Not really picaxe...

Q

QuIcK

Senior Member
I was reading cooldude's post on dimming etc around the house, and the talk got into 'sine wave' dimming & all that, I didn't want to hijack his thread, but...
could you use a (cheap) audio amplifier, say 100w, to drive 12v halogen lamps? they dont operate at very high output voltages (when not dealing with transients)
then you could generate a sine wave, vary the amplitude and have silent dimming. no harmonics from chopped AC etc.

you would have to limit the amplitude of the sine-wave, as the voltage output would be relative to the power being driven. perhaps as a bespoke dimming solution?

or am I missing something here?
 
D

Dippy

Moderator
46 views and no reply.
Obviously no-one wants to make a nit of themselves replying. :)
So, working without a safety net, here I go.
Once you get one reply it ususally opens the gates...

As I understand it:-
Your plan is to use a PICAXE to generate a 50Hz- 60Hz sine wave.
Your routine can adjust the amplitude of that sine wave.
Then run it through a fixed gain power amplifier.
Yes, it'll probably work to a degree but very inefficient.

Better would be to synthesise the sine-wave using power components.
This is partly how some speed controllers for induction motors works although they are VF.

But for 12V Halogens why bother?
Use Switched Mode techniques instead on a 'DC' signal.
Far more efficient and cost-effective (i.e. cheaper and better).

Why do you say "bespoke"? Would it be? :confused:
 
G

geoff07

Senior Member
You could try but would need beefy wiring. The usual trick of one electronic transformer per lamp means you can get away with 1mm wiring at 240v (i.e. a normal lighting circuit). Try putting say 3A (per lamp) through 1mm wiring at a nominal 12v (=35w). You would lose illumination due to the voltage drop.

If you were to try 12v leds (e.g. 8w MR16 units) then that can work but the waveform is very odd due to the leds so that in fact PWM works much better than amplitude variation. But even PWM is not always good due to the weird things they put in the units (like tiny fans in some of mine).
 
Q

QuIcK

Senior Member
sorry to drag this up. i forgot about the thread.

Just to clear some things up...
wiring would be a pain, thinking about it. would also need a beefy dc supply, which drives costs up quite a lot, compared to a simple 220-12v traffo.
I would probably use a sine-wave oscillator type chip, then a digital potentiometer getting control from a picaxe.

the idea is for a recording studio. there is an awful lot of sensitive audio equipment that HATES noisy mains. we have power conditioners which makes a massive difference, but we have tried normal triac dimmers, and we get a lot of noise.

I came up with this idea for dimming lights without any RFI troubling our outboard gear.

apprently there is AC PWM dimming... i think its just pwm driving a beefy transistor of some sort. then a low-pass filter set around 50-75hz. the result is, instead of a pwm signal with constant amplitude, the pwm sig is modulated by the pwm.

maybe a really fast SSR or something?
 
premelec

premelec

Senior Member
Watts up?

How many watts maximum are you trying to control? 100W amp is refered to in first post but for all I know you were going to use dozens of amps... If only 100 watts and power efficiency is not very important you could use a DC supply and MOSFET controlled current source [big heat sink on that MOSFET]... if you choose the DC supply to be the right voltage at full brightness then it's low loss at full brightness.
 
hippy

hippy

Ex-Staff (retired)
In t'good old days lighting was controlled by rheostats and quite possibly variac and auto-transformers would work. A rheostat is basically putting a potentiometer across the mains and driving the lights from the output. A [ what I thought was a 'variac' transformer but isn't ] is similar but is a variable tap on the secondary side of a transformer. There's also perhaps the possibility of using variable capacitance to alter its reactance to act as an AC-resistor.

Best added : Don't put a potentiometer across the mains!
 
Last edited:
premelec

premelec

Senior Member
ah.... the reostat goes in SERIES with the load... :) Variable auto transformers still work very well with very good efficiency - and you can feed transformer primaries with them with no problem...

If you're going to use 12v halogen lamps you could use an auto battery for very low electrical noise source and only slight gurgling of sulfuric acid...
 
D

Dippy

Moderator
Sounds like you are flitting from one method to another...:confused:

What do you mean by a fast SSR?
You have to switch the AC in time/phase controlled manner for standard AC dimming.

You should list the possible methods and list the disadvantages and advantages for each.

In your app it'll be mostly noise vs efficiency.
Generating a sine wave and chucking it through a linear amp may work but very inefficient.
Maybe you could use synthesise AC and vary the amplitude digitally and filter.

If you start using standard burst dimming techniques for AC then it'll work but noise may be an issue.
Maybe you can filter it good enough?

Maybe you could use DC and PWM for brightness. Networked lighting - piece of cake.
Plus some filtering.

Have a good long think, do some research on nerdynet and have a play.

Budget?
 
hippy

hippy

Ex-Staff (retired)
What about varying frequency ? Motor from mains driving a generator / dynamo with variable gear ratio.
 
V

vttom

Senior Member
Put a variable DC regulator IC at each light fixture, with the output driving the lamp. You need to send 3 wires to each fixture, though... GND, +12V, and a control voltage.

