Author Topic: Fried Tristar TS-60 (Read 9847 times)

dlenox · « **Reply #33 on:** January 02, 2009, 02:46:45 PM »

Maybe others can chime - but to summarize the circuit requirements:

this is not meant to trigger a dump load but to shut down the turbine!
have power light
have status light showing state
for simplicity sake, lets say it drives a 12vdc solenoid/relay/contactor/SSR/whatever so limited to 20amps or less.
means to make it automatic/manual reset, if manual provide for external spst switch
response time < .25 sec once desired trigger voltage is achieved
triggered at 'tuneable' voltage
not everybody is 12/24 volts please! work within 48v system, and in my system max of 60v for trigger
I don't care if kit or pre-assembled
cost should be less than C40

Other please chime in!

Dan

dlenox · « **Reply #34 on:** January 02, 2009, 02:49:54 PM »

11) If need basic power for circuit it can be 'tapped' from a single 12v battery.

ghurd · « **Reply #35 on:** January 02, 2009, 02:55:30 PM »

We are on the same page.

High volt P-fet drives a 48V relay. A ~30V 5W ZD.

Latch or otherwise, same result.

Small DC SSRs get hotter than fancy fets. The data sheets show I looked at they leak much more.

I mostly used DC controlled AC SSRs a few times for stray items with grid power (dust collectors, fans, other devices needing remotely controlled). Works fine, but always below 15A at 120VAC 60HZ.

I have done larger things that worked fine. They never got serious use, and I do not recall any numbers. I do recall they became much hotter than I expected.

G-

ghurd · « **Reply #36 on:** January 02, 2009, 03:34:23 PM »

I was thinking my circuit with a very large hysteresis. Huge. 6V?

The huge hysteresis will greatly decrease the power used during normal conditions (latent characteristics get crazy complex).

High voltage P-fet driving a 48V relay gets around some of the high voltage issues.

Multiple series 5W ZDs if necessary for the SSR Vin.

Thus avoiding the 60V issues. Sloppy, but a dandy 'worst-case' relay driver.

Disclaimer #2.

Yes. I have a great distaste for relays.

This circuit will not switch fast or often. Hopefully it will never be needed.

This is not the same as 5 times per second for 20 years.

G-

dlenox · « **Reply #37 on:** January 02, 2009, 09:47:28 PM »

Ghurd,

another sleepless night, 3 weather forcasts called for 7-10mph winds we had up to 40. had again to shut turbine down to high winds for the umpteenth time in the last couple of weeks - and winter has just started.

in looking over your dump controller, one really/really nice thing to add would be the ability to use a DVM on a test point to set the trigger voltage. this way I would not have to replicate the trigger voltage to be able to get close to setting the desired value.

not a requirement, mearly a nice accessory. mind you I am simply an end user and not all of us are as inventive at setting dump/trigger voltages by replicating he situation until we are in the middle of the event that we wish to 'control'. after all control is an illusion anyways.

one nice selling point about this circuit is that it could be used to trigger a dump/diversion load, shutdown relay, or what ever sort of 'stuff' that you want to have happen at a set voltage. for me it would be a nice extension from your work that has already been laid

my thanks up front to all that contribute, and those that make things happen, it just so happens that electrical circuitry is not my 'thing'.

If there is anything software-wise that I can help with - including programming of PICAXE programming, c#, asp, website - whatever I am more than happy to return the favor, as software development is my bartering chip and is what I am best at.

Dan lenox

dlenox · « **Reply #38 on:** January 02, 2009, 10:58:36 PM »

Ghurd,

Here is a thought that I had as I was going to sleep for the 2nd time tonight.

If you are familiar with the PICAXE family then we could collaborate on the development of the circuit. You design the circuit and I could provide the code to program the chip! It really is something that counld ultimately simplify the interfacing, at a very low cost, and give the end user a lot of 'adjustability' to suit their needs.

The environment is simple enough to control so it is just interfacing the incoming trigger voltage range that is desired as well as controlling the outputs! You handle the interfacing circuitry and I handle the software implementation!

