input needed. VHO testing...

[DJ88]

Ok,

It is starting very soon. Before hand I want to ask for any ideas on what I should test for and post online about the VHO testing I will be doing. Here is what I was going to do so far.

Power(watts) consumed by ballast.
Power(watts) consumed by lamps.
Temp of lamps.
Temp of ballast.
Lumens of lamps.
Hours run.
Model of ballast.
Lamps will be URI aquasun 10,000K.

Each ballast will be tested for one month. Starting with the 660 I am buying tomorrow.

I am building a frame for my light meter so that I measure the lumens at the same distance each time.

Any input is appreciated.

Ballasts being tested are the following;

Icecap 430 & 660
Fulham workhorse
PFO VHO
Tar VHO
possibly a lampsnow VHO(not sure yet) expensive to order from the states to just test and then try to sell afterwards.

[ 19 February 2002: Message edited by: DJ88 ]</p>

[smokinreefer]

just a FYI, there is someone selling a 660 in the buy/sell forum

[DJ88]

Thanks Shao.. images/smiles/icon_smile.gif Missed that post..

[JoJo]

Hi Darren, the Fulham Workhorse you are going to use, is it a Workhorse 7? If so, was it from Albrite? and how much is it? I emailed Albrite, they have the Workhorse 5 and 7, but won't tell me the price. Thanks!!

Danny images/smiles/icon_confused.gif

[DJ88]

Danny,

Last time I checked the 7 was around $70.

[titus]

Hello,

Oh my my Darren go Darren. Questions:
1) Are the measured parameters going to be logged 24/7 into a computer and then you calculate the mean, std dev, max and min values? I wonder any of us here on this BB have a copy of LabView and their interface card to do this.

2) Ideally, we should also monitor power line voltage to take into account of power flutuations. If we go nuts, we can also measure acoustical and RF noise generated by these guys.

3) Maybe we can get the retailers or manufacturers to sponsor by either lending you the ballast, or selling it at a discounted price. If you need to, tell them you are doing it for this board and that their names will be mentioned and banners posted in the article.

4) Can you please elaborate on the frame? Is it for mounting the lights at a fixed distance above the light meter?

[StirCrazy]

Darren, are you using the same bulbs for the tests on different ballasts? if so you should burn in the bulbs firstt theen you will still have to take into account that they bulbs intensity will have a slight change after 1 month of use so each test will have to have a diminishing factor built in to compensate for this..

Idealy there would be a new set of the same type of bulbs for each test and they would be "burnt in" for 1 month befor you started the test (which would make each test take 2 months). even doing this you would have to also have some way to quality test the bulbs, maby a base line balast that you would pop the bulbs into after the burn in to make sure the output is with in a acceptable amount of deveation.

I am probably just stating the obvious as you probably have already thought of this, but a test like what you are planning (a direct comparason of ballasts) you have to ensure everything other than the ballasts is constant or you will not have valid results.

having said that,
Power(watts) consumed by ballast.
Power(watts) consumed by lamps
Temp of lamps.
Lumens of lamps.
Hours run.
Model of ballast.
Lamps will be URI aquasun 10,000K.
this should be fine, I don't realy think "Temp of ballast" is important unless you are going to put a heat sink on each ballast.

Steve

[DJ88]

Titus,

I will be taking measurements of the lumens, lamp temp and ballast temp twice a day. IF anyone has any kind of data logging software tht woudl be cool. Other wise I was just goign to log it in excel so I can show tha data in tabular and graphical form.

If I had the equipment to measure noise and power fluctuations I'd go for those too.

I have tried approaching Icecap but have not ever recieved a response. If anyone can help me get in touch with any of the manufacturers I'd appreciate it. This is coming out of my pocket.

The frame will be made out of acrylic and will be a box that sits overtop of a 12" length of tube. I was thinking about making it opaque so it will only measure the light emitted. Any thoughts?

Steve,

New lamps will be put in with each test. I will start the measurements from day one and this will demonstrate the decrease of lamp intensity over time. We may find that some ballasts "burn in" a lamp faster than others.

