Both compression drivers above are arguably the most popular for DIYers. D220Ti is the darling of Econowave, and DE250 is used extensively in Gedlee Speakers.
Having built the Econowave I wanted to get that "extra smoothness" so popularised by DE250 reviews. After finding out that they had good polar response with Econowave, I bought a pair and lived with it for a while. Indeed they sounded very smooth.........But after a while I thought D220Ti is more resolving and DE250 is too "mellow" ?
Anyway after multiple swap between these two drivers I decided to test them using tone burst (because distortion test is boring and DE250 won the CSD.. he..he..). Read about Tone Burst test here.
SO the key summary:
1.5khz - 6khz: DRAW, they are practically the same although DE250 had small win. But not enough to differentiate at all.
6.5khz - 8.5khz: Winner: D220Ti. DE250 is having a hard time to return to zero amplitude. I think this is a major factor why D220Ti sounded more resolving (being metal?). At 7.5khz the DE250 output is quite deformed.
9k - 13k: DRAW, D220Ti a bit better
14k - 15k: Winner: DE250 by miles D220Ti falling apart at 15khz
16k - 20k: Winner: D220Ti picked up and left DE250 behind. The tests were not audible to my ears anymore.
So the conclusion is D220Ti performs better in Tone Burst test. The test itself could be improved by less increments of frequencies to cover more differences but this one should be enough.
D220 is confirmed to be more resolving, and while DE250 is smoother it is at the expense of resolution and accuracy.
The good news is D220Ti is much cheaper. The bad news is ... I've invested in the DE250s
Having built the Econowave I wanted to get that "extra smoothness" so popularised by DE250 reviews. After finding out that they had good polar response with Econowave, I bought a pair and lived with it for a while. Indeed they sounded very smooth.........But after a while I thought D220Ti is more resolving and DE250 is too "mellow" ?
Anyway after multiple swap between these two drivers I decided to test them using tone burst (because distortion test is boring and DE250 won the CSD.. he..he..). Read about Tone Burst test here.
SO the key summary:
1.5khz - 6khz: DRAW, they are practically the same although DE250 had small win. But not enough to differentiate at all.
6.5khz - 8.5khz: Winner: D220Ti. DE250 is having a hard time to return to zero amplitude. I think this is a major factor why D220Ti sounded more resolving (being metal?). At 7.5khz the DE250 output is quite deformed.
9k - 13k: DRAW, D220Ti a bit better
14k - 15k: Winner: DE250 by miles D220Ti falling apart at 15khz
16k - 20k: Winner: D220Ti picked up and left DE250 behind. The tests were not audible to my ears anymore.
So the conclusion is D220Ti performs better in Tone Burst test. The test itself could be improved by less increments of frequencies to cover more differences but this one should be enough.
D220 is confirmed to be more resolving, and while DE250 is smoother it is at the expense of resolution and accuracy.
The good news is D220Ti is much cheaper. The bad news is ... I've invested in the DE250s
Last edited:
Could you elaborate on your measurement setup (distance, SPL, etc.) so we can reproduce this test with similar conditions?
The advantage of tone burst is that you can "see" the effect of reflections by comparing the first bursts to the next ones. The other advantage is also that you don't even need to eq the driver flat, you just need to normalize the recorded signal and you can compare any driver without needing to eq them, because the signal is so narrow in frequency (in fact high Q variations could have an impact, but eq would likely not help here...).
That makes it a great tool for comparing different drivers, and easy to replicate in different situations!
Square waves are also a great test, testing all the frequencies up in one measurement, and with the same property regarding reflections, but you need to eq the driver flat for that.
I will try to measure my 1.5" drivers the same way, on a similar wave guide (1.5" 90x50 rectangular PT waveguide from JBL) and compare them with yours.
Concerning the problem at 15khz for the selenium driver, could it be that you hit a specific resonance of the driver. Looking at the FT, the signal essentially consists in 3 pics, close in frequency. For example the 15khz signal is 13.5khz/15khz/16.5khz.