Or, you can just buy commercial DC lighting kit.
 
B

BeanieBots

Moderator
What's with all this fascination with and talk of AC:confused:
I thought the objective was to reduce noise. If yes, then it's a no brainer to use DC.
Use PWM to generate a DC demand voltage by filtering and buffering.
Use the demand voltage to drive an emmiter follower or similar to drive the lamp.
et voila, Steady state DC lamp control with no noise.
 
M

Michael 2727

Senior Member
I thought Halogen lighting didn't like to be dimmed, or do they make better ones now.
What about LED replacements, prolly save some power at the same time :cool:
You can get less noisy dimmers, they cost around 3 times as much and larger.
 
B

BeanieBots

Moderator
Correct, halogens don't like being dimmed, or to be more precise they need to be run at the correct running temperature for the halogen cycle to work. Otherwise, the filament will suffer from premature evaporation. For simialr reasons, they prefer running from AC rather DC though I've never been able to quatify the difference in life expectancy between AC and DC drive.
 
D

Dippy

Moderator
If each PWM controll was at the light node you could use network techniques to control and S/mode techniques to provide the filtered variable power output to the light.
Anyone above vegetable status could ensure there is b-all noise on the power and digital line. And as the distance from modulating 'controller' to lighting device is short the noise would be low. The bottom line is to absorb those nasty high harmoncs within the device.
Sorry, I really don't see why this is being made SO difficult...:confused:
 
Q

QuIcK

Senior Member
it was just a thought. i dont exactly need 255 steps of dimming, its not a tv studio!
basically, i'm just wandering if you guys could think of doing silent dimming.
Ive looked into IGBT and sine-wave dimming. they are expensive, which eliminates them for the minute, but maybe not in the long run.

another thought i had would be a 1xpri : 4xsec transformer.
so, 220v in, the 4 output taps of (for example) 1v, 2v, 4v, 5v. this would give a cut-down binary type output (like the inner workings of a DAC), giving a range from 0-12v in 16 steps.
then use relays, eg.
 

Attachments

Q

QuIcK

Senior Member
i thought about a variable dc reg with local control (my plan is to do local control, with each 'dimmer' only running 1 or 2 lights).
unfortunately, by the time you get the 220vpri, 12vsec transformer at 50va, variable reg able to do 5A (50w lamp), then control for the reg (i was thinking digital pot in it's feedback loop) its about 60 quid a channel, which is an enormous amount for.
I cant imagine its very efficient either.
I'm going to look into the price of a 3-tapped transformer that will work. if they are 100 quid, then the variable reg will be the way to go!
 
hippy

hippy

Ex-Staff (retired)
QuIcK said:
ianother thought i had would be a 1xpri : 4xsec transformer.
Click to expand...
I'm not sure that's going to work the way it's wired as you'll be shorting secondary windings out by the looks of things.

It could perhaps be done if each secondary had both ends of the coil coming out, then you place in series however many secondaries to increase volts ( or remove to reduce volts ).
 
B

BeanieBots

Moderator
I'm losing track of the objective here.
Is it to produce the most complex and expensive lamp dimmer possible:confused:
Second to flashing an LED, dimming should come close to the simplest of functions.

How about a large variac with a motor and gearbox to adjust it. You could use an RC ESC to control the motor driven by 'servo' from the PICAXE.
Maybe a motor driving a generator which supplies the lamp. The PICAXE controlling the motor speed would then control the lamp brightness. Maybe others can come up with more obscure and diverse methods.
 
Last edited:
Q

QuIcK

Senior Member
ah, ofcourse.

i didnt think about the coils between the taps. i'd originally whipped up the schem with individual transformers, then realised it was basically a multi-tapped transformer. now i realise that was wrong. i need to physically disconnect a coil, not just bypass it :rolleyes:
 
Q

QuIcK

Senior Member
nah, the point is not dimming. but dimming without RFI on the mains.

analogue audio equipment is sensitive to the RFI that chopped up mains (read triac dimming) produces.
i'm looking for a clever alternative to IGBT dimming (which is about 100-200 pounds diy, without many examples).

I thought it would be easier working with 12v, as that is within the power of a power amp or op-amp type thing.

just looking for thoughts.
 
E

eclectic

Moderator
QuIcK said:
nah, the point is not dimming. but dimming without RFI on the mains.

analogue audio equipment is sensitive to the RFI that chopped up mains (read triac dimming) produces.
i'm looking for a clever alternative to IGBT dimming (which is about 100-200 pounds diy, without many examples).

I thought it would be easier working with 12v, as that is within the power of a power amp or op-amp type thing.

just looking for thoughts.
Click to expand...
Just a wild / whacky thought.
A couple of mates who are radio hams
don't get mains hum on their transmissions.
They use big 12v batteries.

e
 
D

Dippy

Moderator
If all the previous suggestions are impossible then just switch on more lights.
I'm genuinely surprised that you aren't capable of modulation methods AND supress the noise.:confused:

"dimming without RFI on the mains".
If you'd read my previous suggestion you would have seen that this wouldn't have been the case.
I wish you luck.
 
You must log in or register to reply here.
Top