Re: Why ST heads?
« Reply #3 on: September 14, 2017, 08:03:08 AM »
How many hours of brush life are you folks getting with these ST heads?

I agree that most things will run on a rotten waveform. Some electronic gadgets have cheap and terrible power supplies, and those are the things that probably wont run on anything but a good sine wave. And the grid gets pretty bad sometimes too, I've heard my microwave buzzing like an old fashion door bell before. Hooked a scope up and found a badly chopped wave coming off the grid!

A correctly wound gen head naturally wants to produce a sine wave. The best winding patterns were vigorously worked out and perfected a century ago, well known to every winding shop, and published in thousands of books on electric machinery around the world. Back in those days, most electricity was produced with independent "microgrids" in various parts of large city's and other locations using huge engines running on producer gas. Eventually the small grids merged, and steam turbines were perfected.

I prefer brushless. No electric noise from brush jitter. From experience with brush machinery, things like slip ring runout, bearing imperfections and any vibration effects can cause brush contact trouble. Perhaps that is where some of the waveform issues are coming from.



BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: Why ST heads?
« Reply #4 on: September 14, 2017, 04:06:45 PM »
My ST-3 is still young at about 3000 hrs. No service issues beyond bridge diode and bearings, both replaced under 100 hrs.  Very few reports of failures in service beyond the usual diode and bearing failures; my neighbor's aluminum wound rotor being the exception. 

I think the racheting, spikes, and harmonic hump on the ST are related to stator winding and the non-skewed rotor. The spacing is regular and similar around the peak on all the units I've seen.  The ST-5s have redundant brushes at 90 degree spacing and show the same characteristic distortions which rules out brush bounce. The distortions vary dramatically in severity between units.  I'm more familiar with the waveform variations than some because I designed a simple AVR before the cheap china units were available and made it available to folks here and they sent me o'scope shots of their ouput.

Cheap 3600 rpm brushless 2 pole generators also can have very lousy looking waveform, sometimes with massive high frequency EMI. 

I prefer brushes for my application because of the problems with Listerflicker; the 10 hz power pulses.  My custom AVR is able to compensate more rapidly without the additional inductive time lag of a brushless system.  It's critical for me as I can't use my 2000 watts of heat lamps in my shop in the winter unless the Listerflicker is tamed.  I can electronically compensate about the equivalent of adding the heavy SOM flywheels.  I'd like both but we have no barns with old CS's here in Arizona and my budget is thin.






38ac

  • Hero Member
  • *****
  • Posts: 2327
    • View Profile
Re: Why ST heads?
« Reply #5 on: September 14, 2017, 04:40:06 PM »
I dont keep exact track of my generator hours but at less than 500 hours the  original brushes are the same length as thebrand new spares I keep. I just checked them after the last run. I have never read that brush life was an issue with anyone? Even if I had to replace them annually it's really a non-issue for me, they are cheap and easy to replace.
Collector and horder of about anything diesel

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: Why ST heads?
« Reply #6 on: September 14, 2017, 09:05:01 PM »
I"ve got over 2000 hrs on my current brushes and they are barely worn. And this is with the ST-3 that only has 2 brushes. If you have a problem with wear or conductivity of the brushes, it because of low quality brushes with contaminates in the graphite.  A good set last and lasts.

I frankly prefer the simplicity of a brushed head; no electronics spinning on the rotor where they are subject to vibration and mechanical stress, and the simplest rotor windings.  I just wsh some vendor would tweek the design of the single phase ST heads to clean up the waveform, consistently. Consistently proper harmonic regulation without distortion would also be nice.  The best of the current ones would suffice, but I'd be happy to pay double for a skewed rotor and a more voluptuous waveform.




