If anything it would do the same as it did for the woofers which is to keep the impedance more constant.
Here's the graph with the zobel connected. 1.1uF 12 ohms
To make it easier to compare here's the graph without the zobel again.
Looking at the graphs, if I did things right it looks like maybe only a 1-2 ohm difference between using the zobel and not using the zobel. Only way to know if that difference is audible is to try it tomorrow at work.
Here's the graph with the zobel connected. 1.1uF 12 ohms
To make it easier to compare here's the graph without the zobel again.
Looking at the graphs, if I did things right it looks like maybe only a 1-2 ohm difference between using the zobel and not using the zobel. Only way to know if that difference is audible is to try it tomorrow at work.
Last edited:
Could you remind me whether you're using any other crossover components at the speaker or it's all at line level (I'm on too many similar threads).
Changing the impedance shouldn't have an effect on the amp other than to make it draw more current.
Changing the impedance shouldn't have an effect on the amp other than to make it draw more current.
I'm not familiar with it, however I've seen mention of audio transformers, L-pads and 12dB/oct slopes in this thread.
RC will not linearise the impedance with regard to the amplifier if it is behind certain other components. It may change the impedance as a load for the filter, thereby changing the response.
RC will not linearise the impedance with regard to the amplifier if it is behind certain other components. It may change the impedance as a load for the filter, thereby changing the response.
Some examples using your impedance plot. First, a generic 1kHz 2nd order high pass for 8 ohms.
Now optimising the values to get closer to the target.
Now doing the same with the RC in place.
Now optimising the values to get closer to the target.
Now doing the same with the RC in place.
Did a test with the midax zobel and couldn't hear a difference.
Proceeded to feed a 5KHz sinewave to the amp and while keeping the volume low I saw a drop of just 1mV so I have determined that the zobel isn't needed.
Proceeded to feed a 5KHz sinewave to the amp and while keeping the volume low I saw a drop of just 1mV so I have determined that the zobel isn't needed.
So I tried this version of the l-pads. It sounded better but I felt the tweeter was too loud.
I then decided to reduce the tweeter level, however before I could make the changes I listened more and decided the midax was a little low.
So I'm going to try this as it requires only adding one resistor.
Now all this is being done by ear.
I then decided to reduce the tweeter level, however before I could make the changes I listened more and decided the midax was a little low.
So I'm going to try this as it requires only adding one resistor.
Now all this is being done by ear.
Instead of messing with fixed l-pads I'm tempted to find a couple 16 ohm variable l-pads and use those.
That's certainly what I would do.
16 ohm L-pads are elusive, but I can buy them here: https://willys-hifi.com/collections/l-pad-attenuators
16 ohm L-pads are elusive, but I can buy them here: https://willys-hifi.com/collections/l-pad-attenuators
Parts express has a 100 watt one, however I don't need one that big.
That said since I use resistors in series with the tweeter to equal 15 ohms, I could likely use an 8 ohm l-pad after the series resistors for that one.
That said since I use resistors in series with the tweeter to equal 15 ohms, I could likely use an 8 ohm l-pad after the series resistors for that one.
So I think that I've found an issue. I connected my scope probe to the - speaker wire at the amp and I got what looks like an oscillation if I turned the volume up beyond a certain point with music playing and if I put the probe on the - speaker wire at the speaker I got an even higher oscillation signal.
So I'm gonna disconnect the amp and connect my scope to the + and - speaker terminals at the amp and see what I get with the speaker connected and disconnected.
The cause could be the long speaker wires and I suppose that oscillation could affect the upper treble and may be why it doesn't sound exactly right. If it is then I'll figure a way to move the amp to the speaker so that the speaker leads are short.
So I'm gonna disconnect the amp and connect my scope to the + and - speaker terminals at the amp and see what I get with the speaker connected and disconnected.
The cause could be the long speaker wires and I suppose that oscillation could affect the upper treble and may be why it doesn't sound exactly right. If it is then I'll figure a way to move the amp to the speaker so that the speaker leads are short.
The long speaker wire was the problem. I extended the signal and power wires to the speaker and made the speaker wire short and there is no oscillations whatsoever.
I did find that if the speaker wire touches ground at a B+ of 44Vdc, the overcurrent protection isn't enough to keep the chip from being damaged.
Luckily I had an identical board modded the exact same way.
I did find that if the speaker wire touches ground at a B+ of 44Vdc, the overcurrent protection isn't enough to keep the chip from being damaged.
Luckily I had an identical board modded the exact same way.
I'm glad that extending the signal and power leads to the amplifier solved your problem.
For the signal wire I used three BNC cables connected together and installed a BNC jack on the chassis.
Now I just need to figure out mounting the amp to the speaker. Will likely use a wood piece with the chassis mounted to it then use two l brackets or 45 degree brackets to mount the wood piece to the top of the speaker baffle.
The chassis has a slightly slanted front and the top goes over the front and extends over the rear with the bottom also being the sides of the chassis. Everything is mounted to the chassis top and back so that all I have to do is remove 6 screws to free the top of the chassis from the bottom of the chassis.
I also plan on mounting the level control from the rear on the top of the chassis opposite the volume control.
I do have a couple banana jacks and I may have some banana plugs. So I may replace the 1/4" jack with the banana jacks which will eliminate the issue of the speaker + or - contacting the chassis.
Another thing I did was to set the current limit on the power supply at 1A as that amount of current is only seen very briefly upon turning the supply on and I don't even get over 250mA of B+ current when the amp is operating at a decent output level.
