Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

June 16, 2024, 02:21:56 PM

Login with username, password and session length

Recent Posts


Light Bulb Limiter

Started by J M Fahey, March 17, 2011, 12:23:04 AM

Previous topic - Next topic


cool - thats what i thought!

so if i turn on this amp thats been sitting around in an unknown condition and start to troubleshoot it and the bulb is glowing brightly (this is all hypothetical right now) it's going to be alright to leave it on as i try and figure out the problem? i was imagining id have to shut it off if there was excess current, but i guess if the bulb has an actual the limiting function things should be all good, riiiight?

J M Fahey

Anyway be prudent.
Turn it on, measure, guess what's wrong, turn off and replace what you think is bad.
Turn on again, recheck, and so on.
And there's many measurements which are made with the amp off, such as continuity, resistance, transistor shorts, etc.
And some damage can be had even with bulb limiter inline.
Example: you put an electrolytic the wrong way.
It won't *explode*  but it may very well overheat and vent.
Or any part which can be damaged if it receives the full 25 W allowed by the bulb.
So use it, it's a tool, definitely not Magic.


Last question:

is the light bulb limiter only useful in a preliminary testing stage - making sure youve got the right voltage on rails and such things, no shorts, no faulty or open spots - but should be removed before actually testing the amp in a playing situation?

the reason im asking: i was thinking about it today and it occurred to me that if the bulb limits current to below 1 mA, the amp cant actually function properly, can it? depending on the amp, of course, but most need more current than that at least through the primaries, right??? in which case if you had the limiter inline, could you actually test the amp beyond taking voltage measurements?

does this question make sense?


The current will not be below 1ma but I think you meant to type 1 amp. The idea is to have no load until the DC offset is at least below 500mv on the output. You would then put a dummy load on the output to make sure the DC offset is still at least lower than 500mv. At that point you can use a scope to look at the wave form on the output and anywhere through the amps signal chains. Say you have a healthy amp and connect it to the speaker then yes you can play the amp, but it will not sound very good. It at least can confirm that it is passing signal, but if you have a scope you will know that already.

Juan Fahey has in the past given advice about connecting the speaker after the unit is powered up. I can't explain his thoughts on connecting the speaker while after turning the amp on... He has his reasons for this and I hope he can explain that part for us. So I guess we both have a question about this topic.
"A person who never made a mistake never tried anything new." -Albert Einstein


Oh yeah I meant 1 amp -

In the case of a tube amp, though, I'm obviously not turning it on without a load, anyway....

But just to be clear, you are confirming my suspicion: an amp with a light bulb limiter inline will not actually SOUND like it should, even if everything in the circuit checks out, correct?


Yes not only does the bulb limit current that can get through, but it is limiting voltage as well. So power rails through the amp will go down by a certain percentage too. Basically the sound of the amp will be limited too. So yes a healthy amp won't sound like it should with the bulb limiter in line.

Edit: I am referring to an example of using the light bulb limiter in regards to a solid state application in my previous post.
"A person who never made a mistake never tried anything new." -Albert Einstein

J M Fahey

The "connect load *after* turn on" bit comes from the fact that most SS amps start "stupid" (imagine yourself being shaken away in the middle of the night and asked something within seconds)

In the case of the amps, normally there are a couple large value electrolytic caps which take a couple seconds to be charged and reach operating voltage.

During that time it's common that the still unbalanced amp places full +V or -V at the speaker terminals, which is heard as a turn on loud Pop.

Which means it's pulling *big* current from one of those rails .... often more than the bulb limiter will allow...... so rail voltage collapses.

On *some* amps, that collapsed rail does not allow some important capacitor to charge, it never stabilizes, bulb keeps shining and amp does not work.

Yet same amp, if turned on without load, does not collapse and in a couple seconds is stable.

So the drill is:
Turn amp on with limiter, no load
Measure voltages, specially no DC on output (less than 200mV)
If normal, connect speaker
Listen for hum/hiss/buzz/whatever
If all fine, play a couple chords at a couple watts volume (which is quite loud inside a small room)
If all fine, plug amp straight without limiter.
Connect load.
Enjoy .  :dbtu:


Be using a Light Bulb Limiter for years , sure saves on fuses lol


Quote from: Jack1962 on June 23, 2014, 12:51:53 PM
Be using a Light Bulb Limiter for years , sure saves on fuses lol

And smoke detector batteries!
Life is what you make it.
Still rockin' the Dean Markley K-20X


Internet/YouTube advice
Twice this week folks have turned up with repairs using a limiting lamp rated at 250 to 300 watts.

