Thank you all, for the good imputs. I am suppriced to see that you belive there is not much difference in output impedance between PP and SE amps. I will probably simulate with a output impedance value between 1 and 2 ohm. Then cablesize and series inducteres can be used to trim the right bas performance, I think.
I would broaden the range just a bit, and say 1 - 4 ohms to be safe because I have several 300B PP 0 fdbk pushpull and SE amps with output impedances of > 3 ohms or so, and this would be typical..
Don't forget you also have to include the DCR of the transformer primary and secondary in the source impedance calculation, and a DCR of 1 ohm on the 8 ohm tap is not uncommon.
Note that the following example should be considered a gross simplification but is anecdotal of the sorts of impedances you can expect with an SE amp.
My 300B SE amplifier has a 3K:8 OPT, the primary DCR is about 100 ohms, and the secondary DCR close enough to 1 ohm for the purposes of this example..
The output tube rp is about 700 ohms.
Primary DCR is 100 ohms
The impedance transformation ratio is 375:1 primary to secondary.
Secondary DCR is 1 ohm
So figuring the source impedance on the primary is approximately 800 ohms then (800/375) + 1 = 3.13 ohms at the secondary for the 8 ohm tap.
Leakage inductance may increase the effective source impedance at higher frequencies, but I ignore it since I am primarily interested in the effect on the woofer.
The amplifier source impedance will be roughly 3 ohms in this example, and is in fact close enough for most purposes.
It is not at all unusual for a 0 fdbk PP amplifier optimized for output power as opposed to source impedance to have a significantly higher source impedance than an SE amplifier using the same tube type.
Note also that class A and class A/B PP amplifiers will have different source impedances even with the same OPT and in the case of the A/B amplifier that source impedance changes significantly as a function of power output beyond the class A region of operation.
Don't forget you also have to include the DCR of the transformer primary and secondary in the source impedance calculation, and a DCR of 1 ohm on the 8 ohm tap is not uncommon.
Note that the following example should be considered a gross simplification but is anecdotal of the sorts of impedances you can expect with an SE amp.
My 300B SE amplifier has a 3K:8 OPT, the primary DCR is about 100 ohms, and the secondary DCR close enough to 1 ohm for the purposes of this example..
The output tube rp is about 700 ohms.
Primary DCR is 100 ohms
The impedance transformation ratio is 375:1 primary to secondary.
Secondary DCR is 1 ohm
So figuring the source impedance on the primary is approximately 800 ohms then (800/375) + 1 = 3.13 ohms at the secondary for the 8 ohm tap.
Leakage inductance may increase the effective source impedance at higher frequencies, but I ignore it since I am primarily interested in the effect on the woofer.
The amplifier source impedance will be roughly 3 ohms in this example, and is in fact close enough for most purposes.
It is not at all unusual for a 0 fdbk PP amplifier optimized for output power as opposed to source impedance to have a significantly higher source impedance than an SE amplifier using the same tube type.
Note also that class A and class A/B PP amplifiers will have different source impedances even with the same OPT and in the case of the A/B amplifier that source impedance changes significantly as a function of power output beyond the class A region of operation.
Last edited:
I totally agree. Aside from your example, a lot of the common budget PP Pentode amps also have a Zout around 4ohms.
That was the range I had in mind when suggesting to design with a 2 ohm source impedance in mind (for an unknown tube amp). It's 2 x 1ohm, but 1/2 x 4ohms. It would compromise the extremes equally and be closer to ideal for amps in the middle range.
The DC resistance of the speaker cables is a factor too, but I don't even want to consider another "cables" debate.
There are many texts and articles on controlling speaker cone motion by lowering the output impedance of the amplifier and associated wiring. It is important to remember that the speaker generates a voltage when it moves. The "cone control" is provided by forcing that voltage generator to operate into a short circuit, or as low an impedance that is practical. What is often overlooked is that the speaker voice coil has a DC resistance too, and it is often several ohms.
Many speaker design programs and texts do not consider this since it is beyond our control.
Hello,
We often hear about speaker impedance, amplifier output impedance and dampening factor.
When we buy speakers and construct boxes we carefully calculate and target a loudspeaker with a Quality Factor “Q” of 0.71.
George by the way that I read your comment s the amplifier (plus cables etc) dampens the motion of the speaker when it acts as a generator. Shouldn’t we be speaking of total Q and tuning the amplifier and loudspeaker to be critically damped as a system rather than speaking of dampening factor (the ratio of speaker impedance and amplifier output impedance)?
DT
All just for fun!
We often hear about speaker impedance, amplifier output impedance and dampening factor.
When we buy speakers and construct boxes we carefully calculate and target a loudspeaker with a Quality Factor “Q” of 0.71.
George by the way that I read your comment s the amplifier (plus cables etc) dampens the motion of the speaker when it acts as a generator. Shouldn’t we be speaking of total Q and tuning the amplifier and loudspeaker to be critically damped as a system rather than speaking of dampening factor (the ratio of speaker impedance and amplifier output impedance)?
DT
All just for fun!
So, I suppose this is the reason the 16 ohm speakers were more common in the tube era..........?
Another related question: If zobels can be used to flatten out the impedance of loudspeakers, why are they not standard equipment in multi-way speakers? Is it because the values for the zobel compensation are highly dependant on the output impedance of the amp and with a low Z solid state amp they are not needed?
Another related question: If zobels can be used to flatten out the impedance of loudspeakers, why are they not standard equipment in multi-way speakers? Is it because the values for the zobel compensation are highly dependant on the output impedance of the amp and with a low Z solid state amp they are not needed?
Zobel values aren't dependent on source impedance. They are purely determined by the driver's parameters. It's fairly common to see them on the woofer in a multi-way system because you don't want the woofers impedance rise to interfere with the low pass filter. But, you will rarely (if ever) see them on a tweeter. This is because a tweeter has no low pass filter and the designer usually assumes voltage drive (amp with a very low Zout).
A speaker with a flat impedance would certainly be more tube friendly. It would take a little more correctional circuitry than just Zobels though.
Actually some of us do design our speakers and amplifiers as a system. In my case the amplifier existed prior to the current speaker system so I designed my speakers knowing that the source impedance of the amplifier was around 3 ohms. This allowed me to tune the enclosure and cross-over design for a non ideal voltage source and achieve reasonable tuning on the bottom end. I also took into consideration the source impedance of the amplifier for the midrange and treble X-O design.
The next amplifier will be designed to provide the same approximate source impedance, and several of my other amplifiers are generally close enough. (all of them are my designs)
Kevinkr,
It is good to hear of you designing speakers, XO’s and amplifiers to work together. I have been working back the other way adjusting the amplifier Zo to work into the 300 ohm impedance of Sennheiser HD 600 headphones. I have been from 120 ohms Zo down to 6 ohms. The Zo really does impact the sound of the lower frequencies. I am enjoying ~ 20 ohms Zo. I do like tight glutes not flabby bottoms. When time permits I will install and do some testing in Audio Tester; FFA distortion testing and frequency response.
DT
All just for fun!
It is good to hear of you designing speakers, XO’s and amplifiers to work together. I have been working back the other way adjusting the amplifier Zo to work into the 300 ohm impedance of Sennheiser HD 600 headphones. I have been from 120 ohms Zo down to 6 ohms. The Zo really does impact the sound of the lower frequencies. I am enjoying ~ 20 ohms Zo. I do like tight glutes not flabby bottoms. When time permits I will install and do some testing in Audio Tester; FFA distortion testing and frequency response.
DT
All just for fun!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.