ST Head Waveform Regulation

[Ronmar]

My 20/2 Listeroid powered ST-15 setup provides near useless power. Sure, it will run the basic electric motors and the fridge. However, so many things have trouble with the power that I'm not sure what to do. Running "off grid" is an exercise in beeping UPS', grumbling electrical components, flickering lights and unsteady power.

Things that won't run, no matter what I do:

1) microwave
2) air conditioner (central)
3) TV
4) Direct TV sat receiver
5) half of my UPS'

My Robin/Subaru/Homelite generator runs everything without problems.

Part of the problem is the unsteady power of the listeroid, along with very poor governing, allowing the frequency to vary. The other part of the problem is the poor waveform. The combo is unsatisfactory electrical power.

Cujet, one issue with the ST series is the rectified harmonic energy is fed right to the field raw.  It is a real ugly waveform.  The addition of some filtration to this rectified signal greatly improves the output waveform.  My ST-5 initially would not feed a UPS.  With the addition of some filter caps to the output of the bridge rectifyer which basically turns the waveform from pulses into DC, it now runs all the UPS's in my house, and I have no other issues with any other device in my home running off the output including all the ones you mention.  The flicker is mainly a function of single cylinder engines, and the 2 cylinder roids don't fire directly in opposition, their power output curve is similar to a single.  The voltage output of the harmonically excited generator being RPM dependant dosn't help matters any.  An AVR takes a lot of the edge off of the perceived flicker in incandescent lights.  Most of my lights are CFL's and they don't show it as bad(hardly notice it) as the incandescents do .  I just posted some old waveform pics showing the output waveform of my ST-5 with different field inputs for Veggie in this thread.

http://www.microcogen.info/index.php?topic=1938.0

[cujet]

I put a SX460 clone AVR on my ST10. Works great, stable voltage, under speed roll off, and no Z winding. I have not had a chance to scope it but no problems yet. The SX460 is designed for brush-lesses generators though. I do not know if it will be harder on the brushes being a SCR regulator. With the higher impeadance of the ST field the inrush current will be reduced somewhat. I believe you must power it from both legs (220v) to get enough voltage for the field, I am.

The only anomaly that I have noticed is that the stability control has little or no effect. Even at maximum gain the generator output is stable. Probably has something to do with the high impedance of the ST head. The clone 4609s are available on Epay for around $70 bucks. Worth a try, worked for me!

Greg

Question: Is the ST head still self exciting without the Z winding in the circuit? It looks like a great idea and I'd be very willing to try it.

One of the issues I have is that my ST-15 head does not produce enough voltage. Under a 6000 watt load, the voltage sags to 106-109. That's a bit low for comfort.

Using an AVR in this manner"might" solve that problem and a few others. My ST-15 head did come with an el-cheapo AVR, but it's been removed as it lowered voltage further still.

Couple that AVR with extra gen head flywheel mass to smooth out the power pulses (the reason the microwave does not work) and I "might" be on my way to use-able power.

Thanks!

Chris


Chris

[Horsepoor]

I've observed my neighbor (Cujet - Chris) struggling with his generator head for several years. Interestingly, I have an almost identical Listeroid setup as follows:

20/2 GTC down rated to 850 rpm with extra heavy flywheels. (As I recall Chris is his running at 750 rpm)
I have an ST 15 generator head, 11 in pulley which equals a 70 lb chunk of metal spinning at 1800 rpm.
I replaced my AVR very soon after installing my gen head. I don't recall if it burned out or just rattled apart. (As I recall Chris had a malfunctioning AVR and ended up removing it). I ended up spending $49 to replace the AVR from Tom Osborne (Central Georgia Generator = CGG).

I replaced the misaligned brush holder in my ST 15 head and the brushes. Had to fabricate a few small parts to get the brushes to lineup properly on the slip rings but now I have a good and flush contact surface.

In fairness to Chris, I bought my ST 15 generator head in 2006 from a shop down in Miami, which is now out of business, and Chris bought his from another vendor up in Orlando, as I recall back in 2007. There is a difference in quality, perhaps I just got lucky or I just had a different set of defective parts.

