Background. I want to design amplifiers by listening the result on loudspeaker. Problem arose when I realized that without good speaker, I wouldn't be able to judge if an amp is great or not, or sound the way I wanted it with my real speaker. And for safety reason I couldn't prepare a great enough speaker for testing with amp prototypes.
But I thought that great/proper headphone would be much cheaper than proper speaker. So my questions:
1) Is it possible to drive a headphone from power amp, lowering the load resistance to around 8 Ohm with some circuit and have the ability to judge the amplifier?
2) Any suggestion with the proper circuit after the output of the power amp?
3) Stupid question but how dangerous is it to the ears when the amp and headphone drivers blow up in smoke?
4) Any better plan to achieve what I want to achieve?
If the above plan fails then I might have to think about serious/fail-proof speaker protection circuit as it will be abused during the tweak, and use my main speaker.
But I thought that great/proper headphone would be much cheaper than proper speaker. So my questions:
1) Is it possible to drive a headphone from power amp, lowering the load resistance to around 8 Ohm with some circuit and have the ability to judge the amplifier?
2) Any suggestion with the proper circuit after the output of the power amp?
3) Stupid question but how dangerous is it to the ears when the amp and headphone drivers blow up in smoke?
4) Any better plan to achieve what I want to achieve?
If the above plan fails then I might have to think about serious/fail-proof speaker protection circuit as it will be abused during the tweak, and use my main speaker.
You CAN drive a resistive load, say 8 ohms, AND attenuate some of the output to headphone levels, for monitoring purposes.
That said, unless grossly wrong designed, "all amplifiers will sound the same" , since any competent modern design will be well flatter and have vanishingly lower distortion compared to what your headphones can offer, so in practice "you will hear no difference".
Modern amplifiers are very close to "a piece of wire with gain" , we have advanced that much.
Personally I would use the best pair of speakers I can get and fuse them well below destruction power level, listening with a them at realistic levels, in a real world room, will show much better detail than a pair of headphones.
Unless you are designing a dedicated amplifier specifically meant to drive headphones, but I guess that is not the case.
That said, unless grossly wrong designed, "all amplifiers will sound the same" , since any competent modern design will be well flatter and have vanishingly lower distortion compared to what your headphones can offer, so in practice "you will hear no difference".
Modern amplifiers are very close to "a piece of wire with gain" , we have advanced that much.
Personally I would use the best pair of speakers I can get and fuse them well below destruction power level, listening with a them at realistic levels, in a real world room, will show much better detail than a pair of headphones.
Unless you are designing a dedicated amplifier specifically meant to drive headphones, but I guess that is not the case.
Last edited:
Use a step-down trafo to feed your headphones. Then you can explore the effect of the PSU on dynamics - listen with a 8R load in parallel with the trafo, and without. Unfortunately headphones don't give you much if any insight into soundstage depth, an important subjective attribute of an amp.
Something like this could be suitable. Adjust the divider to suit the amp and headphone impedance.
DC safe and the zeners will clip the audio at the chosen Zener voltage + the ordinaray diodes volt drop. Use say 4v7 or 6v8 Zeners. Caps are fitted back to back to make a non-polarised one.
DC safe and the zeners will clip the audio at the chosen Zener voltage + the ordinaray diodes volt drop. Use say 4v7 or 6v8 Zeners. Caps are fitted back to back to make a non-polarised one.
Attachments
Thanks gentlemen. Looks like i have enough info to go further with the project.
JMFahey - this project is based on a premise that amps do not sound the same
I will make change to the amp and monitor the result from a headphone. I agree that a real good speaker might be better but if the amp oscillate it can be costly. A correct speaker doesn't have to be expensive, but it is rare and if you have it you will keep it safe.
Abraxalito - trafo! I didn't think of it. It can be effective in assessing the dynamic quality often ruined by voltage divider. May be transparency will suffer but i think it is not what i assess the amp for.
hifly - I have just checked the ESP. Nice but doesn't have the protection.
Mooly - I will try your circuit and compare it with a trafo. Nice protection features.
Anti - I didn't understand about the 40 ohm impedance and the capacitance but not difficult to try both ways and see if there is difference to the sound?
Jameshillj - this is for amp prototype testing. What benefits have you got from using headphone for testing? And is there something special with manganese?
