Hey guys. Finally, I got around to assembling my own console. I already have ideas for channels and circuits for them, but the most important element is missing - the summing circuit. A mixer without a summing circuit cannot be a mixer. Yes, I certainly know many passive circuits of the simplest summing circuits based on potentiometers and 5 resistors. But this is not what I need. I wanna get vintage warm and flash sound and I am looking for active or passive summing circuits from SSL, API, Neve, Neumann, Studer. If you have such circuits or know where to find them - please tell me. Best wishes)
or can you tell me how the adder works, not necessarily the actual diagram, the main thing is to understand the essence and create something of your own based on it. I would be very grateful!
I've also heard about such a scheme:
Each channel has a neve 1272 or Neumann preamp at the input of the combiner, and then a fader and resistor based circuit that combines all channels into one signal. The idea as to me is not bad, but only for the section of groups, that is the sum of 4 groups in one channel or master channel, and then stereo isolation and output amplifier. But I need a summarizer for the channels as well. To be able to select the group to which the summarized signal will be sent. What can you say about it?
Each channel has a neve 1272 or Neumann preamp at the input of the combiner, and then a fader and resistor based circuit that combines all channels into one signal. The idea as to me is not bad, but only for the section of groups, that is the sum of 4 groups in one channel or master channel, and then stereo isolation and output amplifier. But I need a summarizer for the channels as well. To be able to select the group to which the summarized signal will be sent. What can you say about it?
So you're trying to build a summing mixer with active inputs? That's not too hard but alot of the flavor these consoles add is created mainly by the EQ circuits, input gain stages and coupling transformers. Not much is added in the summing circuitry. Thats typically done with a traditional inverting current summing circuit that mirrors the individual input resistance to the output.
You can also do passive summing, which doesn't add any noise. You can also use NE5532s in unity gain to build an inverting current summing stage with an inverting buffer behind it to fix the phase.
I have some older schematics for Neve, Trident and API stuff which I'll check. I used to service these many years ago when I lived back east. Now I just re-cap them and update some of the gain stages.
You can also do passive summing, which doesn't add any noise. You can also use NE5532s in unity gain to build an inverting current summing stage with an inverting buffer behind it to fix the phase.
I have some older schematics for Neve, Trident and API stuff which I'll check. I used to service these many years ago when I lived back east. Now I just re-cap them and update some of the gain stages.
I am planning to assemble a mixer with a large set of different devices on board. At the input, as a preamplifier, there will be a redesigned Neva 1081 circuit, after which there will be an 1176 compressor, then an equalizer again, another compressor, a reverb, a saturator, and then the signal will go to the sum and into the groups. I have already thought out a summing system for the groups. I will simply put 4 summers connected to all 16 channels. And on each channel, you can choose which specific summer the signal will go to, and it is logical that it will also be selected which group the signal will go to. If you have Neva and API summer circuits, can you share them? I would like to look at their circuitry and assemble my circuit based on their summation. I do not know yet whether to make an active or passive mixer. I want to get closer to the sound of the 9000 SSL and Neva consoles to completely color the sound and make a cool sound. But for now without lamps, because I haven’t worked with high voltage yet and I’m a little worried about such circuits.
Hi, main resource for this stuff is groupdiy.com.
If I remember API has been using virtual ground summing, while at least the old Neve is passive sum type. I'd do it with just regular ne5532 or modern equivalent, virtual ground, just because I'd want lots of buses and it would get expensive fast 🙂
Do what you dream about, it's a huge project, financially a drain, and likely takes some years to accomplish. It's great fun though, so go ahead do good background study, perhaps some protos to see whats the sound, hownmuch work there is and how much equipment you need in your workshop and so on, before investing huge amounts of money into the parts.
Google "steve dove console design pdf" and read that few times over at first, to get good overall glimpse.
