I'm looking forward to it!
I'm not sure if this helps, but something like this would probably be perfect.
Of course, I'm not sure how close you are to having yours ready, it might be better
Edit: From his homepage:
Now, THAT'S what I'm talkin' about!
Last edited:
OK, so let's talk about the design. Please provide your input.
POWER SUPPLY:
Considering convenience and the need to be compatible with various global mains voltages and frequencies, supplying power to the unit from a switching DC wall transformer is probably best. Internally, I would like to have +/-15VDC rails for the op-amps. Typically the switching PS supplies a single positive rail, and I will generate the other one using a special IC. So, I could specify that a 15V DC switching supply has to be used. Alternately, I could specify that an 18VDC-24VDC switching supply is used - this would require an additional linear regulator like the 7815 plus a couple of extra components and the IC to generate the negative rail. This latter solution might be better, however, because I can make sure that the positive rail is adequately filtered and at the correct voltage. I did a quick check of a US supplier (Jameco) and they carry both 18VDC and 24VDC switching wall supplies for about $15ea.
I can pretty easily add on a 5V-12V DC output that can also be used to power the 2x4 crossover modules. This can probably also be done for the 2x8, etc. but I need to double check the required current capability.
GAIN STAGE:
It seems that a good choice for this is the LME49740 quad op-amp. This is an excellent, modern, low-distortion op amp for audio applications, and 10x (20dB) gain should be no big deal. I might just make this a fixed gain circuit, and the builder can change the gain as they wish by using different resistors.
The input will have a basic RFI LP RC filter, with F-3 something like 80kHz-100kHz. The outputs of the MiniDSP crossovers are AC coupled so there is no DC coming in to the gain stage. I will have to do some modeling and try to optimize the gain resistors for low DC at the output.
OUTPUT
I will use 47R-100R isolation resistors on the unbalanced outputs. An unbalanced to balanced line driver option might also be useful in case the user wants to drive a pro amp. This could be left unpopulated, or the unit could have both unbalanced and balanced outputs. This would also make it easy to bridge any amp that does not already have that capability. Extra cost, but handy. I would probably use THAT1646 or TI's DRV134.
Thoughts?
-Charlie
POWER SUPPLY:
Considering convenience and the need to be compatible with various global mains voltages and frequencies, supplying power to the unit from a switching DC wall transformer is probably best. Internally, I would like to have +/-15VDC rails for the op-amps. Typically the switching PS supplies a single positive rail, and I will generate the other one using a special IC. So, I could specify that a 15V DC switching supply has to be used. Alternately, I could specify that an 18VDC-24VDC switching supply is used - this would require an additional linear regulator like the 7815 plus a couple of extra components and the IC to generate the negative rail. This latter solution might be better, however, because I can make sure that the positive rail is adequately filtered and at the correct voltage. I did a quick check of a US supplier (Jameco) and they carry both 18VDC and 24VDC switching wall supplies for about $15ea.
I can pretty easily add on a 5V-12V DC output that can also be used to power the 2x4 crossover modules. This can probably also be done for the 2x8, etc. but I need to double check the required current capability.
GAIN STAGE:
It seems that a good choice for this is the LME49740 quad op-amp. This is an excellent, modern, low-distortion op amp for audio applications, and 10x (20dB) gain should be no big deal. I might just make this a fixed gain circuit, and the builder can change the gain as they wish by using different resistors.
The input will have a basic RFI LP RC filter, with F-3 something like 80kHz-100kHz. The outputs of the MiniDSP crossovers are AC coupled so there is no DC coming in to the gain stage. I will have to do some modeling and try to optimize the gain resistors for low DC at the output.
OUTPUT
I will use 47R-100R isolation resistors on the unbalanced outputs. An unbalanced to balanced line driver option might also be useful in case the user wants to drive a pro amp. This could be left unpopulated, or the unit could have both unbalanced and balanced outputs. This would also make it easy to bridge any amp that does not already have that capability. Extra cost, but handy. I would probably use THAT1646 or TI's DRV134.
Thoughts?
-Charlie
Sounds good to me, but I'm not really an electronics major ¬.☻
If, however, the unit could run from 12V, that might open it up to other applications such as car audio, if you plan to sell them.
Also, if the OP amps are replaceable, just in case someone craaaazy enough decided they wanted to replace them with some of the Burson units... *runs away*
If, however, the unit could run from 12V, that might open it up to other applications such as car audio, if you plan to sell them.
Also, if the OP amps are replaceable, just in case someone craaaazy enough decided they wanted to replace them with some of the Burson units... *runs away*
I can make it so that the first regulator (for the positive rail) can be omitted when using it in a car. But this would limit both rails to whatever DC voltage the car electronics are supplying. Also, I don't have any practical experience with automotive applications, but I believe the electrical system is seriously noisy and that you have to take special measures against this. If anyone can chime in on that please do so.
I am planning to use thru-hole components so that your typical "I'm OK with a soldering iron" DIYer can populate the board. This will likely be a PCB+parts kit unless there is a real screaming demand for something more "finished".
I'm definitely NOT planning to make any special accommodations for $280 (for quad amplifier) op amp so-called "HD upgrades". But I can't stop you from wasting your money!
-Charlie
afaik, dc-dc adapters are typical for equipment not designed for cars. In a car, the power supply is usually 13.8v.
Soldering station is only a power switch away!
The only accommodation that would make it easier is having the OPamps on a socket, but it really doesn't make much difference to me, if I ever did anything like that, I'd desolder / solder new components...just thinking of different ways to make it more flexible
The Peavey IPR3000 is close, at 75X (39dB) gain. It's rated as needing 0.775V input for 840W/4Ω per channel.
Presumably, if it's available yet, the IPR3000's nicer-looking clone, the Crest ProLite 3.0, has similarly high gain.
But, again, if it can't be driven from a phone charger anyway, why not just use the balanced one? I'm with Davey. I don't get this discussion at all.
Because a lot of people have already purchased the units after reading about them and not even considering that driving amplifiers could even be an issue with a line level device.
They see this:
And they buy it. It's even worse when you consider the example shown is a turntable.
So, now I'm ~$240 invested in these two boards, it's not as simple as "just use the balanced one", unless you're suggesting that re-purchasing new units is a reasonable solution.
Well, due diligence...the output limitations of the unbalanced 2x4 have been known and discussed for quite a while.
That said, one can light up the clip lights of an IPR-3000 using the unbalanced 2x4. I've done it. Ditto with a Dayton Class G rack amp. So it's not like there are no insanely powerful subwoofer amps that can run just fine on the unbalanced 2x4.
What you say is true, but just as true is the fact that the 2x4 (unbalanced OR balanced) has insufficient gain with some amps, and when significant boost or EQ is used in the crossover functions. I am simply arguing for something that provides additional gain when needed, something that the user could add on to the existing boards.
Dear All,
As I'm reading these thread, it doesn't seem like anybody is going to budge so I'm pretty sure we don't want to get involved in it as it doesn't seem like we'd be able to change anybody's mind... It's the spirit of DIYaudio, everybody will have a different setup, some of you will find the product fitting their needs others won't with a specific board and will need an add on. It's not like we have a "magic" wand here to change the specs of a product as stated on spec sheet. Hearing ideas to implement add on boards is always great and we certainly welcome it!
So here is what I propose in the spirit of keeping everybody happy!
- Charlie started this good thread with the intention to build what I guess I could sum up as a 4 ch line driver (let me know if I missed that point...
- I think that the thread went out on different path since then. Best to get back to that discussion? We're looking forward to seeing his board which I'm sure will help people in the community.
DevTeam
As I'm reading these thread, it doesn't seem like anybody is going to budge so I'm pretty sure we don't want to get involved in it as it doesn't seem like we'd be able to change anybody's mind...
It's the spirit of DIYaudio, everybody will have a different setup, some of you will find the product fitting their needs others won't with a specific board and will need an add on. It's not like we have a "magic" wand here to change the specs of a product as stated on spec sheet. Hearing ideas to implement add on boards is always great and we certainly welcome it!So here is what I propose in the spirit of keeping everybody happy!
- Charlie started this good thread with the intention to build what I guess I could sum up as a 4 ch line driver (let me know if I missed that point..
. - I think that the thread went out on different path since then. Best to get back to that discussion? We're looking forward to seeing his board which I'm sure will help people in the community.
DevTeam
I understand that, though I disagree re: the balanced/x8 units. However, there's an elephant in the room that your solution might address but you haven't mentioned: noise.
My experience trying an external line driver downstream of an unbalanced miniDSP (I tried two, an AudioControl Four.1 car-fi box powered by a 12V wall wart, and a Samson S-Convert) was a complete failure, because of the noise introduced. Perhaps your solution will be superior.
Hmmm, I'd like to know more. Did you do some measurements or use a scope to check into the problem at the time? If so, what did you find?
No, I have acoustic measurement tools but not electronic measurement tools. I just chalked it up to too many amplification stages in the chain.
My crude but effective solution was simply to switch to an amp to ones with more suitable input sensitivity. Solved the problem well enough. For most subs, it probably wouldn't be an issue, because the noise was IIRC fairly high up in frequency (hissing). With the Aurasound NS10-794-4A it was an issue, because the driver has way more extended top-end bandwidth than a typical subwoofer driver: underhung coil, copper-sleeved motor, top end actually extends to ~10kHz on axis, counting cone breakup!
Ugghhhhhh.....
So, I went and read a good portion of the minidsp thread over at avs and it would appear that noise is indeed an issue for anyone who has tried to use a gain stage after a MiniDSP board.
Maybe I should just sell these boards on the swap meet and save up for a 4x10HD, that's the one I really want anyway.
So, I went and read a good portion of the minidsp thread over at avs and it would appear that noise is indeed an issue for anyone who has tried to use a gain stage after a MiniDSP board.
