Designing a mixer: the Formula Sound FF6.2 (by Barry Penaligon)

[Podéis encontrar la versión en castellano de este artículo es este enlace/You can find the Spanish version of this article following this link: https://zitroztr.wordpress.com/2017/06/27/formula-sound-ff6-2-historia-del-diseno-de-uno-de-los-grandes-mezcladores-analogicos-de-nuestra-epoca-cortesia-de-barry-penaligon ]

Hi all.

I am pleased to publish this second article written by the great Barry Penaligon. Except this little intro, it is NOT written by me. I would like to show how much I appreciate this excellent material provided directly from the most reliable source. Hope you can enjoy this interesting information (probably you will end loving this mixer), and I wish I can provide even more in the future.

If you are interested in the history of Formula Sound, you can take a look on this article: https://zitroztr.wordpress.com/2017/06/09/a-brief-history-of-formula-sound-by-barry-penaligon (this version is in Spanish: https://zitroztr.wordpress.com/2017/06/09/historia-de-formula-sound-cortesia-de-barry-penaligon ), in this very same blog.

A big THANKS, Barry.

Designing a mixer: the Formula Sound FF6.2

FF6.2L Front Persp4

The FF6.2 ( http://www.funktion-one.com/products/ff62-dj-mixer/ ) started following a discussion with Funktion one, we were talking about the FF6000 and mentioned that Formula Sound were working on a new mixer. There was a suggestion about doing a follow on from the FF6000 ( http://www.funktion-one.com/products/ff6000-dj-mixer/ ) and so a discussion started on what a new Formula Sound/Funktion one mixer should look like.

Mike Igglesden took the lead for Funktion one, I produced some initial drawings and we had a couple of meetings at Funktion one, David Bruml and John Newsham of Funktion one joined in discussions.

From the outset it was to be a follow on from the FF6000 so a table top mixer with the connections at the top edge as the FF6000. We discussed what we liked and did not like about the FF6000 and concluded the effects loops would go.

 

FF-6000

The well-known FF-6000, predecessor of FF6.2

 

I had a vision of a mixer that could be used for creating music (home studio) and be used for live performance so a compressor in channels 1 & 2 and AUX facility. I felt the compressor could be useful both home mixing and second/third microphone.

Funktion One thought a 4 band EQ tailored to their speakers would be a good idea also we had previously showed them a variable frequency filter we had done some work on and Mike thought variable frequency high and low pass filters would be a great feature. I felt a 4 band EQ would need more height than the 7RU of the FF6000 and thinking of the SYS2000 I decided it would need to be 8RU.

We had done a special full kill EQ for a few FF6000 customers and Funktion one liked its performance so we agreed the 4 Band EQ would be based on this.

We felt the console mic was good but could benefit from 3 band EQ, otherwise make the output section similar to the FF6000 without the effects loops. Mike thought a full 12 LED channel meter would be good and phones monitoring that was switchable pre/post EQ.

So I did a drawing of a 8RU front panel, channels with A/B input switch, gain control, 4 band EQ, 12 LED meter, variable frequency HF and LF filters with switch, CUE, 60mm faders and 2 AUX sends. Channel 1 & 2 with compressor switches and Channels 3 – 6 with A/B cross fader assignment.

The output consisted of Master, master insert switch, Booth, Booth balance, Booth mono switch, Fader curve AUX 1 & 2 level, with phones and console mic as FF6000 (console mic with 3 band EQ). Metering would be similar to FF6000.

I had the idea for some time that the Zone output should be remote controllable (as a Zoner would), the DJ does not want to be bothered for adjusting the sound level for a separate area so having a Zone control that could be mounted in the separate area seemed a good idea so there is no Zone volume control on the FF6.2.

There were a few iterations of what the front layout should be, after some discussion the AUX went from 1 rotary and 1 switch to 2 switches, initially ch1&2 had a mute switch but that was dropped.

We started with a connector panel that was very similar to FF6000, later we removed the RIAA/Line switch from the back so stereo channels were RIAA or line on the A/B switch, also the line input on channels 1 & 2 on phono connectors became balanced line inputs on jacks.

FF6.2L Connections panel 1

We discussed a removable fader panel as the FF6000, I felt the layout should be a bit more traditional with cross fader below the output section, also I felt we should move back to an anodised aluminium front panel so we settled on this with no removable fader panel. The intention at this stage was that the mixer would be purple.

