I should clarify what I just said about N/P pairs. Since all the drains are tied
together, the real criteria for a 0V offset is to have the total current of all
the P devices equal to the total current of all the N devices.
together, the real criteria for a 0V offset is to have the total current of all
the P devices equal to the total current of all the N devices.
Andrew - the question was about voltages. I answered that question. Never mentioned the word 'current'.
You quoted only part of my post, I suggest you read the whole post and in context of the question that was asked I did also say that:
Even in single pair F5 the kits supplied by Jack used 5% source resistors and getting a 50mV difference between was not considered to be a problem as long as the offset was acceptable.
You quoted only part of my post, I suggest you read the whole post and in context of the question that was asked I did also say that:
Even in single pair F5 the kits supplied by Jack used 5% source resistors and getting a 50mV difference between was not considered to be a problem as long as the offset was acceptable.
Sangram, I did read your whole post.
I have re-read it in the light of your comment.
Your statement is still wrong.
I have re-read it in the light of your comment.
Your statement is still wrong.
The upper and lower currents are the same. The currents through the source resistors are the same. The voltage drop across the source resistor appears different if the source resistor values are not equal. That difference in voltage drop is due to unequal source resistor values. It is NOT due to mosFET tolerances.
Because it's stable without it?
A good question, I hope somebody more knowledgeable can chime in with good info.
A good question, I hope somebody more knowledgeable can chime in with good info.
Yes, but is a Zoble not supposed to be an impedance matching device?
Wikki says:
When used to cancel out the reactive portion of loudspeaker impedance, the design is sometimes called a Boucherot cell. In this case, only half the network is implemented as fixed components, the other half being the real and imaginary components of the loudspeaker impedance. This network is more akin to the power factor correction circuits used in electrical power distribution, hence the association with Boucherot's name.
------------
So I suppose that the Zobel would depend on the speaker?
Wikki says:
When used to cancel out the reactive portion of loudspeaker impedance, the design is sometimes called a Boucherot cell. In this case, only half the network is implemented as fixed components, the other half being the real and imaginary components of the loudspeaker impedance. This network is more akin to the power factor correction circuits used in electrical power distribution, hence the association with Boucherot's name.
------------
So I suppose that the Zobel would depend on the speaker?
If you know your load, you can due without, but, and a big but, you run across a load that will start an oscillation, then you can kiss your amp or speakers good bye.
Better to play it safe and not worry about it , include one.
Regards
David
Better to play it safe and not worry about it , include one.
Regards
David
test your amp and if it's stable , go without zobel
but , if your amp is acting as hooker , changing speaker(s) every hour .........
but , if your amp is acting as hooker , changing speaker(s) every hour .........
Some amplifiers change gain with the load variation.
Some amplifiers change gain at HF and/or VHF when the load has a very high resistance.
In some situations this change of gain can result in stability margins changing to interfere with the output such that the amp performance is severely degraded.
One method of controlling what load the amplifier sees is to add a Zobel to the output so that the "change" in load at HF/VHF is reduced to what does not change the stability margins.
For the Zobel to work it MUST PASS HF signals and above.
It cannot do this if there is significant impedance @ these very high frequencies.
i.e. the Zobel ROUTE must tap into the output and feed back into the Return such that the total ROUTE inductance is kept VERY LOW.
The Zobel cannot work at very high frequencies if the ROUTE is many inches/centimeters long.
Some amplifiers change gain at HF and/or VHF when the load has a very high resistance.
In some situations this change of gain can result in stability margins changing to interfere with the output such that the amp performance is severely degraded.
One method of controlling what load the amplifier sees is to add a Zobel to the output so that the "change" in load at HF/VHF is reduced to what does not change the stability margins.
For the Zobel to work it MUST PASS HF signals and above.
It cannot do this if there is significant impedance @ these very high frequencies.
i.e. the Zobel ROUTE must tap into the output and feed back into the Return such that the total ROUTE inductance is kept VERY LOW.
The Zobel cannot work at very high frequencies if the ROUTE is many inches/centimeters long.
So what you are saying is that the zobel needs to be placed directly at the output of the amp, and not at the speaker terminals.
Yes, an effective Zobel works with high frequency signals.
It must be located in a route that is capable of passing high frequency signals
But, there is another Zobel that comes to our aid.
The speaker cables are interference pick up aerials.
They will inject interference into your amp box.
A Zobel at the chassis speaker terminals can be fitted to attenuate this cable interference.
This is the Pi filter that you see being referred to.
R+C at the amp,
L||R in the cable between amp and terminals.
R+C at the terminals.
It must be located in a route that is capable of passing high frequency signals
But, there is another Zobel that comes to our aid.
The speaker cables are interference pick up aerials.
They will inject interference into your amp box.
A Zobel at the chassis speaker terminals can be fitted to attenuate this cable interference.
This is the Pi filter that you see being referred to.
R+C at the amp,
L||R in the cable between amp and terminals.
R+C at the terminals.
I'm planning to build a First watt f5 V2 amp, with 4 output devices/channel.
Being a cheapskate, i'm wondering if I can use a pair of 600W/2x49V toroid transformers which I have modified by removing half of the secondary turns, so it's now a 2x25 volt transformer?
Is it still a 600W transformer?
Or do I have to replace all secondary windings with thicker wire? (which is very easy to do)
Being a cheapskate, i'm wondering if I can use a pair of 600W/2x49V toroid transformers which I have modified by removing half of the secondary turns, so it's now a 2x25 volt transformer?
Is it still a 600W transformer?
Or do I have to replace all secondary windings with thicker wire? (which is very easy to do)
wind all that copper back on until you have EXACTLY the same number of Turns on all the windings.
Then you can parallel the windings to get back the missing VA.
Then you can parallel the windings to get back the missing VA.
Parralelled windings can cause problems. If I were you I would use one set of windings for each channel.
If done correctly paralleled windings do not cause problems.
It's the standard way of all dual winding transformers, both for Dual Primaries and for Dual Secondaries.
It's the standard way of all dual winding transformers, both for Dual Primaries and for Dual Secondaries.
All my life🙄, I've lived in the misconception that it's the size of the core that predicts the power transfer ability....