Headroom Voltages Vs Sonics - More headroom isn't always a good thing?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Electrobumps said:
people getting picky about spelling and spouting off about how many consoles they designed.  ???

The same kind of people that are here teaching you for free!
This is a respectful community, be humble , Learn and ride on the BUS

 
Whoops said:
You should know, it's next to Doug's nickname, 7066

To be fair,  I guess you'd have to know the forum a bit. Lot's of prolific posters on other fora are full of it. This is not that kind of place.
 
I only caught up with this thread now , but yes very interested in the evolution of  Dickies Helios console  ,
If you listen back to the records that were made with that console you can hear the headroom expansion as it goes on , Stones records circa 68 were fizzly at all levels really, Tommy the original too  , then by the time  Who's Next  that charachteristic sizzle only touched the peaks . Of course the Studer A80 added an extra level of fidelity too and track count .
The Floyd took it to another level again with Dark Side and Wish you were but that was  EMI's consoles , its all been down hill from there  sonically  to my ear, Ic's may have done plenty for the spec sheets and the reliabillity factor but sonically they overload like a brick hitting a shit house wall  ::) gracious overload is mans best friend in studio .
 
I've been recently trying to get the gain-staging sorted for my ongoing (looks to be lifelong) project of building a rudimentary 16 channel (12 channel, 4 returns) console based on Helios amplifiers and routing topology (more or less). I'm trying to use the 2128 amp as a fader buffer(24V)  and found that at first I was smacking the buffer way to hard with the output from the 22113 line driver (using nominal 34 - 36V)that  I was using in the mic pre. I dialed the 22113 back a bit (feedback resistor) and it seems to be behaving better. The  little 2128 buffer in the router will definitely contribute to the sound if you are not careful (many will of course argue that there isn't supposed to be a sound if you are doing it right). I'll probably try and negotiate my way carefully along the edge with 22113 gain. I don't want to raise the feedback  resistor back too high on the 22113.

The router module has a pan, mono and cut switches, meter select (pre/post fader) and 2 sends with pans (one of the sends switchable pre/post). Like I said, rudimentary.

I had to think a bit about how to not load the 2128 output too much, remains to be seen whether I have done that adequately. I only have one Helios schematic showing fader makeup and the amps are not identified. I ended up doing the sends a bit differently to the schematic (for better or worst).

At the moment I'm deciding whether I should use bus transformers (free gain but bandwidth limiting) or stick with the original transformerless philosophy (potentially greater electronic noise).

The whole balancing act with headroom that I have and will continue to be going through is an interesting exercise for someone like me who only has a basic understanding. I'm probably going to be doing it more by feel that numbers, although I will check the numbers where I can.

 
Squeaky said:
The whole balancing act with headroom that I have and will continue to be going through is an interesting exercise for someone like me who only has a basic understanding. I'm probably going to be doing it more by feel that numbers, although I will check the numbers where I can.

The basics of the balancing act are fairly straightforward. With a 24V rail the maximum output is about +18dBu. It is normal to operate with about 26dB of headroom in a desk so your nominal internal operating level should be around -8dBu.

So the output of mic pre should be -8dBu -> output of EQ likewise. The only place the level returns to normal line level is at the line outs where you use a transformer with the required 6dB gain.

Cheers

Ian
 
Squeaky said:
I've been recently trying to get the gain-staging sorted for my ongoing (looks to be lifelong) project of building a rudimentary 16 channel (12 channel, 4 returns) console based on Helios amplifiers and routing topology (more or less). I'm trying to use the 2128 amp as a fader buffer(24V)  and found that at first I was smacking the buffer way to hard with the output from the 22113 line driver (using nominal 34 - 36V)that  I was using in the mic pre. I dialed the 22113 back a bit (feedback resistor) and it seems to be behaving better. The  little 2128 buffer in the router will definitely contribute to the sound if you are not careful (many will of course argue that there isn't supposed to be a sound if you are doing it right). I'll probably try and negotiate my way carefully along the edge with 22113 gain. I don't want to raise the feedback  resistor back too high on the 22113.

The router module has a pan, mono and cut switches, meter select (pre/post fader) and 2 sends with pans (one of the sends switchable pre/post). Like I said, rudimentary.

I had to think a bit about how to not load the 2128 output too much, remains to be seen whether I have done that adequately. I only have one Helios schematic showing fader makeup and the amps are not identified. I ended up doing the sends a bit differently to the schematic (for better or worst).

At the moment I'm deciding whether I should use bus transformers (free gain but bandwidth limiting) or stick with the original transformerless philosophy (potentially greater electronic noise).

The whole balancing act with headroom that I have and will continue to be going through is an interesting exercise for someone like me who only has a basic understanding. I'm probably going to be doing it more by feel that numbers, although I will check the numbers where I can.

I feel you on all of this.
 
Bus noise should not be an issue with modern low noise summing amps.

There is a nominal insertion loss, or noise gain of N+1 but this too is not problematic.

