Monday 18 December 2017

Fat ones, thin ones, tall ones, short ones - Module Mechanics Revisited

At the start of this blog I mentioned the reasons for using a module enclosure, particularly the mechanical integrity and screening properties and also that the Fischer modules were reasonably priced. Except they are no longer so cheap. SInce I started using them the price has gone up by 40 percent. In addition, after a couple of years experience using them, their shortcomings are beginning to become noticeable.

Tube modules tend to be be quite large to make enough room for the tubes. In turn, this means tube mixers tend to be pretty big (Holgers 12 channel one is one metre wide). To put this in perspective, the modules are nearly twice as fat as 500 series modules. So I have been spending some time looking at ways to squeeze more functionality into a given space and also how to squeeze existing functionality into a smaller space. Put another way, I can get just six tube modules into a 19 inch rack space but you can get eleven 500 series modules in the same space.

The first option I looked at was stereo or two channel modules. These are the same width as the existing modules and fully compatible with the standard mechanics and motherboard interface, but just have two channels of electronics in them rather than one (these were mentioned earlier in the Mark 3 6U modules discussion). They are line level only modules so the active electronics is essentially the Twin Line Amp design. Many people now seen to be moving to line level only mixers and use external preamps for tracking either direct or via the line level mixer. This basically splits the functions of a classic mixer in two. From my point of view, the advantage is I can now get twelve channels in a 19 inch rack space instead of just six. The 8 tracker becomes a 16 tracker overnight.

However, the Fischer modules are beginning to become a limiting factor. The type T 6U module I use places the top surface of the PCB 14.2mm from the left hand edge of the front panel. This relatively large offset makes front panel design more difficult for a two channel module. The controls of the left half are 14.2mm plus their height from the left edge of the front panel. For the twin channel module to be symmetrical about the centre line, the right half controls must be the same distance from the right side of the front panel. So the controls are pushed 14,2 x 2 0 28.4mm towards the centre making the centre look cramped and the edges bare. Maybe a different mechanical scheme could reduce the 14.2mm offset sufficiently to fis these limitations.

These modules operate at line levels so screening is not so critical as it is for microphone preamplifiers. Maybe a simpler mechanical scheme can be used that will be more cost effective?

The second option I looked at was a making modules half the existing width, that is 1.4 inches wide instead of 2.8 inches. This means the tubes cannot be mounted directly on the PCB as they are at present. Instead they need to be mounted vertically which means they use up more PCB space. It turns out there is probably not enough space to house the three tubes used in the Eurochannel mic pre or the Twin Line Amp plus the input transformer and all the required passive components, but there is room for the two tubes used in the Classic mic pre design. So this was the first PCB I laid out after first designing an adaptor PCB to allow the tubes to be mounted vertically.

At this point it became clear that 1.4 inch wide Fischer modules, although available, were not going to have enough room. Most models of Fischer modules are quite wasteful of width. For example, the T types used for 6U modules mentioned earlier, where the top surface of the PCB 14.2mm from the left hand edge of the front panel. In a 1.4 inch wide module we only have just over 35mm to play with. The tubes are 22mm in diameter and if we lose another 14.2mm due to the module type we have already used more than 36mm. There was only one Fischer module that wasted much less width. This was Design I illustrated below:


This version only loses 6mm leaving an available space of 29.56mm above the PCB which is plenty for the tube. The module kit consists of:

  • A plastic insulating panel that fits to the rear of the PCB (good idea)
  • A standard blank front panel
  • A piece of bent aluminium that forms the screen
  • Four steel standoffs
Considering it includes a blank front panel I don't need (all my front panels are custom made), a piece of bent aluminium, four standoffs and a bit of plastic, it is not very cost effective. I did ask but Fischer will not sell any of the component parts separately.

In the meantime I decided to try a simple construction of my own. This starts with the original Eurocard method of fixing the front panel to the PCB which makes the top surface of the PCB just under 4mm from the left hand edge of the front panels (10mm better than current modules). Here is a drawing that shows how this works:



A neat little die casting is used. The PCB is fixed by a screw into a tapped hole in the die casting and the front panel is similarly fixed using another tapping. The result is the top surface of the PCB is just 3.97mm from the left hand edge of the front panel. Here is the fixing in detail:




It occurred to me that by using a longer screw we could fix a tapped standoff to the die casting and use this to attach a side plate at the right of the module. At the back of the PCB is the 32 way connector which is attached to the PCB in a similar manner. Again using a longer screw we could attach another tapped standoff to attach the rear of the other side plate. Here is a picture of the basic assembly:


This early version also includes a screen on the left side. This is held 3mm away from the bottom of the PCB by spacers. The screen is made from 1mm mild steel and is very rigid. A single long screw goes through the left screen, through the spacer and the PCB, through the threaded hole in the die casting and into the threaded standoff. At the back the stand off is fitted similarly using the 32 way connector fittings. On the right is the right side screen, also made from 1mm steel and held on by screws into the tapped stand off. The whole assembly is very rigid and strong, even though the front panel has not been fitted.

This then needed to be mated to the new PCB using vertical tubes, shown here:


You can see the two small boards on which the tubes are mounted. These are fixed to the main PCB using right angled headers. When the mechanics and PCB are mated they look like this:


The up side of this scheme is it is easy and cheap to build. It is also very strong so there is no likelihood of front panel controls being damaged by flexing of the PCB as in the EZTubeMixer design. The down side is there is only screening on three of the six sides of the module. However, the sub-racks I use have top and bottom screens fitted so the only side of the module not screened is the back where the 32 way connector is.

This scheme is easily extended to wider or taller modules. Wider modules simply need taller standoffs. Taller modules simply need larger sheet steel screens. The only concern might be whether the larger steel screen flexes sufficiently in the centre of the screen to be a problem. If it is, another standoff could be used.

In practice we can probably eliminate the screen beneath the PCB. The left most PCB in a sub rack has its PCB right beside the aluminium side of the sub rack. Its right side is screened by its right  hand steel screen. But this also screens the bottom of the PCB of the module to its right and so on up to the other end of the sub rack. This may seem like a further compromise but it is a screening scheme I have seen recently in two professional broadcast mixers, one made by Glendale and one made by EELA, both for BBC local radio. If it is good enough for the BBC it is good enough for me.

The mechanical design could be refined further. For example, the the standoffs at the front could be eliminated if another pair of diecastings were used to mount the screen to the front panel on the right. This may or may not be an advantage. The scheme with stand offs does mean you can test a module without the front panel but it is not clear if this is an advantage or not. The general idea though is definitely worth exploring further.

No comments:

Post a Comment