In 2008, I ended up replacing both generator head bearings with Timken (I know this is like putting lipstick on a pig but I don't ever want to have wrestle those heavy ST head components around again, and to do so with damage was a real effort). While I was inside the gen head, I placed a coat of Glyptal onto the windings and replaced the 1960s Chinese vintage diode bridge with a (CGG) 40 AMP large heat sink version: This is also over kill but I do not want to be messing around with trouble shooting a generator failure during hurricane recovery operations. 

I am currently in the process of mounting the solid-state diode bridge rectifier, and the AVR on to a power meter panel / board away from the Listeroid to isolate it from vibration. I hope to finish this process this week.

The results of what I have been accomplished thus far include:

1.   Virtually eliminated light flicker, even in florescent lights, I believe this was primarily accomplished by engaging lots of heavy spinning metal flywheels and pulleys. Perhaps my 850-rpm prime mover speed, as opposed to 750-rpm, is just enough to smooth out the power pulses where the flicker cannot be seen. Perhaps placing a little more spinning metal on Chris's generator head could push the pulse flicker into the not noticeable category.

2.   My existing UPS, microwave, TV, computer, and everything else have no problem digesting the power output from my ST. I have invited my other neighbor (Ray Carlson) to stop by with his O scope because I really want to see what the  wave form looks like. While we are at it, I am going to check my Metro 6/1 800 rpm ST 7.5 gen head which is setup the same as my ST 15. And then the $99 Harbor Freight two stoke.

3.   This week I load tested my 20/2 GTC with ST 15 head on a hard 5 hour run on six year old treated waste jet fuel. The generator was loaded continuously to 9 & 10 kW for most of the time with occasional loads up to 11,200 watts (52A @ 215 V). The 11,200 W load is an absolute maximum, as the frequency started dropping (58 Hz) while the Listeroid struggled to maintain rated speed. I held the 11 kW load for 20 minutes just to test the over temperature shutdown system. I am extremely pleased with the unit, now that I have been able to fix all the original defects. 

Chris has an impressively designed and setup 20/2 GTC, ST 15 all on a steel wheeled cart. His cooling system and compact setup are very nice. It is a real chick magnet. I know his job really prevents him from doing all the tinkering I have done, but as hurricane season approaches, we both need to ensure operational effectiveness of these power units.

Chris - I've got a spare AVR and a spare 20 A solid state diode bridge if you want to test them out.

Forum – I really would like to know more about ST DC filtering via capacitors. I do not have a deep electronics knowledge, so if someone would break it down in simple terms I would really appreciate your help. I need to know what size capacitor works well and exactly where to insert the C in series with what. I have traced and labeled all wires on my ST heads so I can now interpret any instructions / advice.   

[Ronmar]

Forum – I really would like to know more about ST DC filtering via capacitors. I do not have a deep electronics knowledge, so if someone would break it down in simple terms I would really appreciate your help. I need to know what size capacitor works well and exactly where to insert the C in series with what. I have traced and labeled all wires on my ST heads so I can now interpret any instructions / advice.  

The filtration cap acts like a battery.  The pulsed output of the bridge rectifyer charges with this energy, and basically turns a pulsed output to a flat(or nearly so) DC voltage.  You use electrolytic capacitors, with a voltage rating higher than the peak voltage.  The capacitance is installed across the output of the bridge rectifyer where the energy is fed down to the brushes.   Rule of thumb for filtration is around 2000 Microfarads(represented an "uF") per amp of power supply current draw.  Too much filtration, however will delay the voltage response, as it takes a little time for the caps to charge and discharge with input changes such as when you load and unload the generator.  In this case you would want to use the minimum ammount of capacitance to get the AC output waveform cleaned up.  Another side effect is the output voltage of the generator increases with the capacitance being added.  This is due to the interaction of the capacitance with the windings, and the more stable power being applied to the field(field is more stable, and not radically expanding/collapsing with the unstable input).  I am at work now, I can post some pics and O-scope shots of the before and after field excitation waveforms later this evening.  