The chip interface is as simple as writing some VB and timing code, all of which if we publish the code the user can modify to their hearts content to satisfy their needs.

What are your thoughts on this?

Dan

Simen · « **Reply #39 on:** January 03, 2009, 12:03:14 AM »

I was about to suggest the picaxe... :-D

These small buggers are excellent to this kind of work; easy to program and simple to construct on a veroboard etc. Just buildt a small volt-logger/simple charge reg.; works like a charm.

RandomJoe · « **Reply #40 on:** January 03, 2009, 05:21:03 AM »

Would it be too much of an overcomplication to add a wind speed sensor to your control sequence? Keep track of max wind speed gusts, maybe track how often they are occurring, then when you trip the turbine off latch in those values. After the battery voltage drops to the point you might want the turbine to restart, watch the wind and wait for it to drop to some more desirable value on the gusts. Tracking time between gusts could allow dynamically adjusting the delay before releasing the turbine.

I don't know if the wind speed indicators are reliable enough to want to use them for that, nor have I played with wind power yet so perhaps there isn't even much point to this... I'm just a control freak who designs and installs building automation systems for a living. Figuring out ways to keep human involvement in regular operation to a minimum has become my favorite part of the job!

dlenox · « **Reply #41 on:** January 03, 2009, 07:52:20 AM »

Joe,

no actually it is not very complicated at all, the only problem is the computer, they fail.

I have over 30+ yrs of software experience and unless you have a very high grade computer then it eventially will fail, typically at the times that you need it the most in a critical control situation. Not to mention that it gets periodic updates, gets rebooted, etc.

computers are quite reliable, but not infallible that's why there are backup systems. IMHO.

Dan

ghurd · « **Reply #42 on:** January 03, 2009, 10:01:41 AM »

I am familiar with picaxe and stamp to the point of knowing what they are. I am too old and grumpy to learn more about them.

I know they have great potential for people who want to figure it out. Only cheap and simple for people who have the hardware and understand the chip.

Seems like a lot of bother just to run a basic dump load, battery box fan, or trigger a relay? Another "sledgehammer to crack a nut"?

My circuit is very accurate. The battery voltage needs to be replicated. A test point would introduce potential errors for most users.

Getting the battery voltage is not so hard. Get a 6V spring-top lantern battery, or 4 Ds in a holder. That will be about 6.5V.

Get the 'real' batteries to 6.5V below where the shorting circuit will activate.

Series them for the trigger voltage.

So 59V trigger, 59-6.5= 52.5V on the batteries. (That's 13.125V for each 12V) Should be easy to get them to there for a few minutes?

Then adjust the blue pot until the yellow LED lights.

Typing that, and the 15 second voltage increase, makes me believe 6V hysteresis is not enough for automatic reset. 9 or 10V sounds better.

Removing the charging amps alone will quickly drop the battery voltage quite a bit, the dumping TS-60 will drop it more very fast. The grid is down so the batteries will be loaded(?) dropping the voltage even faster. Could lead to cycling faster than I would want.

Another glitch may be cold batteries, or automatic equalization. The TS60 can decide the batteries need to be 60V or more. If the shut-down circuit is set for 59V, then the batteries will never (automaticaly) get to the voltage the TS60 wants them to be.

Add the unfamiliar to me XW into the mix... and now I have a headache.

Perhaps in the interest of bandwidth we should email, then post whatever it is we think up to see if anyone thinks of potential problems?

G-

fcfcfc · « **Reply #43 on:** January 03, 2009, 02:20:54 PM »

Hi All:

Well, I am sure getting dizzy at this point... What ever happened to K.I.S.S....

The way I see it is as follows:

You have a 48VDC battery bank that under wind turbine charge conditions you do not want the voltage across the bank to exceed around 58VDC or so.

You had hoped the settings in the inverter would do as a trip point for this and go into sell mode and take down the voltage until a low point was reached on the battery bank, maybe 44VDC or so, or what ever voltage represents the DOD you want.

