I come to California somewhere in the next two weeks. I send you a mesage if i have the schedule.
We have sometimes DC on the powerlines in Germany.
American High End components had trouble in the past because their 60Hz transformers started to hum. I know from Krell and others that they make special transformers for Germany now that do not have that problem.
I have seen DIY circuits that can remove that DC from the line.
The frequency of the German power line is very stable but the sinewave can have harmonic distortion. Power Generators from PS Audio and other are all the rage now here.
Flipping polarity of the plug (that is very easy here because the jack is symmetric) can have a big effect on sound. A friend of mine measured harmonic distortion in poweramps and one connecion gave less distortion. One reason is that the stray capacitance of the main transformer can couple noise into the ground and the powerline.
We have sometimes DC on the powerlines in Germany.
American High End components had trouble in the past because their 60Hz transformers started to hum. I know from Krell and others that they make special transformers for Germany now that do not have that problem.
I have seen DIY circuits that can remove that DC from the line.
The frequency of the German power line is very stable but the sinewave can have harmonic distortion. Power Generators from PS Audio and other are all the rage now here.
Flipping polarity of the plug (that is very easy here because the jack is symmetric) can have a big effect on sound. A friend of mine measured harmonic distortion in poweramps and one connecion gave less distortion. One reason is that the stray capacitance of the main transformer can couple noise into the ground and the powerline.
Here is a circuit from LC Audio and here is the expanation
This filter is inserted between the mains grid and the primary winding of the transformer. This way you can avoid mechanical humming from the mains transformer, caused by a varying DC component on the mains grid's AC voltage. The DC component comes from unbalanced loads on the mains grid, such as hair dryers, industrial machinery etc. with only half wave rectification. One half of the AC sine wave is loaded while the other is not. This cause imbalance, and a DC.
With this filter module, which also doubles as a safe mains distributor for your DIY power amplifier, you can remove the DC component and keep your power transformer quiet. It can be used for up to 500 VA idle load, and 1000 VA max load.
This filter is inserted between the mains grid and the primary winding of the transformer. This way you can avoid mechanical humming from the mains transformer, caused by a varying DC component on the mains grid's AC voltage. The DC component comes from unbalanced loads on the mains grid, such as hair dryers, industrial machinery etc. with only half wave rectification. One half of the AC sine wave is loaded while the other is not. This cause imbalance, and a DC.
With this filter module, which also doubles as a safe mains distributor for your DIY power amplifier, you can remove the DC component and keep your power transformer quiet. It can be used for up to 500 VA idle load, and 1000 VA max load.
Attachments
Hello Joachim
Interesting but most good quality products which includes Japanese amps dont suffer from noisy transformers. Well made power transformers don't need this circuit described by LC Audio to make them quiet.
AR
I know, but some people here could have that problem with older amps.
There was a time where makers of ring core transformers tried to squeeze as much VA out of them as they could so used very high permeable materials that where easy to saturate.
I heard more then one amp from abroad humming loud.
Nowerdays that is a non issue, i agree because what they want is sell and you can not sell a humming transformer to a typical German High End Customer.
There was a time where makers of ring core transformers tried to squeeze as much VA out of them as they could so used very high permeable materials that where easy to saturate.
I heard more then one amp from abroad humming loud.
Nowerdays that is a non issue, i agree because what they want is sell and you can not sell a humming transformer to a typical German High End Customer.
That is correct for uncorrelated signals, what happens when there are multiple sources of distortion all created by the same mechanism?
Hum or buzz is a problem with DC, as is designing transformers for 60 hertz and then using them at 50 hertz. A number of folks have made that mistake.
As to a reversal switch it is interesting to see when it has an effect and then why.
"If I can hear it, I should be able to measure it!"
The second question is if I change the phase and leave it where it sounds better, does the better position change after a while. I. E. does the DC offset stay constant and magnetize my RF filters and transformer core? Are there other pieces that are also affected?
I had that experience. You change the direction of a component and it sounds better or say "different". After i while you change back to the old direction and again it sounds ... ......
When i make REAL High End stuff for customers that are prepaired to wait i run in all components on much higher voltage then they actually get in the circuit.
For expample i do a string of capacitors and connect them to the mains for one week, but now we are getting into muddy waters. Can not wait to get the hatemail that this is totally stupid.
When i make REAL High End stuff for customers that are prepaired to wait i run in all components on much higher voltage then they actually get in the circuit.
For expample i do a string of capacitors and connect them to the mains for one week, but now we are getting into muddy waters. Can not wait to get the hatemail that this is totally stupid.
Hate mail from idiots, oh is there any other kind?
There actually is a good explanation for that happening. It also is measurable. It is also correctable!