If I remember API has been using virtual ground summing, while at least the old Neve is passive sum type. I'd do it with just regular ne5532 or modern equivalent, virtual ground, just because I'd want lots of buses and it would get expensive fast 🙂
Do what you dream about, it's a huge project, financially a drain, and likely takes some years to accomplish. It's great fun though, so go ahead do good background study, perhaps some protos to see whats the sound, hownmuch work there is and how much equipment you need in your workshop and so on, before investing huge amounts of money into the parts.
Google "steve dove console design pdf" and read that few times over at first, to get good overall glimpse.
Last edited:
maybe I'll make a passive adder section. I also wouldn't mind ready-made adder circuits on and without microcircuits, if you have them) The main thing is that the console doesn't cost more than $1000-1200. The circuit will still be greatly changed and in most places I'll replace transformers and microcircuits with more modern ones or remove them altogether (I'll only remove transformers, because in my country there aren't any necessary ones at all or they're no longer being produced). I'm planning to first release a home compact version of such a console for 8 channels of hand-made production and probably make such ones to order on my website. I hope by June I'll be able to present something interesting and analog to the world. By the way, what devices do you think can be implemented in the console to make a cool recording with preliminary processing?
To make cool recordings you'd likely want whats in SSL console 😀
Well, most of work is handled by experience of the engineer, console is utility. I'm hobbyist though, so do not take my word. I find it very cool to have hands on faders, pots, especially doing live sound, but these really aren't necessary today. Summing bus is a product I never understood, summing in box is fine, I can patch in a conpresor if I want, but compressors in box are fine too.
Mixing and recording in box is the game of today. What you really need is fine mic pre amps, collection of mics, good stands and cables and keep them well maintained, and your quality is not depending on the gear. If you want to expand on this a monitor section and a controller with bunch of faders and pots which control a DAW.
If you want to make an analog board, make sure it has all utilities you need. I'd start from monitor section and metering, then preamps. Not sure how you want to work, whats your workflow with it, but likely you'd need at least few sum buses for monitoring, then couple for effects, and perhaps groups for processing, VCAs could come handy, and it approaches an SSL, and soon it will be as good as a DAW, if you add automation to it 😀
Sorry I do understand your dream as I have had it too, I just want to give a realistic perspective on it. However, if you have capacity to pursuit your dream you should do it. Start from what has biggest utility value to you so you can start using it right away, and expand fron there. I hope you get lots of fun time withvthe project and the end product!🙂
Well, most of work is handled by experience of the engineer, console is utility. I'm hobbyist though, so do not take my word. I find it very cool to have hands on faders, pots, especially doing live sound, but these really aren't necessary today. Summing bus is a product I never understood, summing in box is fine, I can patch in a conpresor if I want, but compressors in box are fine too.
Mixing and recording in box is the game of today. What you really need is fine mic pre amps, collection of mics, good stands and cables and keep them well maintained, and your quality is not depending on the gear. If you want to expand on this a monitor section and a controller with bunch of faders and pots which control a DAW.
If you want to make an analog board, make sure it has all utilities you need. I'd start from monitor section and metering, then preamps. Not sure how you want to work, whats your workflow with it, but likely you'd need at least few sum buses for monitoring, then couple for effects, and perhaps groups for processing, VCAs could come handy, and it approaches an SSL, and soon it will be as good as a DAW, if you add automation to it 😀
Sorry I do understand your dream as I have had it too, I just want to give a realistic perspective on it. However, if you have capacity to pursuit your dream you should do it. Start from what has biggest utility value to you so you can start using it right away, and expand fron there. I hope you get lots of fun time withvthe project and the end product!🙂
SSL are ( were) the ultimate mixing tool. I never really enjoyed tracking on them (4000), that's why most studio's had a A room (tracking rooom) with Neve or Ams/Neve or Api and a B room ( mixing room) with SSL.
Anyway, Matthewmathews you are expecting things which are way different than what you think: SSL 9000 are totally different beast than older Neve, Api or even SSL 4000E/J.
9000 have an analog sound but it's really closer to digital than anything else ( read very very clean), the reason it's craved by new school R&B and hip hop artists, not much by other genres.
