I'm just lazy in this regard, and have thus learnt my lesson. Now I don't make those mistakes anymore... 😉
But yes, absolutely. Once you are geared up, it's a cheap exercise to experiment and build different amps. Especially if you use strip board. (I will rather pay the money to get a PCB for every amp though, even if it's just an experiment. Working on strip board tires me out.)
You use the resitor and check if heat up? its the same as use an bulb in serie?
Can use any resistor value in serie? 1R, 2R, 4R, 8R or 10R will be the best? i get 50W instead 25W will be better?
(sorry for too many questions!)
You don't want to use too low a value as then it doesn't serve the safety function of limiting current that can melt an OPS transistor. I might even try 20R as 10R works fine to make some pretty loud music. You can check for heat but mostly have a DMM strapped across and monitor voltage when you flip the on switch. It should have a voltage comparable to what you expect quiescent bias current to be - say no more than 300mA or so on startup because bias pot may be in high position. So 0.3amps across 10R is 3volts. If you see something like 12v or 29v across that resistor - immediately shut off power and debug. You can at that point pull out the driver stage and monitor the base voltage to the driver - that way the OPS are kept safe. You will probably find something causing the VAS to hit the rails and fix that before reconnecting.
WOW thanks! what about size? i mean 50W instead 25W will be the same? i think that price are similar
You don't need a 50w resistor but doesn't hurt to use it. Unless you intend to run at full output swing continuously, 10R is fine. Once, it got so hot I burnt myself touching it to test to see if it was hot. Dumb thing to do 😱 Just use a DMM and trust that 25V through 10R for 30 seconds will make it toasty hot.
High power resistors can be of several types. Most metal clad resistors of 25-100W rating are designed for use with heatsinks (note any fins and tabs with screw holes). If you don't use a suitable heatsink, they cannot dissipate more than 5-10 watts and you'll waste your money because a cheap wire-wound resistor will be just as good when air cooled. Of course, either will become extremely hot at >5W anyway. Metal clad types are expensive, as you find, but are unnecessary for the job.
I find that cheap, lower rated film resistors of 1-3W that burn like a fuse, are better insurance than large resistors that survive whilst your output transistors burn. Small amplifiers of 100W rating are fine with these, using even 22R resistors. Note that the quiescent current of many class AB amplifiers is quite low at only 50-100 mA or so for a single pair of output devices. Only large amplifiers for disco sound levels need more current.
However, the only safe protection device is the lightbulb limiter and for beginners, this is going to work every time - recommended. I still use one with any doubtful customer job or general testing, not just amplifiers.
I find that cheap, lower rated film resistors of 1-3W that burn like a fuse, are better insurance than large resistors that survive whilst your output transistors burn. Small amplifiers of 100W rating are fine with these, using even 22R resistors. Note that the quiescent current of many class AB amplifiers is quite low at only 50-100 mA or so for a single pair of output devices. Only large amplifiers for disco sound levels need more current.
However, the only safe protection device is the lightbulb limiter and for beginners, this is going to work every time - recommended. I still use one with any doubtful customer job or general testing, not just amplifiers.
I prefer bulb limiters too. They are much easier to connect and are foolproof. Its easier to see a light come on out of the corner of your eye than stop and read a volt meter. This is especially handy with random oscillation.
thanks guys, the resistores that i talked about are this http://g02.a.alicdn.com/kf/HTB1G0Er...Power-Metal-Shell-Case-Wirewound-Resistor.jpg come with heatsink
I always used the bulb, but is possible to use the light+resistor, will be ok to upgrade my bench 🙂
I always used the bulb, but is possible to use the light+resistor, will be ok to upgrade my bench 🙂
Use 100 or 60 watt mains filament bulb wired in series with the mains to the amplifier. I
🙂
The bulb rating will depend on the designed idling current of the whole amplifier. You haven't stated what you want to build but assuming a true 80W/8R rated amplifier (not promotional figures) uses one pair of TO3P output transistors and in class AB, seldom draws more than 100 mA/channel from the secondary windings. A 60W bulb will be plenty for such a stereo build.
If your amplifier is going to be bigger, you need to allow for the increased idling current so that the bulb is only just glowing or only just dim when idling with no speakers, no signal. The 100W bulb might suit up to a 150W/8R amplifier but only you will know what the current is likely to be according to class, the total operating current of each amplifier board and all the support circuits at idle.
A bulb limiter is very flexible in that you can't get more current than the lightbulb draws normally when lit directly from the mains. What makes it necessary to get the bulb size about right though, is that the amplifier should have sufficient current to operate normally with no signal, so that fault conditions don't arise on account of not having enough power to operate the support circuits as well, allow you to test and perform adjustments correctly and even drive some small PC speakers at low level, without bringing the bulb up to full brightness and then clipping it badly because of the inherent extreme supply sag.
Conversely, if you use a large bulb, it may not do enough protecting and never glow brightly enough to show useful details like signs of oscillation, bias instability, support circuit operation etc. You may have to calculate a few currents for the transformer primary rather than the load on the secondary windings or find a few larger bulbs at the right price but this isn't hard to understand or all that expensive.
