The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

[Ryan]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

Great build documentation, I have a feeling these will be duplicated by many, hopefully someone will veneer a set!;)

[jsr]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

Bob,

Thank you very much for taking that data on the amp filter. Can you explain how you came to 90Hz for the filter corner? It looks like full level is about 134-135dB, so 3dB down from that would be about 131dB which looks like 100Hz. I'm likely reading it wrong so hope you can clarify it for me. Thanks.

[BOBinGA]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

Remember that the effects of the filter is actually the difference between the two curves. Since the driver I used also starts rolling off, you can't just look at the absolute level of the measured output. The blue line is the level without the high pass filter and the red line is the response with the filter engaged. The difference between the two curves doesn't get to 3db until somewhere between 90 and 95hz. Likewise the difference between the curves doesn't get to 6db until somewhere between 75 and 80hz.

I just checked WinISD and it looks like when you add a second order high pass filter on the EQ/Filter tab, it give you a filter that is 3db down at the specified frequency. So if you want to model this amp in WinISD, a second order high pass at 90hz looks like a very close match to the actual filter in this amp. This reduces output by 3db at 90hz and 6db at 75hz - exactly what I measured.

I don't have any way to measure FR directly, so this is as close as I can come, but I don't think I'm more than 1/2db off. That should be close enough for government work.

-Bob

[jsr]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

Thanks Bob. Much appreciated. That's good to know the HPF is around 90Hz. Much more useful than the 150Hz soec.

[lostboy77]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

Would surface mounting the woofer (rather than flush mounting) help with the Z-offset and allow these to run "without feet", or does that throw off the XO too much?

[BOBinGA]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

I actually cut a baffle to try that out. But unfortunately, I didn't save my measurements. As I remember, even surface mounted it still needed a little tilt back. It also introduced some ripple into the response. It wasn't much and it did reduce the "feet" needed to a half inch or so. After dinner tonight, I'll do a little geometry and see what difference surface mounting would cause. But there are other ways to put a tilt on it. You could extend the baffle down another 1 and 1/16 inch or build a base wedge that's 11.5*. (I think that's what it turned out. Again, I'll try to investigate later tonight.)

-Bob

[BOBinGA]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

By geometry, the tilt back angle is 11.83 degrees. I measured with a protractor and it was 11.5 or so. Close enough. This requires the front of the speaker to be 1 - 1/16 inch higher than the back. The 1" feet are perfect for this, but if you prefer some other solution like a wedge base or making your own boxes, just make sure the front is higher than the back by this amount or close to it. It doesn't have to be exact. After all, when was the last time you put your head in a vise to listen to music. I'll post some FR measurements later that show the effects of differing angles, but its not huge.

If I did the geometry right, even surface mounting doesn't get you much. You still need 3/4" feet. But again, it doesn't change the FR that much that it would be very noticeable - it just won't be "perfect". More to come.

-Bob

[chrisn]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

Good job, they look nice!

[BOBinGA]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

Thanks everyone for your interest in my little project.

One recent question asked about using the PEDS without the feet, so I took another look at that. Here is a response chart with the mic on axis with the speaker tilted and then again with the feet hanging over the front edge of their stands so that I was measuring perpendicular to the face. (The design axis is half way between the center of the two speakers. Also, since there is no degree symbol on my keyboard, I will use a * instead.)



You can see that it really makes little difference. Yes, if you tilt them back and listen directly on axis, you get a very flat response. That's the darker line in this graph. But if you eliminate the feet, you still get an easily acceptable response shown by the red line. Using them without feet does cause a small peak around 1500 hz leading down to a dip around 9k, but it only 3db and still quite acceptable. (I should note here that I used one octave smoothing for the graphs in this post to more easily show the differences. Had I left the smoothing at 1/6 octave, it would have been very difficult to show any changes.)

While I had the mic set up, I also measured the polar responses. First, here are the horizontal changes with listening angle:



Its not Geddes perfect out to 60*, but its within a 5 db window from 40hz to 10K hz, which is still pretty darn good. If you stay within 45* it even stays in a 3db window. You can see that by 3K, the woofer is starting to beam a little bit but the tweeter starts to take over by 5K and you are back to normal response levels.

The vertical polars are not as pretty, but they are still very good within a 15* window.



These are measured perpendicular to the face of the speaker - not with them tilted back. So I included a measurement at -11.8* (the red line) that corresponds to the tilted position using the feet. But the big point here is that the response gets quickly out of whack if you go above the box. Even at 15* above perpendicular, it starts losing treble. But even out to 20* below axis, you still get very good response. This corresponds to lying on the floor about six feet in from of them on a three foot stand. Your kids might do that, but seated in a chair, there are no problems.

The moral of this little story is that yes the feet help a little bit if you are right on axis. Off axis to the sides, they perform well way out to 45*. The vertical response shows that when choosing a stand or a shelf to put them on, its better to have them a little above ear height than below. Ear height with the feet is still your best bet, but don't worry if this is not possible in your room.

Well, the honey-do list is calling, so that's all I have time for today. Next time, I will try to take them apart and show individual driver responses. But that's going to take some time, so I won't be doing that this weekend.

Oh, one last thing. I set all the sub and satellite knobs on the amp to 1 o'clock as a starting point. You can adjust to taste from there, but most of the measurements above were with them set that way.

-Bob

[joeybutts]

Re: The PEDS - a 2.1 speaker system for your desk, den, bedroom or office

Hi Bob! Quick question which I don't believe I read in the thread, do you happen to know the F3 of the satellites? Got a buddy looking for a pair of speakers for his film history class and looking to keep it budget friendly. He can't utilize a lot of bass due to poor room construction and flanking, but I think if I can get him something respectable around 80hz that would be good. I've looked at a few designs and figured this may be another option for him.

Thanks!

[BOBinGA]

I can't believe its been three years since I posted this little project. My set of PEDS have been doing a great job as the mains for my modest living room TV.

I had a request to see what changes might be needed if the newest version of the DS90 were used in place of the old ones the PEDS were built from. I swapped the old and new DS90s into my PEDS without any changes to the crossover and the difference is so slight, I wouldn't change anything. If I were to change anything, I might try a 4 ohm resistor for R2 in front of the tweeter instead of a 5 ohm resistor. This is measured with my mic compensation inadvertently turned off and in a different position and location than the measurements in my PEDS description, so its not quite as bright on the very top end, but it still shows a very flat response using either woofer +/- 2db from 150 to 20,000 hz other than that one small dip at 16,000 which you are not likely to hear. But that is just a property of the ND16 tweeter. I also have newer versions of the ND16 tweeters and those are identical to the ones I used originally.

[donradick]

I'd forgotten that you used the DS90s - they sounded great.
I'm finding ported subs not to my liking, and coming around to your preference of sealed subs.

[Kootskid]

Hi Bob,

I am thinking of building the PEDS as I have previously bought some DS90's and some ND20FA tweeters. I would not build the sub as I have one already. I would like to build a center channel using the DS90 and the ND20FA in an horizontal mtm configuration. Would you have any idea's on why this may not work?. If no negatives, I will start playing with some free design software. Did you measure the T/S parameters of the DS90 or just used the published data on the PE site. Thanks for the help, I tried to PM you, but was not allowed for some reason. I am investigating with the Web Manager.

Myles

[BOBinGA]

Hi Myles.

I actually measure everything first. I measure T/S parameters, response in room with a calibrated mic and in the box I intend to use and base my XO based on those measurements. For the most part, in my experience PE's T/S parameters are good, but unless you measure in box and in room, you cannot easily predict the baffle step you need to account for in your XO.

I also have a mtm horizontal center channel with the DS90 and ND16 built and in use for the past year. I am on vacation with only my tablet and don't have the design at hand, so I will have to ask for a few days until I can post the design next week some time. I do remember that I had to design a couple of dB hot in the 2k to 4k octave so that it sounded more natural 15 to 30 degrees off axis where my wife and I sit in the living room. These are the kind of things that just using the measured parameters can't tell you.

Give me a few days to get back home and look up everything on my home laptop. I will post the design and a short write up of my reasoning.

-Bob

[Kootskid]

Bob,

Thanks a lot. I would be interested in your equipment that you use for measurements. I am in the market to buy some equipment, but not sure what to buy as I am just a hobbyist having some fun. Would still like some quality measurement gear to get going. Would be nice to have a new post for reference. Thanks again

Myles