If you know what the DC magnetization current is on the primary what happens if you unbalance the secondary or induce a current in a tertiary?
How would you measure the results and make the unbalanced load self adjusting?
Last edited:
Is it the same Richard Heyser who built an amplifier that went to the moon?
Questions over questions. I usually do high resolution distortion measurements like you do but not the same method. My resolution floor is -150dB. Did not try long term avaraging so far and i simply BUILD ( with my hands) a lot of stuff and later listen to it.
I also know a lot of peolle i trust that give me advice.
My motto is not perfection but fun anyway. Can´t hide my German Engineer heritage though.
I also know a lot of peolle i trust that give me advice.
My motto is not perfection but fun anyway. Can´t hide my German Engineer heritage though.
JPL is probably Jet Propulsion Laboratories: Jet Propulsion Laboratory - Wikipedia, the free encyclopedia which preceded NASA and than was incorporated into it.
Designing a power transformer isn't simple and a lot of aspects are involved. You need more metal for lower frequencied to keep the sme flux level. At the same time most commercial transformer vendors try to run their transformers pretty close to saturation. They get higher efficiency for both power and dollars, but the distortion in the core goes up, noise from loose lams and magnetostriction goes up, heat goes up and usually the sound degrades. Further most audio designers end up with very small conduction angles, creating lots of harmonic energy which in turn makes more noise both magnetically and radiated. The small conduction angles also require much larger transformers because the VA requirements depart a lot from the watts needed.
Add a little DC (.5V) and you can push the transformer over the edge.
The circuit posted above I find a little (or more) scary. You need a cap with the necessary peak voltage rating (the transorb will vaporize if there is any real voltage across it, they don't handle a lot of power) and a high continuous ripple current rating. It should also be fused since a short in the cap has no current limiting and a big cap like that can explode with amazing force. Use with caution . .
Add a little DC (.5V) and you can push the transformer over the edge.
The circuit posted above I find a little (or more) scary. You need a cap with the necessary peak voltage rating (the transorb will vaporize if there is any real voltage across it, they don't handle a lot of power) and a high continuous ripple current rating. It should also be fused since a short in the cap has no current limiting and a big cap like that can explode with amazing force. Use with caution . .
Actually its only relevant for correlated signals. Uncorrelated signals won't have relations in their distortion artifacts.
In the old days when distortion analysers weren't too good and amps weren't either this cancellation/peaking was common. The classic rule was that the analyzer had to be 10X better than any signal it was working with. however those only had a meter and looked at everything all at once. With a spectrum analyzer its less of an issue.
Complex signals and their distortions are more the province of IM and multitone testing. Look at the spectral contamination tests for more meaningful and useful stuff. (for those with AES access: AES E-Library: Spectral Contamination Measurement)
I use this a lot for transducers. Less for electronics but its still important.
Beware of the posted DC-filter! It's outright dangerous.
Incredible that this is posted here and goes uncommented.
An electrolytic cap may see reversed polarity, but no more than about 1.5V - more than that, electrochemical processes start, heat the cap and boom. For references see datasheets from cap manufacturers.
Companies like Bryston use in parallel to the cap a bridge with each two diodes in series (and antiparallel to eachother).
Have fun, Hannes
Incredible that this is posted here and goes uncommented.
An electrolytic cap may see reversed polarity, but no more than about 1.5V - more than that, electrochemical processes start, heat the cap and boom. For references see datasheets from cap manufacturers.
Companies like Bryston use in parallel to the cap a bridge with each two diodes in series (and antiparallel to eachother).
Have fun, Hannes
This is an interesting circuit. I would expect it to actually GENERATE a DC component in the signal across the xformer. The zenerdiode assymetrically limits the voltage across the cap, which has the effect to 'lift' the ac component, that develops from the cap impedance, way from the zero volts base line. The net effect is a DC component across the cap that is subtracted from the ideal sine wave, so generates DC.
Did anybody sim this?
jd
Last edited:
Demian,
On this we disagree. This may be a case of practice versus theory.
If the distortion is uncorrelated then the total distortion is calculated by the RMS method or the square root of the sum of the averages (mean) squared. On this we seem to disagree.
In some cases the distortion adds linearly when it is phase coherent. If I take two resistors each with 1 volt of signal across them creating 1nV of distortion the total distortion is 2nV across both and of course the signal is now 2 volts so the % of distortion remains the same. It does not go down by the square root of sources as random distortion or noise would. That is because the distortion is generated by the exact same mechanism and signal.
That is why when I parallel instrumentation amplifiers for lower noise, the noise does go down but the distortion does not.
- Status
- Not open for further replies.