XRK971 - So you guys have used hp to debug amps. I'm surprised that I'm the only one who have not Probably due to previous experience that voltage divider at amp output didn't sound good with hp. But I just found a cheap hp that I think can show the balanced frequency from top to bottom, which triggered this project.
JMFahey - this project is based on a premise that amps do not sound the same
I will make change to the amp and monitor the result from a headphone. I agree that a real good speaker might be better but if the amp oscillate it can be costly. A correct speaker doesn't have to be expensive, but it is rare and if you have it you will keep it safe.Abraxalito - trafo! I didn't think of it. It can be effective in assessing the dynamic quality often ruined by voltage divider. May be transparency will suffer but i think it is not what i assess the amp for.
hifly - I have just checked the ESP. Nice but doesn't have the protection.
Mooly - I will try your circuit and compare it with a trafo. Nice protection features.
Anti - I didn't understand about the 40 ohm impedance and the capacitance but not difficult to try both ways and see if there is difference to the sound?
Jameshillj - this is for amp prototype testing. What benefits have you got from using headphone for testing? And is there something special with manganese?
XRK971 - So you guys have used hp to debug amps. I'm surprised that I'm the only one who have not
Probably due to previous experience that voltage divider at amp output didn't sound good with hp. But I just found a cheap hp that I think can show the balanced frequency from top to bottom, which triggered this project.As 'X' mentioned, make sure there's no dc voltage on the output
I use the headphones &/or attenuator to check the balance of the sound when driving a 'dummy load' at both normal/low power and then nearly full output power. Some speakers present quite avreactive load to some ampliers and can substantially colour the sound, IME. It's just an active 'ear-check'.
Manganin resistors have a 'very detailed sound without the increased brightness' than the usual foil resistors like the TC 2575, for example, or 'naked Vishays, etc. They are mainly used in power engineering and/or instrumentation
They have a different sound to say, Dale, Caddocks, Sisumi, etc. - some people say they're an audiophool component at high $ cost and don't make any difference - some people say ….
Mn is a metal composite of 85% copper, 12% manganese, and about 2% nickel - they don't sound the same as pure copper resistors (many versions).
I use the headphones &/or attenuator to check the balance of the sound when driving a 'dummy load' at both normal/low power and then nearly full output power. Some speakers present quite avreactive load to some ampliers and can substantially colour the sound, IME. It's just an active 'ear-check'.
Manganin resistors have a 'very detailed sound without the increased brightness' than the usual foil resistors like the TC 2575, for example, or 'naked Vishays, etc. They are mainly used in power engineering and/or instrumentation
They have a different sound to say, Dale, Caddocks, Sisumi, etc. - some people say they're an audiophool component at high $ cost and don't make any difference - some people say ….
Mn is a metal composite of 85% copper, 12% manganese, and about 2% nickel - they don't sound the same as pure copper resistors (many versions).
I have a roll of Mn wire. To get to certain resistance it must be long. Winding it to become a coil is what made me unsure. But if it's 'good' I might want to try.
Strange that MOX is not recommended? I liked it in crossovers so I use it in amp output too. The resistors and passive parts are the last thing I want to compare by ears.
Last night i tried several compensation schemes and got confused of a new option that sounded transparent but unreal. I wish i can listen to them thru bigger speakers, or at least with the headphone, if the circuit were ready. May be i should put higher priority on it, but i have many new prototypes ready on the tables.
Strange that MOX is not recommended? I liked it in crossovers so I use it in amp output too. The resistors and passive parts are the last thing I want to compare by ears.
Last night i tried several compensation schemes and got confused of a new option that sounded transparent but unreal. I wish i can listen to them thru bigger speakers, or at least with the headphone, if the circuit were ready. May be i should put higher priority on it, but i have many new prototypes ready on the tables.
I think speakers is where your effort should go. You don't really need big speakers though. Using locally available components and a competent DIY design to match their parameters to their optimum enclosure and crossover, you could probably build your own Hi-fi loudspeakers for an affordable price. The problem is first to find the competent loudspeaker designer and suitable quality drivers. Look locally - there is always someone with such specialist design skills in a big city.
.
At a young age, you can afford a lot better sounding headphones ($30) than you can afford speakers. BUT I test my amps with $8 salvage TV speakers or car radio speakers in the beginning. Make sure you've got the 8 to 10 ohm kind or measure them at first. Resistance is about 80% of impedance, so 8 ohm speaker measures about 6.5 ohms typically.Jameshillj - this is for amp prototype testing. What benefits have you got from using headphone for testing? And is there something special with manganese?