EDIT:  Here are some images to show you what the signals look like.

The first is the raw output of the Z winding.


The next is what it looks like after it is rectified.  This is what is fed to the brushes and field on a stock ST.


Here is what it looks like with some filter capacitance installed(basically DC):

[Horsepoor]

Thank you, this just the type and scope of information I was looking for. So, to save me some time, what size capicator would you or someone else recommend for an ST 7.5 producing 4,000 watts ? And what starting recommemdation would you have for an ST 15 producing 10,000 watts? I figure I dont need to reinvent the wheel because their must be sevral forum members who have refined the selection to a narrow band of choices. Thank you again for the pics and post reply.

[Ronmar]

Thank you, this just the type and scope of information I was looking for. So, to save me some time, what size capicator would you or someone else recommend for an ST 7.5 producing 4,000 watts ? And what starting recommemdation would you have for an ST 15 producing 10,000 watts? I figure I dont need to reinvent the wheel because their must be sevral forum members who have refined the selection to a narrow band of choices. Thank you again for the pics and post reply.

You need to put an amp meter in circuit with your field, and measure the AC amps being fed to your field winding with the respective generator under load.  If I recall correctly, I think my ST-5 pulls about 2.5A of field current with 3000W of electric load.  I am currently running Four 1300uF 50VDC electrolytic caps(each about the size of a D battery) in parallel(caps in parallel are additive) for a total of 5200uF, connected right across the bridge rectifyer output.  Electrolytics, like batteries, are polarity sensitive, so be carefull how you hook them up.  Once you measure your field current with an amp meter, I would say start at 1000uF per amp and work up from there till you get the minimum ammount of capacitance in circuit that achieves your output waveform goal.  Too much is bad, as like a battery, once charged, the cap acts like a UPS feeding the field to carry the load across a slump in the input, which is how it transforms the ragged input into nearly flat DC.  If you are running a lot of capacitance, and are running the generator at a very high load(lots of Z winding input), and suddenly remove a large load, the output voltage WILL spike.  Under heavy gen load the caps are being fed at their highest rate from the Z winding.  When the load(and z winding input) goes away, the caps still have a lot of energy feeding the field, but the output load has dissappeared, so that field energy turns into an output voltage increase untill the energy in the caps bleeds down to the level that the Z winding is feeding under the reduced generator load...  Same is true with sudden load increases, it takes a little time for the caps to charge with increased input, so there is a little output voltage drop untill they catch up.  The more capacitance you run, the worse the voltage excursions are with major load changes, so use only what is absolutely necessary.

Good luck

[WStayton]

Ronmar:

  So would I be correct in assuming that a cheap-o AVR is just an elegant collection of capacitors and some mechanism to provide an appropriate voiltage for the field?

  Conceptually, a very smart mouse with a voltage meter and a variac who ALWAYS watches to output voltage?

  Yea, I know, it couldn't POSSIBLY be that simple!  <grin>

Regardz,

Wayne Stayton

[Ronmar]

Yep, only the cheep-o ones still use mice:)

You are correct, the AVR monitors the output voltage and inveresly regulates the current flow into the field to maintain the output voltage.  A typical AVR dosn't use many caps, it is more chips and transistors...

The use of caps on the ST head dosn't do anything for voltage stability, in fact, it probably adds lag/overshoot to it if too much is added as I described above.  the Z winding relationship to the field is where that type gets it's relatively stable voltage.  The caps only help clean up the waveform.  there are capacitive regulated generators, but they are wired much different, and can use a change in capacitance to change the output voltage.  There are quite a few ways to skin this cat...

