I bult a pair of ESLs using 1/8 spacers and a 72x20in area. I can deal with the beaming. However, I'm having trouble understanding what value to use for the load resistor. I imagine the capacitance is fairly large because with 1 ohm on the transformer primary, the treble starts to roll off. If I remove the resistor, the treble has a peak and the amp goes into protection mode on high freqs.
Here's my limited understanding (spare my EE noobness 😀). Please correct me if I'm wrong
Even with an ideal transformer, I only see two variables (load resistor and panel capacitance). The panel capacitance depends on the size and spacing (effectively fixed). The only thing I can change is the resistor value to shape this low pass filter for a flat freq response.
Does that mean my amp is stuck with whatever value the panel resistor combo produces a flat response?
I must be forgetting something. The huge full range panels can't possibly have no load resistor and just expect the amp to drive a near-zero impedance load.
Am I forgetting about transformer inductance, or just got the whole thing wrong?
thanks
Here's my limited understanding (spare my EE noobness 😀). Please correct me if I'm wrong
Even with an ideal transformer, I only see two variables (load resistor and panel capacitance). The panel capacitance depends on the size and spacing (effectively fixed). The only thing I can change is the resistor value to shape this low pass filter for a flat freq response.
Does that mean my amp is stuck with whatever value the panel resistor combo produces a flat response?
I must be forgetting something. The huge full range panels can't possibly have no load resistor and just expect the amp to drive a near-zero impedance load.
Am I forgetting about transformer inductance, or just got the whole thing wrong?
thanks
It depends on the stepup trans. you're using. If I remember correctly, the stray inductance is directly/partly responsible for the resonance, and the series resistor is there to dampen it. You can play with values, don't think you should assume just 1 ohm will do it. I use direct drive myself and don't deal with such issues. You also might Google and look for a series of papers that Menno Van der Veen (??) wrote on such topics. He came up with a toroid stepup that is supposed to be about the best approach and performer,..will leave that for you to decide. Those papers explain quite a bit and are educational.
I believe that the panel capacitance and leakage inductance of the transformer are resonating. The resistor "de-Q's" the resonance (spreads it out over a broader frequency band) and makes the thing easier for your amplifier to drive.
If you hear HF roll off when you use a 1 Ohm resistor in series with the primary, you may want to get transformers with lower leakage inductance. You could also try putting resistors on the secondary (the ESL side) of the transformer instead of on the proimary side. They will have to be much higher resistances ( start at 1k Ohm or so) and put one between each stator and it's transformer connection.
I_F
If you hear HF roll off when you use a 1 Ohm resistor in series with the primary, you may want to get transformers with lower leakage inductance. You could also try putting resistors on the secondary (the ESL side) of the transformer instead of on the proimary side. They will have to be much higher resistances ( start at 1k Ohm or so) and put one between each stator and it's transformer connection.
I_F
Thanks for the replies. I'll try putting some 1K resistors on the secondary. Is there an equation for calculating which values to use (as a guideline, i understand nothing works exactly like theory)?
I read some people putting a resistor in parallel with the primary in order to present a resistive load to the amp. Wouldn't the signal at high freq still short across the capacitative load (and ignore the resistor)?
I read some people putting a resistor in parallel with the primary in order to present a resistive load to the amp. Wouldn't the signal at high freq still short across the capacitative load (and ignore the resistor)?
hozo said:
Probably. You can spend hours trying to derive it, or you can spend a few minutes with some resistors and clip leads and find out exactly what you need. In this case, I vote for the clip leads.
I have not seen that before. It doesn't seem like a good idea because it would make the dips in the impedance that much lower. When it comes to amplifier stability it is a little hard to predict what will happen. DIY ESLs are all different and amplifiers have different phase margins. If you have an amp that is well protected against weird loads, it might be easiest to try it and see if it works for you.
I_F
In case of significant DC-offset from amplifier, you should consider a primairy resistor instead of a secondairy. This should be a high quality metal film resistor (not wire-wound).
You could achieve the same results as with a secundairy resistor (impedance and freq) while blocking the offset to some degree.
You could achieve the same results as with a secundairy resistor (impedance and freq) while blocking the offset to some degree.
Resistor
Hi,
a series resistor has several effects.
- it reduces the stress on the amplifier, by raising the impedance and reducing the phase angle at the upper freq-range.
- it dampens the resonating circuit that´s made up from the speakers capacity together with the trannies stray-inductance. Thereby it can linearize the frequency and phase-response
- it forms a voltage-divider circuit with a -3dBpoint where the resistors impedance and the (reflected to the primary side) speakers impedance equal. Because of the falling nature of the speakers impedance towards higher freqs its a filter with lowpass-character. This is often the most obvious effect of all
So there is a optimum value for this resistor with a ´standard-value´ around 1Ohm. You can get the same results by using series resistors on the secondary side by multiplying the primary resistors value by the square of the turns ratio. I´d prefer the use of a primary resistance.
The use of a parallel resistor has the only reason to linearize the impedance generally in a easy, low part-count style. The typical gaussian-like curve with impedance values of sometimes 100Ohms or more in the mid of the audio freqs can be lowered to more useable values with an high-wattage parallel resistor between 8 and 22Ohm. But it reduces the even low speaker impedance a bit more. Sometimes You find both, a series low ohmic value resistor and a higher ohmic parallel resistor.
jauuCalvin
Hi,
a series resistor has several effects.
- it reduces the stress on the amplifier, by raising the impedance and reducing the phase angle at the upper freq-range.
- it dampens the resonating circuit that´s made up from the speakers capacity together with the trannies stray-inductance. Thereby it can linearize the frequency and phase-response
- it forms a voltage-divider circuit with a -3dBpoint where the resistors impedance and the (reflected to the primary side) speakers impedance equal. Because of the falling nature of the speakers impedance towards higher freqs its a filter with lowpass-character. This is often the most obvious effect of all
So there is a optimum value for this resistor with a ´standard-value´ around 1Ohm. You can get the same results by using series resistors on the secondary side by multiplying the primary resistors value by the square of the turns ratio. I´d prefer the use of a primary resistance.
The use of a parallel resistor has the only reason to linearize the impedance generally in a easy, low part-count style. The typical gaussian-like curve with impedance values of sometimes 100Ohms or more in the mid of the audio freqs can be lowered to more useable values with an high-wattage parallel resistor between 8 and 22Ohm. But it reduces the even low speaker impedance a bit more. Sometimes You find both, a series low ohmic value resistor and a higher ohmic parallel resistor.
jauuCalvin
MJ Dijkstra said:
If your amp has more than a few mV of DC at the output you need to repair your amp, not add a resistor to the output.
I_F
Hi ,
Why wire wound resistors are not recommended ? I understand that their inductance is still very low , compared to transformer's leakage inductance.Are there are other bad side effects ?
Why wire wound resistors are not recommended ? I understand that their inductance is still very low , compared to transformer's leakage inductance.Are there are other bad side effects ?
While many modern amps have servo control to prevent dc offset, several designs have not.
It's worth checking.
It's worth checking.
- Status
- Not open for further replies.