Electronic Scoreboard

[scoreman]

I am new to this so please bare with me
I am building an electronic scoreboard and it works more or less
The score is entered on computer and transmitted to a controller 25 / 50 yds away. The pic controller receives the input and outputs the details to 6 x (8inch) 7 segment displays I have made. The controller and 7 segment displays are driven by a 12v leisure battery.

In order to save power, the 7 segment displays are cycled through for 1 - 3 ms each . However they are not as bright as I expected. Is this normal given that they are being sequenced? How do you recommend I increase their brightness.
Each segment is 4 parallel sets of (4 leds in series)

The displays are common anode driven by BC147 / BD140 combination and the segments are earthed P16NF06

 

[colin55]

Increase their ON-TIME.

 

[scoreman]

I have tried this but it doesn't make a lot of difference
Whilst increasing the On-Time helps a little, once I get to 4/5 ms per display you start to notice the flicker
I guess increasing the voltage might help or changing some of the circuit components

 

[ericgibbs]

I have tried this but it doesn't make a lot of difference
Whilst increasing the On-Time helps a little, once I get to 4/5 ms per display you start to notice the flicker
I guess increasing the voltage might help or changing some of the circuit components

hi,
Please post the LED driver circuit section.

 

[colin55]

There are 6 displays. 4mS per display max and hit them with 60mA. Do all the house-keeping when one of the displays is illuminated.

 

[scoreman]

View attachment 62963
Here is the main part of the circuit - A small change - I was going to use PN100s but switched to BC147s due to difficulty in getting what seems to be obsolete parts

When the score has been received and is being output there is no housekeeping other than a quick check on Rx

Information on displays each segment is calculated to use 80ma

 

[ericgibbs]

Here is the main part of the circuit - A small change - I was going to use PN100s but switched to BC147s due to difficulty in getting what seems to be obsolete parts

When the score has been received and is being output there is no housekeeping other than a quick check on Rx

Information on displays each segment is calculated to use 80ma

hi,
I dont any resistors in series with the LED segments, are they on the chopped off section of the drawing.? if so what are their value.??

 

[scoreman]

Yes - they are chopped off
Each segment has 16 leds connected 4 sets of 4 leds (typ 2.1v 20ma) connected in series with 180 ohm resistor
I have also tried replacing PN100s with BC548

I will run a few tests with one digit giving it all and part cycle times - this should provide another general check

 

[ericgibbs]

Yes - they are chopped off
Each segment has 16 leds connected 4 sets of 4 leds (typ 2.1v 20ma) connected in series with 180 ohm resistor
I have also tried replacing PN100s with BC548

I will run a few tests with one digit giving it all and part cycle times - this should provide another general check

hi,
If I understand correctly, [assuming the all the LED's do have a 2.1Vfwd drop] thats 4*2.1V = 8.4V... You have a 12V supply and assuming a 0.1V drop across the driving transistors, thats 12 - 8.6v= 3.4V across the 180R.

Thats only ~19mA LED current. For a multiplexed LED display that is far too low. Try a 92R series resistor [ or another 180R in parallel with the existing 180R]

Lets know the increase in brightness.

Note, I assume your power supply can supply the increased current.??

 

[scoreman]

View attachment 62997
Hi ericgibbs
Thanks for your help
On the calculation side I arrived at the following
Each segment has the Leds connected as in picture
4*2.1 = 8.4v (assuming the leds are as the packet says 2.1v and 20ma)
with a 12v supply and assume .1v drop on the transistors, this would give 3.4/0.02 = 170ohm say nearest is 180ohm
But as each segment is made up of 4 parallel strings I make the total current 80ma but I may have misunderstood the formulae

The test power supply can provide 2 amps at 12v so this should be OK

Just run a modified program on the pic for my latest test ie 1 digit for 1ms works! the leds are at full brightness - proves a bit of hardware (I wasn't sure about the size of the transistors)
I will add back 1 digit at a time and run more tests tomorrow
could be hardware or software

 

[ericgibbs]

hi,
Looking at your diagram, each string of 4 LED's has its own 180R resistor, so thats only ~20mA, add a 180R across each of the four 180R's or replace with 92R's.

Do you follow ok.?

