Hi Jmascreen,
Sorry this took so long but I finally had a chance to look at it today. Your XO topology is interesting but I do not like how high the breakup mode peak is - it will make the tweeter sound very muddy. I also don't like how the tweeter falls off a cliff - your will lose your shimmer and air. The problems are circled in red:
Here is how I would do it - and I see that the DC28F tweeter has issues and requires a complex XO to solve - and you may actually be better off getting a more expensive tweeter that is more well behaved and save on XO components. Try a RST28F-4 for example.
Anyhow, here is the XO schematic - it is 2.5khz and both drivers are in positive phase. This has the effect of great time alignment for transients. Inverting the tweeter shows a pretty deep null at 2.5khz.
Here is the predicted Freq Response:
Here is the predicted Impedance, it dips to a minimum of 5ohms but is otherwise smoothly varying:
And here is the predicted step response, not the leading edges are very close together:
Plus the absolute phase is positive and well aligned for great transients and attack.
The Xsim .DXO crossover schematic and all data files and plots is attached in a Zip archive below.
Hope this helps and good luck!
Sorry this took so long but I finally had a chance to look at it today. Your XO topology is interesting but I do not like how high the breakup mode peak is - it will make the tweeter sound very muddy. I also don't like how the tweeter falls off a cliff - your will lose your shimmer and air. The problems are circled in red:
Here is how I would do it - and I see that the DC28F tweeter has issues and requires a complex XO to solve - and you may actually be better off getting a more expensive tweeter that is more well behaved and save on XO components. Try a RST28F-4 for example.
Anyhow, here is the XO schematic - it is 2.5khz and both drivers are in positive phase. This has the effect of great time alignment for transients. Inverting the tweeter shows a pretty deep null at 2.5khz.
Here is the predicted Freq Response:
Here is the predicted Impedance, it dips to a minimum of 5ohms but is otherwise smoothly varying:
And here is the predicted step response, not the leading edges are very close together:
Plus the absolute phase is positive and well aligned for great transients and attack.
The Xsim .DXO crossover schematic and all data files and plots is attached in a Zip archive below.
Hope this helps and good luck!
Last edited:
The RST28F would work better of course. I am not sure why you tried to make the DC28F with this TL then. That was a lot of work to get the XO above. Do you want to at least try it? It should sound very good.
You will need to mount the RST28F, make FRD and ZMA files and do this all over again. Are your woofer and tweeter measurements time aligned? Because I don’t see any time offsets in the Xsim model. That would only be true if data has already been time shifted and aligned. If above data was not time aligned then the sim will be off.
Otherwise you need to take 3 data sweeps:
Do all this at most 1/48 octave smoothing. It needs to show all little peaks and bumps.
1. Woofer only
2. Tweeter only
3. Woofer and Tweeter System in parallel
Take it at 0.5m distance on mic axis with tweeter. Use 2.0Vrms (calibrate your SPL meter next to mic).
All done without touching microphone or speaker or any mechanical alignment. We need to do audio interferometry to get the woofer / tweeter acoustic offset down to the mm. This is done be slowly varying the delay in one driver until the prediction matches the parallel System sweep precisely. Each little peak and dip.
You will need to mount the RST28F, make FRD and ZMA files and do this all over again. Are your woofer and tweeter measurements time aligned? Because I don’t see any time offsets in the Xsim model. That would only be true if data has already been time shifted and aligned. If above data was not time aligned then the sim will be off.
Otherwise you need to take 3 data sweeps:
Do all this at most 1/48 octave smoothing. It needs to show all little peaks and bumps.
1. Woofer only
2. Tweeter only
3. Woofer and Tweeter System in parallel
Take it at 0.5m distance on mic axis with tweeter. Use 2.0Vrms (calibrate your SPL meter next to mic).
All done without touching microphone or speaker or any mechanical alignment. We need to do audio interferometry to get the woofer / tweeter acoustic offset down to the mm. This is done be slowly varying the delay in one driver until the prediction matches the parallel System sweep precisely. Each little peak and dip.