Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

BCD Bus design questions

Status
Not open for further replies.

chrisathome

New Member
I'd like some help with ttl logic, my electronics is well rusty. I need to explain what I have, and what I want to do with it, sorry for the length.
I volunteer as an engineer for a small UK hospital radio charity (= no money!) and need to rebuild their programme switching circuits.
I have a turnkey audio switching unit, built for me in another life, that has 10 individual removable relay cards, each of which can select one of 4 possible card inputs to its output. Some of the inputs are commoned to each card. Cards can be removed from the unit without affecting operation of the other cards. Each card has 4 latching set/reset relays which survive power loss, and each relay has a signalling contact, active ttl high (e.g. +5v through 150 ohms) when operated. Card logic ensures only one relay per card can be latched on at a time. The cards were built around 74 logic not LS or Cmos.
Card source switching is achieved using a separate select card with 4 momentary buttons and IN4148 diode matrixes to set ttl high on a 3 bit BCD bus common to all cards. A momentary set button on each card transfers the bus state to the card logic when pressed, using AND gating. When no buttons are pressed the bus lines are pulled to 0v by 100k resistors on each card. On each card the bus connects to a 4028 BCD to decimal decoder, then other driving ic's. Both the select card button assembly and the set button on each card are wired to +5v through 150 ohms.
I have brought out the 3 bit bus, card set line and the relay status lines from each card, together with +12v +5v and 0v to an external connector on the unit as I need to extend the switching ability and relay status to other areas in the studio complex. The 3 bit bus obviously has to be bidirectional, the status doesn't.
The station has overall screened multipair cable to all the areas concerned which I can use for this purpose. Is any buffering or level shifting required on the long bus lines which will effectively be star wired out from the router position? Worst case is about 15 metres. I have a couple of 74HCT04 hex inverters available for test purposes but my logic knowledge is 20 years out of date. The multipair will only be carrying steady state voltages, no serial or parallel data.
Also, I need to pick up the +5v relay status lines at the remote end then use them to drive 12v led + resistor indicators in an existing display unit that I cannot modify. I can supply the 12v from the router and thought to use a ULN2003A for this as the each led in the display assembly require grounding pin(s) to turn the led(s) on.
Hope this is enough for someone to understand!
Thanks

Christopher
 
It's it a little hard to follow but I think I understand your setup. But I missed why you say the BCD control line is bidirectional?

For DC levels you should not need particularly buffer the lines. How will you be remotely configuring the BCD commands -- just switches similar to those on the unit?

To convert the 4 push buttons to 3 line BCD you can use a 74147 or 74148 priority encoder in place of the diode matrix (you may have in invert the outputs to get proper polarity since the outputs are active low).

Does the LED indicator input pin carry the LED current when grounded? Do you know how much current the LED control input must sink to ground?
 
Thanks for the info, have no idea why I said the bus was bidirectional - put it down to a senior moment!

At the remote end of the switching bus I will be using the same technique as on the crate switch card - momentary push buttons and diodes to set the data up. Each studio has to insert itself into the programme chain instead of the automated playout pc and revert at the end, so I just need two buttons plus indicator leds at each location. The decoders look for BCD 001, 010, 011 and 100, any other codes are invalid and ignored. So it is relatively easy to set each switch up with a few diodes. I was really concerned about long lines hanging on the inputs of the 4028's, although they are held at logic 0 by the 100k's.

In operation, a remote unit will ideally power the card set line some milliseconds after the bus is set up, to ensure correct triggering. Is there an easy way to achieve this? It's not really essential, as if the card doesn't set properly the status leds will indicate this, but in the interests of accuracy it would be nice.

The original crate used 10 cards for audio and 10 for composite video. I have stripped out the video circuits and only plan to use about 5 cards so powering the remote leds should be no issue, I can use the status +5v directly from the relay. The only issue is the main studio which uses a 3 button illuminated switch assembly supplied by the desk manufacturer, the momentary buttons have 12v leds and resistors built in. I would hazard a guess that the current requirement for each button led is 20-30ma - they are presettable to red, green or orange by a jumper, so presumably use a bicolor led. I just need to apply +12v to the panel then take the relevant buttons "led" pin to ground. So will the 2003A be ok for this? The studios will only have three status lines - themselves and the playout pc, they just need to know which is on air at any time. Each studio will only be able to switch themselves in and out, to make life simple for the volunteer presenters.

Three other areas will have comprehensive audio monitoring of various station sources with appropriate indicator leds, which doesn't present an issue.