 

[scoreman]

Hi Ericgibbs
A whole new area of electronics opens!
I understand what you are saying and I think I understand the reasoning but I am just not sure about the theory

First if I understand correctly my calculations only work if there is a constant 12v but not in a cyclic configuration
Why do you suggest doubling the current as opposed to any other figure - is it dependent on the 'on-time' in a cycle?
I don't know the max current that the leds are supposed to handle
Is there a formulae for this situation or is it just rule of thumb?
Is there a web-site you can recommend for me to read up on this?

Secondly on the cabling from the control box to each digit I am using 8 x 24AWG wire. In a constant 12v dc environment, in theory, this is well within limits and works in practice
However as it is a cyclic system do I have to take other factors into account?

 

[ericgibbs]

hi
Can you post the LED type number, in that way I can check the LED datasheet for the optimum current based on the On/Off duty cycle.

 

[scoreman]

Hi ericgibbs
I think I have a problem - I couldn't find the LED Type number
I have been back to the supplier and he says it is not a flash LED, so, it have not On/Off duty cycle (lesson there somewhere)
and that the best current is 20-25mA
I suppose I can run a few tests pulsing leds with different resistor values
Anything I should consider ie limit the pulse length to 100us?

 

[colin55]

Don't waste your time. You can pulse the LED with 60 - 70mA max on a duty cycle of 15% with pulses up to 5mS.

 

[ericgibbs]

Hi ericgibbs
I think I have a problem - I couldn't find the LED Type number
I have been back to the supplier and he says it is not a flash LED, so, it have not On/Off duty cycle (lesson there somewhere)
and that the best current is 20-25mA
I suppose I can run a few tests pulsing leds with different resistor values
Anything I should consider ie limit the pulse length to 100us?

hi,
I didn't say they are a flashing type LED.

Just try the 92R or another 180R in parallel with the existing 180R,,, if thats still not bright enough try 68R's

 

[scoreman]

Hi
Thanks for all your input
Ok I have run several tests with sets of 4 leds (all compared to a base constant 12v set with 180R)
The wife did the viewing as a blind test

Also I have run several tests comparing pulse times, and 1ms seems best (not a lot of difference from 100us) at 5 ms there was definite blinking

With 120R (30ma?) Leds are more lemon coloured than the amber leds and not as bright as the base set
at 90R (2 x 180) (40ma?) slight improvement
73.3R (3 x 220) (50ma?) again slight improvement
68R (53ma?) further improvement and quite good
60R (3 x 180) (60ma?) colour practical the same as the test base although not chosen as the brightest compared to 68R
57R (10 + 47) (64ma?) - the best of the test batches

Conclusion - I should use a 56R or one of the other two (68R or 60R) unless you have any further thoughts (68R change all resistors or add 92R to current 180R)
Question 1) - If I use 56R the current per segment is in the order of 260ma or 1.8 amps (if all 7 segments lit)
I had designed the system to use 24AWG wire what are your views on the wire carrying this current (the individual segments should be OK and in theory the Anode should be OK but are there any special wire factors to take into account as it is pulsed)
Question 2) I realise the Leds won't last as long do you have any experience as to their likely life (eg few hours / hundred hours / thousand hours)

Having done some statistics, the number of segments likely to be 'on' in an average period per digit is 5 (ie 30 segments)
Therefore likely average power if 56R is 1.3amps
Compared to an always on system with 180R is 2.4 amps

 

[ericgibbs]

hi
Assuming each LED is is on for say 1mSec and there are 6 DIGIT's thats roughly a 17% duty cycle.

I would choose 60R. If the LED's are operated within their current/power limits they should operate for many thousands of hours.

 

[scoreman]

Thanks everyone
I will add an extra 91 ohm in parallel to give 60ohm to each string
Approx this will result in an average 240ma per segment - 1.68 amps per 7 segments when fully lit
However on average only 5 segments are on resulting in 1.2 amp requirement

For the future I would consider using an always-on system as it is slightly brighter than cyclic and use 15 leds to save power
The average amp requirement would be 1.8 amps rising to 2.5 amps if all segments on
It would mean a redesign of the control board but nothing too difficult

Thanks again for your help

 

[ericgibbs]

hi
Assume a V drop of 8.6V across each 4 LED segment string, with a 60R series resistor and a duty cycle of 1/6 ie: 17% and there are 4 strings per segment.

Each string will have an average of 57mA * 0.17 =9.5mA, so for a full segment of 4 strings thats 4*0.0095 = 37mA.

For a fully LIT digit of 7 segment thats an average of 0.26A.

Note: only one DIGIT will be enabled for a 1 in 6 period.

E.