Well, like some TV detectives I don't believe in co-incidence, so I wasn't too surprised to discover not one but several YouTube videos on how to build a limiting lamp (or dim bulb tester) that advise using globes of such high wattage.  In one of those videos Uncle Doug even says of the globe "the bigger the better", so we have people specially seeking out a globe of 300 watt or more rating.

Since this is horribly bad advice for anybody trying to repair a solid state amplifier I decided to collect some actual numbers to show why.

Temperature co-efficient of resistance
The first thing we need to be aware of is that a tungsten filament lamp is not a pure fixed resistance, its resistance changes with temperature, and since it goes from room temperature to about 2500-3000ÂșC when running it shouldn't be too surprising that its resistance also changes, almost as dramatically as its temperature, typically increasing by a factor of between x10 and x20 times.

It is this fact that we make use of with a filament lamp as a power limiter.  If we are careful with our choice of lamp then it will be cool and low resistance if the amplifier is operating normally, but will light and go much higher resistance if the amp has a fault.  This means that when cool most of the voltage will appear across the amplifier, but if hot most of the voltage will be dropped across the lamp.  The lamp will operate somewhat like a self-resetting fuse.  This is why other types of lamps such as fluro's or LED's won't do.

So a filament lamp is a non-linear resistor, and as such we can't simply apply Ohm's Law.

But just as a fuse must be selected to be large enough not to blow during normal operation, yet small enough to blow when there is a fault, so our limiting lamp wattage must be selected to pass enough current for amplifier idle operation, but to light and limit if the current gets too high.  You wouldn't say "the bigger the better" with a fuse, and the same applies with a limiting lamp.

Typical amplifier load
I measured the idle consumption of my homebrew Twin-50, two identical 50W s.s. amps of conventional design in a single case.

My energy meter told me they were drawing about 100mA at 240VAC and a power of 15 watts.  Well these do not exactly compute, but we can get the general idea that one 50 watt amplifier draws about 10 watts at idle, that's around 50mA at 240VAC or 100mA at 110VAC.

This is for idle or very low output, and the amplifier mains input "looks" like a fairly high resistance, but if we try to drive the amplifier harder it will appear to be a lower resistance and draw more current.  This increased current must come through our lamp which causes it to heat up and increase resistance.  Actually this is an unstable flip-flop situation so that when the lamp starts to light the change in its resistance causes a regenerative or positive feedback effect where the bulk of the voltage drop quickly shifts from the amplifier to the lamp.

If we stop or reduce the drive to the amplifier the reverse will happen and the bulk of the voltage drop will shift back to the amplifier from the lamp as it cools.

So what we need is a lamp that will pass the idle current with minimal voltage drop, but if the amp tries to draw much more current than this, say due to a shorted output transistor or power supply rectifier, the lamp will light up and drop most of the voltage across itself, saving the amplifier from further damage.

Possibly the most important application is when testing a solid-state amplifier that has just been repaired.  If a damaged component, such as a driver transistor, has been overlooked then just plugging the amp in and switching on can result in all the newly fitted transistors being destroyed in milliseconds.  The correct limiting lamp will prevent that very annoying disaster.

This table shows the marked wattage and implicit nominal resistance and current for a range of lamps.  It also shows the measured cold DC resistance, and the ratio of hot to cold resistance.

While all of these lamps, with one exception, are for 240VAC operation, the same will be true for 110VAC lamps although the resistances will be about half and the currents roughly doubled, however the wattages divided between the lamp and load will be the same on either supply voltage.

WattsOhms(hot)Amp(hot)Ohms(cold)R hot/coldRemark
3516460.146121.113.6Arlec QI downlight
4014400.167104.313.8Philips golfball
4014400.167104.213.8HPM golfball
5011520.20882.414.0GU10 QI downlight
6010420.24074.114.1Performer (250V)
1503840.625Par 38
2752091.146Heat lamp
3751541.563Heat lamp

Choosing the right wattage lamp
We can't just apply Ohm's Law to a non-linear resistance, but a bit of electrical commonsense and intuition leads us to a rule of thumb that the limiting lamp needs to have about the same power rating as the amplifier being tested.  You can go lower or higher depending on what is to hand but reasonable limits would be between half and double the amplifier output rating, tending to the low side out of caution.