This discussion took about 3 months but now we had the features of the mixer and I could start the design, some of the design was already set as mechanically the base and back panel would be similar to the FF6000, the front panel would be anodised aluminium with no removable fader panel but otherwise it would mount in a similar fashion as the FF6000. The PCB’s would be vertically mounted on the front panel (as all our mixers), the inputs would be pairs of PCB’s of which the left would carry the traditional input functions and the right would carry the variable frequency HF & LF filters plus the 12 LED channel display.

I had drawn the front panel with a mix of rectangular and round switch buttons, broadly in line with what we had done on the FF6000 (rectangular A/B input select, round CUE) and to give best aesthetic appearance (so HP/LP variable filter switch is round to differentiate from AUX and cross fade assign switches), the switches that take these buttons are LED switches (as FF6000) and originally we were to have 3 or 4 colours (usual red & green plus yellow and maybe blue) at the suggestion of Funcktion one. We produced some prototypes with the different colour LED switches are concluded (both Formula Sound and Funcktion one) that keeping to red and green was best.

 

While we were considering the outline of the mixer I also was thinking of how to get the best noise performance, the addition of the HP/LP variable frequency filters alone would push up the current consumption (more current and potentially more noise) so I concluded that it would need 4 different ground routings. My ground routings would be 1 for input/output connections, 1 for audio signals through the mixer, 1 for non audio signals and 1 for control circuitry (variable frequency filters and remote zone control).

I started designing the CH1&2 inputs, I saw no reason to change the use of ribbon cables to carry the signals and power between PCB’s, so I allocated signals to the ribbon cable and concluded that I would need a 34 way ribbon cable.

It quickly became clear that even with 8RU it would be difficult to design PCB’s with leaded (as opposed to surface mount) components in the space available, not wanting to compromise on using leaded components (because our products have always been serviceable) I considered how we might still use leaded components by increasing the density and concluded that plug in daughter boards would allow me to use leaded components and get it all on the PCB’s.

I sketched out the circuits I would need for channels 1 & 2, identifying where a new circuit would be needed and where a circuit would be re-used from elsewhere. The full kill EQ would be a 4 band version of a 3 band full kill EQ we used on a few FF6000 and FSM600 specials we have done and it would go on a daughterboard. Funktion one specified the frequencies and I calculated the values and laid out the EQ on a daughterboard.

I wanted to keep the microphones on a separate mic buss because I was not sure if the channel 1 & 2 mics should go in or be left out of the booth, I also needed separate pre and post EQ headphone monitoring. These two requirements meant switching mic/line signals and pre/post EQ headphone monitoring, both required similar switching but could not be accommodated within the existing switches which meant either relays or FET switching. Although we used relays in the FF6000 to do the loop swap I have never liked this approach so decided it would be FET switching. [Note: about loop swap, from FF6000’s manual: “LOOP SWAP – This switch allows the effects processors to be swapped e.g. if just one processor was inuse and connected to FX1 ( routed from X-FA) it could be used in loop 2 (routed from X-FB)just by depressing LOOP SWAP. The red and green l.e.d.s also change over as a reminder that Swap isselected.”]

In 2009 I revised the mic buss switching on the PM100 to use a FET switch so I decided to put two of these on a daughterboard for a stereo FET switch. This FET switch could be operated by the A/B input switch for mic/line and the pre/post EQ switch for headphone so fulfilling both requirements.

 

PM100

The PM-100 mixer in 8 and 4-channel configurations

 

The IC’s commercially available for LED meters have been going obsolete so I decided we would use a transistor based 12 LED meter on the inputs, this approach also has the merit that it uses a single current for all the LED’s in series rather than separate current for each LED so the current used is lower. The PM80R uses such a design so I re-used the PM80R meter circuit.

Originally all of the LED’s were to be 5x2mm rectangular, but Ian McDonald (Formula Sound) felt round LED’s looked better, we had used round LED’s on the PM80R. After some short discussion with Funcktion One they agreed and we settled on round LED’s (except for mute).

FF6.2L Detail1

The variable frequency high and low pass filters were supposed to be based on a circuit we had developed a few years previously however the device it was based on became unavailable so I had to start almost from scratch with a new design. The frequency of the filters is varied by varying a resistance, with a digital resistor the resistance is varied by digital control while allowing the signal to remain analogue so we use a digital resistor or digi pot to vary the frequency of the HF and LF filters while keeping the audio analogue throughout the whole signal chain.

The compressor is based on the compressor in the FSM600, although this one operates on both MIC and line input of channels 1 & 2. I decided that I would bring threshold and compression controls out of the base of the unit so the user could change the settings (as a set up – not in live use).

