I've built several kind of amps, from fully symmetrical design to single differential design. All have turn on and turn off thump.
Searching here, the most encouraged cure is to use output relays.
But some good commercial amps do not have these thumps, although they do not use delay relay at all.
1. What is the mechanism of this Thump generation?
2. Is there any trick if we design an amp so it will not have thump?
Searching here, the most encouraged cure is to use output relays.
But some good commercial amps do not have these thumps, although they do not use delay relay at all.
1. What is the mechanism of this Thump generation?
2. Is there any trick if we design an amp so it will not have thump?
Those thumps are usually caused by capacitors placed in the signal plath or in biasing networks. At power-on those capacitors are discharged and they take some time to charge to its steady state conditions. Most circuits tend to show undesired behavior until these capacitors are charged
Sometimes clever design may reduce thumps without adding output relays. Symmetric topologies tend to be less likely to produce thumps. Slowly raising and dropping the supply rails tends to dramatically reduce the thump magnitude
Sometimes clever design may reduce thumps without adding output relays. Symmetric topologies tend to be less likely to produce thumps. Slowly raising and dropping the supply rails tends to dramatically reduce the thump magnitude
In the same vein as what Eva was saying, I recently had cause to pull out an old Dyna PAS tube preamp to build a small system for my wife's use. I had never used the preamp; it had been donated years ago by a friend. I thought that it would be wise to check it out on the bench before pressing it into service. I replaced the aged electrolytics and turned it on.
No thump.
The secret? Tube rectification. The rails come up very slowly, indeed.
(The preamp worked fine, by the way.)
It's not difficult to design regulators that incorporate a slow start. That way you get a fringe benefit by bringing up the rails at a rate that allows the circuit to settle.
Grey
No thump.
The secret? Tube rectification. The rails come up very slowly, indeed.
(The preamp worked fine, by the way.)
It's not difficult to design regulators that incorporate a slow start. That way you get a fringe benefit by bringing up the rails at a rate that allows the circuit to settle.
Grey
At first I tought this is due to which supply gets the value first, the +VCC or -VCC. Any imbalance voltage in rising or turn off will make thump.
It seems that making the supply ramping slowly is the answer. If the supply is +/-40V, we make it ramping 0-5-10-15-----35-40V slowly?
This needs special PS design. Any design available to look at?
What about opamp ps. Once I make this slowly ramping supply for opamp (manually), but they still exhibit thumps.
It seems that making the supply ramping slowly is the answer. If the supply is +/-40V, we make it ramping 0-5-10-15-----35-40V slowly?
This needs special PS design. Any design available to look at?
What about opamp ps. Once I make this slowly ramping supply for opamp (manually), but they still exhibit thumps.
i do not have sure lumanauwn, but maybe a cheaper way
Thank you Eva to clarify the thump in a simple way to understand.
Also thank you Lumanawn to introduce this question.
I am thinking if we can left the circuit in "stand by mode", something alike remote controled television set...but much more simple.
I imagine one resistor in series with transformer primary....resulting in half voltage, or less than half voltage... circuit always on. To switch circuit to power work, you make a short circuit over the resistor leads, making a switched jumper on it. Hey!, i never did that!....do not have sure if will work!
Other simple, and cheap idea is to increase the condenser value making voltage divider to supply driver transistor and bias circuit....making the big cap charge slowly.... this way, the thump will be there, but will be interesting and slow...you will see the speaker move slowly up and down till stop in a stable form.
The third one may be the best.... use a relay and a big condenser in parallel with the coil.... feed this relay througth a resistor.... speaker will be conected to output post when the rellay became on.... the RC will produce the time needed to amplifiers stabilize... when really ready...no more fluctuations....this one seems, to me, to be very good....because no complicated circuits, and you can also use other relay contacts to switch on a fan and to make some led show you when ready or not ready.... many people love those lamps in the amplifier panel, but not my case, ligths flashing to me, only in Christmas time.
regards
Carlos
Thank you Eva to clarify the thump in a simple way to understand.
Also thank you Lumanawn to introduce this question.
I am thinking if we can left the circuit in "stand by mode", something alike remote controled television set...but much more simple.
I imagine one resistor in series with transformer primary....resulting in half voltage, or less than half voltage... circuit always on. To switch circuit to power work, you make a short circuit over the resistor leads, making a switched jumper on it. Hey!, i never did that!....do not have sure if will work!
Other simple, and cheap idea is to increase the condenser value making voltage divider to supply driver transistor and bias circuit....making the big cap charge slowly.... this way, the thump will be there, but will be interesting and slow...you will see the speaker move slowly up and down till stop in a stable form.
