TOPIC: Does Amp Gain matter?

Hi Team,

So if we biamp with a DDRC24 before the amps, does the gain of the amps matter? Can they be different? Can the input voltage required for max power out be different?

In some respects I’m expecting you to say “yes it matters - gains need to be matched” but for those saying that explain this... if I have a stereo amp driving speakers conventionally with a powered sub also in the system, who’s to say what the gain of that sub’s amp is? Powered Subs are designed to be used with many many different stereo amps, they can’t all be exactly gain matched even if we try to do our best with the little dial on the back of them. So is gain matching really that important?

Thanks

It doesn't really matter.

If you use amplifiers with different gain you can easily adjust the channel output level of the DDRC to account for the difference. Even if you have amplifiers with the same voltage gain you will still likely need to make output level adjustments to account for differences in driver sensitivity. For example in my system my mid and tweeter are powered by amplifiers with identical voltage gain but I adjust the level of the tweeter down by -3.5 dB to account for increased sensitivity compared to the mid. If the tweeter was powered by an amplifier with 3.5 dB less voltage gain I could leave the tweeter output level at 0 dB and have the same result.


They can be different.

If your amplifiers have different input sensitivities it just means you will run in to clipping on one amplifier sooner. Take for example two amplifiers with identical voltage gain (lets say 20 dB for simplicity) but one amplifier has an input sensitivity of 1 volt and the other has an input sensitivity of 2 volts. Because they have identical voltage gain for a given input both amplifiers will have identical voltage output. At an input voltage of 1 volt both amplifiers will have an output voltage of 10 volts (25 watts in to a 4 ohm load). However, once you exceed an input voltage of 1 volt the amplifier with 1 volt input sensitivity will be clipping while the 2 volt input sensitivity amplifier will not clip until an output voltage of 20 volts (100 watts in to a 4 ohm load).

To take that example further let's say the 1 volt input sensitivity amplifier is powering a tweeter that is 6 dB more sensitive than the mid being powered by the 2 volt input sensitivity amplifier. And let's say as a result we have adjusted the DSP output level of the tweeter to be -6 dB compared to the mid. Now when the mid amplifier is receiving an input voltage of 1 volt (10 volt output voltage) the tweeter amplifier will be receiving an input voltage of 0.5 volts (5 volt output voltage). The tweeter amplifier will still clip when it sees an input voltage of 1 volt (10 volt output voltage) but now this will not be reached until the mid amplifier is seeing an input voltage of 2 volts (20 volt output voltage).

Hope this helps.

Michael

Michael,
A comprehensive answer -thanks.

Can I add a complication: What if the DDRC24 is running a stereo pair plus a sub? Outputs 1 & 2 will be L & R respectively and have Y-Splitter cables to send the same signal to a pair of stereo amps. One for tweeters one for woofers. Output 3 goes direct to a powered sub.

Now does gain matter or will Dirac sort out any imbalance of levels between tweeters and woofers?

thanks again

asx77 wrote:
You didn't ask me but in that case I think you'd want the amps to be gain matched really. Where this matters in the crossover region because all Dirac could do is make the total level higher or lower at a particular frequency but not change the relative contributions from each driver. If one amp has a higher gain then the frequency range over which that driver is dominant will be extended, which may well adversely affect sound quality. How significant a problem this may be would depend on the speakers and the difference in gain.

I'd question whether using two amps was worth it in the first place though, particularly if the frequency range sent to the main speakers is being limited by sending more to an active subwoofer.

You could use inline attenuators (e.g. Rothwell) to compensate for large gain differences though.

Hmmm... yes, the crossover region becomes problematic.

How closely would they need the gain to match do you think? Attenuators tend to come in set values. Is being within a few dB ok?

Do you know how far off the gains are? If not I would measure it.

Michael

Is there a simple and accurate enough way to measure it?

I assume you have a DMM and REW?

Turn off all filters on the DDRC and send a 1 kHz -30 dB sine wave from REW to the DDRC (with volume at 0 dB) via USB. You can measure the RCA output voltage with the DMM but it should be around 66 mV (DDRC full output voltage is ~2.1 V, 2.1*10^(-30/20) = 0.066 V).

Then measure the amplifier output voltage with the DMM (I would have a speaker hooked up while taking this measurement so the amplifier has some load). You can then calculate the gain as 20*log (Vout/Vin). Let's say you measure 1.28 V at the amplifier output when driven by the -30 dB sine wave, the gain would be 20*log(1.28/0.066) = 25.8 dB.

I recently used this method on a Hypex NC252MP amplifier and the results match the data sheet gain of 25-26 dB so it seems to work reasonably well (numbers in example above are my actual measurements).

Michael

Forgot to mention this but at the end of the day what you are really interested in is acoustic level between the woofer and tweeter. As such you could also take gated acoustic measurements and see how far off the woofer and tweeter levels are (and also compare to not biamping).

At the end of the day I am not sure that this is really worth the effort and I think I would forego biamping.

Michael

That’s great. When I get a chance I’ll give that a try. Many thanks.

Wrt bi-amping and on whether to or not. I just remember about 15 years ago on a more ‘basic’ but still good system, I added a power amp to an integrated and biamped. The improvement may have been particular to that system but £ for £ it was one of the best upgrades I’ve ever done. I’ve always been tempted one day to try again.

Now since getting the DDRC24 I’ve swapped my solid state for a valve amp! Something else I’d always wanted to try. Sound improvement yes. The main difference is the sound is no longer brittle or harsh. I can listen for hours (not that I often get the opportunity) without fatigue.

Downside, I’ve gone from 200W down to 30W I don’t listen at big volumes and it’s a small room, but I do wonder whether now is the time to try biamping again with the 30Wpc on the woofers and another 15Wpc on the tweeters. I’ve not got the second amp but have seen one that might fit the bill.

For years I avoided the upgrade bug but since the DDRC24 I just find I want to squeeze out more! Maybe 2020 has something to do with it too!!

Hi,
So my multi-meter couldn't go that low so i borrowed a friend's. Set REW to generate 1000Hz sine wave at -30db. Default output L+R. DDRC-24 master vol 0db. Routing left and right to output 4, I removed filters, turned dirac off made sure everything g was 0db and measured... Wait for it.... T-RMS 0.183 V AC. What am I doing wrong?
Thanks

Are you measuring 0.183V at the amplifier output or at the DDRC output?

By routing left and right channels to a single output the DDRC output voltage will be double the value I quoted.

Does the DDRC show -30 dB on the input meters and -24 dB on the channel 4 output meter?

Michael

Hi Michael,
OK so just doing Left now.
Input on Dirac 1 showing -27db
Output 4 show g -27db.
DDRC-24 measuring 0.094v ac

Changed REW to -32.5 to get the input/output on the ddrc to show -30db and measured again. Reads 0.070 so getting close now.