Short circuit protection

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The idea of using a circuit breaker is good from a safety standpoint (helps prevent fires and such), but don't expect this to save the amp itself from damage.

Some designs use fuses in the output signal path, but this often won't protect your output devices from damage during a short and can cause audible distortion. You absolutely do want fuses in your +/- DC supply (to protect the transformer from a shorted output transistor). I buy power transformers that have a thermal cutout switch built into them to make sure that expensive hunk of iron doesn't melt down if I really mess up.

The most reliable way to protect against a shorted load is to use what is called foldback current limiting. What happens is that the amp senses the voltage drop across the output transistors load sharing resistors to determine what the load resistance is. The protective circuitry then reduces the output signal to stay within the safe operating range of the output devices no matter what the load is doing. When you remove the short, the amp starts output full power again (assuming the supply fuses haven't blown).

Some people don't like the effect that fold back current limiting has on the sound when it engages. Personnally I enjoy replacing output transistors a lot less. In my opinion, if the current limiting circuitry ever engages during normal use then you either have way too low impedance speakers or else you need a more powerfull amp.

Check out Randy Sloan's book on building High Power Audio Amplifiers. He goes into great detail on how to design an amp to prevent damage from a shorted output.

Phil Ouellette
 
Web site wrong

Sorry guys I got the wrong site The # for SEAL Electronics is(606)452-4135 in weeksbury KY.

Phil in Understanding Electricity and Electronics by Randy Slone on page 183 The purpose of an amplified diode circuit is to provide forward bias for output driver transistors to eliminate crossover distortion as it pertains to the amp on that page . I guess I answered my own question oh well ( I bet i'll remember it ) :)
 
Wow! A Blast from the Past. I rememmber that Note from years ago. Indeed it gives good Treatment on the Subject.While Your at that site also Check out AN1308 100 & 200 Watt Wideband Low Feedback Audio Power Amp. Another great read. BTW I also do not like Foldback Curent Limiting and I use rail Fuses and Bypass the Fuse With a Film cap. As far as Protecting the Transformer a Mains fuse on the Pri. Side of the transformer is the Best overall protection and is required by alot of Gov.Reg's ie UL, VDE, SA,
 
Hi ppl,

I agree that a mains fuse or circuit breaker is definitely required to prevent a fire hazard, but given the need to handle inrush currents, it is not likely to provide much protection for the transformer itself. A thermal cutout switch built into the transformer absolutely prevents overheating damage and should be standard practice for power amps (plus it's cheap, Plintron only charges a dollar or two extra for this).

Fuses in the supply rails are needed to protect the rest of the amp from a blown output transistor, but don't provide any real protection for the transistors themselves. By the time a rail fuse blows (assuming it does), the transistors themselves are usually long gone. After all, you can easily exceed the SOA for a bi-polar transistor without ever exceeding the max current capability of the amplifier (the SOA is load impedance dependent). Once secondary breakdown begins, it's by-by transistors.

I know that many DIY amp builders don't bother to include any kind of output protective circuitry (nor do they do anything to protect the speakers from DC), but this isn't any reason to recommend against taking extra steps to build reliable amps that can survive abuse.

We don't cut corners about things that often have at best a minimal effect on sound quality (like the differences between two brands of top notch caps or resistors), so why then be unwilling to go the extra mile and take some steps to prevent all that hard work from going up in smoke.

Phil
 
I had problems with distortion from protection circuits at output power stages in my last project.

It happened because the speaker impedance have expressive dips at some frequencies.

I solved the problem reducing the lower impedance limit of my protection circuit, but it´s is not the correct way, I think, since I didn´t change the power supply and the output and driver stages too.

I think the best way to deal with this problem is using some kind of soft clip limiter at the input of the amp and leave only the short circuit protection to the output circuit.
 
Haldor: I agree with your Views that Protection is Needed and It is True that any General DIY amp as well as Mass market Consumer Amps Maby should have Every Protection Scheme Knowen to man. However I Build my amps along the Lines of the old Threshold amps I was convinced From Many Cases of Consumer and DIY Power Amps i have delt with, That MR. Pass Views at that time Regarding Current Limiting and the Like do Cause Premature Clipping and limit the True Power into Reactive (Real World Speakers)Loads.However i also use His Methods of Lots of output transistors and massive Heatsinks. Thermal cutouts are a good idea not only on the Transformer but the Output transistor Heatsinks as well. Yes the Mains fuse in concert with some Kind of Inrush Current Limiting via a Thermister or Soft Start relay work Rather well. The SOA of a transistor will Change with tempature as well as how mutch Voltage is across it and Foldback Current Limiting is a Constant type of Protection Method and is most often set for worst case conditions thus limiting Amplifire Performance at levels way below the Rated Power under normal operating Conditions. I have also removed the Current Limiting on lots of Consumer Amps and by adding Thermal Cutouts upon the Heatsinks and rail fuses Resulted in an Amp That sounded Better at all volume levels and would servive a shorted output with Hot output transistors. if Foldback type of limiting is wanted then My fav method is what is done on some Marantz solid state amps rather than Cut the Drive off fron the VAS stage thay connect the Transistors used to Current limit to a relay to shut down the amp or disconect the speakers. Another Problem with some Amps is that thay become Unstable when into Current Limit and I have seen quite a few Tweeters Blowen Because of this.