I am reading the temp of the ballasts as it is mentioneed quite regularly as a concern with any type of ballast purchase. I know there will be differences as each ballast will be different with it's efficiency and will produce different amounts of heat. I want to demonstrate where the lost watts of power are disappearing. At least some of them.

Each ballast test will be run on a timer and for he same duration in time of days. 30 days. 60 is getting too long as I'd be looking at this taking a year or more possibly. 30 is fine.

[ 20 February 2002: Message edited by: DJ88 ]</p>

[Canadian]

Darren can you mock up a pic of the contraption you're going to use for the lamp and meter mount (I can't picture it)? I agree with your idea of using a completely sealed and opaque container to eliminate any ambient light skewing the results. The only difficulty I forsee is, in using a sealed box for the lamps and meter, you're going to have to allow for some ventillation with a fan or something so the temp doesn't get crazy-hot inside there.


I don't really see any need to test the lamps and ballasts for an entire month...unless of course your motivation for this testing has changed. If you're still just testing to see if one of them there ballasts, which will remain unnamed, is underdriving VHOs then you should be able to just run them for 24-48 hours and ensure that you do the same for each test and ballast. This way you can do the same test 3 times for each ballast and take the average rather than doing only one test which could give erroneous results for one reason or another.

[DJ88]

Hey Andrew,

My plans for the box are so that it is only placed over the tube for measurements. Then teh lid is closed once it is taken. imagine a box 12" high 6 inches wide and 8" deep. A notch will be cut to fit around the tube on the sides. The meter will be placed on the top of this only for measurements.

I want to run this for a longer period of time so as to see what kind of "burn in" each ballast has on a lamp. As well I want a large amount of data collected. So as to minimize any erroneous data. two measurements or more will be made each day to also factor out changes in power to the ballast etc.

I do see your point about running them for a longer period in the day to make the test runs shorter. ie 18 hours a day and test three or more times. instead of 12 hours with two measurements.


Titus & Victor(or anyone else who has the following),

Do have, know anyone who has a good quality Oscilloscope I can use for this? I want to measure Amps drawn adn well as voltage into the balalst but need to know phase angle to calculate Watts and Not Volt Amps.

Victor, if you have a scope I could borrow for a bit, I am sure I can let ya use the lux meter for an extended period of time for your own testing of Overdriven NO's.

If no one has that what about a good wattmeter???? up to 400W or so.

Awaiting any responses.

[ 20 February 2002: Message edited by: DJ88 ]</p>

[titus]

Hello Darren,

1) Draw out your plan for the box and send it over.
2) Look at this for data logging of the parameters you want:
pH,Temp,Light Data Logging (http://www.drdaq.com)
3) PC based oscilloscope:
PC Based Oscilloscope (http://www.picotech.com/40_42_100.html)

[ 20 February 2002: Message edited by: Titus ]</p>

[DJ88]

Little bit of an update.

After some thought I will now also be meauring the PAR as well as lumens for each ballast. I have ordered a PAR meter and hopefully it will be here in time to start this all at the end of the month.

As well I amlooking into getting a power meter(watts) if anyone has acess to one I could rent for say six months I'd appreciate it.

[StirCrazy]

don't you have a multi meter Darren? you can get the watts with that. also how much was the Par meter? did you find one cheeper than 500.00?

Steve

[DJ88]

Yes I have a multi meter. problem with that is it measures RMS voltage and RMS amperage. Those don't give you true watts. They give you Volt Amps(VA). In AC circuitry that is. In DC you will get true watts.

There are three types of power measurements in AC circuits.

Volt Amps (VA)
Watts
Volt Amp Reactance (VAR)

I'll post more if ya want.

I need a meter that does the calculations to get Watts in an AC circuit. If I dont get a wattmeter I need an oscilloscope to find out phase angle..

yeah I found one much less than $500

[DJ88]

Tho I could buy a meter that measures power factor and then calculate Watts fom VA..

hmmmm

[StirCrazy]

um, you going to share how much the par meter was and whare you got it with us? sorry I didn't come right out and ask I thought you would have caught it images/smiles/icon_smile.gif

Steve

[DJ88]

Apogee instuments. $90 it is just the sensor.

