DIYMobileAudio.com Car Stereo Forum banner

Setting Gains w/o Oscilloscope

1 reading
75K views 48 replies 19 participants last post by  UNCLEDAN23  
#1 ·
Okay, here's a write-up based on the notes I took while Mark Eldridge was setting gains on ItalynStalyn's car. I'm sure there are some errors here and there. Please point them out, so i can edit this. I plan on adding pics when I can get them.

By properly adjusting gain setting in an audio system, the user can maximize the amount of clean power available from their equipment. This reduces noise, adds headroom, and prolongs the life of the equipment being used.

The first thing to do prior to setting gains is to set all pre-amp levels and EQ’s to flat (zero) on HU and processors. Secondly, any electronic crossover settings should be adjusted to include the frequency of the test tones that will be used to set gain. For example, my subwoofer output is usually set to low-pass at 63 or 80Hz. For setting gains on the sub channel, I will use a 100Hz test tone, so I will adjust the low-pass for my sub output to 200Hz.

Once the adjustments have been made, you’re ready to pop in a gain setting CD such as Autosound 2000 Disc 104. Gain settings should be adjusted from the beginning of the signal chain to the end. So, you’ll want to start at the HU(s) and work your way through processor(s), then to the amplifier(s). The gain for each channel (or channel pair) should be set using an appropriate test tone. By appropriate, I mean the test tone should be a sine wave at a frequency within the usable range of a particular channel (see note on crossover settings above).

I use a 3-way active HU, so I have High, Mid, and Sub outputs. I will set the gains of these outputs using 100 Hz, 1 kHz, and 4 kHz respectively. I will also be using a test tones recorded at -5dB. This will allow some minor clipping when driven to maximum levels, but any distortion should still be inaudible. The Autosound 2000 Disc 104 includes multiple tracks that will allow for overlap. Gain overlap will allow the user to extract every bit of undistorted power out of their system. The table below illustrates the pros/cons of using overlap when setting gains.

Note: The mini amp can’t play the 100Hz tone with much output at all, so it’s likely necessary to place the speaker of the mini amp closely to your ear before setting the gain. The change in pitch that occurs when the signal begins clipping should be audible.

0 dB Overlap No-clipping, but good amount of power unused +/- 0.1% THD
5 dB Overlap Minor, inaudible clipping (Good compromise) +/- 0.3% THD
10 dB Overlap Clipping is audible (Max. overlap to set gain) +/- 1.0% THD
15 dB Overlap Noticeable clipping is present +/- 10% THD

With all of the preliminary stuff out of the way, it’s time to determine at what volume level our HU starts to distort. The mini amp is connected to one of the mid channel(s). The min amp is to be powered on, but the volume should be kept to it’s minimum (turned all the way down) to avoid hearing the internal amp clip, as opposed to the incoming signal. With a 1kHz test tone playing, the volume of the HU is increased until the pitch of the tone from the mini amp changes. If the HU reaches full volume without this occurring, that HU doesn’t clip. If there’s a slight change in pitch of the tone from the mini amp, the volume of the HU should be reduced to the point just prior to the change in pitch of the test tone. This is the HU’s maximum, unclipped output (MUO) setting.

Note: The change in pitch heard from the mini-amp, indicates the point at which the signal begins to clip. The observed change in pitch comes from second, third, and fourth-order harmonics which occur with clipping. So instead of hearing a single-frequency tone, we hear multiple-frequency tones playing together.

This process is to be repeated, in order, down the signal chain. So any processors between the HU and amplifier(s) are to be tested with the mini-amp and test tones. Remember to use an appropriate test tone for each channel being used (i.e. 100Hz for sub, 1kHz for mid, and 4kHz for the high channel(s)). When setting gains on processors in the signal chain, the HU’s volume should be set to the MUO, as determined previously. The gain controls of the processor being set should be set for the MUO. If the output of the processor clips even when its gain is set to the minimum, the HU’s volume may need to be reduced below the MUO.

Once the gains have been set to MUO on the HU and any processors in the signal chain, the amplifier is ready to be adjusted. Since the output of the amplifier is much greater than the pre-amplified signal, its voltage must be reduced prior to feeding it into the mini amp. This is accomplished by using a voltage divider. I am using a 10:1 voltage divider to set gains for my amplifiers. Depending on the output of the amplifiers being adjusted, a larger voltage divider might be necessary.

The voltage divider is placed inline, prior to the mini-amp. The inputs of the voltage divider are connected directly to the speaker outputs of the amplifier. The HU is once again, set to its MUO. The amplifiers gains are adjusted using the same test tones as before. Once the gain pots on the amp have all been adjusted for MUO, you’re done.

NOW, you can go set you EQ, but remember that any boosts will cause the signal in that range to exceed MUO earlier. To avoid this scenario, only use “cuts” in EQ, as opposed “boosts”. Good luck!
 
#2 ·
From Mark Eldridge's response in the 12V Events section:

Hey everyone,

In as many years as I have been teaching the level setting and gain structure technique, I've never seen this much discussion on it. Glad to hear so many of you are interested in getting it right. It is the most important tuning step we can do after everything is installed.

I wihs we had been able to spend another hour explaining the process, because when you understand the theory behind it, everything makes more sense, and it's easier to approach without any confusion factor. But, that's what we do at the seminar, so come to one, and you'll learn a whole lot more than just gain structure... :)

As far as making two voltage dividers, why not just put a switch in the 10:1 divider you have, with an additional resistor in to switch between 10:1 and 20:1. You could use a DPDT switch, and add a third setting for 5:1 for very low powered amplifiers. You can get even crazier and add outputs to connect your volt meter to check the actual voltage, an o-scope, and some of those really cool LED lights that dance with the music... :)

Following are the steps for setting gains that are in the Advanced SQ Seminar notebook. There's not a lot of explaination with them, just the basic steps listed so it'e eeeasier to follow when working on a car.

System Level Setting and Gain Structure – The Most Important Tuning Step!!!

What is it?

Adjusting the input and/or output level controls of each component in the audio system so that the output performance of each component is matched to the input of the following component

All system component input and output levels are set to clip at the same point

Done properly, it will ensure that the system’s overall signal-to-noise ratio and dynamic range are maximized


Why is it so important?

It allows for peak system performance with minimal distortion and unwanted noise. It also allows full use of system controls (i.e. the volume control will not be limited to only the first three clicks before onset of distortion and magic smoke)


The Important Factors

Dynamic Range – The ratio between the highest and lowest audio
signal levels, measured in decibels

Noise Floor – The noise power generated internally within the system
components in the absence of any input signal, usually measured
in decibels referenced to a specific power level

Signal-to-Noise – The ratio of the signal power to the noise power in decibels

Max Output Level – The maximum output level attainable from a device
with no distortion, measured in decibels or as voltage

Distortion – In theory, it is the addition to or modification of a signal caused by a
piece of equipment. In our reality, it is the undesirable effect of
clipping, causing harmonic distortion.

Tools Needed

Test CD – Autosound 2000 CD #104
Radio Shack Mini-amplifier w/ RCA input adapter
Oscilloscope (optional)
10:1 high-power voltage divider (20:1 if amplifier outputs are over 60V)


The Steps on How to Do It

1. Ensure access to inputs and outputs of all components in the signal chain
(you can, for example, get head unit output access at input of first processor). Also ensure all speakers are disconnected from the amplifiers.

2. Set all tone and equalizer controls to flat if possible

3. Insert CD in head unit, set track to Track 11 (1 kHz @ 0 dB down) on repeat

4. Connect mini-amp and/or o-scope to head unit output

5. Turn up head unit volume to maximum undistorted output level (watch o-scope
and/or listen to mini-amp)

6. Leave the head unit set there

7. Connect head unit output signal to next component input and connect mini-amp to
it’s output

8. If a component has both input and output level controls, set the component’s output level below 1/4 max output temporarily

9. Adjust its input level to max undistorted level

10. Now adjust the output level for max undistorted output

11. Connect its output to next component’s input, and adjust the next components input
and output levels accordingly

12. Continue through all pre-amp level components

13. For crossovers, choose appropriate 0 dB down frequency tracks for the crossover output band frequency range being adjusted (At very low frequency ranges, lower than 200 Hz, and unclipped signal will not be audible through the mini-amp. However, clipping of the fundamental will be audible if the amp is held close to your ear. So you will hear nothing with the gain set for no clipping, and as you turn it up and start to clip, the distortion will be heard on the amp. Looking at a scope makes it easy to see visually)

14. At the amplifier, connect the scope and/or the mini-amp with voltage divider to the amplifier output

15. Choose an appropriate frequency and recorded level track on the CD, recommend -5dB or -10 dB for a 3:1 voltage gain overlap

16. Adjust amplifier gain for maximum undistorted output

17. Now, adjust relative gains for each frequency range (i.e. pull down any channels that are too loud relative to the others, using the amp gains first, and the crossover outputs second.

18. THAT”S IT!!!



Hope this helps.
 
#7 ·
There is a similar tutorial using the Radio Shack mini amplified speaker and disc 4 in one of the Autosound 2000 tech briefs. If I remember correctly, it shows how what parts you need to attach it to the speaker leads with everything contained in a small project box. Unfortunately I lent my briefs to someone many moons ago and never got them back.:(
 
#16 · (Edited)
#17 ·
Has anyone here tried the mini-amp gain setting method described in the first two posts?

I'd really love to avoid renting a scope, and this seems like the most feasible alternative. However, I have a few concerns regarding the equipment. The mini-amp is the Radio Shack SA-155 or SA-150. It is designed to have home equipment as its source, evidenced by its input sensitivity of 160mV for a CD player, versus a car amp's input voltage of 2V or higher.
SA-155 Mini Stereo Amplifier 310-1957 Specifications

My preouts are 4V, and I'm sure many others do 8 or 9V. It would seem this would overdrive the mini-amp into clipping. Does this happen? From Eldridge's description it seems he just plug and plays this method to all systems without worrying about it.

Could an external level control like the PAC LC-1 bring the voltage down to the appropriate level? And would the test still be worth doing after the signal is changed so much from its orginal value?
Pac-Audio.com Product Details | iPod Integration for your car and More by Pac-Audio - Connecting you to the future
 
#18 · (Edited)
Reasonably new to the forum, but wanted to contribute a drawing of the amplifier voltage divider circuit and wiring diagram to this thread. This is a 20:1 (21.3:1 to be precise) divider and works well with any amplifier. In fact, it's identical to the one I use myself. The only modification I'd make if I were to make a new one would be to replace the spring-loaded speaker terminal with a 6-foot 2-conductor speaker cable, simplifying connection to the amplifier. Using the speaker terminal, though, makes wiring up the resistors cleaner and faster, and there's no worry about the speaker cable getting snagged on something, pulling the circuit guts out. The choice is yours.

I've included Radio Shack part numbers so everyone interested can source the parts in their neighborhoods.

--david
 

Attachments

#19 ·
I can tell you from firsthand experience that, yes, when you're connecting your pre-amplifier components to the Mini Amplifier's input, the amplifier's circuitry can be driven into clipping, tricking you into thinking your upstream component is clipping. I've mitigated this by keeping the volume dial at its lowest setting. It works, but a better solution, I think, is the PAC LC1 B-dole mentioned. It's purely a variable resistor and will simply lower the signal before going into the Mini Amplifier. Think of it the same as the voltage divider in my post above, because that's what a potentiometer is actually. It will not in any way degrade the signal preventing you from hearing when the upstream pre-amplifier component is clipping. Of course, you could construct a similar device for under $10 using a 10k potentiometer, a project box and the spare panel-mount RCA jacks you'll have if you make the amplifier voltage divider shown in the thumbnail image in the post above.
 
#21 ·
Yup. My point in providing the drawing is to help those people looking to jump into this hobby but are a little new or unsure how to construct devices such as this. I've seen a few people post on the site who expressed a desire to delve into the more technical aspects of the hobby but lacked the technical training or know-how when it comes to these sorts of things. Hopefully this will help one (or more than one) person along the trail and encourage them to "peel back the curtain" so to speak. We live in a physical world in which systems and devices work for very specific reasons; when we can get to the root of WHY and HOW they work, that's when the real fun begins.
 
#25 ·
That's like assuming a transformer in any application is 1:1... Just because it isolates does not mean it's 1:1.

There's transformers that will take speaker level down to mic level. They are called DI boxes. Yes, some are dividers to get it down to line level then the transformer takes it to mic level but some go straight into the transformer at speaker level and poop out mic level. Speaker level out of an Ameg SVT no less ;)

So speaker level out of a car headunit to line level, isolated is no big deal...

WELCOME TO JENSEN TRANSFORMERS, INC.

go crazy.
 
#24 ·
Related question (i think). I didn't really follow the process in the original posts, but i'm guessing its very similar to the approach I'm attempting.

I'm using a voltage divider to feed the line in on my laptop and then looking at the signal in Audacity (which displays sort of like a scope).

I'm pretty sure my voltage divider is ok (i have 10:1 and 33:1 with a switch). The output checks out when tested with a battery and voltmeter.

I get a trace on the pc ok, but seem to get clipping from my HU (i'm checking speaker outs not pre-puts) earlier than i'd expect. The trace is showing clipping at 18 out of 30 volume. I have turned down the levels on the input (in software) but still shows a clip.

I have not been able to check the signal voltage output from the voltage divider as my multimeter doesn't have a suitable range for measuring an alternating signal at this level.

Any ideas on how to improve?

Thanks
Stuart
 
#26 ·
Okay...just wanted to get clarifcation on which transformers you had in mind. Yes, a Line-to-mic level transformer will offer 20dB of attenuation on average, which should be equivalent to the "pad" circuit I drew above. And many of the line-level transformer primaries measure 20k ohms, there should be only a fractional amount of power flowing through the input side. I handn't considered using a matching transformer, but it seems it would work well.