starfire

  • Sr. Member
  • ****
  • Posts: 282
    • View Profile
Re: Why ST heads?
« Reply #7 on: September 14, 2017, 09:45:19 PM »
Pretty much all electronics these days use a "chopper" power supply, meaning the 230v mains is rectified to 380 VDC, or 120v to 170v, filtered and chopped at high frequency (using squarewave PWM) so as to use very small transformers to arrive at the low voltage DC required.
Therefore it matters not the shape, or frequency even DC of the input mains waveform, and even the voltage over very loose limits. Many modern devices are designed for the global market to happily accept an input AC voltage between 80 to 260 volts with no changes needed, or even DC for that matter.
CFL bulbs will run on AC or DC too and are voltage tolerant. Many mains LED bulbs require AC, as the cheapies use a capacitance dropper. DC wont harm these, they just wont work. Running CFLs on DC reduces radio interferrence dramatically and can increase brightness on some..
Induction motors will run warmer on square waves, as will anything using the old fashioned mains transformer.  The inductance of the windings heat as they attempt to convert the square wave to a sine. Any inductance will oppose a voltage change.
My electronics workshop runs completely from a 600 watt sinewave inverter as does most of my test gear, signal generators, oscilloscopes and spectrum analysers use linear power supplies, inbuilt switched mode choppers are just too noisy for this sensitive precision equipment. My lighting is DC 200v
My  6kva generator unit is 240 volt AVR'd,  the cheaper brushless types have voltage surge issues ,when a large load is suddenly disconnected, the output voltage will spike badly because of the slow time constant of the phasing capacitor.... thats the round white thing that fails regularly in these designs.
 The only effect mains frequency will have, is on the speed of induction motors, and any clock that uses mains frequency as a clock timing signal, ie, very few appliances these days.
Things like vacuum cleaners, electric angle grinders etc using brushed "universal" motors will be voltage and frequency tolerant, and will run the same even on DC.. ie, a frequency of zero.
Washing machines, refrigerator compressors using induction motors just want AC, 45 to 65 Hz you wont notice much difference.
Most modern caddy inverter type MIG/MMA/TIG welders will also run on AC/DC as they too directly rectify the mains, and care not as regards mains frequency.  Running stuff on DC if you can  makes sense, as the appliance is more efficient as there are no power factor issues , but care is needed, DC will quickly fry an induction motor if accidentally connected.
Oddly, and contrary to accepted opinion, modern electronic devices , because of the cheap chopper type shitty power supplies they have built in, are now  more tolerant of shitty mains power than ever, as they have very good  post filtering and extra MOV/VDR protection in place to mitigate the shitty power supply they use. If in doubt, hold an AM radio near a SMPS power supply brick and listen to the harmonics.
Anything thats a pure resistance, ie water and space heating, cooking elements etc are also voltage and frequency tolerant, and give identical performance on DC.... there are no power factor issues even when used on AC.
So, spending great effort to govern an engine to give exactly 50/60 Hz is really a waste of time and money.
Worrying too much about absolute voltages is also not necessary, just avoid fast voltage spikes and changes.
The problem witth aluminium windings remains.
Cheap welding equipment also use aluminium wire, and over many years of repairing this stuff for a living, have found soaking the coils/windings/transformers in old fashioned boiled linseed oil, or any vege oil for that matter will prevent the corrosion that eventually causes shorted turns as the metal swells and ruptures the insulation. This will also work on armatures and stators. It takes just one shorted turn to destroy the equipment. The oil will seep into the windings, stay semi flexible with movement and cover the wire with an air/moisture  barrier
Alumimium is  a relatively poor electrical conductor and will heat more in use, this causes relative expansive movement between windings and will stress fixed insulation over time.
The Chinese have also invented a way to copper plate aluminium, so a simple visual check may reveal nothing unusual. This deception is even worse, as electrolytic corrosion takes place at a rapid rate.... dissimilar metals. 
The old and tested way to judge quality in a generator is LOTS of copper and LOTS and LOTS of iron... ie, choose the heaviest one. Go for larger diameters rather than longer length. Large diameters also indicate more poles, and lower rotation speeds.
Waveform discontinuities/distortion are mostly caused by core saturation, older generators had tertiary windings to advance/retard the magnetic flux under different loading but were expensive needing extra poles on the rotor, thats why they were much larger, and exaggerated by requiring a seperate exciter.... A good modern AVR will usually sort these out, oddly more usual at light loading.
These are just my observations over time.

cujet

  • Hero Member
  • *****
  • Posts: 960
  • Lister power rules!
    • View Profile
    • www.cujet.com
Re: Why ST heads?
« Reply #8 on: September 15, 2017, 01:00:19 AM »
I'd like to say that my ST head has been a good choice, but it never worked right. It made only 109v at 61HZ. Dropped to 90V under heavy, but not max load. The Z winding cannot drive the gen head to make enough voltage. I ended up using an aftermarket VR, hooked to the main winding to power the field. Now makes 125V.

However, the flicker remains. Maybe a good part of that is due to the listeroid's power pulses. But the flicker lessens by 80% after 2 hours. I will study the waveform again hot and cold and see what I can.

The ST head is heavy, robust and strong. The rotor is also very heavy. The inertial mass does help dampen the listeroid a bit. Using the HF 10KW gen head is another cheap ($399) option.

My ST 15 was $600 including dual speed pulley. The Mecc Alte 15 gen head is high quality, but $1200 and runs at 3600 RPM. 

  http://www.brandnewengines.com/S20FS-230BD.aspx?gclid=EAIaIQobChMIxpOR1fCl1gIVikoNCh2TUQYWEAkYDyABEgJYYvD_BwE
« Last Edit: September 15, 2017, 01:02:25 AM by cujet »
People who count on their fingers should maintain a discreet silence

32 coupe

  • Hero Member
  • *****
  • Posts: 531
  • "Doc Johnson" on call !
    • View Profile
Re: Why ST heads?
« Reply #9 on: September 15, 2017, 01:29:38 AM »
My 15kw head is direct coupled.to a Changfa 1115 and I have good luck with it.
I recenty had to replace the original Chinese diode after about  100 or so hours.
The voltages have always been correct according to my kill o watt meter.