[StirCrazy]

90.00US or canadian, damn to bad you didn't tell me I would have given you the money to get me one and we could have split the shipping.. oh well

Steve

[reefburnaby]

Hi,

Unfortunately, I can't find my scope. Its somewhere in my house...but I don't know where I put it.


You may find a big difference between real power and VA on the Tar ballast. Usually, those tar ballasts have a poor power factor. But, the rest of the other ballasts should have very close VA and real power measurements. Since power factor rarely change over time, you can make simple VA measurements for the time being and then borrow/rent a power analyzer to measure power factor.

- Victor.

[ 21 February 2002: Message edited by: reefburnaby ]</p>

[DJ88]

Victor,

I figured that most of the electronic will be quite good for a power factor. I'd still rather get all the numbers fom start to finish. I am not leaving any part of the data out if I can measure it.

Anyone have a Fluke model 41 around? images/smiles/icon_wink.gif

[titus]

Hello,

My personal way of hitting this experiment will be to sample all data sets 24/7 for the entire length of experimentation. Fortunately, we can do this if we have a ADC card for the computer. There are a few out there that are around the $100 price tag. The website I posted previously would have these device on hand.

[DJ88]

I'd love to have automatic uninterrupted logging of data. But it's all the extra expenses. I am incurring a lot of them for this right now. Ballasts alone will hit almost $500.00 or more. images/smiles/icon_eek.gif So far its $200.00 on light meters. images/smiles/icon_sad.gif

If I could set up something to measure and record mV on my PC I'd be laughing. It's that damned cost again. A rough guesstimate for exchange $Cdn to £ is 2.2. So for a £120 board I am looking at $300.00 once shipping is included. If not more. Then there is duties and taxes. Grrrrrrrrr.

Tho I may talk to the head of my program at BCIT and see if he can get me an interface card somehow. They hav eto exist. It is BCIT after all. All I need is one to measure mV.. all I need.. Then I could run it 24/7 over a couple of weeks and we could see a graphical log of the output decay in PAR for each ballast. Set up a sealed enclosure with one lamp inside and the PAR meter attached inside. Fan for cooling and circulation and you can do 30 days of testing in 1/2 the time. All with some pretty sweet graphs showing the decay.

Hmm.. damn.. more ideas..

Thanks Titus. images/smiles/icon_wink.gif images/smiles/icon_eek.gif

Still looking for a Fluke 41 tho. Be nice to measure the wattage with only one meter. images/smiles/icon_wink.gif

You know. Even if I was only able to get a board made to measure higer voltages and not accurate with the mV, it would be a snap to make up an amlification board the PAR sensor feeds into then the board sends the amplified signal to an interface card for logging. Then it is the software. grrrr..

I do have an extra PC here I could set up to run the collection software only. Hmm..

TITUS! Why'd ya have to bring this up! lol

Just took a closer look at the Data logger Titus.. It is only accurate down to 5mV. I need to 1mV to be accurate. The sensor is accurate to 1mV = µmol m^-2 s^-1. Plus it only measures to 5V. Sigh.

Grrr.. double Grrrrr hmm amplifier.. sigh.. I gotta try n sleep now.. Not gonna happen.

[ 22 February 2002: Message edited by: DJ88 ]</p>

[DJ88]

Found this one in New Jersey. This is a definite possibility. Calling tomorrow for pricing.

PC Voltage recorder. (http://www.ertco.com/volt101_voltage_recorders.html)

It has an accuracy of .5mV and can do real time data logging. Comes with software that is easily modified/set up for sepcific uses.

ie. recording 1mV as 1 µmol m^-2 s^-1.

Hmmm..

[titus]

Hello,

My thoughts:
-wait a month or two so we can find better data logging solutions
-maybe in a month or two I finish the trouble part of my thesis and can do the acrylic part for you (I already have an idea how), as long as someone is getting the material
-also once I'm done I can actually call up people like in PFO to see if I can talk them into loaning you a ballast.
.... but again it's time. This is a long project that deserves to be done right and proper.