(Ironically, this is exactly the way I plan my setup to work just with the 4500 series XW...) anyway, the problem is that you were getting inconsistent results from the inverter in both the voltage it kicks in at and the time it spends over the trip point voltage before engaging.

And a last issue, a max voltage trip point to engage the turbine brake system until a certain voltage or off time had been achieved......

Is this right...??..

.....Bill

RandomJoe · « **Reply #44 on:** January 03, 2009, 03:05:49 PM »

Well, that would depend on what sort of computer and what time frame one is interested in. I wasn't thinking about having a full-on PC with HDD, rather a low-power dedicated system with a pulse input for the speed indicator. I figured it could fit into the picaxe that was mentioned, hadn't heard of those before now. Or worst case a small SBC with no moving parts. I figured those would be fairly reliable, as opposed to a standard PC.

Some of the HVAC control systems I've worked on have been up and running for close to 20 years now (since before I was working) without any issues, many living in unconditioned boiler rooms. Granted, they aren't realtime control, but they have enough power to run large buildings full of equipment, collect trend logs, and execute custom programming. I think I might want/need much closer to realtime control for what we've been talking about here, although some of the kit I work with currently will get down to 1-2 second response. (The older stuff was at best 5 seconds, more commonly 10-20, and worst case once a minute.)

And certainly I'd want a hardwired safety - can't count the number of control jobs I've come in on and found the previous installer relied solely on the controller to handle safety issues. A disaster waiting to happen, and in a few cases they did. (Which was usually why I was there...!)

dlenox · « **Reply #45 on:** January 03, 2009, 07:00:33 PM »

Bill,

Welcome to the world of XW... it's a good inverter series, but not perfect.

On a side note: has Xantrex actually released the 4500 series yet? And when you want monitoring/logging software, from all I know I am the guy you want to talk to (http://www.briery.com)

As it turns out I live in a potentially violent wind environment. The issues become compounded by being grid tied, and using a relatively (220ah) battery bank - which is 2x the recommended minimim from Xantrex.

One of the issues that is getting buried here is the fact that my 17' turbine is typically highly stalled. Saying that I know from experience that when the wind hits 36+mph it turns into a different animal, and instead of highly stalled it does not furl properly (to be rectified) and begins to produce a tremendous amount of power. The other day it peaked at 7540watts, on a 6kw inverter, for a wind turbine that basically is designed for 3kw.

I am thinking of this circuit as a backup safety valve, and would think that probably a lot of other wind turbine owners may also want as a fall-back safety device - assuming it was cost effective. My primary defense is my dump load, however when things bad (huge winds, and a grid outage) this is may be my last defense as worst case scenarios will eventially happen.

So there is a lot of background that you may have missed, and my learning is yet to be completed.

Dan

dlenox · « **Reply #46 on:** January 03, 2009, 07:03:36 PM »

Joe,

But don't these systems usually use an embedded operating system? I know when I worked for Lucent they did - and certainly not Windows...

I have been a huge fan of Linux since the early 90's, but 90% of the people in the world seem to settle on their OS.

Dan

dlenox · « **Reply #47 on:** January 03, 2009, 07:17:58 PM »

Ghurd,

No problem with the picaxe - just a suggestion. Let me know if I can be of any assistance.

Believe me I can live with anything better than my current no-fallback situation. In my case the TS-60 is used as a diversion controller only - no battery charging.

I would like to see a range of adjustability not simply a single setpoint trigger voltage. Not sure if that is what you were saying in your post.

Any way of setting the trigger without having to actually achieve the trigger voltage? Like I said the C40 has a point that you can measure some voltage on the board, adjust a pot to adjust the reference voltage of the desired trigger.

Ignore the XW behind the curtain, trigger voltage is just that. If for some outside reason the voltage increases past this set point then so be it, the event still happens. Equalization does not last forever and has to end at some point! Even then I manually would equalize, as it would also trigger my diversion load as well.

Dan

ghurd · « **Reply #48 on:** January 03, 2009, 07:31:01 PM »

Range of adjustability for the relay driver? I am confused. ("trigger voltage is just that")

Can set the trigger voltage with a calculator. And ohm meter.