The harmonic or Z winding in the generator does the same thing naturally, as it's output is based on the current flow thru the secondary winding.  More current with increased generator load = more z winding output being fed to the field to maintain output voltage under the increased load.  It is however a tricky process to determine exactly how many windings are needed in the Z winding to give the desired slope to the output to maintain voltage under load...  unfiltered, it makes for a distorted and unstable field waveform, which translates to a distorted AC output waveform.  Also since you are robbing from the generator field to produce this field energy it also probably adds to the output waveform distortion. 

[sailawayrb]

1. Virtually eliminated light flicker, even in florescent lights, I believe this was primarily accomplished by engaging lots of heavy spinning metal flywheels and pulleys. Perhaps my 850-rpm prime mover speed, as opposed to 750-rpm, is just enough to smooth out the power pulses where the flicker cannot be seen. Perhaps placing a little more spinning metal on Chris's generator head could push the pulse flicker into the not noticeable category.

Well, I am glad that I am not the only one who doesn't notice light flicker WITHOUT using a AVR.  

I suspect (but can't prove) that just maintaining a good steady nominal voltage (I run 122 +/- 3 VAC) and frequency (I run 60 +/- 1 HZ) by always keeping electrical loads balanced, not pushing your engine or ST near their limits, and minimizing undesirable low frequency engine/ST dynamic effects (with resilient mounting, increased flywheel weight, etc.) goes a long way to reducing perceptable light flicker.  It sounds like Cujet's problems may have much to do with simple voltage sag than anything else.

Bob B.

[mbryner]

Using capacitors to filter the excitation current is just 1/2 of an LC circuit.  You could use an inductor such as 1/2 of a transformer instead,  or calc the LC circuit.  No?

[rcavictim]

Using capacitors to filter the excitation current is just 1/2 of an LC circuit.  You could use an inductor such as 1/2 of a transformer instead,  or calc the LC circuit.  No?

Any L you place in series with the exciter circuit is going to make the excitation voltage drop like a stone.  It will also slow the response to load shift as was mentioned that too big a C will do.

[mobile_bob]

the rotor field winding already is the "L" in the tank circuit

i suppose that and the Z winding comprise the "L" component

no need for more inductance?

bob g

[Horsepoor]

Finally got some time to take measurements, and add some photographs concerning my ST 15 Gen head. Built a meter panel with a no voltage emergency shut down system for about $15. I wanted something that would trip if a belt breaks, rectifier blows, etc. Here as some measurements off the top of the ST 15.

At 230 Volts with 35 amp load I read the following (No AVR in the circuit for these measurements):
           55 VDC coming out of the diode bridge at 3.75 A
           74 VAC coming off the Z winding. (Used a clamp meter here to get a questionable reading of 0.45 A)

These 11 pin Potter & Brumfield relays are cheap and allow several trip circuits to be created in one relay trip event. I step up a combination of 12 VDC and 110 VAC switches based upon this relay.

Next we attached an O-Scope to examine what happens to the wave form when the new AVR is removed. As cited elsewhere on the forum, the wave form is slightly better with the AVR in the circuit. In the attached O-Scope photo, I illustrate the worse wave form we could create by intentionally lightly loading the ST 15 at about 1,000 to 2,000 watts. This wave form is still not bad, and when heavily loaded it is close to perfect. But who wants to see perfect. Next step is to explore adding just enough large capactors to raise the voltage a little.

Here is an unresolved problem: With the AVR engaged, I get the following reads at the generator head: 110 VAC line #1, 115 VAC line #2, 226 VAC across both lines, and this holds pretty constant across all load ranges 3 KW to 10 KW. Problem is up at the main house 250 feet way via the gound calbes (4 AGW) I get 105 VAC on some house outlets. A little low, but serviceable. I just wonder why one identical line off the ST 15 always has a voltage 5 or 6 volts lower than the other leg. The AVR measures the 226 VAC line. ? Any ideas?

[Tom Reed]

My ST5 will do that depending on which leg is loaded heavier.

[Horsepoor]

Tom,

Even when very lightly loaded? For example 1,500 watt load on a ST 15 head. Very interesting, learning all the time. Still, I love this stuff because I can repair, fix, and understand just about all of the componets.

Bruce