The flicker is not as bad after I changed all my lights to LED's.

I haven't had the privilege of seeing anyone else's run so I don't really have anything to compare
it too.

i run mine pretty well loaded most of the time.
Mine is an older unit with copper windings.
Can't complain considering the cost.

Gary

Metro 6/1 turning a ST 7.5 KW gen head
Changfa 1115 turning a ST 15 KW gen head
Ashwamegh 2/25
John Deere 110 TBL
New Holland TC 30

"I was sitting here reading this thinking what an idiot you are until I realized it was one of my earlier posts !"

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: Why ST heads?
« Reply #10 on: September 15, 2017, 02:20:27 AM »
Variations in harmonic windings is another common ST problem.  Either high or low.  It seems stator winding is done by help that can't count turns or position windings reliably.

The larger ST heads (above 5KW) are not going to be able to respond fast enough for good flicker correction; the larger the rotor, the greater the inductance and the slower it can respond to a big input for excitation.

Also, most AVRs are NOT designed for flicker compensation and in some cases may actually make it worse, as the correction comes to late, then overshoots, then undershoots....  I had some interesting effects when I was tuning my own AVR design.

The better quality LED bulbs (like Phillips- which regulates to 90VDC internally) should have zero flicker.

 I can't stand the spectrum of the light from  CFLs or LED bulbs myself. White LEDs are really just fluorescent lights...UV LED excites a dab of phosphor, and the light spectrum is just as bothersome to me as any other fluorescent.  Pity as for off grid living LEDs are otherwise very appealing, though not the crappy EMI generating switching supplies in each one.



MachineNLectricMan

  • Newbie
  • *
  • Posts: 25
    • View Profile
Re: Why ST heads?
« Reply #11 on: September 15, 2017, 09:47:35 AM »
The Chinese have also invented a way to copper plate aluminium, so a simple visual check may reveal nothing unusual. This deception is even worse, as electrolytic corrosion takes place at a rapid rate.... dissimilar metals.

Why that would cost way too much for the Chinese to cheat with. This is likely a "Chromate Coating" has a gold to coppery appearance, improves the surface conductivity of aluminum, is very inexpensive to apply.

Pretty much all electronics these days use a "chopper" power supply, meaning the 230v mains is rectified to 380 VDC, or 120v to 170v, filtered and chopped at high frequency (using squarewave PWM) so as to use very small transformers to arrive at the low voltage DC required.
Therefore it matters not the shape, or frequency even DC of the input mains waveform, and even the voltage over very loose limits. Many modern devices are designed for the global market to happily accept an input AC voltage between 80 to 260 volts with no changes needed, or even DC for that matter.


I've seen lots of zener regulators on modern "smart" appliances. Some even use a capacitor and diode in place of the first stage step down transformer. I think that even in PWM regulators, some of the transients and harmonics of a poor AC waveform get coupled to the output and play havoc with the function of the device. Even high frequency PWM's usually cannot respond fast enough to entirely compensate for semi-square wave AC forms and transients.

The only way to get rid of all flicker is likely the battery charge/ inverter type system where the engine only charges a large battery bank. But then goes away some more fuel efficiency in the added system losses.

starfire

  • Sr. Member
  • ****
  • Posts: 282
    • View Profile
Re: Why ST heads?
« Reply #12 on: September 15, 2017, 02:33:46 PM »
The Chinese have also invented a way to copper plate aluminium, so a simple visual check may reveal nothing unusual. This deception is even worse, as electrolytic corrosion takes place at a rapid rate.... dissimilar metals.

Why that would cost way too much for the Chinese to cheat with. This is likely a "Chromate Coating" has a gold to coppery appearance, improves the surface conductivity of aluminum, is very inexpensive to apply.


Unfortunately not. Cut open almost any Chinese sourced IEC connector cable or even mains extension cords and see for yourself, this is becoming very common. The consensus seems to be  "skin effect"  where AC currents only flow on the outer surface of a conductor, the center being largely redundant.  The second  was to allow easier  soldering, crimping is unreliable with "plastic" and unstable metals such as alluminium ...  a few cents of thin copper coating is likely cheaper than the extra labour to install crimps.
This first effect is real, but occurs at higher RF frequencies, NOT at 50/60 cycles...... there is some merit in the second reason .given alloys ability to corrode.  I suspect the real reason is cost reduction, cheaper wire,, cheaper assembly.


