4-channel gain stage with DC-servo for MiniDSP

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
... it's a mystery why they don't get to it, when many of their customers are begging for it and lack of output voltage is one of the main complaints about their platforms??? It's a big fail on their part IMHO.

-Charlie

They have the MiniAMP, perhaps it could be another similar product, MiniLD or something, offering a jumper selectable gain from 3 or 4 db up to 20 or so, and it could just piggy back onto the MiniDSP.

Hey, there's an idea, create a gain stage that's the same size as the MiniDSP and put one of those long-pin style connectors on the edge of it so it could be piggied onto the MiniDSP!

:cool:
 
Account Closed
Joined 2001
The ADAU1701 operates on a 3.3VDC supply, and I see on the data sheet there's an internal 1.8VDC regulator. That pretty much explains the limited voltage output capability of the on-board analog outputs. The only way to achieve more voltage is to add external circuitry. That would necessarily complicate the design.....which obviously they have in the balanced version.

I don't see why a person would need more than 2VRMS capability. The power amplifiers must have very low voltage gain?

Cheers,

Dave.
 
It's the same logic that could be applied to saying we don't need a pre amp...because any cd player can output 1vrms..which can drive an amplifier...all you really need is a 100k pot and we should all be good to go!

Realistically, if you want to be able to drive (the majority) of amplifiers to full power, you need a gain stage between the source and the amplifier, especially with record companies like Chesky, that have taken a stance against the "loudness wars" and started producing recordings that are quite low level, comparatively. If one uses a MiniDSP as advertised / pictured, it doesn't offer a single DB of gain...a situation which becomes a problem if you try to drive multichannel receiver inputs as I am, or amplifiers which require decent voltage to reach full output....which quite a few are, considering they're designed under the assumption that 99% of users will have a gain stage between the source and their amplifier.
 
Yeah, I understand, but that's an input signal level issue (primarily) and not an output voltage capability issue of the interim device. Do you see what I mean?

Cheers,

Dave.

I do. I just think that the unit should offer a little gain, as it's intended use is between a source and and amplifier.. they should make it more clear that it's really just a buffer..so if a customer applies a Linkwitz transform, a heavy bass note will clip the unit at even a relatively quiet setting.

Really, the problem is two fold and goes beyond amplifiers that take a bit of juice to get revved up...the problem is also as Charlie said earlier, applying a Linkwitz transform or low shelf...effectively robs you of that amount of gain, so using the unit to EQ an open baffle speaker essentially results in a loss of gain in whatever amount your highest boost is set to. Need 10 db of boost at 20hz? Congrats, you now have a -10db loss-stage.

There are two places in the MiniDSP where you can add 12db of gain to any frequency.. Wouldn't it be magical if the unit had 24 additional db of gain beyond the maximum setting of the overall volume control?

Meh, I dunno. The whole thing stems from them wanting the unit to be able to be ran from USB power...flexible, but inflexible at the same time.
 
Account Closed
Joined 2001
Remember, equalization only serves to extend/flatten the response of a driver. If you think about it another way, the amount/type of EQ required is essentially irrelevant to this issue.

Yes, if you applied a Linkwitz Transform with much boost you would be asking the miniDSP to apply a voltage gain above zero db at certain frequencies. However, you would normally be adjusting the gain of that channel lower by an amount approximately equivalent to the gain you applied to achieve the same SPL.

The ultimate limitation still remains regardless of the amount of EQ applied. You have 0.9 volts available to a power amp with X-db gain, to a speaker with X-sensitivity and X-displacement capability.

Regarding programmed gain in the miniDSP......You can certainly program gains greater than 12db.....even more if you want. :) Thinking of the device as limited to unit-gain is not literally correct.

Cheers,

Dave.
 
