Audio Power Amplifier Design book- Douglas Self wants your opinions

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
the LED bulbs give about 5 to 7 times the light output for the same wattage as a Halogen bulb.
The long fluorescents give about 2 times the light output for the same wattage as a Halogen
The compact fluorescents give about 50% more light output for the same wattage as a Halogen.

The Halogen gives about 30% more light output for the same wattage as the cooler filament in an ordinary incandescent.
Photofloods give out more light but have a very short lifetime. That's where the Halogen technology comes in: longer lifetime at the higher temperature (=more light).
 
the LED bulbs give about 5 to 7 times the light output for the same wattage as a Halogen bulb.
The long fluorescents give about 2 times the light output for the same wattage as a Halogen
The compact fluorescents give about 50% more light output for the same wattage as a Halogen.

The Halogen gives about 30% more light output for the same wattage as the cooler filament in an ordinary incandescent.
Photofloods give out more light but have a very short lifetime. That's where the Halogen technology comes in: longer lifetime at the higher temperature (=more light).

Really interesting data.

Amazing that CFLs only give out 1.5X the light of a Halogen. I hate CFLs, and my experience is that they don't last.

Cheers,
Bob
 
Really interesting data.

Amazing that CFLs only give out 1.5X the light of a Halogen. I hate CFLs, and my experience is that they don't last.

Cheers,
Bob

They have also been known to burn the house down due to overheating of the electronics in the base. This is aggravated if used with a dimmer, even if the dimmer is set to max, due to the fact that the CFL has to be restarted after every pulse.

Unfortunately, it appears that LED replacements could have a similar problem.
 
The long fluorescents give about 2 times the light output for the same wattage as a Halogen
AndrewT data is way off, better stick to audio, time to do a bit of research with examples from mfg data sheets and not come up with inaccurate data.
For instance,
A F54T5HO lamp puts out about 4700K mean lumens at 54W lamp power, mid 90's lumen per watt eff. At $5/lamp retail!
The F28T5 is probably the most eff fluorescent lamp made, putting out close to 3000 lumens at 28W, so over 100 lumen/W.
A Philips GU10 50W halogen puts out ~430 mean lumens, which works out to 8.6 lumen/W
So above example is a >10x ratio not 2x of a linear FL lamp as stated earlier.
What does a LED of equivalent lumen cost per W? Come on now do your homework, how about 10X maybe of a T5?

T12 lamps are basically banned for new fixture construction due to lamp inefficiencies!!
CFLs are a poor design, almost half the light is trapped inside the twisty and only are rated for around 8-10K hours with appreciable lumen depreciatuion.
Industrial CFL's are a bit better such as the ones that have 3 long sections and a plugin base. They also have a separate ballast, so you do not have to throw away the electronics, if the lamp dies. Still a CFL is around the 40-50 lumen per W range = poor compared to a T5 linear.
 
Last edited:
A F54T5HO lamp puts out about 4700K mean lumens at 54W lamp power, mid 90's lumen per watt eff. At $5/lamp retail!

I keep my plants alive with these. Just bought 2 (warm +cool)...
Cost 7$ each at home depot (philips).

WOW ! .. they are bright on my SS ballast . :cool:
Found the whole rack/ballast on the street corner - F54's were going bad.
Whatta light ! - brighter than sunlight (1 meter).
OS
 
Found the whole rack/ballast on the street corner - F54's were going bad.
Whatta light ! - brighter than sunlight (1 meter).
Good to hear OS, you have caught on.
Philips even have a F54WT5HO/EA (49W) lamp if you want a bit more eff, but more $.
I have boxes of ballasts that were sent back for warranty, saying they did not work, well they have been working for me for years now!!
The trick is to keep the ballast cool, most fixture mfgs fail (Philips/Lithonia) by putting the ballast(s) in with the lamp/heat source = early failure, ecaps mosfets go bang!!
If you want a halogen spot replacement, use a 35/39W Metal halide lamp ANSI M130 (PAR20) and a electronic ballast. The ballast is ~40-50 $ and the lamp is $20 for an Asian brand.
T5HO also come in circular lamps too, F55T5HO 2GX13 base.
Good score on the fixture. I have used Tremclad white water base rust paint(no more stink), it has a nice glossy shine to it.
Time to start my pepper plants going, T5HO works great for seedlings. You have to get the lamps close, say 6-10" and they are nice strong plants ready to go outside in mid May for me.
I try to tell everyone about T5HO lamps!!
For seedings use more blue light (more branching), use 5000-6500K color temp lamps (F54T5HO/850,865), but anyone will work.
 