The third one may be the best.... use a relay and a big condenser in parallel with the coil.... feed this relay througth a resistor.... speaker will be conected to output post when the rellay became on.... the RC will produce the time needed to amplifiers stabilize... when really ready...no more fluctuations....this one seems, to me, to be very good....because no complicated circuits, and you can also use other relay contacts to switch on a fan and to make some led show you when ready or not ready.... many people love those lamps in the amplifier panel, but not my case, ligths flashing to me, only in Christmas time.
regards
Carlos
If you regulate the rails for the driver and amp circuit(excluding the output collectors) then you can design a transistor swicth to switch both the + & - rails on and off using say a couple of comparators or op-amps. You can then use a logic +5V to turn on and off the whole amp circuit. The collectors of the outputs will not have any current flowing in them when the drivers are off. Then you can have another transistor switch to switch on a "bleeding resistor" to discharge the filter caps slowly so there will never be any loose current to create a 'thump'. You may need to have a very small AC power transformer to run the digital circuitry.
BTW: Using this tecnique you can create a delay when turning the amp on and allow the filter caps to fully charge when the rails are switched on. If flip-flops are used in digital circuit, cheap reliable momentary switches can be used. Also you may include a 'fault' circuit in your design to switch off the amp if say a fuse blows or current sensor is triped or something like that.
I have melted my fair share of outputs and some are quite expensive. I have developed some circuits that have saved me money many times and needs not a relay.
BTW: Using this tecnique you can create a delay when turning the amp on and allow the filter caps to fully charge when the rails are switched on. If flip-flops are used in digital circuit, cheap reliable momentary switches can be used. Also you may include a 'fault' circuit in your design to switch off the amp if say a fuse blows or current sensor is triped or something like that.
I have melted my fair share of outputs and some are quite expensive. I have developed some circuits that have saved me money many times and needs not a relay.
Hi lumanauw,
The basic trick is to have the voltage amp stabilise before the output stage is activated.
As long as the voltage amp runs after the power to the output stage dies there will be no turn off thump either.
My personal believe is that an output relay is a great way to protect your speakers. Therefore, not a bad thing. Relay contacts in good condition affect the sound less than a fuse. (one more contact and a thermally variable resistor)
-Chris
The basic trick is to have the voltage amp stabilise before the output stage is activated.
As long as the voltage amp runs after the power to the output stage dies there will be no turn off thump either.
My personal believe is that an output relay is a great way to protect your speakers. Therefore, not a bad thing. Relay contacts in good condition affect the sound less than a fuse. (one more contact and a thermally variable resistor)
-Chris
And not complicated too... a problem with the special minds
Normally they are produce complex ideas, because their mind full of options...never go to an easy solution.
Cuningham is one famous man, have wonderfull knowledge, is respected as one of the bests we have in forum.
But unfortunattely they do not use the straigth line as the smaller distance between two points anymore... they have so big knowledge that they already realise the "straigth line", do not exist in our universe.... everything is curved, have turns, because a straigth line over the earth is never flat, because earth is rounded...all spatial travels made considered that aspect, from gravity atractions of big masses in space..... they considered the non possibility of a straight travel line...... also a flat line made with a pen in a paper is not a line....is a lot of points over the paper "peaks"..... when you go closer to a line, you will se that is not straigth....nothing is perfect aligned in our universe.
Those deep knowledges, makes people ideas always more complex... this way, Mr. Cunningham, beeing an expert in electronics, will never use a damn relay with stupid contacts... this is too much simple.... will use many parts to make this job...will make it better, maybe a little bit more expensive, but will make it better.
Unfortunattely, in audio amplifiers, each stage create its problem to the final result.... some noise, some disturbance.... and the more crazy is that.... the biggest the knowledge, the more complicated the circuit, with too much stages creating their problems, and the next stage trying to compensate something damaged in early stages... this way...enormous circuits, and the sound....normally strange.
Those crazy things in the life... the best the enginner.... normally, with many exceptions, i believe Cuningham can be one of them, but normally their products sound strange.
The techonologie, is so much advanced that is creating some problems..... we can see 2 milions transistors to control some switching functions and lamps flashing on panels.... or displays...
You go on deep, and you will see that the signal patch, normally not many transistors doing the real work..... and a lot of complications come together... with one with their probability of failure.... resulting some miracle working... we never know...how long!.
Strange those things.... by this reason i am running from Medicine Doctors....in special the best doctors.....they will complicate all my life... some of them can kill me..... i am running fast as i can from medicine doctors.
Why not a capacitor or a relay... to delay condenser charge, making infrasonic speaker movement?.....because is too much simple.
regards,
Carlos
Normally they are produce complex ideas, because their mind full of options...never go to an easy solution.
Cuningham is one famous man, have wonderfull knowledge, is respected as one of the bests we have in forum.