... However, you would normally be adjusting the gain of that channel lower by an amount approximately equivalent to the gain you applied to achieve the same SPL.

mmm..I'm not sure I follow on this...I was under the impression that the dipole rolloff is a loss @ 6db/oct below the first dipole peak, so equalizing it requires the reciprocal boost. I imagine that you could EQ it inversely, so that it's all done with loss to achieve a flat response, but then, aren't we just losing anyway?

I guess what I'm getting at is..the unit has a set amount of voltage it can supply, which is really the same as the input sensitivity..meaning the unit is incapable of actually applying any gain at all...it would make more sense if the I/O overall volume sliders maximum position was at 0 like it is, but the meters extended a further 24db (as what's graphically indicated by any of the EQ settings) and the unit had enough gain to cover...gain...?

I could be completely wrong in my thinking, it just seems more natural that the unit's butt should be able to cash whatever checks it's EQ curves can write :)

I love to be wrong, and I learn from being wrong, but I'm having a hard time with this one, lol.
 
Account Closed
Joined 2001
The miniDSP IS capable of applying gain. (I'm not sure where the misunderstanding is here.)

Yes, the maximum voltage output is a hard limitation, but if you program the unit for 12db gain it can/will achieve it. BUT, you'll have to reduce your maximum input to approximately 0.22 volts (-12db from 0.9 volts) to avoid possible clipping.

Dipole roll-offs don't necessarily need to be programmed with boost (above unity) settings. With the miniDSP you can usually achieve your goal by selecting the opposite slope shelving filter and applying a negative gain.

Cheers,

Dave.
 
BUT, you'll have to reduce your maximum input to approximately 0.22 volts (-12db from 0.9 volts) to avoid possible clipping..
Right, this is exactly my point. I guess its just semantics.

If 0 is the maximum level on the indicators before clipping occurs, you have to think of the filter scale with 0 db at the top, meaning any positive adjustment at any frequency represents a loss of output capability at all other frequencies..we appear to agree perfectly about what it actually means..and what its actual capabilities are...the main difference is I'm of the opinion that it could have been done better..perhaps even so that the scale literally provides what it does visually.
 
I really just have one idea, after thinking about it, the best way to describe it is....the output voltage capability should be able to exceed the input voltage capability by whatever tomfoolery the unit can be visually programmed for.

It really would be a better unit if that was the case. Just because we can say "well if you think about it like this...."

Input level meters should be in the red at 0...output level meters should be in the red at +24.
 
Account Closed
Joined 2001
Well, what you can program and what you can achieve are two different things. :) That "conflict" would apply to many different gadgets....not just the miniDSP.

Even if the hardware could support 3.0VRMS output levels, you'd only need to program a boost of 10.5 db and you couldn't achieve your target.

Cheers,

Dave.
 
Well, what you can program and what you can achieve are two different things. :) That "conflict" would apply to many different gadgets....not just the miniDSP.

Even if the hardware could support 3.0VRMS output levels, you'd only need to program a boost of 10.5 db and you couldn't achieve your target.

Cheers,

Dave.
Well, whatever the db conversion would end up being...if I could rub a lantern and ask for one MiniDSP wish, it'd be 1v input sensitivity with input sliders and meter that peg at 0 db, with output sliders that also peg at 0db, yet the indicator lights indicate all the way to +24 covering the same range as the adjustment graphs, with however much voltage that required. I'd rather clip an amplifier because I ran into headroom than clip a line level device (something I've never done in 20 years of audio tinkering until I bought a minidsp)
 
Dear all,

Just reading through this interesting discussion and the feedback. As usual, happy to answer questions and see how we can improve.
Let me try to answer few questions:
- I can understand that some of you need higher level for some amplifier with low sensitivity. Building a DC-DC step up and allowing higher level out of the miniDSP "unbalanced" basic kit with an op amp would be quite tricky knowing we're constrained by space. Not sure if you've see the latest layout but there isn't an inch of space on the top side on that 3" square board.
We "could" modify the input supply (E.g. 12V like balanced) but we'd loose the USB self powered trick a lot of people enjoyed for the past 4 years. So it's not like we're going to have a magic answer for everybody. :)

- Now with all the above said (price/size/simplicity of design), I'm not 100% sure if I understand why the balanced DSP board wouldn't work though. I may have missed that part in the thread though. Maybe someone can let us know?