The rest of the input circuits are very similar to FF6000, the microphone circuit is identical.

The stereo channels (Channels 4 – 6) are based on channels 1 -2 with compressor removed, mic buss FET switch removed ,A/B assign added and Line/Mic replaced by RIAA/Line inputs. The RIAA phono input has been put on a module because I may use a different RIAA circuit in future.

With the inputs designed I turned to the outputs, the output consist of 4 PCB’s mounted vertically and a display PCB mounted horizontally on the back of the front panel (the FF6000 has 4 PCB’s + display).

The console mic and phones PCB is very similar to its FF6000 equivalent, the only real difference being 3 band MIC EQ instead of the 2 band on the FF6000. The pre/post EQ switch is just that, it switches the pre/post EQ FET switches in the input channels via the ribbon cable. This was the easiest to design because essentially it was a copy of the FF6000. I would just mention that in 2011 I changed the headphone drive in the FF6000 to a transistor based one (also changed the FSM600 and PM100) because the IC we used in the previous circuit changed its performance. Anyone with an early FF6000 can get a little PCB from Formula Sound that plugs into the IC socket if they feel the headphone output is not as good as it should be. The transistor based headphone drive is as good as any headphone output I have heard so I have used it again.

The balanced outputs on the FF6.2 are very slightly better than the FF6000, the common mode rejection is slightly better because its individually tuned on each balanced output. The Zone output is controlled in the same way as we remote control the output of the ZMR80 zoner, the gain of an op amp (not a VCA) is varied to adjust the volume using the remote panel.

The mic buss has a switch on the base of the unit so the user may route the channel 1 & 2 mics into the booth or out of it, again its only accessible on set up.

Most of the rest of the output PCB’s are circuits that were used in the FF6000, the X Fader curve control is the same as the FF6000.

The display was based on the FSM600 display, I am currently redesigning it using the transistor based 12 LED meter circuit I used in the inputs. The reason for originally using the FSM600 circuit (an IC) was space but I have managed to fit it onto the PCB however it does use some surface mount resistors.

FF62 Technical for Website

The power supply was supposed to be the same as the FF6000, but the first prototype showed the current consumption was too great so I had to design a new power supply PCB, not particularly difficult but the layout of the PSU can affect the noise floor however the new PSU pcb was fine.

The FF6.2 has 2 pcb’s for each input so twice as many input pcb’s as any of our previous mixers, also the input pcb’s have daughterboards. The output pcb’s also have daughterboards so the complexity of this mixer is far greater than any previous mixer we have done.

The first full prototype unit we built revealed a few mechanical issues which were easily resolvable and a number of electrical issues. Some of the electrical issues were simple and easy to fix, however there were 2 noise issues (at first we thought it was just one but it transpired it was 3 separate problems).

Initially we thought the noise was a single earthing issue and by diagnosing where the problem was arising we could easily fix it, our initial diagnosis suggested it was in the output so we removed the inputs and grounded the signal at the first output stage and quickly found the earth point for the output was in the wrong place which we fixed. However when we reconnected the inputs the noise re-appeared, after much investigation (a lot of it trial and error, low level noise is difficult to track) we established that the resistors that take the audio to the buss had tracks that were too long and the tracks were acting as aerials in picking up noise. We shortened the track and the noise improved in most modes but with RIAA phono input selected it was still too high, again we found the cause was a track that was too long in the phono stage and shortening that finally gave us the noise performance we were looking for.

While looking for the noise problem on the first prototype unit we found the mixer was getting hotter than we anticipated, too hot to be acceptable and the only solution was a pair of fans, one each side of the unit.

Much of the work on the protoypes was done by Ian McDonald of Formula Sound. The variable frequency filters and remote Zone control both use digital control of resistor values which enables us to keep the audio signal analogue while exerting the control of digital. Ian McDonald wrote the programs for both filter and remote panel PIC controllers.

While we were working on the prototype Funcktion one were working on the artwork for the mixer front panel, we also sent them a selection of knobs to choose from. Function one produced an artwork with a layout of legend that probably would not have occurred to us, placing some of the input labelling in a band next to the output simplified the text for the inputs and the font chosen gives the mixer a distinct look. Funcktion one gave us their choice of knobs (which was already one we liked) and that completed the look of the mixer.

FF6.2L Front 1

While Funcktion one were finalising the artwork we started discussing what the mixer would be called, we generally felt there should be some link to the FF6000, FF had originally been chosen to denote Formula Funcktion so it was already assumed it would have the FF prefix.

We talked about FF600, FF60 and I do not quite remember who suggested FF6.2 (my vague recollection is in a discussion on the name with Ian) but we all agreed FF6.2 felt right for this mixer.

Originally the front panel was to be anodised purple but we could not find a suitable purple so we decided traditional black.

The first production units were built and sent to Funktion one for evaluation and field trials, this threw up a few technical issues, a power supply issue and a noise issue that seemed to appear on some units. We made several changes to improve noise performance and each time the issue re-appeared, eventually we found the cableform could move during transport of the mixer and sometimes stopped close to the transformer which was cured by tying down the cableform which removed the problem.

The power supply problem was similarly confusing, initially we thought it was how the PSU started up and we made some changes which seemed to cure it but it would randomly re-appear. Eventually we discovered one of the power regulator had a limiting protection at a particular current and voltage so a slight reduction in voltage was sufficient to cure this final problem.

Funktion one field trials produced a lot of good feedback on the mixer, particularly with the variable frequency filters and the overall audio quality. We started shipping the FF6.2 in late 2016 and followed this with the FF4.2 and rotary versions of the FF6.2 and FF4.2 this month (June 2017).

The price of the FF6.2 was set after we designed the mixer, I had not been looking at cost while designing the mixer (concentrating on design and performance), when the design was finished I added up the cost of parts and build and was slightly shocked that it was so high. The selling price reflects the cost of building a mixer in England.

The FF6.2 looks a little different to previous Formula Sound mixers but look inside at the analogue signal chain, vertically mounted PCBs using leaded components and you can see it is recognisably a Formula Sound mixer. When you listen to the audio quality and use it for the first time it has everything you expect from a Formula Sound mixer.

How does the FF6.2 compare with previous Formula Sound mixers ?, the audio quality on test equipment is very slightly better. The EQ is a full kill EQ, all our previous mixers (excluding custom FSM600 and FF6000) have had a softer EQ and so it will sound a little different when using the EQ. Some may not like 4 band EQ although I suspect anyone who likes the SYS2000 will like the EQ on the FF6.2. The variable frequency HP/LP filters work extremely well, as usual it is a matter of taste as to whether you want to use that feature or not.

 

DSC02300

The widely known System 2000 in a customized 7-channel configuration (not made by Formula Sound)

 

At present the phono circuit is the same as most of our previous mixers all the way back to the PM80, although I have a new design that I think may perform very slightly better. People may say the PM90 sounds slightly different, that will be because the PM90 includes a rumble filter.

The FF6.2 is designed and built to last as long as any of our previous mixers, we have had to use a few surface mount components in the FF6.2 but we have confined to locations where there is no choice (because of space or component availability) so the mixer is still serviceable and maintainable.

Having finished the FF6.2 my thoughts turn to what next, I feel there is still some room for improvement in analogue mixer design, I would like to do a mixer that has the best performance with vinyl and I have some ideas on how to achieve that. I have some ideas for a slightly better microphone circuit and slightly better output circuit, both improvements would be small and difficult to notice the improvement but I believe in trying to make mixer performance as good as it possibly can be (usually as measured by test equipment).

The FF6.2 is the mixer it is because of the collaboration between Formula Sound and Funktion one, the electrical design is all Formula Sound but the appearance, 4 band EQ and variable frequency filters are because of Function one and in particular Mike Igglesden. Together we have produced a superb mixer with a great sound.

P.S. ADDED BY ME:

I’d need to find out if the production models of FF6.2 finally include the windowed LED pushbuttons or the transparent ones, recently incorporated to PM-100 and manufactured (the clear cap) by Formula Sound themselves. All the photos included in this article have been extracted from Thomann UK, as far as I remember, but I think I’ve seen somewhere pictures with those clear transparent buttons. I’m talking about these ones, which my new PM-100 incorporated:

Formula Sound PM-100 - LEDs-2 (1920x1280)

Anuncios

Responder

Introduce tus datos o haz clic en un icono para iniciar sesión:

Logo de WordPress.com

Estás comentando usando tu cuenta de WordPress.com. Cerrar sesión / Cambiar )

Imagen de Twitter

Estás comentando usando tu cuenta de Twitter. Cerrar sesión / Cambiar )

Foto de Facebook

Estás comentando usando tu cuenta de Facebook. Cerrar sesión / Cambiar )

Google+ photo

Estás comentando usando tu cuenta de Google+. Cerrar sesión / Cambiar )

Conectando a %s