If one of these 3 frequencies finds a resonance in the 220ti the result will be quite bad, but maybe the same signal at 15.2khz would be much better, by chance (these resonance usually have a very high Q).
The advantage of tone burst is that you can "see" the effect of reflections by comparing the first bursts to the next ones. The other advantage is also that you don't even need to eq the driver flat, you just need to normalize the recorded signal and you can compare any driver without needing to eq them, because the signal is so narrow in frequency (in fact high Q variations could have an impact, but eq would likely not help here...).
That makes it a great tool for comparing different drivers, and easy to replicate in different situations!
Square waves are also a great test, testing all the frequencies up in one measurement, and with the same property regarding reflections, but you need to eq the driver flat for that.
I will try to measure my 1.5" drivers the same way, on a similar wave guide (1.5" 90x50 rectangular PT waveguide from JBL) and compare them with yours.
Concerning the problem at 15khz for the selenium driver, could it be that you hit a specific resonance of the driver. Looking at the FT, the signal essentially consists in 3 pics, close in frequency. For example the 15khz signal is 13.5khz/15khz/16.5khz.
If one of these 3 frequencies finds a resonance in the 220ti the result will be quite bad, but maybe the same signal at 15.2khz would be much better, by chance (these resonance usually have a very high Q).
Last edited:
I hunted high and low a while back, and Parts Express in the USA was the best source for the d220ti - even with the shipping charges. There are some Australian vendors that have other Selenium products, but not the d220ti.
Here are Gainphile's burst signals merged in one single file.
I also removed the clicks in the beginning and end of each burst to ensure that we were not measuring an impulse response...
There is a 0.25sec silence between each one of the 35 bursts.
You can use audacity to record and compare your measurements.
The burst signals are, in order:
1500Hz
1600Hz
1700Hz
1800Hz
1900Hz
2000Hz
2200Hz
2400Hz
2600Hz
2800Hz
3000Hz
3500Hz
4000Hz
4500Hz
5000Hz
5500Hz
6000Hz
6500Hz
7000Hz
7500Hz
8000Hz
8500Hz
9000Hz
9500Hz
10000Hz
11000Hz
12000Hz
13000Hz
14000Hz
15000Hz
16000Hz
17000Hz
18000Hz
19000Hz
20000Hz
I also removed the clicks in the beginning and end of each burst to ensure that we were not measuring an impulse response...
There is a 0.25sec silence between each one of the 35 bursts.
You can use audacity to record and compare your measurements.
The burst signals are, in order:
1500Hz
1600Hz
1700Hz
1800Hz
1900Hz
2000Hz
2200Hz
2400Hz
2600Hz
2800Hz
3000Hz
3500Hz
4000Hz
4500Hz
5000Hz
5500Hz
6000Hz
6500Hz
7000Hz
7500Hz
8000Hz
8500Hz
9000Hz
9500Hz
10000Hz
11000Hz
12000Hz
13000Hz
14000Hz
15000Hz
16000Hz
17000Hz
18000Hz
19000Hz
20000Hz
Tell me how much they charge with shipping. Maybe I can ship them a little cheaper for you, not sure - depends what they charge. (I would buy them from Parts Express then toss them in a Priority Mail box and ship them to you).
I don't mind doing that for a couple folks.
The "silence" is not actually a silence but rather a modulation of the waveform, so this should not be changed. We want to maintain this modulation fidelity.
I think SL favor this type of signal as it can drive tweeters at very high level safely. For example in this test he drove the tweeters to 30Vpp (!)
Spatial distortion
Why thank you kind sir. I will do my homework and get back to you.
Other Aussies interested, maybe we can do a deal?
edit:
GAK! The drivers cost US$97.98 a pair plus US$88.63 postage from PE. They still come in under 100 bucks each but that postage is a killer.
Last edited: