I got caught up in the excitement of the 3886 power amp chips only costing about $7 ea, not that I regret it, but the whole project still added up to several hundred dollars by the time it was done right. Although I have no personal experience with switchmode amp chips, I did research and design a few switch mode power supplies for a digital video sweep generator we produced back when I worked at Tektronix. They generate Rf, and need to have very thought out protection circuitry. These chips claim to not need an output filter, if I'm not mistaken, but I'd consider using some output filtering anyway. Grounding method would be critical, the circuit board probably needs to have a ground plane on one side, all traces as short as possible for the most critical circuitry. If I ever get around to building a small battery operated stereo for taking camping, I'd seriously consider this chip, mostly for the sake of it's high efficiency, and because it seems to be able to work on a very low power supply voltage. I'd probably use two 12 volt lithium batteries, each bypassed with large electrolytics for that.
Power shading is no good with 3 inch drivers for several reasons, but the main one is that you need more cone surface area operating at full amplitude to get enough acoustic energy below about 500HZ. At 30HZ, 25 per side is arguably barely enough. The full amplitude drivers would run out of dynamic range too soon with amplitude shading. Since there's no use for a passive crossover or EQ in this line array, there's probably no real advantage in using separate poweramps for every two drivers. It ends up being more to go wrong, since so many more parts on the bottom line. I would want to have at least 100 watts rms for each side though. I'd stay with analog for a home system largely because I have a very good knowledge of how they work and how to test them in about every way. Switch mode amps seem to have evolved a lot in recent years. Some are probably fine enough. When they do go bad, some have been known to blow up pretty badly, taking the speakers with them. If you burn out some of the drivers in a line array, how easy is it to replace them? That's my thinking on that.
From what I've learned so far, amplitude shading is only possibly desirable when the line array only goes part way to a room boundary. If it gets within the last 15% or so of the distance to the ceiling or floor in this case, I've read that you're better off without the shading. Shading is usually used with giant pro sound PA systems, that are hung above a stage, and don't get real near any acoustic boundaries, and the drivers are much larger and much more efficient than these Vifa TC drivers. With some amplitude shading, they still have enough acoustic power to do the job they need to do.
Last edited:
Bob,
Ok, maybe power shading not a good idea but you should not fear class D amps, especially the TPA3116D2 - I have built it and it sounds fantastic - no noise issues at all. The newly available ones made in China somehow manage to sell for only $17 and have been tested by others on this thread and reports confirm that they sound great. I like it not only for the price but also for the sound quality. The clarity and bass extension beat many other amps at this power level. Uses cheap laptop 19v SMPS bricks for power.
http://www.diyaudio.com/forums/class-d/237086-tpa3116d2-amp.html
Texas Instrument also makes a 300 watt class D for about $110 including 50 v SMPS which would probably be perfect for this speaker.
http://www.aliexpress.com/store/product/TAS5630-AND-SMPS-AII-IN-ONE-BOARD/601158_290293124.html
Ok, maybe power shading not a good idea but you should not fear class D amps, especially the TPA3116D2 - I have built it and it sounds fantastic - no noise issues at all. The newly available ones made in China somehow manage to sell for only $17 and have been tested by others on this thread and reports confirm that they sound great. I like it not only for the price but also for the sound quality. The clarity and bass extension beat many other amps at this power level. Uses cheap laptop 19v SMPS bricks for power.
http://www.diyaudio.com/forums/class-d/237086-tpa3116d2-amp.html
Texas Instrument also makes a 300 watt class D for about $110 including 50 v SMPS which would probably be perfect for this speaker.
http://www.aliexpress.com/store/product/TAS5630-AND-SMPS-AII-IN-ONE-BOARD/601158_290293124.html
Last edited:
I think I will document the full process some time in the near future. If I made the right choices only time will tell. It is a bit crazy to do this using only hand tools. But I'll get there. Not nearly done yet!
Keep up the good work mate
What do you guys think of the Behringer FBQ1502? It looks like it is all analog - no DSP. Indepedent left/right sliders. For $100 it sounds awfully tempting just to have for all my other speakers to play with.
My one concern for use as the main IDS-25 EQ is whether or not it has enough boost at 30 Hz (+20 dB IIRC). +12dB may be the most. However, one can offset everything by -8dB to get +20dB relative - will need to boost overall level out (or into EQ).
My one concern for use as the main IDS-25 EQ is whether or not it has enough boost at 30 Hz (+20 dB IIRC). +12dB may be the most. However, one can offset everything by -8dB to get +20dB relative - will need to boost overall level out (or into EQ).
Last edited:
Wesayso,
You are right, the DEQ1024 has a lot more features for an extra $150. Perhaps all the bad analog business that Bob had to fix can simply be avoided by running the CD player's SPDIF into the DEQ and then using a third party DAC on the output? More boxes and requires a digital source though. Staying in the digital domain for all processing prior to DAC and power amp is a good thing though. Nonetheless, the $100 pricepoint and sheer simplicity of hard sliders and good old op amps to effect the EQ'ing interests me for my usual stuff. How is speaker tower coming along? Have you installed drivers yet?
You are right, the DEQ1024 has a lot more features for an extra $150. Perhaps all the bad analog business that Bob had to fix can simply be avoided by running the CD player's SPDIF into the DEQ and then using a third party DAC on the output? More boxes and requires a digital source though. Staying in the digital domain for all processing prior to DAC and power amp is a good thing though. Nonetheless, the $100 pricepoint and sheer simplicity of hard sliders and good old op amps to effect the EQ'ing interests me for my usual stuff. How is speaker tower coming along? Have you installed drivers yet?
I still got a long way to go. The second tower is about halfway finished. Will take me the rest of this week. After that a week of sanding, another for a preliminary coat and only then will I be able to get my baffles CNC cut. That gives me some time to build a test box for some damping material testing.
Nice name though; Lord of the Waves, The Two Towers.
Nice name though; Lord of the Waves, The Two Towers
.
In the past I've seen a lot of graphic equalizers that gave you a rather non-smooth FR when the controls deviated significantly from being centered. With the controls up fully, the FR looked like a triangle wave on a scope. Pretty bad. Used with a IDS25 clone you would be pushing controls close to their limit. A Parametric should be a better way to go if it has enough range, but you might well need more range than +/- 12dB according to one IDS25 project guy (see attached pic).
The DEQ2496 has optional compressor/limiter functions that might be useful if you like to play music real loud. If you know how to use the DEQ2496 with strictly digital in and out it might be pretty good. They may have cut the corners on the analog circuitry in the DEQ2496 just so you could hear an improvement when using the digital only interface instead...(?)
I'd worry about the TI switchmode amp "huge oscillation" issue when the output goes open for any reason. There could be a rather large transient when the output connection is reconnected. Hopefully an intermittent speaker cable wouldn't result in blown drivers...
One idea for that is to put a load resistor soldered right in the amp, that is maybe 30 ohms, so the output never goes completely open. Max power out would be reduced by this idea, but maybe it's still a good idea.
Tube guitar amps have a similar problem, sort of, where they can blow up if the speaker gets disconnected while the signal level is high, since the output tranny then turns into a spark coil. I've seen that issue blow output tube sockets apart physically (arcing between pins), damage output tubes (internal arcing) and blow open output transformers. All my guitar amps have built in output load resistors that switch in when the speaker plug is pulled out, just for this reason (which may or may not even be good enough since the switching isn't instantaneous).
I expect to have an all analog EQ circuit designed by the end of this week for the IDS25. Then you could use other input sources as well, such as a tuner or TV that didn't have digital outputs.
The DEQ2496 has optional compressor/limiter functions that might be useful if you like to play music real loud. If you know how to use the DEQ2496 with strictly digital in and out it might be pretty good. They may have cut the corners on the analog circuitry in the DEQ2496 just so you could hear an improvement when using the digital only interface instead...(?)
I'd worry about the TI switchmode amp "huge oscillation" issue when the output goes open for any reason. There could be a rather large transient when the output connection is reconnected. Hopefully an intermittent speaker cable wouldn't result in blown drivers...
One idea for that is to put a load resistor soldered right in the amp, that is maybe 30 ohms, so the output never goes completely open. Max power out would be reduced by this idea, but maybe it's still a good idea.
Tube guitar amps have a similar problem, sort of, where they can blow up if the speaker gets disconnected while the signal level is high, since the output tranny then turns into a spark coil. I've seen that issue blow output tube sockets apart physically (arcing between pins), damage output tubes (internal arcing) and blow open output transformers. All my guitar amps have built in output load resistors that switch in when the speaker plug is pulled out, just for this reason (which may or may not even be good enough since the switching isn't instantaneous).
I expect to have an all analog EQ circuit designed by the end of this week for the IDS25. Then you could use other input sources as well, such as a tuner or TV that didn't have digital outputs.
Attachments
Last edited:
Bob,
I am really looking forward to your EQ circuit - thanks for doing this. Regarding the intermittent open circuit issue with the TI amp - never had any issues with it. I have in fact, connected and disconnected my tc9fd drivers from this amp while it is still on and playing music - not a good idea for sure but I forgot to turn it off since that requires pulling the wall plug on the ps brick. I don't have a turn on or turn off thump problem at all, it might be a built in soft start logic?
I am really looking forward to your EQ circuit - thanks for doing this. Regarding the intermittent open circuit issue with the TI amp - never had any issues with it. I have in fact, connected and disconnected my tc9fd drivers from this amp while it is still on and playing music - not a good idea for sure but I forgot to turn it off since that requires pulling the wall plug on the ps brick. I don't have a turn on or turn off thump problem at all, it might be a built in soft start logic?
Would it help to use a bass reflex enclosure to augment the bass a little so that such massive equalization will not be needed in the low end? Tuning it at at about 70 Hz migjt work real well to bring about a fair bit of gain in the lower register. The cabinet will have to be bigger of course and separate, say every 5 drivers into own chamber with a rear firing port. Something narrow and deep and port could be a slot even. I know this changes the group delay and is this a huge problem for the bass? Even as a sealed cabinet, the dimensions of the stock IDS25 cabinet seems small for how large the Qts is for the Vifa. I think sealed cabinet calls for something closer to 5 liters per driver.
3 liter would be considered enough according to s test from 3 years ago by Klang und Ton (see PM). But look at the difference between 19 liter and 2.8 liter. Now picture 25 of them and ask yourself if the enclosure would make that much difference to the final EQ. Would there be a difference in a bigger enclosure? Sure, but how much...
I went sealed although I admit playing with simulations to see what ported could do. In the end it was not worth it to me.
I went sealed although I admit playing with simulations to see what ported could do. In the end it was not worth it to me.
Last edited:
Bass reflex cabinets usually need significantly more internal volume to work right, than an appropriate sized sealed box. More importantly, damping in a vented box is weak below the tuned frequency, so forcing the low end up with active EQ is more likely to cause the drivers to bottom out. I would never combine active EQ that pumps up the low end with a vented box for this reason.
The SPICE models for the analog active EQ for a IDS25 clone is done. I just need to draw up a final complete circuit to present. This should be done in another day or so.
The SPICE models for the analog active EQ for a IDS25 clone is done. I just need to draw up a final complete circuit to present. This should be done in another day or so.
That was my reasoning. I believe Linkwitz even boosts open baffled woofers and if you lowered the tuning of the ports enough you wouldn't boost below the tuning frequency. But why bother if the EQ power is available for the sealed enclosure resulting in possibly better transient response?
IDS-25 Clone Active EQ Circuit
If I were building an IDS-25 Clone with the 3 inch Peerless/Vifa TC or TG drivers, this is the Active EQ circuit I would use. I haven't actually built this, but it should work fine. The graphs are the result of the SPICE models. I had to do each section separately due to limited page size in the free version of SPICE. If you can't get an exact R value, use one that's as close as you can. Same with caps. Verify operation on bench if possible before connecting into Hi-Fi system. After clicking on a graphic to make it big, you can click again on an icon in the lower left corner to make the image even bigger..
If I were building an IDS-25 Clone with the 3 inch Peerless/Vifa TC or TG drivers, this is the Active EQ circuit I would use. I haven't actually built this, but it should work fine. The graphs are the result of the SPICE models. I had to do each section separately due to limited page size in the free version of SPICE. If you can't get an exact R value, use one that's as close as you can. Same with caps. Verify operation on bench if possible before connecting into Hi-Fi system. After clicking on a graphic to make it big, you can click again on an icon in the lower left corner to make the image even bigger..
Attachments
Last edited:
Bob,
Wow! Looks very nice, and not too complicated. You recommend poly caps on the EQ components, could ceramic caps at 5% tolerance work? If I were to miniaturize this as SMT, ceramics would be the way to go. I have to start saving for the 50 drivers now. This circuit looks like a fun build. Thanks so much for doing this and sharing.
Regards,
X
Wow! Looks very nice, and not too complicated. You recommend poly caps on the EQ components, could ceramic caps at 5% tolerance work? If I were to miniaturize this as SMT, ceramics would be the way to go. I have to start saving for the 50 drivers now.
This circuit looks like a fun build. Thanks so much for doing this and sharing.Regards,
X
That was quick, Bob! Now I just need to learn how to translate that into something real .
A side note, found while surfing the net: if you're wondering how to do a 25 driver Bessel array, I found this interesting site: Bessel arrays
.A side note, found while surfing the net: if you're wondering how to do a 25 driver Bessel array, I found this interesting site: Bessel arrays
That was quick, Bob! Now I just need to learn how to translate that into something real
A side note, found while surfing the net: if you're wondering how to do a 25 driver Bessel array, I found this interesting site: Bessel arrays
I need to learn how to read French first!
(I managed to turn on translate function - very useful) I used to do this with optical components and I would send laser light through a Bessel function aperture which provides a beam of light that propagates over large distances with minimal diffraction effects. So I suppose this will make the line array "throw" the sound over a long distance with minimal loss of perceived power and comb interference ('self-diffraction' in a way) at the sweet spot (middle). Note how flat the sweet spot is according to the simulation with 25 drivers - over an area of about 14 drivers it is very flat. Very interesting and a good reason to go with a 25 driver fan-out and 25 cheap class D amps. The 8 watt $10 TPA3110 from Sure jumps out as a candidate. For more power, the 50 watt $17 TPA3116D2 is the way to go.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.