Last edited:
rsavas,
400 watt metal halide lights work real nice for growing plants indoors but they sure do get hot! I'll have to look into some of the model numbers you listed, I am not up on light technology these days. I agree that the compact fluorescent bulbs aren't that nice. I hate the light itself and they don't last anywhere near the time they advertise they do. And as far as being more environmentally safe that is terrible information as far as what is inside those tubes and the fact you are tossing the ballast out with them at the same time. I always mixed incandescent lights with my fluorescent as the flicker from those lights alone would drive me batty after awhile.
 
T5HO usually replaces MH HID in most applications, due to cost and efficiency. This is what they seem to be specifying on new commercial construction.
4-6 T5HO (210-320W,19K-28K lumen) is used to replace 1 of 400W MH (450W, 33K lumen). Even with a electronic HID ballast it maybe 425W. Some use 6-8 T8 Hi-BF too.
There are two different 400W ANSI MH lamp types, probe/regular start M59(old) and Pulse/Hi-v start M135/M155. The later are more efficient and have better properties over-all, They are good to use as well, for certain applications.
Check out this ref.
High Intensity Discharge Lighting | Quartz Metal Halide | Elliptical | MVR400/VBU/XHOPA | 12642 | GE Lighting
Flickering in a CFL, they usually switch in the 20-30KHz & >40 KHz range.
Ever since I got into lighting, I am looking up, checking out peoples lighting setup :) People wonder, what are you looking at :) I get the sure, are you okay, look !!
Getting mercury out of lighting is like getting gas out of your car, venture to guess which will happen first?
 
Last edited:
www.hifisonix.com
Joined 2003
Paid Member
CFL volumes are still growing dramtically beacuse in China, Indonesia, Africa and India people are urbanizing/have more disposable income and need light. They are relatively cheap, efficient and last a long time.

However, LED's are the new kids on the block, and volumes in the USA and EU are ramping very fast (50% YoY growth). Still expensive, but prices are dropping fast.

Unfortunately, most CFL's are not good quality and some can be very noisy (EMI). Best to stick with a named brand IMV.

The most exciting thing for me about LED lighting is the ability to network lamps and then control them via your mobile or energy harvesting switches. This is the hot application area. In the USA you can check out Greenwave and TCP who collaborated to bring smart lighting to some of the big retail chains. I find it pretty cool that you could sit in a cafe in Paris and control your lights in California via your smart phone!
 
AndrewT data is way off, better stick to audio, time to do a bit of research with examples from mfg data sheets and not come up with inaccurate data.
..............
Tha data led me to expect 90l to 100l per watt.
Some leds are now being developed that are getting around 160l per watt.

On that basis I took the printed data as a guide and bought LED replacements for Hologen bulbs.

I now have 9 different types from 7 different manufacturers.

I can assure you NONE get anywhere near the light output from a 10times higher wattage Halogen, No matter what claim the manufacturer makes.

I know that our eyes are scaled a bit like our ears in that they react to a log scale of brightness/power.
I estimate that in general LED replacement bulbs give about half the light of a 10times higher wattage Halogen bulb.

I stand by my original post
the LED bulbs give about 5 to 7 times the light output for the same wattage as a Halogen bulb.
 
the LED cost trap!!
The bottom line is cost/economics,
A F54T5HO puts out ~4500 lumens for single piece cost of ~$5. Find any LED source that comes close to putting out the same light for $5?
As a working example,
Say a LED puts out 100 lumen, you need 45 of them!! so they have to be ~11cents each. Where do you find a 100 lumen LED for 11cents a piece? Now you have to mount all these LED's on pcb assemblies etc. Compare that cost to a simple lamp socket.
I can not see a LED ever competing on a cost bases as I described above.
Bonsai, you can control any light source remotely, it is not just for LEDs.
 
rsavas,
You can also have a screen in your refrigerator door today but it seems rather silly what some of the proposals are today. Do we really need to turn our lights on remotely from around the world? Though we can automate and control many things these days most of the application are unnecessary to say the least. I go hiking in the hills near my house and it is amazing how many people can not put down their phones, they are connected 24/7 and don't even see anything around themselves. We have become to immersed in this technology to enjoy what is already there.
 
Disabled Account
Joined 2012
Like everything else, convenience and longer life cost more. Like with batteries... longer life ones will cost you more. These alternative bulbs are mostly about longer life and the convenience they give in not replacing them as frequently. Especially if the lights are high up in a ceiling etc. With some types, I can install them and pretty much forget about them for 10 years. AND, I can get 'white' light (6500 degree light... close enough) which i cannot get with incandescent bulbs. Are they over-priced? yes, of course... for awhile.