I'm familiar with the Jensen transformer company. Thanks for the link.

Do you, by chance, know this company, who also make high-quality transformers--(all manner of them, in fact)?
 
#27 ·
20B is a LOT of pad for car audio. Especially when a lot of amps can take the input levels many if not most stock head units put out.

Never used the company you linked, Usually go straight for Jensen, also used a bunch of Sescom. I'll do more looking at them. So a return thanks for the link :D
 
#29 ·
20B is a LOT of pad for car audio. Especially when a lot of amps can take the input levels many if not most stock head units put out.
Remember, the 20dB pad is used only when setting the amplifier gains and is not wired into the system during normal use. We're merely dropping the speaker-level voltage coming off the amplifier to a level that doesn't overdrive the Mini Speaker's input stage, which is quite sensitive.
 
#30 · (Edited)
No point in using an amplified speaker to "set gains" with, and for that matter an O-scope. What are you "setting gains to"? A random source volume setting using a worst case signal that won't ever be nowhere near your listening medium. You set your amp to clean output with that solid tone but then you play a track which only has 1 sample hitting 0dB. Great, now the rest of the track is not loud enough, and all the sub 0dB portions aren't "set to" anything.

I don't get it.

Find a range at the source that gives you a volume range that you are comfortable with, leave some gain on top of that for low level dynamic recordings, and turn the gain up by ear till just below you start to hear it strain. It's nothing more than a rough approximation since the primary device used to evaluate is an imprecise one.

Setting gains that precisely is like a tailor taking your measurements with a precision ruler.

Just look at a home setup which doesn't give you the option to be silly like that. Do we ever thinking about "oh is this volume setting where my preamp or amp starts to hit .123% THD?". No, you just set it to where it sounds clean for the material and accept the "rough approximation" used to get that outcome. Never heard of even the most elitist of audiophiles questioning even that. Magic crystals on RCA cables and Bob's barricades wire elevators get WAY more attention even.
 
#32 ·
In theory I'll agree with you. As you know in home audio we have two pre-amplifier/amplifier system architectures. We either have an integrated system, with the home theater receiver being the most well-known example, where the pre-amplifier and the amplifier(s) are in a common chassis. Or, we have separate components with a stand-alone pre-amplifier and a separate amplifier.

In the case of the integrated system, the signal level between the pre-amplifier stage and the amplifier stage is handled by the design engineer and none of the signal levels between the stages can be adjusted by the user. They’re fixed. In the case of the separate pre-amplifier and amplifier, the situation is different, where the signal level on the pre-amplifier’s outputs varies based on the volume control’s level setting. Most all separate amplifiers have a set input sensitivity, and that level is almost always set to a 2.0 V RMS level on the unbalanced inputs, meaning the amplifier will go to full power just below clipping when it receives a signal of that level. There are exceptions to this input sensitivity number, but not a great deal. Many pre-amplifers have a signal output level of between 5.0 and 9.0 V RMS on their unbalanced pre-amplifier outputs. Thus, we have a “gain overlap” potential of 3:1 on average, and this works well where the user is completely capable of getting the amplifier to go to full power, while providing them the ability to find a comfortable listening level.

In a car audio system, as you know, we have a different situation with the amplifier because they have a variable input sensitivity adjustment. Why? Because unlike home audio, where the pre-amplifiers have a fairly standard output voltage that’s more than capable of bringing the amplifier to full power, the power amplifiers are designed with the knowledge that pre-amplifier components have a wider range of output voltage levels, and the amplifier needs to be designed to be adaptable to these levels. If that weren’t true, the amplifiers would be designed with a fixed input sensitivity. There was a time when car audio looked very much like home audio “separates” systems. For those of you who were into car audio back in the 80s, you may remember when Alpine and Kenwood made their amplifiers with proprietary DIN signal jacks designed to work exclusively with their tape decks. Those amplifiers had fixed input sensitivity settings because the manufacturers knew the signal levels coming from their pre-amplifiers.

That said, I fully believe in the approach of needing to match the amplifier’s input sensitivity to the upstream pre-amplifier components. In simple systems with just a CD player and an amplifier, yes, I agree that setting the amp’s gain setting can be quite simple and can be done by ear with reasonable accuracy. The only downside is the system’s signal-to-noise level isn’t optimized. But when the system includes one (or more) pre-amplifier components in the signal chain, optimizing the system’s signal-to-noise performance gets exponentially more complex and therefore requires a more technical and science-based approach.