XRK971 - So you guys have used hp to debug amps. I'm surprised that I'm the only one who have not
I put the speakers in cardboard boxes (from the parts house) so they will have some bass. Naked speakers have no bass.
When the amp has <200 mv of DC offset and you can prove is not ultrasonically oscillating, then you can risk your $30 headphones: with the attenuator circuit Mooly suggested. Putting it parallel an 8 ohm 100 watt (or whatever wattage) resistor makes sure the amp doesn't run away when output impedance is high. Trigger for such HF oscillation runaways is often a radio transmitting nearby, like from a maintenance man's FM radio or a passing CBer in a car. Scope image of quiescent waveform out of a silent amp is great. I don't have a scope so I detect RF oscillation with an analog VOM on AC scale with a series capacitor on the negative probe. Silent amp, significant voltage on the AC scale, probably oscillation. You can prove the frequency by changing the cap from .047 uf (normal music detection) to 390 pf (ultrasonic only if the AC is still there).
After you've listened to the amp for hours & hours on car radio speakers you're sure a solder joint is not going to pop loose and create DC on the output jacks, then you might try $100 speakers. Personally I don't trust DC coupled amps and have built speaker capacitor ones up to now, but there are speaker protection boards on e-bay etc that can disconnect the output if DC occurs. At least they are advertised as doing so.
My high school friend built 4 way speakers with drivers from Alliedradio and they sounded horrible. Piano is hard to reproduce, and that is how I test speakers. I play piano, I know what they can sound like, especially Steinway grands. I had $120 LWEIII 3 way speakers for 4 years then the ST70 amp or something burned up a tweeter. I had them stored while I was in the Army and Dad threw them away. So I had $80 KLH23 2 way speakers for 25 years. Then as my salary went up, $115 used PA speakers, then right before I quit working, $800 a pair used speakers that retail new for $700 each.
Hope your career takes off and you can afford the good speakers before age 56. Or get smart & build some. Just putting drivers in a ported box doesn't make a good speaker IMHO. There has to be a sound deadend room, a calibrated microphone, a measurement program, then lots of tweaks to get the peaks & valleys smoothed out.
Last edited:
I agree that piano is an excellent test of distortion on an amp. One of my first kit amps from eBay sounded ok but terrible on piano. You can tell if the notes are clear or not clear and discordant on a song you know. AC coupled test speakers is great idea. You don’t need expensive speakers. Use for example a $10 TC9FD in a simple 5L box sounds great as test speaker. A $18 Dayton PA130-8 is also excellent as it is sensitive so let’s you hear hum and hiss at lower levels. It can take some power too and sounds excellent as an all around full range.
An easy box to build and great sounding as attested by all the happy builders is the 0.40x scale foam core Karlsonator for TC9FD and 0.53x scale Karlsonator for PA130-8. They will keep you listening for a long time. Plans in post #1
Mini Karlsonator (0.53X) with Dual TC9FD's
An easy box to build and great sounding as attested by all the happy builders is the 0.40x scale foam core Karlsonator for TC9FD and 0.53x scale Karlsonator for PA130-8. They will keep you listening for a long time. Plans in post #1
Mini Karlsonator (0.53X) with Dual TC9FD's
Yes, this is what I need, thanks. I have just looked up Ayrton-Perry but still haven't yet understood how to do it. Later I probably will search for youtube tutorial
I already have speakers but for prototyping with some risks I don't want to use them. This amp prototyping is about to find the best sounding amplifier (in the world). That's why not many speakers qualify for the monitoring.
Beside 'quality' which is often subjective, for best system it is necessary to have low frequency capability most system don't have.
Hi Indianajo, I did build my own speakers since childhood. My uncles built, repaired and sold audio stuffs. That's how I was exposed to DIY.
I invested a lot of time into my speakers. I might have forgotten how to do it again. Beside the usual measurements, too many recipes and secrets
Like you and Indianajo, I used to use piano for testing, but now I didn't do it again. I try not to be trapped in a situation where the system only suits certain musical instrument or music genre. Now I just listen with various songs and I will know if it is any less than my current best.
Last night I tried changing output transistors in one of the amps. Comparing a matched pair with hfe=33 with a matched pair with hfe=105. Like my prediction, my ears preferred the lower hfe.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.