Some simple pan circuits also introduce insertion loss (10dB for a popular topology).

Last century I explored lower noise bus topologies, but IMO not worth the trouble today. If you want truly noise-less combining use digital.

JR
 
I wish my noise/headroom issues was to do with a Helios mixer.  Some of the Helios mic preamps/eq's I built had headroom/noise issues. 36V made an excellent mic preamp if kept at lower gain.  At higher gain settings noise became noticeable.  I too used 2128 and 22113 cards (I should redo the 22113 as i could have made it much smaller). Input is a Jensen (or Cinemag) 1:10 input transformer, while for the output I used a Cinemag quadlifier, like used in the API type builds.
The ones I used 24V with had no noise issues but not a ton of headroom either. That said they are still good mic preamps and have a distinct sound that I definitely like.  Recording guitars through U87 and the Helios with eq engaged, the mid and high range really place the guitar in that sweet spot that makes the work I put into them pay off.
I am still working on the 36V ones.
 
Rocinante said:
I wish my noise/headroom issues was to do with a Helios mixer.  Some of the Helios mic preamps/eq's I built had headroom/noise issues. 36V made an excellent mic preamp if kept at lower gain.  At higher gain settings noise became noticeable.  I too used 2128 and 22113 cards (I should redo the 22113 as i could have made it much smaller). Input is a Jensen (or Cinemag) 1:10 input transformer, while for the output I used a Cinemag quadlifier, like used in the API type builds.
The ones I used 24V with had no noise issues but not a ton of headroom either. That said they are still good mic preamps and have a distinct sound that I definitely like.  Recording guitars through U87 and the Helios with eq engaged, the mid and high range really place the guitar in that sweet spot that makes the work I put into them pay off.
I am still working on the 36V ones.
There is little headroom improvement in going from 24V to 36V (just 3.5dB). If you run internally at -8dBu your headroom will be fine on a 24V system. Noise always increases with gain. At high gains, layout, screening, general construction  decoupling and wiring are are as important as design in determining noise. There is no reason why those old designs should not comfortably exceed and EIN of --120dBu.

Cheers

ian
 
Thanks Ian. Yeah my layout for the 22113 is amateur at best. The wiring is a mess too. I am going to redesign the pcb's. I wish pahstah hadn't disappeared.
 
If anyone was considering building a Helios type pre/eq from scratch, I would probably recommend that they use the schematic that David Hinson posted somewhere recently on this site as a basis, and just change the voltage supply for the 22113  line amp on that schematic from 24V to (about) 36V (which I think was about the next Swettenham iteration anyway - although certainly not the final). I think that a 680p capacitor was used on that schematic in a LPF? I think 68p was typically used previously? Not sure if that was a small but significant design change. If building one of these I would personally never exclude having the 2128 (or equivalent) upfront. 

Such a Helios module would be pretty good without using an output transformer.

I was a bit obsessed with converting the Helios mic pre to include output transformers. Less so now, except perhaps for the sends, monitors and main outputs (if building a console that is). I have built Helios-type mic pre modules without transformers on the direct outputs. Plenty good. 
 
I'll add the caveat that my above-suggested approach would improve the headroom available over the earlier 2128-mic preamp to 2128 line driver (or equivalents including -ve voltage supply equivalents) designs, which in their defense were themselves used to make a lot of well-appreciated records. What I mean is that I think the 2128 - 22113 approach may be better for a stand-alone module (i.e. to be used outside of the console architecture). Anyway, this has been discussed ad nauseam.
 
Yeah I don't want to drag on another Helios thread either.  I believe that my resistor values for the gain stage may be at fault.  I did have a 36v Helios that worked fine. I am sure I f'ed up somewhere in the others.
 
ruffrecords said:
For maximum headroom you would normally aim to bias the output transistor to close to half the rail voltage
You would do that for a symmetrical output stage, where both sides are capable of delivering enough current to the load.
In a typical Single Ended stage, one side (the active device) is capable of swinging close to the rail, but the other side, usually a simple resistor can swing only to a fraction of the other rail, by consequence of the voltage divider effect. So typically, a SE stage will be biased such that the idle voltage is closer to the rail that is attached to the active device. For a single-ended/single rail NPN stage, the collector voltage will be at about 1/3rd of the rail voltage.
The half-voltage rule of thumb is true only when the load is much lighter than the collector/anode/drain resistor.
Similar reasoning applies to Constant-Current load.
 
Electrobumps said:
1073 pre I will often crank the gain on a snare for some crunch and back the output right back as not to clip my DAW,  "technically" on paper stupid.  sonically this can sound amazing.    Do this on an SSL mic pre, it will sound like a disaster.

In rock records I always did the same thing with our SSL 4K G series Mic Pre,
we had quite a few Neve preamps at the studio, But on snare and bass if I wanted some crunch (saturation/clipping) I much prefer the SSL mic pre than the Neve (1073).

So a "Disaster" for one person is the "Grace" for another
 
Back
Top