It would have to experience the trigger voltage to be certain it works.

G-

RandomJoe · « **Reply #49 on:** January 03, 2009, 09:12:18 PM »

Yes, the HVAC panels themselves use a proprietary firmware, but the programming and operating interfaces are all (proprietary, often poorly written) Windows-based apps now. The only reason I even use these panels to play with is I can usually come up with "free samples" from time to time - someone burns up an output or input and needs theirs replaced, but the controller still works fine otherwise. Or they upgrade to newer tech and don't want the old items. Stuff like that.

If I wanted to build something separately from the work-related sources, I've always figured on using an SBC with an IO header, running a Linux stack. RTLinux or such if I really need sub-millisecond timing but I think any 'ol Linux would do fine for most everything I've contemplated. A couple hundred milliseconds either way probably isn't going to affect me.

I keep toying with the idea of picking up a stack of PC/104 cards and see what I can build out of those. I'm not good at doing something without a real purpose though, perhaps between my solar stuff and possibly some future wind stuff I'll find a worthwhile reason!

And you won't catch me in Windows if I can help it. I have to use it at work (all that nice, proprietary software) and I do have an old Win2K session in VMWare for some scanner programming software, but I'm otherwise all Linux since around 1998!

Dave B · « **Reply #50 on:** January 04, 2009, 02:25:05 AM »

Hi Dan,

Still following this thread with great interest and would like to add that I too went through (and still some what after changes) the "different animal" syndrome of the machine taking off out of stall and producing whopping amounts of power.

I rewound a new stator, had new blades carved to 16' from 18', we reversed the pitch (rotation) and lessened the angle of the tail mount all to get this under control and a better match to my alternator and load. It's furling now just before and still sometimes a bit late of my intended 3-3.5KW max @ approx. 300-350 RPM. but I can sleep better now knowing I've let it run in up to 60+ MPH winds.

Proper furling is everything and along with my shorting switch and mechanical brake I am back in control. After burning up 2 previous machines and 1 blade strike I am very close to getting it right.

You have a hybrid also and will eventually get it the way you want. If I remember, you climb and do not have a tilt tower. I sure can appreciate your helpless feeling when that here we go again with the high winds happens time and again with a late furling machine.

Keep us posted, I am probably in for adjustments all over again when I set up for battery charging and will be looking for assistance. Fun all over again. Dave B.

Flux · « **Reply #51 on:** January 04, 2009, 03:21:46 AM »

Dan, Ghurd's controller will be fine if you can persuade him to do it.

Without access to the voltage reference it is a bit tricky to make a test point for setting trip voltage but if you fit a calibrated knob to the set volts pot you can calibrate it to nearer than a volt and that should be adequate. You can play with the setting with the controller relay not actually shutting the machine down to fine tune the voltage.

I would have done something with a voltage reference, comparator and a relay but I don't have time to do this for you. Ghurd's device does the same thing except that the voltage reference is buried in the chip.

Young'uns do it digital, they find it easer, good luck to them. I do it analogue, I find it infinitely easier and at my age I can't change . I find that much of the digital stuff is really a sledgehammer to crack a nut but it comes down to what you understand. The next generation will never be able to use simple analogue solutions.

That is not a problem, what does worry me is that they may loose sight of first principles and we shall continue to re invent the wheel. I see much discussion here that may have helped Faraday but it was sorted in about that era.

Flux

dlenox · « **Reply #52 on:** January 04, 2009, 07:40:46 AM »

DaveB,

I keep pretty tight control on the speed of my turbine, I know that I can be making more power, but until I get furling under control it stays this way.

Most of the times the rpm's are pretty low, and it rarely goes over 125-150rpm. I actually like this as in the long run a low rpm machine should equate to less maintenance issues.

I have seen it spinning at times at 225rpm and it's a bit scarey, I just can't imagine a 16-17' machine spinning reguarly at 300-350 rpm! That has to be a sight (and sound) to see! Too bad we are not geographically located closer.