But unfortunattely they do not use the straigth line as the smaller distance between two points anymore... they have so big knowledge that they already realise the "straigth line", do not exist in our universe.... everything is curved, have turns, because a straigth line over the earth is never flat, because earth is rounded...all spatial travels made considered that aspect, from gravity atractions of big masses in space..... they considered the non possibility of a straight travel line...... also a flat line made with a pen in a paper is not a line....is a lot of points over the paper "peaks"..... when you go closer to a line, you will se that is not straigth....nothing is perfect aligned in our universe.
Those deep knowledges, makes people ideas always more complex... this way, Mr. Cunningham, beeing an expert in electronics, will never use a damn relay with stupid contacts... this is too much simple.... will use many parts to make this job...will make it better, maybe a little bit more expensive, but will make it better.
Unfortunattely, in audio amplifiers, each stage create its problem to the final result.... some noise, some disturbance.... and the more crazy is that.... the biggest the knowledge, the more complicated the circuit, with too much stages creating their problems, and the next stage trying to compensate something damaged in early stages... this way...enormous circuits, and the sound....normally strange.
Those crazy things in the life... the best the enginner.... normally, with many exceptions, i believe Cuningham can be one of them, but normally their products sound strange.
The techonologie, is so much advanced that is creating some problems..... we can see 2 milions transistors to control some switching functions and lamps flashing on panels.... or displays...
You go on deep, and you will see that the signal patch, normally not many transistors doing the real work..... and a lot of complications come together... with one with their probability of failure.... resulting some miracle working... we never know...how long!.
Strange those things.... by this reason i am running from Medicine Doctors....in special the best doctors.....they will complicate all my life... some of them can kill me..... i am running fast as i can from medicine doctors.
Why not a capacitor or a relay... to delay condenser charge, making infrasonic speaker movement?.....because is too much simple.
regards,
Carlos
Re: And not complicated too... a problem with the special minds
Hi Carlos
I have found that over time relay contacts tend to corrode causing resistance at the contact point. This = voltage drop X currnet, and can cause bad contacts and horrible sound. This can be fixed by plating the contacts with gold. Unfortunatly, this is expensive. I have done many experiments with NOT using a relay for two reasons.
1) I already designed my amp circuit with a regulated power supply for all but the output drivers(collectors) so making it switchable was not that much more complicated. Just turn off the regulators.
None of these added componants are in the signal path. Tis not that complicated, just simple electronic circuits put together in redundancy.
2) 5 surface mount transistors, one quad comparator chip, and a few small surface mount resistors are cheaper than buying a relay large enough to break a peak of 7A and not fuse contacts together. We are talking just a couple of dollars. Maybe the circuit construction is more complex, but I am only making one unit for my self and complexity really doesn't bother me too much.
I started with a simple amp circuit and just have continuously figured ways to improve. However, on this path, I may never actually finish it.
destroyer X said:
Hi Carlos
I have found that over time relay contacts tend to corrode causing resistance at the contact point. This = voltage drop X currnet, and can cause bad contacts and horrible sound. This can be fixed by plating the contacts with gold. Unfortunatly, this is expensive. I have done many experiments with NOT using a relay for two reasons.
1) I already designed my amp circuit with a regulated power supply for all but the output drivers(collectors) so making it switchable was not that much more complicated. Just turn off the regulators.
None of these added componants are in the signal path. Tis not that complicated, just simple electronic circuits put together in redundancy. 2) 5 surface mount transistors, one quad comparator chip, and a few small surface mount resistors are cheaper than buying a relay large enough to break a peak of 7A and not fuse contacts together. We are talking just a couple of dollars. Maybe the circuit construction is more complex, but I am only making one unit for my self and complexity really doesn't bother me too much.
I started with a simple amp circuit and just have continuously figured ways to improve. However, on this path, I may never actually finish it.
You know, the only time you would break a current of many amps is in a fault condition. Who cares about the cost of the relay then? Unless of course you don't turn the volume down before turning the amp off. Then you cause your own problems. Replacing a relay every few years is not a great hardship.
On really big amplifiers, I've seen triacs to short the output to ground. This works well for the speaker(s) but is hell on the outputs. No series element to worry about. I guess we'll call it the purists path. It's simple as well.
-Chris
On really big amplifiers, I've seen triacs to short the output to ground. This works well for the speaker(s) but is hell on the outputs. No series element to worry about. I guess we'll call it the purists path. It's simple as well.
-Chris
That's OK Cunningham, i turned myself nervous, a bad mood.
I am sorry by the agressive post.
And you broke my mood with so much kindness, i thank you very much for that.
You can not realise what was going inside myself, a terrible revolution when i perceived the thinks (designs) are turning more and more complicated.