Pretty much all electronics these days use a "chopper" power supply, meaning the 230v mains is rectified to 380 VDC, or 120v to 170v, filtered and chopped at high frequency (using squarewave PWM) so as to use very small transformers to arrive at the low voltage DC required.
Therefore it matters not the shape, or frequency even DC of the input mains waveform, and even the voltage over very loose limits. Many modern devices are designed for the global market to happily accept an input AC voltage between 80 to 260 volts with no changes needed, or even DC for that matter.


I've seen lots of zener regulators on modern "smart" appliances. Some even use a capacitor and diode in place of the first stage step down transformer. I think that even in PWM regulators, some of the transients and harmonics of a poor AC waveform get coupled to the output and play havoc with the function of the device. Even high frequency PWM's usually cannot respond fast enough to entirely compensate for semi-square wave AC forms and transients.

Zeners with capacitor droppers are only seen in extremely cheap and poorly designed products, and these by their very nature a short lived anyway and deserve little mention.


Modern PWM choppers run at frequencies from 30kC to over 1 MHz. The number of chopper cycles to one mains cycle  is many thousands, and can therefore compensate any transients very easily. The primary High Voltage smoothing capacitors mitigate most of these transients before they even reach the chopper. The chopper only sees the DC derivative of the input voltage whatever the input waveform. A square wave input is less likely to harbour  mains spikes per se, but has the possibility of overshoot on leading and trailing edges. These are known, predicatable, and therefore easily accounted for in the circuit design. Sine inputs can suffer many distortions of a more random nature.

The only way to get rid of all flicker is likely the battery charge/ inverter type system where the engine only charges a large battery bank. But then goes away some more fuel efficiency in the added system losses.

Flicker is caused by a cyclic voltage change.  A well setup generating system has little to no voltage variation with frequency (rpm.) The time constant of the L/C ratio of the field coils dictate/limit  the regulating ability of the AVR as does the loop gain.  At 50  cycles, each cycle is 20 milliseconds.
With incandecents, we have thermal inertia, with CFLs we have DC capacitor smoothing integral  to the "ballast". both inertia mechanisms reduce luminosity change.
 With these, it is generally sufficient to  regulate the fields over a few 10s of cycles ( 200 milliseconds say ) to eliminate rpm/ load variation  flicker.
LEDs  usually have a very fast response to any voltage/current  change so these can be  the worst..  The cheapies use the above mentioned capacitor droppers, little can be done, these actually strobe at 50 Hz, as does a normal flourescent tube.. The better quality versions have a chopper type  regulator and are in a similar category the the CFLs mentioned above where the internal DC level supplies a constant current to the LED strings due to internal capacitance.. A capacitor is the electrical equivalent to a flywheel.
I do agree though, the best way is to run lighting from battery backed inverters. This gives constant and immediately available lighting.

BruceM

  • Hero Member
  • *****
  • Posts: 3054
    • View Profile
Re: Why ST heads?
« Reply #13 on: September 15, 2017, 02:47:45 PM »
"Even high frequency PWM's usually cannot respond fast enough to entirely compensate for semi-square wave AC forms and transients"

This is utterly and outrageously incorrect and shows a complete lack of understanding of DC power supply design. I suggest you read up on the subject matter before spewing misinformation.

The only typical compatibility issue with square wave or modified square waves is the peak voltage and the technique used for bootstrapping (start up) ; some designs may not allow for that.  Most modern designs do, allowing a broad range of peak input voltage especially PF corrected ones.  Even primative designs are very tolerant of typical power line transients since the first stage is bulk DC conversion.

Flicker shows up only in  cheap LED/CFL designs where there isn't true regulation. 

EMI issues can occur when electronics were expecting relatively clean DC and instead get a load of high frequency EMI.  A practical example example is a PV charge controller taking rectified AC with lots of EMI instead of the clean DC PV input it was designed for.  Adding 4- 0.1 uF snubbers to the bridge diode and perhaps a common mode choke will typically solve that. 

Asking questions or reading is a good way to learn.  Spewing misinformation isn't.  You are likely a troll, so I won't bother to respond to your posts again.







MachineNLectricMan

  • Newbie
  • *
  • Posts: 25
    • View Profile
Re: Why ST heads?
« Reply #14 on: September 16, 2017, 09:15:55 AM »
""Even high frequency PWM's usually cannot respond fast enough to entirely compensate for semi-square wave AC forms and transients"

This is utterly and outrageously incorrect and shows a complete lack of understanding of DC power supply design. I suggest you read up on the subject matter before spewing misinformation."

Your forgetting that the rate of rise for the squares in the semi-sine wave is going to be in the multi-megacycle frequency range, v. 30 kilocycles for the PWM. That will get through. Your forgetting about capacitor ESR too. Most commonly used "filter" capacitors in today's cheap electronics will not filter out this effect. I once designed some computer motherboards that operated in very noisy environments and have been thoroughly exposed to these effects and the problems they caused in the logic circuits and memory.

Have you really designed any of these circuits?