DISCLAIMER: The above are what i have found Works for me and your results may varry. The above concepts are for the Advanced DIY'er and should Not be attemted by those unable to Properly Calculate all the Elements Required to implement the System or Subsystems mentioned above.
 
Hi PPL,

I hope I'm not being too much of a boor. I'm new on this site and am probably shooting my mouth off more than I should. I'll confess, when it comes to DIY projects I have turned into a belt and suspenders kind of guy, probably because of all the smoke I've released over the years doing dumb things.

I'll never forget my first real electronics project, a 50W tube based Ham radio transmitter I built from a kit in high school. I was so proud of that baby, carefully following the directions, dressing the wires nicely, clean solder joints. Everything looked so perfect, right up until I plugged it in and the smoke started rolling out of the transformer. Turns out I'd wired the filter caps in backwards and when the transformer overheated the windings shorted and it just toasted everthing, I think I actually saw the windings glow for a second or two. It's amazing I wanted to have anything more to do with electronics after that.

Now, I always power up every new project with a variac and even then I don't stand too close until it's had a chance to warm up for a few minutes (1812 overture with exploding tantalums anyone). In my own case you're right, there can't be too many protective devices.

Peace,

Phil
 
Halder: No i do not find your Post Boring at all. Like I sed i tend to agree that for people just starting out should Use lots of Protection Methods untill thay get comfortable With Building and using the Equipment. I can relate to your Tranceformer Problem. I would have Thought that an AC mains Fuse would have saved that transformer.Like yourself i can remember the Tube era quite well. I had lots of problems with DIY Articals in the Likes of Popular Electronics, Radio Electronics ect. Most had lots of errors that later were Corrected in Print and quite a Few lacked Basic Protection Schems like the AC mains fuse.One Protection Method that i use Both in concert with a variac or as a stand alone Idea is to place a standard Household light Bulb in series with the Transformers Mains input. The Wattage of the Lamp being about half of what the Amp is Rated to output Total. This will light the lamp and save lots of expensive parts in the evbent of an error.Under normal operation the lamp lights Bright when power is first applied then gets dim as the Supply caps charge. If the Lamp stays bright then a problem is happening. This method is for Class Ab and class B amps since Class A amps Draw alot of Current the wattage of the Lamp will have to be raised. I can't say what woud be a Good Rating since I havent Built alot of Class A amps.Yes there are Times when just a Variac alone will allow to mutch Current into a circuit and thus i use this with the lamp for First time Circuits and Troubleshooting.
 
I studied the possible ways of protection circuits a while back and actually the best method in my mind would be to use a specialised IC.

There seem to be three that are still available from some parts sources. These are

TA7317
HA12002
upc1237

One can find the datasheets from
http://www.freetradezone.com

These are a good solution as the total part count is quite low and at the same time you get
*power on delay
*dc voltage protection for output
*over current protection
 
ppl said:
light Bulb in series with the Transformers Mains input. [/B]

A label printer we use to make used a lamp as a current limiting element. The lamp lets a quick pulse of current flow until the filament starts to warm up and increases the series resistance. Back when I worked in the technical support group I once got a call from a technician wanting to know why the lamp didn't light up when he opened the cover. I guess he thought it was like the light in your refrigerator. ;^) That really made my day.

Phil
 
I think that simple current limiting circuits on the output are perhaps a better solution than foldback current limiting... with these, you shouldn't risk the possible instability of a foldback circuit, and the effect on sound quality should be nil if it's properly implemented... of course the trick is making sure that your load isn't going to draw enough current to reach the current limit. Now, of course with a simple current limit circuit, you're still going to deliver the max current into a short circuit, but if your output stage is robust, at least you'll be within the SOA, and your transistors will survive.

The GM75 amp by Neil McGann has a simple current limiting scheme which works well... you can find a carbon copy of Neil's extinct website complete with the schematics on my fossilized website (if it's still there...): http://members.nbci.com/_XMCM/hifiZen/kaa2.html

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