Now I just need to figure out mounting the amp to the speaker. Will likely use a wood piece with the chassis mounted to it then use two l brackets or 45 degree brackets to mount the wood piece to the top of the speaker baffle.
The chassis has a slightly slanted front and the top goes over the front and extends over the rear with the bottom also being the sides of the chassis. Everything is mounted to the chassis top and back so that all I have to do is remove 6 screws to free the top of the chassis from the bottom of the chassis.
I also plan on mounting the level control from the rear on the top of the chassis opposite the volume control.
I do have a couple banana jacks and I may have some banana plugs. So I may replace the 1/4" jack with the banana jacks which will eliminate the issue of the speaker + or - contacting the chassis.
Another thing I did was to set the current limit on the power supply at 1A as that amount of current is only seen very briefly upon turning the supply on and I don't even get over 250mA of B+ current when the amp is operating at a decent output level.
Last edited:
Concerning the long speaker wire I suspect either it's the .47uF and 4.7 ohm resistors being the specified values for the TDA2030 chip that's the problem or it's the 16 ohm impedance or maybe the chip just doesn't like the capacitance of the long speaker wire (likely might have been ok with an 8 ohm load) or maybe having the speaker wire run right next to the copper screen of the screen room was the problem due to capacitance between the speaker wire and the copper screen.
Here's some photos of the board and modified amp chassis.
Cut a hole in the board to match the one I cut in the chassis. Plus I drilled four holes in the chassis bottom and wood so that the amp can be secured in place. Next time I order something from Amazon I may order a metal fan grill to put on the bottom of the board.
Bottom of chassis with the hole cut out.
Top of the chassis with the level control mounted in its new location.
Inside the chassis. I tried to find a 50k pot, but all I could easily find were 10k pots, however I did find a 100k dual gang pot which will be wired in parallel for 50k. I'll also rewire the pots so that the 10k volume control is first, whereas now it is where the volume control is last. I haven't found the banana jacks I have.
Hole was put where it is so that air would flow over the other heatsink. Plus it gives the fan a proper air intake instead of just relying on the small holes in the chassis.
One thing I need to do is figure an easy way to cover the terminals on all the drivers and crossover in such a way that the covering can be easily removed for if I need to service the crossover or a driver. That will do two things. Make it better using it at work as nothing can really be said then about exposed terminals. Make it harder for the terminals to short against the copper screen and damage the amplifier. It's a lot different working inside a room where the walls are at ground potential.
Here's some photos of the board and modified amp chassis.
Cut a hole in the board to match the one I cut in the chassis. Plus I drilled four holes in the chassis bottom and wood so that the amp can be secured in place. Next time I order something from Amazon I may order a metal fan grill to put on the bottom of the board.
Bottom of chassis with the hole cut out.
Top of the chassis with the level control mounted in its new location.
Inside the chassis. I tried to find a 50k pot, but all I could easily find were 10k pots, however I did find a 100k dual gang pot which will be wired in parallel for 50k. I'll also rewire the pots so that the 10k volume control is first, whereas now it is where the volume control is last. I haven't found the banana jacks I have.
Hole was put where it is so that air would flow over the other heatsink. Plus it gives the fan a proper air intake instead of just relying on the small holes in the chassis.
One thing I need to do is figure an easy way to cover the terminals on all the drivers and crossover in such a way that the covering can be easily removed for if I need to service the crossover or a driver. That will do two things. Make it better using it at work as nothing can really be said then about exposed terminals. Make it harder for the terminals to short against the copper screen and damage the amplifier. It's a lot different working inside a room where the walls are at ground potential.
Got the amp mounted and it all works good.
If I were to install a 7 ohm resistor in series with an 8 ohm l-pad and put a 16 ohm resistor in parallel with the midax, would the crossover then still see 15 ohms and would the resistor in parallel with the midax affect how it sounds?
If I were to install a 7 ohm resistor in series with an 8 ohm l-pad and put a 16 ohm resistor in parallel with the midax, would the crossover then still see 15 ohms and would the resistor in parallel with the midax affect how it sounds?
L-pad controls are intended to work on their own. I've never come across a situation in which fixed resistors are added to a variable L-pad control.
You'd be putting 7 ohm in series with the varying series resistance of the L-pad and 16 ohm in parallel with the varying parallel resistance of the L-pad.
I suspect this would result in effective series and parallel resistances that no longer vary in the manner required to present the required constant impedance to the crossover.
Perhaps AllenB would care to comment.
You'd be putting 7 ohm in series with the varying series resistance of the L-pad and 16 ohm in parallel with the varying parallel resistance of the L-pad.
I suspect this would result in effective series and parallel resistances that no longer vary in the manner required to present the required constant impedance to the crossover.
Perhaps AllenB would care to comment.
Last edited:
What it would do is given the l-pad is 8 ohms the crossover would see a 16 ohm load and the l-pad would see an 8 ohm load.
That said I got it right just using fixed resistors. Went back to this version of the fixed l-pads and the speaker sounds better. I honestly think I was used to the bloated bass of the cardboard box the speakers were in and tried to get a similar sound by reducing the level of the midax and tweeter.
That said I got it right just using fixed resistors. Went back to this version of the fixed l-pads and the speaker sounds better. I honestly think I was used to the bloated bass of the cardboard box the speakers were in and tried to get a similar sound by reducing the level of the midax and tweeter.
I don't agree with that as an "8 ohm" L-pad doesn't consist of two fixed 8 ohm resistors.
Rs would range from 8 ohm to zero while Rp would range from zero to typically 20 ohms.