-RNM
 
Last edited:
rsavas,
You can also have a screen in your refrigerator door today but it seems rather silly what some of the proposals are today. Do we really need to turn our lights on remotely from around the world? Though we can automate and control many things these days most of the application are unnecessary to say the least. I go hiking in the hills near my house and it is amazing how many people can not put down their phones, they are connected 24/7 and don't even see anything around themselves. We have become to immersed in this technology to enjoy what is already there.

Ah , you speak of the "little black box people ". They don't know where they
are without GPS. No knowledge of geography,direction, (moss on the N
side of trees). So reliant on that tech they would be frantic and
stressed without it.

I actually have to memorize the city/area ... I still know exactly
where I am by landmarks (mountains/ buildings/rivers). Remember
addresses , phone #'s ,commitments with just little notes or memory.

We are forced to have a "dump" of info in this tech age. The home PC
is a good repository for the mass of useless required info that we
must have at our disposal.
I have no (dumb)phone ... !! But at least I won't be getting Alzheimer's with
lack of use. :D

OS
 
Last edited:
OS,
I have resisted the smart phone upgrade, my kids have them and I don't need to be connected to the internet every minute of the day. You are so right it seems people do not know how to read a map these days, need a gps nanny to tell them which way to go and where North is. Asking someone for directions to somewhere today is a joke, they want to pull up map quest and then they don't know what they are looking at anyway. I guess the old Boy Scout in me has learned a few things that still come in handy.We had Heathkit and could solder and build something and today they have Arduino and just plug things in. And tell someone to machine something on a non CNC machine these days and they wouldn't know what to do. I guess this old mind is still working pretty good for being from another Century!
 
Tha data led me to expect 90l to 100l per watt.
Some leds are now being developed that are getting around 160l per watt.

And as the reviews show, barely any commercially available lamps reach any of those numbers at this time: LEDBenchmark - Unbias LED lighting reviews.

Most LED downlights, especially drop in ones, suffer from only having sufficient light at the hot spot. The drop in body sizes of MR16 and GU10 are only good for dissipation up to around 6W given the practical constraints of LED emitter temperature. Consequently most 'drop in' bulbs at present miss the mark on replacing a 50W halogen altogether. You need about 10-12 Watts of LED power at least.

Really for the price and results nothing beats fluorescent tube lighting for workshop illumination. It gives you the minimised shadow flood effect, retains fairly good colour and energy performance (as long as you don't go for bottom shelf parking lot grade tubes), solid life expectancy and is quite cheap to replace versus any equivalent in LED.

The advantages in practice going for LED for this are not really that major at this time.
 
Last edited:
Cherry stated that these caps are crucial for stable performance. So if it is not (a mild) shunt compensation, what else it could be?

Cheers,
E.
It is a negative feedback capacitive path, generally to compensate for the input pole of the gain device being compensated.

It's the same kind of capacitance you put at at opamp's output to negative input, to increase stability, by bypassing anything that's in the global loop at a certain frequency (that darn outputstage that's a lot slower than the driving stage)

When you look at a transistor as a gain device, you can look at one this way:
-Base/Gate: Input(+)
-Emitter/Source: Output(+)
-Collector/Drain: Output(-)

A transistor or FET is the smallest unit of (inverting) gain device; a stage comprising a number of transistors most often also yields an inverting stage to which this feedback can applied. Think VAS input, output stage output.

The usual miller cap is also just that, a capacitive negative feedback path from the (Multi)transistor's output to input for stabilizing purposes.

Hence, I've termed capacitors that compensate gain devices by connecting the output inverted to the input (either by an inverting input OR output) "local feedback" capacitors. Because it's feedback around a single device.

Now, when you tink gain devices made up of several transistors, you can keep thinking the same way: Local to that device. That's what "cherry" compensation is: a capacitive negative feedback path that includes the output into the VAS. The output, with respect to the VAS input is inverting.

These caps one finds all over the place in nested loop systems to keep inner loops sane and stable. To coin differently each and every feedback cap configuration, while all they are is mere negative feedback caps to the gain unit (one or more transistors /stage / stages) for stabilization purposes, is a bit uhm... ridiculous? Also given this view (a simple capacitive negative feedback) it can't be called anything special on its own?

TPC describes just one feedback path, the rest attempts to describe various combinations of local and extra-local (next nesting level, not always the global loop) compensation.

So with respect to the "Cherry" compensation and mandatory "shunt" compensation; if by "Cherry" you mean you take out the basic miller compensation and "include" the output stage instead, you've effectively taken away stability from the inner, local loop. That's why those "shunts" are needed, It looks odd, a cap from supply rails to the base/gate, but in effect, they are local NFB loops to the gain device.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.