Hope this makes it clearer

Christopher
 
Since you stated the control wires are screened (shielded?) I don't see a particular problem with interference on the control lines. If necessary perhaps you could reduce the value of the pull-up resistor to, say 10k ohm, to increase noise immunity.

What signal ground connection will there be between the two consoles? Is there a possibility of a ground loop condition?
 
What do you mean by powering the card's set line after the bus is set up. Do you mean a power-on reset type of signal?
 
Signal grounding is no problem, master distribution area is Properly audio grounded and peripheral areas are star connected with this ground only. Mains power is isolated from this and care has been taken when installing to avoid ground loops via the mains driven gear. All audio is balanced individual screened (shielded) pairs. All control cables are overall shielded multi pairs.

Re the delay on the set lines, was just concerned that using one button to set the bus and trigger the card using diode encoding might cause problems with the bus data being ambiguous when the trigger was processed on the card, i.e. bus state not settled. Might be irrelevant, long time since I had to think of these things.

Am I right about using a 2003a to drive the 12v LEDs? The 3 button switch has a common 12 -15v supply pin and individual pins for each led which are taken to 0v to turn the led on. I assume I can attach an output of the 2003a to this pin and the relevant 5v status line to the corresponding 2003a input (0v will be common), to turn that led on? As I said, it's a long time since I had to do anything like this.

Christopher
 
Is the timing for the additional circuit different from that on the receiver audio switching unit? If not then I see no timing problem.

Yes, the 2003a would appear to do what you want to control the LEDs.
 
Since you stated the control wires are screened (shielded?) I don't see a particular problem with interference on the control lines.

That depends on whether every control line is shielded, or just one shield on the whole cable. In that case, maybe not from outside sources, but at 15m, 45 feet? You would have to worry about cross talk. What you would need would be twisted pairs, where every control line has a ground twisted around it. Then shield the whole cable from outside noise.

Also, it is not good practice to connect a TTL pin directly to the outside world. That's what they make buffers/drivers for. Since you are not coming off logic but a diode matrix, then you probably don't need to buffer your inputs. I would suggest a 100 ohm resistor in series with the input and a .1uF cap to ground to help filter out noise spikes.

I just saw this as recent as two years ago, on a piece of test gear that was supposed to latch if a unit under test gave a fault output. Even though all the cables leading to the units under test in the chamber were shielded, just the short run inside the test fixture to the unfiltered TTL inputs caused all circuits to latch when one triggered.
 
Mike

The cable in use is overall screened twisted pair, probably 20 pairs, may be more (I'd have to look). I can arrange for one leg of each pair to be grounded without a problem, and the overall screen would be grounded at one end as usual. The cable would be dedicated to this function i.e. not sharing switching functions with anything else, and I will have +12, +5v and ground also on the cable . I'm quite happy to add ttl driver chips at the remote ends, I'm just completely out of touch with current practice so not sure which would be suitable. The cards I will be using are 20 years old, and I have 20 of them so spares and replacements are no problem, I need 5 max. Unfortunately the cards didn't use ic holders.

The in crate switching just makes +5v through 150 ohms and in4148's to the 4028 BCD to decimal. I am hoping to simply parallel my remote switches across the crate ones and can supply the remote end with either +5 or the +5 through 150 ohms, whatever is best. Each of the BCD lines is taken to ground right at the 4028 on every card with 100k ohms. I was thinking I might have to do the same at the remote end as well? Don't mind what I do, just need some guidance!
 
no, one pull down right at the input to the chip is all that is needed for x number of remotes.
I would put the 150 ohm right at the input, and a .1uf cap parallel to the 100k pull downs. I would piggy back it right on the pull down resistor. Then all your remotes can pull to +5 and you only need the one resistor and you use it as a filter as well. The only problem with that approach would be if the cable is sliced and the control line set to +5 is shorted to ground. Of course, if you're generating that +5v with a 7805, it'll turn off on high current so you won't burn it out. That's assuming the +12v feeding the +5v is still intact. As long as you can reach the minimum input voltage on the 4028 (3.5v), you won't need a buffer.

You also don't have to tap into the inputs at the chip, unless those points are already brought out to a connector. If you are adding a wire from a connector to the inputs, then you can bring your inputs right to the anode of the diode driving the 150 ohm resistor, and feed anode to anode and use the same resistor to the input of the 4028.
 

Attachments

  • matrix.bmp
    1.6 MB · Views: 170
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top