Maybe I should just sell these boards on the swap meet and save up for a 4x10HD, that's the one I really want anyway.
I hate to burst your bubble, but I experienced the problem using a 4x10, which I described in this post:
http://www.diyaudio.com/forums/minidsp/237510-4-channel-gain-stage-dc-servo-minidsp.html#post3524103
The 4x10 has slightly more output voltage capability, but not much gain from input to output. I had connected a CD player to the SPDIF inputs. With the particular tracks I wanted to listen to, the large dynamic range (large crest factor and low average level) combined with the amount of boost I was using on some of the drivers (in an OB speaker application) made the max SPLs that I could reach to be way too low. I could not supply enough average voltage to the amp and I could not "turn up" the gain anywhere, because there is no preamp where you can do that. You can only attenuate using the 4x10 volume control.
Good luck.
-Charlie
http://www.diyaudio.com/forums/minidsp/237510-4-channel-gain-stage-dc-servo-minidsp.html#post3524103
The 4x10 has slightly more output voltage capability, but not much gain from input to output. I had connected a CD player to the SPDIF inputs. With the particular tracks I wanted to listen to, the large dynamic range (large crest factor and low average level) combined with the amount of boost I was using on some of the drivers (in an OB speaker application) made the max SPLs that I could reach to be way too low. I could not supply enough average voltage to the amp and I could not "turn up" the gain anywhere, because there is no preamp where you can do that. You can only attenuate using the 4x10 volume control.
Good luck.
-Charlie
Honestly, if you need mega-power just get the Peavey IPR/Crest ProLite amps, or a Dayton SA1000. Those will work fine with your unbalanced miniDSP.
@ Charlie,
8Vrms (20dBu) output on the balanced section of the 4x10 and it was still limited output??
Unfortunatley I'm indeed afraid that we wouldn't be able to work higher levels than any of these levels. Pro Audio I/O are already circa these levels and you're already talking of a poor gain structure for most consumer amp (i.e. very hot signal coming out of the DSP, amp running at low SNR as you're going to lower the vol control all the way). You'd need a serious line driver to be able to reach higher levels than 8Vrms and keep a good SNR.. it's going to be tricky I think.. but maybe I misunderstood.
FYI, a little trick on the miniDSP, you can indeed add gain to a channel if needed using an advanced biquad. It's a bit of a trick but it's there. You'd have to pay attention at your levels though.
Hoping this info helps.
DevTeam
8Vrms (20dBu) output on the balanced section of the 4x10 and it was still limited output??
Unfortunatley I'm indeed afraid that we wouldn't be able to work higher levels than any of these levels. Pro Audio I/O are already circa these levels and you're already talking of a poor gain structure for most consumer amp (i.e. very hot signal coming out of the DSP, amp running at low SNR as you're going to lower the vol control all the way). You'd need a serious line driver to be able to reach higher levels than 8Vrms and keep a good SNR.. it's going to be tricky I think.. but maybe I misunderstood.
FYI, a little trick on the miniDSP, you can indeed add gain to a channel if needed using an advanced biquad. It's a bit of a trick but it's there. You'd have to pay attention at your levels though.
Hoping this info helps.
DevTeam
I'm sure 8vrms would be plenty for me, that's more in line with what I was getting with my old analog preamp.
As long as you don't need 8vrms of input to get there.
@ Charlie,
8Vrms (20dBu) output on the balanced section of the 4x10 and it was still limited output??
Unfortunatley I'm indeed afraid that we wouldn't be able to work higher levels than any of these levels. Pro Audio I/O are already circa these levels and you're already talking of a poor gain structure for most consumer amp (i.e. very hot signal coming out of the DSP, amp running at low SNR as you're going to lower the vol control all the way). You'd need a serious line driver to be able to reach higher levels than 8Vrms and keep a good SNR.. it's going to be tricky I think.. but maybe I misunderstood.
FYI, a little trick on the miniDSP, you can indeed add gain to a channel if needed using an advanced biquad. It's a bit of a trick but it's there. You'd have to pay attention at your levels though.
Hoping this info helps.
DevTeam
Now that you mention it, I probably could have used the balanced outputs but I didn't think of it (I was using my amp in a bridged configuration). I was using the unbalanced output to drive an amp that was in bridge mode. I already described the problem in detail earlier in this thread:
http://www.diyaudio.com/forums/minidsp/237510-4-channel-gain-stage-dc-servo-minidsp.html#post3524103
Let me state, again, that the problem was compounded by source material with large dynamic range and low average signal level PLUS the use of around 12dB of low frequency boost paired with -12dB of input gain to keep internal levels below 0dB.
Sure, I could have added gain using a biquad or turned up the input gain, but this would increase the chance of digital clipping, which sounds horrible and is readily obvious. I would rather have the occur in the analog domain, where it is less offensive or could be at least partially accommodated by amplifier headroom.
If this is such a non-problem, why are other people complaining about it?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.