4000 (E/G) and AMS/Neve (VR, VR legend, VX) are a bit different: they have more 'character' on their own. Vseries have more low end, sound a bit more polished to me. 4000G are closer to V but like the E they are way more mid focused and a bit metallic sounding. Which makes them great for pop, rock, metal and old school hip hop. Ams/Neve are great for things loaded with lows imho.
What i really miss from those desks are the eq, which all have their own things to add...
I suggest you take an hear to Brainworks digital channel recreation i find them to sound close to behavior of the real ones, at least they give an idea of the overall sound of each consoles ( including Focusrite and Amek consoles too).
Older Neve and Api are different sounding. Those have the most character because they use transformers all along the signal path ( which is not the case for previous consoles): mic input, at chanel ouput, summing amplifier input and ouput ( for Neve-1272-, Api used only on output) and at line ouput stage ( main mix but not only, sends, aux,...).
This is what 'makes their sound' mostly. Doa on Api too which are parts of the sound.
Summing is done differently on this older ones: passive rather than active. The most extreme being Focusrite where Rupert Neve decided a max of 8 channel by summing stage then another one to sum the previous summer...
If you want this kind of sound forget about active and use passive summing into full mic preamp. This have gone a bit outfashioned but 15 years ago it was all the rage. I still use this approach.
DA to passive summing to mic preamp. As i already tracked through the same mic preamp the color adds up... i use some 1290 preamp 'clones'- Neve preamp from same serie as 1073-, and some Api style pre -different from 312 but with original transfo and DOA) i have older RCA desk summing amplifier too ( Ba21a, tube based)....
I'm awaiting impatiently recreation of Focusrite ouput tranfsormer to add this option too... or maybe a more Amek circuit, i'm a fan of 9098 sound ( i'm a Rupert Neve fanboy). A lot of option possible.
For summing i use NYDave's circuit ( one of the old Groupdiy hero designer) which you can find here:
https://www.diyrecordingequipment.com/blogs/news/15851740-how-to-build-a-diy-passive-summing-box
Works great and enable use of daw (and plug ins) and hardware seamlessly which i like ( as much as i like hardware, digital tools including comp and eq i can't live without).
I will link a Fred Forsell article about the two different summing approach and the pro and cons. Steve Dove series of article Tmuikku talked about is worth a read too. And as him i suggest you to join Gdiy too.
Last word, i totally get your dream but thats a HUGE amount of work. Only doing the independant box is a bit of work... integrating them into a console is even more as you'll have to deal with ground issues, logic control, etc,etc,...
It's not impossible but expect a fulltime job and a huge amount of money to spend... with your share of frustration if you go that way. 😉
Anyway, Matthewmathews you are expecting things which are way different than what you think: SSL 9000 are totally different beast than older Neve, Api or even SSL 4000E/J.
9000 have an analog sound but it's really closer to digital than anything else ( read very very clean), the reason it's craved by new school R&B and hip hop artists, not much by other genres.
4000 (E/G) and AMS/Neve (VR, VR legend, VX) are a bit different: they have more 'character' on their own. Vseries have more low end, sound a bit more polished to me. 4000G are closer to V but like the E they are way more mid focused and a bit metallic sounding. Which makes them great for pop, rock, metal and old school hip hop. Ams/Neve are great for things loaded with lows imho.
What i really miss from those desks are the eq, which all have their own things to add...
I suggest you take an hear to Brainworks digital channel recreation i find them to sound close to behavior of the real ones, at least they give an idea of the overall sound of each consoles ( including Focusrite and Amek consoles too).
Older Neve and Api are different sounding. Those have the most character because they use transformers all along the signal path ( which is not the case for previous consoles): mic input, at chanel ouput, summing amplifier input and ouput ( for Neve-1272-, Api used only on output) and at line ouput stage ( main mix but not only, sends, aux,...).
This is what 'makes their sound' mostly. Doa on Api too which are parts of the sound.
Summing is done differently on this older ones: passive rather than active. The most extreme being Focusrite where Rupert Neve decided a max of 8 channel by summing stage then another one to sum the previous summer...
If you want this kind of sound forget about active and use passive summing into full mic preamp. This have gone a bit outfashioned but 15 years ago it was all the rage. I still use this approach.
DA to passive summing to mic preamp. As i already tracked through the same mic preamp the color adds up... i use some 1290 preamp 'clones'- Neve preamp from same serie as 1073-, and some Api style pre -different from 312 but with original transfo and DOA) i have older RCA desk summing amplifier too ( Ba21a, tube based)....
I'm awaiting impatiently recreation of Focusrite ouput tranfsormer to add this option too... or maybe a more Amek circuit, i'm a fan of 9098 sound ( i'm a Rupert Neve fanboy). A lot of option possible.
For summing i use NYDave's circuit ( one of the old Groupdiy hero designer) which you can find here:
https://www.diyrecordingequipment.com/blogs/news/15851740-how-to-build-a-diy-passive-summing-box
Works great and enable use of daw (and plug ins) and hardware seamlessly which i like ( as much as i like hardware, digital tools including comp and eq i can't live without).
I will link a Fred Forsell article about the two different summing approach and the pro and cons. Steve Dove series of article Tmuikku talked about is worth a read too. And as him i suggest you to join Gdiy too.
Last word, i totally get your dream but thats a HUGE amount of work. Only doing the independant box is a bit of work... integrating them into a console is even more as you'll have to deal with ground issues, logic control, etc,etc,...
It's not impossible but expect a fulltime job and a huge amount of money to spend... with your share of frustration if you go that way. 😉
Last edited:
Why two eq in series? And if you redesign a 1081 it's not one anymore. 😉 The sound 'is in the iron' anyway ( transformers), so you'll have to find convincing clones, originals you'll have to sell a kidney ( or both and arms and legs!)..
You want to implement reverb? Which kind? A digital one is going to be some serious work ( lifetime work)!
hmmm. This is going to be complex to implement.
Study the Forsell paper.
Then you'll have to study Neve 80XX consoles schematics. They are availlable if you search... i really suggest again to join Groupdiy, it will ease the search ( go to 'technical documents' section once you have joined...).
For Api, take a look at CAPI products... Jsteiger already done the work. I should have the basic circuit principle availlable too, let me take a look into my archive...
Tubes? Not used on any SSL ( SSL stand for Solid State Logic) and except for it's first small scale mixer Rupert Neve never used tubes either in any design.
Last edited:
Api schematics, here we go.
Look for 'lossless combination amplifier' schem.
If you want the whole 'bus' thingy you'll need a pair of 2603 transfo on 'booster' cards ( see CAPI) and a pair of 2503 transformers to drive outside world ( in fact you put a 2603 on combining amplifier output, a fader ( 600ohm constant load) and then schem 2 and you have a 'booster' card...). 2520 DOA too ( many clones availlable, Scott Liebers 'red dots' are the undistinguishable from vintage originals, 'blue dots' modern variants as real API ones on offer).
Look for 'lossless combination amplifier' schem.
If you want the whole 'bus' thingy you'll need a pair of 2603 transfo on 'booster' cards ( see CAPI) and a pair of 2503 transformers to drive outside world ( in fact you put a 2603 on combining amplifier output, a fader ( 600ohm constant load) and then schem 2 and you have a 'booster' card...). 2520 DOA too ( many clones availlable, Scott Liebers 'red dots' are the undistinguishable from vintage originals, 'blue dots' modern variants as real API ones on offer).
Attachments
Whether you use passive or VE summing is a little academic TBH. You can get a slide rule out and calculate which is quietest for a given number of inputs, but you'll be looking at fractions of a dB difference between the two once you factor in the make-up amp for the passive mixer. A nice thing about VE summing is that you can disconnect nodes for best SNR when not needed, whereas if you do that with passive you'll need to change gain on the make up amp. Passive will potentially add more buffers if you want to do it properly. It's a misconception that passive has a lower parts count, as this only applies if you do it in an amateur fashion. Remember that VE summing can appear very similar to passive if you aren't looking at the whole schematic... I suspect people mistake VE stages as passive, as they don't realise that the amps invert.
Rupert left the company in the mid 70s, and a lot of the famous consoles were designed by a team of other designers. His famous diode-bridge limiter was actually designed by his cousin (as an example). The V series were called so as a pun, on account of being designed by David Pope (V = Vatican, geddit?). Rupert had a lot of interesting philosophies, but he was not an academic. He worked with a team of gifted designers and they often fail to get the credit they ought to have. There are plenty of retired engineers out there who can back this up.
The famous 1073 is based around BC184C and 2N3055. The tricky parts are the transformers, as the firm (Marinair) went under in the 70s. Carnhill make replicas, but people argue as to how authentic they are.
The Neve VR is almost entirely based on NE5534 ICs. It was the company's answer to SSL. Rupert had long left the firm when it was designed. The last of the iconic 8000 series, the 8078, utilised NE5534s for the output stages. The Montserrat Neves that were made to order for George Martin had copious amounts of NE5534 in them.
NB - the last of the really big SSLs (before they went SMD) used really clean silicon such as the SSM2017 (IIRC). Not easy to find. A lot of studios prefer the E Series simply because it's easier to keep on the road and has less unobtainium ICs.
oh yeah - do study the Forrsell doc. It nails things nicely.
Rupert left the company in the mid 70s, and a lot of the famous consoles were designed by a team of other designers. His famous diode-bridge limiter was actually designed by his cousin (as an example). The V series were called so as a pun, on account of being designed by David Pope (V = Vatican, geddit?). Rupert had a lot of interesting philosophies, but he was not an academic. He worked with a team of gifted designers and they often fail to get the credit they ought to have. There are plenty of retired engineers out there who can back this up.
The famous 1073 is based around BC184C and 2N3055. The tricky parts are the transformers, as the firm (Marinair) went under in the 70s. Carnhill make replicas, but people argue as to how authentic they are.
The Neve VR is almost entirely based on NE5534 ICs. It was the company's answer to SSL. Rupert had long left the firm when it was designed. The last of the iconic 8000 series, the 8078, utilised NE5534s for the output stages. The Montserrat Neves that were made to order for George Martin had copious amounts of NE5534 in them.
NB - the last of the really big SSLs (before they went SMD) used really clean silicon such as the SSM2017 (IIRC). Not easy to find. A lot of studios prefer the E Series simply because it's easier to keep on the road and has less unobtainium ICs.
oh yeah - do study the Forrsell doc. It nails things nicely.
Thank you for your opinion. I realise that the world today is a studio in a box so to speak, but that's not close to me. I have 9 years experience in sound, and almost always with big consoles and analogue equipment. Plus I got my engineering degree just in time and I want to fulfil myself not only as a mix engineer but also build something of my own. My goal is to make analogue accessible, because it's alive. Yes, this topic is still very heatedly discussed on forums and on the internet in general, but as a person who has worked with consoles and hardware for several years, I want to make analogue equipment accessible and at least not occupy huge rooms without using digital technologies. My idea is to take the console, the most famous hardware that are really useful in recording and mixing, and combine them into one (so far small for 4-8 channels, but still) analogue device that will please musicians and engineers with its efficiency in recording and possibly mixing. For now, of course, I'm only focused on cool recording functions and devices to improve recording, as I understand that without tube technology in the form of the same equalisers from Pultec, you may not get the information you want to get. In short, I want to make a recording in a box only without digital technology. Most likely I will only make an automation circuit to make the console remember the values of potentiometers and can restore their position to simplify work and save projects, but this is still in the future. I will make my first independent console completely analogue!
Montserrat used 5534 with discreet buffer to increase current in order to drive the 'pumpkin' transformers ( toroidal).
All Rupert following designs ( Focusrite, Amek, RND) followed this approach introducing a tertiary feedback winding on the toroidal.
I did a bit of archive study about it some month ago. Very interesting to see he eventually decided to use transformer's core saturation as an effect on 'Portico's' design...
And how he makes each generation of design an evolution over the previous.
You are absolutely right about the the engineering crews surrounding him. Very talented people.
And yes AMS/Neve are totally different design and designers. Pope did a great job about the V series imho, routing capability were beyond any other consoles of their time despite way more complex to understand ( DAW made all this caduque but with a Reel to Reel it did make a real difference to other brands and why they were used in tracking room mainly). Eq are great on them. Most US people dislike them, we mostly prefer them in EU.
All Rupert following designs ( Focusrite, Amek, RND) followed this approach introducing a tertiary feedback winding on the toroidal.
I did a bit of archive study about it some month ago. Very interesting to see he eventually decided to use transformer's core saturation as an effect on 'Portico's' design...
And how he makes each generation of design an evolution over the previous.
You are absolutely right about the the engineering crews surrounding him. Very talented people.
And yes AMS/Neve are totally different design and designers. Pope did a great job about the V series imho, routing capability were beyond any other consoles of their time despite way more complex to understand ( DAW made all this caduque but with a Reel to Reel it did make a real difference to other brands and why they were used in tracking room mainly). Eq are great on them. Most US people dislike them, we mostly prefer them in EU.
Last edited:
UPD: In my opinion, the recording process with my console should look like this: First comes the preamplifier-harmoniser, then the signal goes through the console, through the channel, the group, then the master, out to the tape recorder. Then from the tape recorder I play back the tracks and bring them to another tape recorder. After I put the track together I transfer it from the reel-to-reel tape recorder to digital where I do mastering and store and publish it. I would like to provide the users of my consoles with the same thing. But at the same time I don't want to put them in a clear analogue framework. Nobody forbids to put a sound card or DAC/ADC instead of a tape recorder and mix in a box. The most important thing is for the console to become the main and indispensable tool in recording and mixing music, and where to write and where to store it is the user's right.
From what I've heard Neve and the older Api's are more suited to recording, and SSL is the benchmark for mixing and mastering. Neves sound very cool, but according to most recording engineers they are not really suitable for mixing. From my console I want a symbiosis of SSL console and old generation Neve and Api consoles. That's why I will use their schemes when designing mine.
You just redifined inline principle in post 16. 😉
This is how it's done for 40 years with 'big' studio consoles ( not only big... Mackie 8 buss was inline too ( as Behringer's clone of the time).
This is how it's done for 40 years with 'big' studio consoles ( not only big... Mackie 8 buss was inline too ( as Behringer's clone of the time).
Last edited:
Nobody does mastering on big inlines: you rarely need more than 16 chanels ( and only working with stems) and we need 'better' speced gear ( let's see how you'll deal with the psu and grounding issues. 🙂 ).. That said we use parts of the 4000 consoles routinely ( the masterbus comp) but the rack version..
Ams/Neve (V series) are way more complex to operate because of their flexibility about routing and their automation is imho a pita to operate ( Flying Faders and Encore) wrt SSL automation which are way simplier in term of routing, that's all and why most freelance prefer SSL. About sound i prefer AMS/Neve and never had an issue working on them either tracking or mix ( but i was paid to teach students how to use them so it might explain things, 😉 ).
Classic Neve design (80XX) are split design ( or split monitors) and easy to use either in mix or tracking, you could do live p.a. with them no issue.
I'm probably going to use passive summing, since I haven't quite figured out active summing. About the transformers. They are not really available to me or cost more than I can spend on them, so most likely I will replace them with matching circuits built on op-amps or capacitors, so it will be cheaper and more practical. I am not against the use of transformers in circuits and will even be happy to use them, but again they are not available to me to purchase, and schemes and drawings of such transformers I have not found to collect such transformers myself. According to my rough calculations so far, there are 11 transformers of different characteristics per channel. One Neve 1081 input preamplifier has 2, my adapted 1176 has 2 transformers, etc. So the transformer issue is something I haven't decided on yet. I will probably go for the option of replacing them as it is more economical.