If your amplifier is going to be bigger, you need to allow for the increased idling current so that the bulb is only just glowing or only just dim when idling with no speakers, no signal. The 100W bulb might suit up to a 150W/8R amplifier but only you will know what the current is likely to be according to class, the total operating current of each amplifier board and all the support circuits at idle.
A bulb limiter is very flexible in that you can't get more current than the lightbulb draws normally when lit directly from the mains. What makes it necessary to get the bulb size about right though, is that the amplifier should have sufficient current to operate normally with no signal, so that fault conditions don't arise on account of not having enough power to operate the support circuits as well, allow you to test and perform adjustments correctly and even drive some small PC speakers at low level, without bringing the bulb up to full brightness and then clipping it badly because of the inherent extreme supply sag.
Conversely, if you use a large bulb, it may not do enough protecting and never glow brightly enough to show useful details like signs of oscillation, bias instability, support circuit operation etc. You may have to calculate a few currents for the transformer primary rather than the load on the secondary windings or find a few larger bulbs at the right price but this isn't hard to understand or all that expensive.
Thanks for the explanation - easy to get incandescent bulbs that have all been replaced by LEDs or CFLs in my house. Just need one of those screw bases with terminal screws.
Hi Guys
With the light bulb power limiter, you use a low-watt bulb when first powering the amp. The initial tests are doen with NO LOAD since you really just want to check voltages and see a clean output wave.
Note that even an unbiased or damaged output stage can pass a clean sine wave with no load. A light load of 1k will show up crossover distortion or other irregularities.
Depending on how much gain the amp has, the output offset may not settle until the output stage is set at least to a few milliamps. If the outputs are 'off' then the frontend will battle with it to get the correct DC feedback.
As you become confident that the circuit is wired correctly, you increase the wattage of the lamp. It is best to have at least two lamp sockets wired in parallel so you can simply screw down a second lamp when you wish to pass more power or unscrew one to pass less power. lamp holders like this are available from most hardware stores. You might have to buy ceramic ones with wires attached, or get surface-mounted ones you can attach to a piece of plywood.
You will need various lamp wattages to be able to test various things. You do not use only big lamps for big amps - you get their incrementally from small lamps.
Have fun
With the light bulb power limiter, you use a low-watt bulb when first powering the amp. The initial tests are doen with NO LOAD since you really just want to check voltages and see a clean output wave.
Note that even an unbiased or damaged output stage can pass a clean sine wave with no load. A light load of 1k will show up crossover distortion or other irregularities.
Depending on how much gain the amp has, the output offset may not settle until the output stage is set at least to a few milliamps. If the outputs are 'off' then the frontend will battle with it to get the correct DC feedback.
As you become confident that the circuit is wired correctly, you increase the wattage of the lamp. It is best to have at least two lamp sockets wired in parallel so you can simply screw down a second lamp when you wish to pass more power or unscrew one to pass less power. lamp holders like this are available from most hardware stores. You might have to buy ceramic ones with wires attached, or get surface-mounted ones you can attach to a piece of plywood.
You will need various lamp wattages to be able to test various things. You do not use only big lamps for big amps - you get their incrementally from small lamps.
Have fun
I'll never build a chip amp, not even it will be cheaper, and probably will perform better, than any newbee discrete amp effort, it is zero, cero, null, zèro fun
Bass authority of a chip amp can't touch a well designed discrete amp around here.
I see you guys use the 60w lamp on the AC mains side. I use two power resistors on each leg of the DC rail side so I know which side has the problem and also know what current draw is. So a little more info with resistor (or lamp) on DC rail side.
I see you guys use the 60w lamp on the AC mains side. I use two power resistors on each leg of the DC rail side so I know which side has the problem and also know what current draw is. So a little more info with resistor (or lamp) on DC rail side.
No, don't try to use bulbs in series with the DC rails. They won't give you sensible voltage differences you can measure as with fixed resistors, since their own resistance will change with the current passing through the filament and hence their temperature.
The plan of using the bulbs at the AC power input side is to protect a complete but unknown amp, which will be the case for a new build or one under repair, where you usually need to test everything together. If you are only checking out a stuffed PCB, it would be wiser to use a dual tracking power supply configured for current limiting, that can be adjusted as needed, from maybe 10 mA upward to the expected idling current level. This doesn't have to be a full pro. bench supply with high (rail level) voltage outputs. A simpler, low current and voltage DIY build would be good enough if it used a reliable current regulation circuit and there was provision to meter the actual current and voltage supplies to both rails.
You could still use fixed resistors of several different values or even rheostats with a pair of DMMs reading current but that isn't quite as flexible and failsafe.
The plan of using the bulbs at the AC power input side is to protect a complete but unknown amp, which will be the case for a new build or one under repair, where you usually need to test everything together. If you are only checking out a stuffed PCB, it would be wiser to use a dual tracking power supply configured for current limiting, that can be adjusted as needed, from maybe 10 mA upward to the expected idling current level. This doesn't have to be a full pro. bench supply with high (rail level) voltage outputs. A simpler, low current and voltage DIY build would be good enough if it used a reliable current regulation circuit and there was provision to meter the actual current and voltage supplies to both rails.
You could still use fixed resistors of several different values or even rheostats with a pair of DMMs reading current but that isn't quite as flexible and failsafe.
- Status
- Not open for further replies.