It does have over 4Vrms (measured between +& - pins) or 2Vrms (measured between + & GND when say you use it unbalanced). Is 4Vrms still not enough? Most amp we use (even pro amp) would have a rated output circa 1.4Vrms. (e,g, QSC amps: QSC - PLX2 Professional Power Amplifiers) That's ProAV amp though, quite a lot of watt.



DevTeam
Please let me know and I'm happy to help answer some question.

DevTeam
 
It does have over 4Vrms (measured between +& - pins) or 2Vrms (measured between + & GND when say you use it unbalanced). Is 4Vrms still not enough? Most amp we use (even pro amp) would have a rated output circa 1.4Vrms. (e,g, QSC amps: QSC - PLX2 Professional Power Amplifiers) That's ProAV amp though, quite a lot of watt.



DevTeam
Please let me know and I'm happy to help answer some question.

DevTeam

2v would probably be acceptable for me, unfortunately, I already own two boards and have constructed the enclosure for them, so re-purchasing new boards isn't really an option. It's really my fault though, I failed to do enough research on the product's capabilities before purchase. I just read the literature and example diagrams. The unit is shown as being in the same place in the audio chain as a pre-amplifier, so I made the dreadful assumption that it could fill that role.

How about a unit similar to your MiniAMP that can be attached via the ribbon cable, but offered 12-20db of gain (whatever the voltage ended up being)?

The absolute best solution would be a hardware modification, imo. If a couple of components could be swapped that allowed 2-3v (it's always better to have some dynamic headroom) of output at the expense of being able to be powered by usb. Maybe such functionality could be built into future revisions and selected with a jumper. :santa:
 
Dear all,

Just reading through this interesting discussion and the feedback. As usual, happy to answer questions and see how we can improve.
Let me try to answer few questions:
- I can understand that some of you need higher level for some amplifier with low sensitivity. Building a DC-DC step up and allowing higher level out of the miniDSP "unbalanced" basic kit with an op amp would be quite tricky knowing we're constrained by space. Not sure if you've see the latest layout but there isn't an inch of space on the top side on that 3" square board.
We "could" modify the input supply (E.g. 12V like balanced) but we'd loose the USB self powered trick a lot of people enjoyed for the past 4 years. So it's not like we're going to have a magic answer for everybody. :)

- Now with all the above said (price/size/simplicity of design), I'm not 100% sure if I understand why the balanced DSP board wouldn't work though. I may have missed that part in the thread though. Maybe someone can let us know?

It does have over 4Vrms (measured between +& - pins) or 2Vrms (measured between + & GND when say you use it unbalanced). Is 4Vrms still not enough? Most amp we use (even pro amp) would have a rated output circa 1.4Vrms. (e,g, QSC amps: QSC - PLX2 Professional Power Amplifiers) That's ProAV amp though, quite a lot of watt.



DevTeam
Please let me know and I'm happy to help answer some question.

DevTeam
DevTeam,

Thanks for joining the discussion.

Let me describe the problem first. I'll take the example you provided of an amplifier, the QXC PLX2 series amps. I followed your link and found that the input sensitivity was, as you states, about 1.1Vrms to 1.3Vrms. However, looking down the table you will see that the input voltage required to clip the amplifier (e.g. the point where distortion starts to rise above circa 1%) is 5.5Vrms for the PLX1104 and PLX1804, and 11Vrms for the PLX1802, PLX2502, PLX3102, and PLX3602. This means that the amplifier has an additional circa 10dB to 20dB of available power that can be accessed if the input can be driven above the input sensitivity. It's this extra headroom that allows for short dynamic bursts of the music to be reproduced, however, your MiniDSP products can not generate these output voltages so the power is inaccessible.

There is another dimension to the problem - audiophile recordings that have a large dynamic range (large crest factor) also have a lower average level. This means that more gain must be applied to bring up the lower average level in the recording to "normal" listening levels. But because there are large dynamic transients in the music that will also be increased along with the average level, their peak levels may now be above digital 0dB and this will cause obvious distortion.

The other "problem situation" that has been mentioned in this thread is when boost is applied via EQ or shelving filters like the Linkwitz Transform biquad. It's not uncommon to have more than 12dB of boost at some frequency and up to 24dB with the LT. This means that the MiniDSP output will be distorting when the "nominal" level gets to -24dB. One solution is to lower the input level by the same amount as the maximum boost, but this just bring us back to the problem of low levels at the output except it's even worse under this scenario. Your customers are experiencing this exact situation when they use your products with loudspeakers that require boost, such as subwoofers with LT, open baffle speakers that require significant loss compensation, or when doing room resonance mode equalization at low frequencies.

Solution:
Your existing products are great, but they just can not produce the necessary output levels that are sometimes needed. A solution is to offer a high-quality analog gain stage product with a few gain options for each channel e.g. +10dB or +20dB. A 20dB gain translates into output voltages of 10x the input voltage, which seems like overkill but remember that the source may have low Rms levels for audiophile recordings and this would allow the user to bring up the average level. At the full 0dB output level of the MiniDSP of 0.9Vrms (about 1.3Vpk) 20dB would result in 13Vpk voltage. If we use an audiophile grade op-amp like the LME49720 (dual) or LME49740 (quad) the maximum rail voltages are +/-17V. A suitable PS can be made using a 24VDC switching PS wall transformer, a linear regulator to create a stable and low noise 15V or 16V positive rail, plus a DC voltage inverter IC like the NJU7662D or LT1054 to generate the negative rail from the positive rail. If you combine the gain stage with a clipper or limiter so that the opamp can not be overdriven you've got a great product. This "MiniGain" unit could be provided in 4-channel or 8-channel versions, as "in-a-box" or kit, and following the existing MiniDSP cosmetics.

I think that this would be a win-win, both for you and your customers, and would not require you to change any of your existing products.

-Charlie
 
Solution:
Your existing products are great, but they just can not produce the necessary output levels that are sometimes needed. A solution is to offer a high-quality analog gain stage product with a few gain options for each channel e.g. +10dB or +20dB. A 20dB gain translates into output voltages of 10x the input voltage, which seems like overkill but remember that the source may have low Rms levels for audiophile recordings and this would allow the user to bring up the average level. At the full 0dB output level of the MiniDSP of 0.9Vrms (about 1.3Vpk) 20dB would result in 13Vpk voltage. If we use an audiophile grade op-amp like the LME49720 (dual) or LME49740 (quad) the maximum rail voltages are +/-17V. A suitable PS can be made using a 24VDC switching PS wall transformer, a linear regulator to create a stable and low noise 15V or 16V positive rail, plus a DC voltage inverter IC like the NJU7662D or LT1054 to generate the negative rail from the positive rail. If you combine the gain stage with a clipper or limiter so that the opamp can not be overdriven you've got a great product. This "MiniGain" unit could be provided in 4-channel or 8-channel versions, as "in-a-box" or kit, and following the existing MiniDSP cosmetics.

I think that this would be a win-win, both for you and your customers, and would not require you to change any of your existing products.

-Charlie

Exactly the thought I was having...something like this, but line level 20db gain instead of a power amplifier for speakers. And not class D ¬.¬
 
Account Closed
Joined 2001
Gents,

I think you're sort of beating up on miniDSP for not providing a product that they've already provided. :)

I understand Charlie's point regarding extra/reserve dynamic range, but I'm thinking this would normally not be an issue for the majority of situations, IF power amps with average voltage gain are used. (Yes, if you have inefficient speakers and a power amp with 20db voltage gain you've got a problem.)

Anyways, if you feel you need more than 2 VRMS maximum output, I would think miniDSP could alter a resistor value in the analog output stage of the 2x4 balanced unit.......assuming the on-board power supply swing is great enough. (I don't have one, so I can't measure it.) Wouldn't that solve your issues without having to resort to an "add-on" external board??

I see the situation with lower recorded level audiophile recordings as a separate/different issue. You can run across this type of problem even with all analog signal chains. That should be solved upstream of the miniDSP.....or any other gadget that might reside between preamp and power amp.

Cheers,

Dave.
 
I understand Charlie's point regarding extra/reserve dynamic range, but I'm thinking this would normally not be an issue for the majority of situations, IF power amps with average voltage gain are used.

I hate to keep having to drive this point home...

Find an amplifier...any 2 channel amplifier... that offers more than 1000 wpc and can reach that output level, plus dynamic headroom with 0.89v of output.

You can't use "majority of situations" here, most serious home theater rigs now usually boast at least that much subwoofer power, especially if they're at all interested in THX reference levels.

So far, none of the subwoofer amplifiers I've shopped for (including the one I already own) can do it..and I've only found one multichannel amplifier for my main speakers that even comes close (Emotiva UPA 700, 870mV).

I'm not trying to beat anyone up, I just think that losing the flexibility of USB power in exchange for enough voltage gain to make the unit universally usable for any amplifier is more than a worthwhile tradeoff.

At least I'm being nice enough here that I'm willing to be satisfied with purchasing something else from dsp4you, or build a gain stage to make the components work with my existing gear, or even modify the unit, rather than my initial reaction, which was to ask for a refund based on the fact that I really didn't want to have to go shopping for new amplifiers...and if I did, I wanted to be able to pick them based on sonics and power delivery, not input sensitivity.

/grumpy.
 
Account Closed
Joined 2001
I hate to keep having to drive this point home...

I understood your point the first time. No need to keep repeating it. :)

What I guess I misunderstood is your actual power requirements! 1000wpc is waaaaaay more power than any application I could think of in a domestic environment. You're on a different end of the audio playing field than I am, so you do need the extra gain. I haven't had an amplifier with more than 100wpc in my system in years. :)

Anyways, we're not making any progress on the topic and Charlie is probably pissed at me for jacking this thread.

Cheers,

Dave.
 
Gents,

I think you're sort of beating up on miniDSP for not providing a product that they've already provided. :)

I understand Charlie's point regarding extra/reserve dynamic range, but I'm thinking this would normally not be an issue for the majority of situations, IF power amps with average voltage gain are used. (Yes, if you have inefficient speakers and a power amp with 20db voltage gain you've got a problem.)

Anyways, if you feel you need more than 2 VRMS maximum output, I would think miniDSP could alter a resistor value in the analog output stage of the 2x4 balanced unit.......assuming the on-board power supply swing is great enough. (I don't have one, so I can't measure it.) Wouldn't that solve your issues without having to resort to an "add-on" external board??

I see the situation with lower recorded level audiophile recordings as a separate/different issue. You can run across this type of problem even with all analog signal chains. That should be solved upstream of the miniDSP.....or any other gadget that might reside between preamp and power amp.

Cheers,

Dave.

Why do you have to resort to "IF power amps with average voltage gain are used. "??? I own what ever amps that I own. I you saying that now I have to go out and buy some more power amps because I want to use a MiniDSP? I've posted several examples in this thread of why you need more gain. That should be all the motivation that is needed for a supplementary gain stage! Do I need to post a list of all the amplifers that can't be driver by the MiniDSP???

Come on, dude, you are in a losing position with your argument. If MiniDSP doesn't come up with this product I will.

-Charlie
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.