Note carefully that we do not expect the amplifier to work normally when a limiting lamp is in series with its mains supply.  Some voltage will be lost across the lamp so the DC voltages in the amplifier are likely to be similarly reduced, and while the output may be capable of something close to full voltage swing with the speaker load disconnected, the amplifier will not be capable of anything like full power output into a load, and any attempt to drive it hard will cause it to sag and the lamp to light.

While a limiting lamp will work fine with the vast majority of guitar amplifiers there are some amplifier circuits, particularly later models with Switch Mode Power Supplies (SMPS), that will not work at all unless the lamp is quite a high wattage.

Valve/tube amps are a bit of a different case because they have a higher idle power draw due to the valve heaters.  They also tend to be a lot more forgiving under fault conditions, inherently soft-starting as the valves warm up.  Where you get instant fuse blowing you can be pretty sure that the fault is somewhere in the power supply, and there is normally much less need to power the amplifier up limited so you can take measurements.

While filament lamps are being phased out for normal lighting applications there are some exceptions, oven lights, fridge lights, and heat lamps.  In the first two applications the high and low temperatures currently prevent the use of SMPS, so that rules out CFL's and LED's.  Heat lamps of course are intended as radiant heaters so they are not "inefficient" for their application.  These specials look like being available into the foreseeable future.
If you say theory and practice don't agree you haven't applied enough theory.

J M Fahey

Agree and add: so called "high efficiency" filament lamps are still being freely sold.

They actually have a small filament quartz lamp inside a standard glass bulb, with standard screw (E27)  base, for replacement purposes:

They work exactly the same as the old ones, just adjust for real power: the "100W equivalent" is actually 75/80W and so on ("high efficiency", remember? ::))  but that's no big deal.

And of course, you can still use a "pencil" quartz lamp

although the smallest one easy to find is usually 100/150W  ... of course you'll need the proper socket ,  I strongly suggest you use it inside the proper light fixture; they get REAL hot and a fire is possible.

Don't ask how I found out  xP


PIC 01
My old (unsafe - no safety ground) Light Bulb Limiter.

Note that it has a 3 prong plug but only because that's what I had when I made it.  ::)

If ROLY were here he would give me a good verbal lashing for such an abomination.
Well... it actually works as long as you are careful... and it has saved me from some mis-wired lashups on the bench.

PIC 02
My newly constructed (safety grounded) Light Bulb Limiter.  <3)

PIC 03
Hey!!! why doesn't the bulb light up?  :lmao:
If it ain't broke I'll fix it until it is.


Yeah I made one on an old wooden checker board that I made in shop class years ago. However, it was wired with a three prong, but still I always thought to build a more professional looking device. Still using the good ole checker board one today, lol.

Quote from: galaxiex on July 26, 2015, 04:44:43 PMIf ROLY were here he would give me a good verbal lashing for such an abomination.
Well... it actually works as long as you are careful... and it has saved me from some mis-wired lashups on the bench.

Speaking of Roly... I have not seen him post here lately and hope all is well with him.

"A person who never made a mistake never tried anything new." -Albert Einstein


Quote from: DrGonz78 on July 27, 2015, 01:37:24 AM

Speaking of Roly... I have not seen him post here lately and hope all is well with him.

Yeah, me too.

Last post I saw from him ended with "I'm getting too old for this" or something like that...

I do hope he's ok.
If it ain't broke I'll fix it until it is.


I just had a feeling something was not right... Just after reading this I remembered that Roly also took care of the Australian Valve Amps site.


Just went there and right at the top of the page it reads "Maintenance of AVA has been transferred to Stephen Bruce, as Roly's health prevents him from continuing. Over the next few months I will be applying the pending updates. Please be patient. Only this index page and the submission details have been updated so far."

Well I hope he is doing okay and just had a feeling something was just not right since he has been absent as of late.  :(
"A person who never made a mistake never tried anything new." -Albert Einstein