Also, and in special, the Eva amplifier.
Those things happened together when i could see clear the problem.... cannot make simple....if did it simple, will not be recognized as good designers anymore...because all the people doing the same....that's some "silent" Techonologic Society Psychological Pression against us.
I wrote an enormous text... critics of complicated designs are there.... and you stopped the hole thing.
Will not put the text anymore...i realise you all, designer, are the product, the result, of our society, not guilty people.
Now my rare moments of bad mood disappeared.
this way, never mind, and he happy Cunny.
Carlos
I am sorry by the agressive post.
And you broke my mood with so much kindness, i thank you very much for that.
You can not realise what was going inside myself, a terrible revolution when i perceived the thinks (designs) are turning more and more complicated.
Also, and in special, the Eva amplifier.
Those things happened together when i could see clear the problem.... cannot make simple....if did it simple, will not be recognized as good designers anymore...because all the people doing the same....that's some "silent" Techonologic Society Psychological Pression against us.
I wrote an enormous text... critics of complicated designs are there.... and you stopped the hole thing.
Will not put the text anymore...i realise you all, designer, are the product, the result, of our society, not guilty people.
Now my rare moments of bad mood disappeared.
this way, never mind, and he happy Cunny.
Carlos
Since, the thread title exists, I am posting in this old thread.
I found that in 3 stage class AB amplifiers, on/off thumps can be avoided by having quick acting power switches AND a capacitor across the VBE multiplier which prevent the output from being driven during the first few milliseconds after turning power on. Similarly, on turning off, if power is turned off quickly, arcking is almost eliminated giving rise to no thumps.
This is the strategy I use, and I have no thumps on power on and power off.
I found that in 3 stage class AB amplifiers, on/off thumps can be avoided by having quick acting power switches AND a capacitor across the VBE multiplier which prevent the output from being driven during the first few milliseconds after turning power on. Similarly, on turning off, if power is turned off quickly, arcking is almost eliminated giving rise to no thumps.
This is the strategy I use, and I have no thumps on power on and power off.
Arranging for all the biasing to be active long before full rail voltage is achieved can mean an amp is pretty civilized at switch-on and switch-off - the feedback loop locks from low voltages.
It also makes testing easier as you can run the amp of +/-10V rails if you want before going to full power testing.
It also makes testing easier as you can run the amp of +/-10V rails if you want before going to full power testing.
Simple circuits don't thump. The dynaco PAS2 & ST70 didn't thump. Tube rectifiers. The dynakit ST120 didn't thump, with only 12 transistors and a capacitor output. Sounded like **** OEM, but didn't thump. An Apex Ax6 replacement board for one PC15 in the ST120 chassis, 5 transistors, capacitors in & out, doesn't thump.
The hammond H100 with 27 tubes doesn't thump on or off. Hammond A100 same way.
The Peavey products I own dedicate 4 to 7 parts per channel to not thumping. Most have a relay to disconnect the output for a time, but the PV-4c and the PV-1.3k don't.
The Allen 300 I worked on had a whole chassis with timer & relays dedicated to not thumping, which caused the organ to go silent after ~20 years as the electrolytic cap used as a timer started leaking too badly to charge up. Disconnecting that caused the amp to pickup sports talk AM radio, as they failed to put a coil in the outputs of the 2 S100 amps. The open frame relay was a disaster at 30 years, points needed cleaning every year to maintain flow of sound. Even at 100 W speaker currents.
The hammond H100 with 27 tubes doesn't thump on or off. Hammond A100 same way.
The Peavey products I own dedicate 4 to 7 parts per channel to not thumping. Most have a relay to disconnect the output for a time, but the PV-4c and the PV-1.3k don't.
The Allen 300 I worked on had a whole chassis with timer & relays dedicated to not thumping, which caused the organ to go silent after ~20 years as the electrolytic cap used as a timer started leaking too badly to charge up. Disconnecting that caused the amp to pickup sports talk AM radio, as they failed to put a coil in the outputs of the 2 S100 amps. The open frame relay was a disaster at 30 years, points needed cleaning every year to maintain flow of sound. Even at 100 W speaker currents.
Last edited:
The simpleton in me likes simple amplifiers. In fact, I wish my largest amplifier was even simpler than it actually is.indianajo said:
Take it or leave it: circuit reliabity goes down as the number of discrete components increases. Reliability is very important for my simple mind, which apparently, for some is unable to discern 'good sounding' amplifiers. For my simpleton of a mind, if a claim in not backed by hard science with mandatory reproducibilty and repeatability and peer review of research results, it is as worthy as the smelliest of trash. I do not want unnecessary filtering in my amp, whose purpose rests solely on opinion. I want facts, and facts, require reproducibilty and peer reviews.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.