Sometimes I certainly do get tower envy, you guys with the ability to lower your turbine to the ground/from the ground. But with the small amount of farm animals we have here I imagine that they would stretch out the guy wires pretty reguarly in scratching on them. And I do like being guy-less.

Dan

fcfcfc · « **Reply #53 on:** January 04, 2009, 07:55:03 AM »

Hi:

Thanks for then reply...

Yes, I understand really and I am not trying to over simplify or simplify a non-simplifyable problem (if such a word).

Sometimes DYI'ers have a tendency to let the Mfg's off the hook too easily for two reasons. One, they don't like "endless like" arguing over the phone and Two, the thought of designing and building a work around is JUST SO APPEALING and fun.... LOL well, you get my drift....

If the XW would kick in consistently and where it is suppose to, a dump controller would not even be necessary once you get the machine to "shut down" physically and have the electrical brake max. trip point on line as a backup...

This would give you two simple basic systems to handle top end and grid out...

The number of times you have wind output power beyond 6KW simultaneously with grid outage is probably less than .1%. A destructive .1% yes, but not so frequent that it really deserves complex solutions with all the associated intended and NON intended results...

That was really it...

My VAWT is still in the plan stage due to lack of time to finish the prototype but eventually I will have my own set of problems to deal with I am sure...

Good luck....

.....Bill

Dave B · « **Reply #54 on:** January 04, 2009, 11:48:38 AM »

Dan,

Mine takes 25+ MPH to get into the 300 RPM range and as you know when it's turbulant it can go higher before the furling catches up. I opened my airgap also and it runs (loaded) so that the voltage is approx. 1/3 the RPM. 300 RPM runs about 100 VAC per phase into 3 seperate heating elements. If your blades are well balanced 300 RPM seems fast but is not extreme for our size machines, other's may argue.

Hard to compare but the Bergey Excel 10KW machine is rated well above 300 RPM for I think it's 23' diameter. I'm getting used to it and it really is not that loud but when it furls it lets me know it's at that speed by the sound.

The power curve doesn't lie and we've seen what the additional couple MPH wind speed does when running on the edge, there is no mercy. I hope you can get the furling tweaked soon, I have the same winds here especially lately and always in the Winter months. It has been a huge relief to know my machine fully furls and limits it's output and speed as opposed to before not really seeing any action until 30 MPH +, that was way too late. Keep us posted, Dave B.

windy hill · « **Reply #55 on:** January 04, 2009, 06:22:21 PM »

I would also be interested in some sort of shorting switch. my system is similar to yours (dlenox).

2 17ft other power design gennies

48 volt grid tie

2 outback gvfx3648 inverters

8 6 volt cart batteries

I live in a very poor wind area and seldom see gusts of 35 to 40 mph. But when they occur my inverters will cut out because of high battery voltage. My original thought was that i was simply over powering my inverters So i tried shutting one generator down. That did not help. Volts seem to be rising faster than the inverter can react. The only thing that i can find that helps is reducing sell voltage to a very low setting (approx 44volts) but that causes stress on batteries and one must be aware of a high wind forecast. Adding more batteries was another thought but i am unsure if this will correct the problem. I am happy to know that i am not the only one having this problem and looking foward to the greater minds that can comment on this problem.

fcfcfc · « **Reply #56 on:** January 05, 2009, 02:27:32 PM »

Hi:

Regarding the size of the battery bank, I think the following would work:

Take one of your 6Vdc batteries charged to a no load voltage of 6Vdc. Hook up a variable voltage dc source to it and an amp meter. Also put a volt meter across the battery terminals. Raise the voltage output of the charger until you reach the desired max voltage across the terminals that you want the battery to have to tolerate. When you have this voltage across the terminals, read the current flow. This current x 48Vdc will give you the number of watts max that the turbine can put out before you go above your max set point. Keep in mind though that this scenario is with no load from the inverters. But, it will begin to give you an idea if the size of your battery bank is in the correct ball park.

.....Bill

News:

Author Topic: Fried Tristar TS-60 (Read 9847 times)