# I need a 82 VOLT DC power supply!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

#### Rox

i thought an easier way for you so you can check again;

if you conect in parallel the diodes 3 & 4, you would spect a forward volt vaue less than 0.760 since they "would" drawn 1,5 amps each. I would say you will meassure 0.760 value exactly just as if it was only 1 diode there.

#### Guy Grotke

face it: I am right after all

I explained why I can't implement your idea: I don't have a high precision ampmeter that can handle 2.5 Amps. But I do have a high precision voltmeter. Even if the absolute accuracy of my voltmeter is off by 3%, the relative measurements are quite useful and completely repeatable. My voltmeter really does measure in individual millivolt steps.

>the resistance is regulating somehow the current, not the diode....
Okay, now you are just making stuff up because you can't admit that I was right.

But even if I could measure the Amps directly by putting an ampmeter in series with one of the diodes, the resistance of the ampmeter shunt would act as a load balancing resistor so the two diode voltage drops would differ. So a reading that showed exactly half the total current would not mean they had the same forward voltage drop! In fact they would be WAY OFF because a 0.1 Ohm meter shunt (for example) is pretty close to the effective 0.1 to 0.15 Ohm resistance in the diode! Even a 0.01 Ohm meter shunt would throw it off by 10%. My method does not introduce any such measurement artifacts because my voltmeter impedance is thousands of times higher than the circuit it is measuring.

I got the 0.9 Volts @ 2.0 Amps and 1.0 Volts @ 3.0 Amps directly from a 1N5404 datasheet graph. But we can look at another one:

http://www.datasheetcatalog.com/datasheets_pdf/1/N/5/4/1N5400.shtml

The Fairchild datasheet has a pretty readable graph and you can blow it up with Adobe Reader. I see 2 Amps @ 0.82 Volts, and 3 Amps @ 0.87 Volts. That is 1 Amp / 0.05 Volts, or 20 Amps/Volt in the region we care about. Using this chart's values, now my worst case diodes 0.009 Volt difference would give you 0.18 Amps difference. So one of the 3 Amp diodes would carry 2.68 Amps, and the other would carry 2.32 Amps. Still works just fine for a lamp that draws 5 Amps average.

#### Rox

Ok, it starts to looking one of those stupids discussions I have some in a while.

you are saing the internal resistance of the diode is making the diodes match just as I did with external resistances, is it?

I didn´t thought about it untill you talked about... then, why is it recomended to match with externel resistances. And It is not me the only one sugesting it, It is general knowledge, diodes zeners... are not welcome in parallel.

I was looking on those datasheets also, but please would like the one exact you used. Sounds strange to me the specs you used...

I an not saiyng your test was wrong, I just am saing it does not prove the basics of what we are discussing.

As you have no the tools to do the propper test, I suggest you to do a sacrifice with two diodes; conect them in parallel and make 5,5 amps go throw the total system, let us know how long does the diode live. (I guess the problem now, is that you can´t have 55W resistance of 1.8 ohm)

Anyway, thanks for your interest, I don´t care what people does, I learned it was not welcome paralelling them, just will follow my convictions.

#### Rox

got another question for you guys... i ordered a diode and it came in today. upon further research, i find out that there's a "transient voltage suppressor" used with this particular diode the part number is: P6KE 400V 600W

http://www.mbelectronics.com/quickreferenceguide/AI_2000.pdf

it's on the last two pages.

do i need this? what is the suppressor doing in this circuit? i'm going to do a little more research, but i value your guy's input!!!

thanks again,

#### Guy Grotke

why a transient suppressor?

This would be used to protect the diode from any voltage spikes present on the line. If the diode is rated for 400 peak inverse volts, then anything over that would damage it permanently. A transient suppressor can be placed across the diode to absorb any brief (ie. a few microseconds) spikes that exceed the suppressor's voltage rating.

Maybe they knew their fan motor was likely to inject such spikes when it was powered on or off. That is where voltage spikes usually originate: Inductors (such as motors) can generate very high voltage spikes when you disconnect them from a low impedance power line. The magnetic field collapses and since the impedance of an open circuit is very high, the voltage can be many times the original power source's.

update:

hello again!

just figured i'd let you guys know what's happening. i got the 82v diode, and everything is working great! i want to thank you for all your help. if it wasn't for forums, i'd be screwed .

right now i'm using the ENX bulbs (360W). i tried the FXL (410W), but i cracked an essential piece of glass ("fish eye lens") due to too much heat. luckily it only broke in 2 pieces, and i was able to epoxy it back together. i'm looking for some UV filtered, tempered glass to put between the bulb and lens so this doesn't happen again. it's pretty bright now, considering it normally has a 250W metal halide bulb. can't afford the price tag on that one- \$400 . all things considered, i'm happy. \$15.50 for the projector, \$7.00 for bulbs, can't beat that!!!

on a side note:
for anyone who plans to use the ENX or FXL bulbs, spend the extra couple of \$\$\$ to get the brand name bulbs (3M, GE, Apollo, Philips, or Sylvania). with the cheap ones, the Dichroic Reflector material breaks down after a couple of hours, and i see faint splotches of red on my screen. that, and the reflection goes down, thus decreasing light from the bulb...

thanks again!

#### Guy Grotke

need more cooling

Going from a 250 Watt MH to a 360 or 410 Watt halogen would generate a LOT more heat. That's probably why you cracked the lens, and also probably why some of your bulbs have burned up their dichroic coatings so fast.

Get some more airflow and get a real IR filter, not just some tempered glass. BH Photo online sells Rosco Hot Mirror and Thermasheild. Either will do a much better job than plain glass.

If you want to keep it real cheap, DIYprojectorCompany has some IR filter glass that works pretty well. I think the piece I got only cost \$11.

#### Scubasteve2365

LOL @ Rox ...... Where do you find the time man, you do this all day on Lumenlab, and I come here to find out you get into these long drawn out debates here to.

What other websites do you frequent, I wanna check out your post, you are entertaining. You are the only person I know, that will debate something longer than me, and I debate for a long time.

Anyway, I havnt read over this entire post, just skimmed through it mostly.

I dont have an experience with higher power diodes.

I can say, that I too have always been taught that diodes in general where NOT welcome in parallel. But engineering school and alot of places dealing with LEDs and IN4001-4006 series diodes are all very low power.

So in the forward bias, you will have your .7V drop on each one, this would split the current and be less effecient, perhaps enough for an unwanted effect in your low level circuit.

But, in a higher power application then any added resistance, or slightly larger forward bias voltage drop is probably negligible. There is a much larger difference in a drop from .7V to 1.0V, then there is a drop from 85V to 83V, likewise 20ma-25ma, versus 2.5A to 2.3A. In low power diodes that 5mA difference can be cruicial, but in higher power diodes 100mA may not matter much. Its all RELATIVE!

My point is that in general, it is a bad idea to wire two diodes in parallel, but there can be some cases in which the application can call for it properly.

Lastly, I believe that there will never be a way to prove either side of this debate.

I now await 3 pages of data sheets from ROX!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.