[DJ88]

Well the volt recorder is outta the picture. As i sone other product I found. The VR-101 is $199.00 US and the Gauge Compu Scope is $4,995 US.

I'll do it as I intended. Manual recordings three times a day.

And after talking to Dana Riddle The box will be the size of a VHO lamp and will hold the meter 3" above the lamp with three spots to measure. As well I will be going with 2' lamps. This way once the testing is done I can use the lamps myself.

[DJ88]

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr> -wait a month or two so we can find better data logging solutions
<hr></blockquote>

With the budget I have set out I won't have much cash to buy anything more than the ballasts and the lamps at this point in time.

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>once I'm done I can actually call up people like in PFO to see if I can talk them into loaning you a ballast.
<hr></blockquote>

Too late. I have already ordered a ballast frn PFO. Within a few weeks I will have a 430, 660, PFO, workhorse and a Tar.

Thanks anyways Titus.. appreciate it.

[titus]

Hello,

Is it easy for you to build a cheap simple low speed ADC interface for the serial port? Matlab has serial access functions and the following code sample serve to illustrate how we can use our computers with a simple interface to do continuous data sampling.

Assuming that we want to take a sample every 60 seconds with a timing resolution of 'res':

% Code Snippet
res = 0.5;
previous_time = 0;
current_time = clock;

if [etime[current_time, previous_time) < 60)
pause[res);
current_time = clock;
else
previous_time = current_time;
% Trigger and sample ADC
current_time = clock;
end
% Code Snippet

[DJ88]

Titus,

I wish I had the time. Midterms starting next week. And in all honesty I want to do this with as little construction as possible.

Another worry for bulding the board is that would provide a source for nay sayers to say my results are worthless. If I can give the stats for all the equipment used, there are less possibilities for people to discredit the data.

Thanks for the idea tho.. images/smiles/icon_smile.gif I'd love to in all honesty.

[DJ88]

Looks like I found what I am looking for Titus. As far as Data Acquisition. take a peek at this. I am waiting to find out if it goes down to .5mV.

DI-194 DAQ (http://www.dataq.com/cgi-local/SoftCart.exe/products/startkit/di194.htm?E+scstore)

For $33.90US with cable and software I can't go wrong. It's my kind of price. Shouldn't be hard to write the code to just log mV measurements.

[titus]

Hello,

Wow! That is inded cheap. Actually the price is only $12.95 USD, isn't it?

The thing has 20Vpp of measurement range with 8 bit accuracy. This is 20V/256 = 78.125mV resolution. Unless they have switches that allow you to make it more precise.

[DJ88]

Titus,

With shipping and the cable it comes to the $33 price.

took a closer look again.. won't work.. crap..

[ 03 March 2002, 00:25: Message edited by: DJ88 ]

[StirCrazy]

ok I am confused :D
you had said you needed to be able to measure 1mv incraments up to 5V (at least that how I understod it smile.gif ) but when Titus worked out the resolution it worked out to 78.125mV.

no basing my assumptions on the deff of resolution being
"Resolution is the smallest difference in measurement that can be read from the instrument."
that means you would have 6 steps from 0V to 5V so you could measure 0V, 78.125mV, 156.25mV, 234.375mV, 312.5mV, 390.625mV and 468.75mV and you would not be able to accuratly measure any valu in between. now if there is something I am missing please tell me as this is confusing me :D

Steve

[titus]

Hello,

The spec for the card Darren posted says its resolution is 8-bit, meaning it has 256 different levels.

[StirCrazy]

so is that 0.78mV steps then?
just to conferm my thinking.. 1 volt = 100mV.. right?

Steve

[titus]

Hello,

No. 1V = 1000mV.
So 10Vpp of input divided into 8-bit (256) of resolution gives you a number I originally posted. Too lazy to pull our my calculator.

[StirCrazy]

ok ya that makes sence.. don't know what I was messing it up with to get that but oh well..

Steve