latigid on 270 Posted February 2, 2017 Report Share Very pretty! May I ask what current you're driving through the LEDs? Do you experience any voltage sag/droop at the 595 anode outs? The story apparently goes, that if you're drawing more than 5-6mA (continuous) through a 595 output pin, the source voltage will decrease. http://electronics.stackexchange.com/questions/77841/how-much-current-can-my-74hc595-handle-on-each-output-pin While keeping in the MIDIbox paradigm of ordinary shift register chips, one thought is to add PNP BJTs as "current sources" (moreso voltage sources due to constant current sinks on the cathodes). Any thoughts on this? Quote Link to comment
Sauraen 38 Posted February 3, 2017 Report Share On 2/2/2017 at 3:16 AM, latigid on said: Very pretty! May I ask what current you're driving through the LEDs? Do you experience any voltage sag/droop at the 595 anode outs? The story apparently goes, that if you're drawing more than 5-6mA (continuous) through a 595 output pin, the source voltage will decrease. http://electronics.stackexchange.com/questions/77841/how-much-current-can-my-74hc595-handle-on-each-output-pin While keeping in the MIDIbox paradigm of ordinary shift register chips, one thought is to add PNP BJTs as "current sources" (moreso voltage sources due to constant current sinks on the cathodes). Any thoughts on this? The current-limiting resistors are 220 ohms, so roughly (5 - 2.1) / 220 = 13mA. This was the third MIDIbox I built with 595's directly driving LEDs (at least the anodes), and I haven't had any trouble with it, so I never looked into it. This picture is actually a year old, I just put it up because I've gotten some interest in other people building these Genesis synths. Quote Link to comment
Psykhaze 22 Posted February 4, 2017 (edited) Report Share No 74HC541 at all ? if yes your pinout should be pretty optimal regarding the huge CS!Edit : seems not after looking the wiki but i see some NMOSFET , so everything is not under direct driving ? Edited February 4, 2017 by Psykhaze Quote Link to comment
Psykhaze 22 Posted February 4, 2017 Report Share On 02/02/2017 at 0:16 PM, latigid on said: while keeping in the MIDIbox paradigm of ordinary shift register chips, one thought is to add PNP BJTs as "current sources" (moreso voltage sources due to constant current sinks on the cathodes). Any thoughts on this? I agree on BJT/ MOSFETs thing . It's -in theory- better to keep the splitting between Power and Control (once again in theory , shift register first paradigm is long term support and easiness of use before everything) The hybrid role that shift register play help to keep things simple , wich is kinda important in the DIY , with lowered BOM advantage etc. and it can be buffered with 74HC541 in case of huge config. I am impressed by the huge load you managed to have on shift registers @Sauraen However for example (individual) PWM is kind of hard to manage with shift registers,it's a bit sad. I saw about Maxim MAX6966 and MAX6960 as nice LED expanders, offering individual PWM Control. But i still have to write a driver , the nicest part ahah x) Quote Link to comment
Sauraen 38 Posted February 4, 2017 Report Share Oh, all three front panels have used row drivers; in MIDIbox ASIDITY it was a bunch of TO220 NMOSFETS, in MIDIbox FM V2.0 / V2.1 it was a Darlington array, and in this one it's eight TO92 NMOSFETS, they're right next to the orange LEDs in teh middle. Obviously the shift register pins can't sink >1 amp! But I don't quite understand the issue--MIDIboxes have been using BLMs with this scheme for 15 years. And in fact, since only one row is on at a time, having 8 or 16 rows from each pin of the shift register is just the same as having each shift register pin drive one single LED by itself. That's what the DOUT module is specifically designed for. You're right about PWM, but I don't use dimming on any of these panels. :) I don't find that it's terribly useful to convey information to the user. 1 Quote Link to comment
Psykhaze 22 Posted February 4, 2017 (edited) Report Share The "issue" i was trying to mention for -pure reference- was the case of Large RGB (matrixed) setup with PWM control. Otherwise everything is fine with Shift Registers on mono/bi color LEDs (with or w/o Power driver),maybe excepted the "low" refresh rate (wich is quite enough to drive a 16x16 blm @some hertz, so that's not a real problem). it's -only- technical improvement thoughts for the talking fun =) Edited February 4, 2017 by Psykhaze Quote Link to comment
m00dawg 23 Posted February 6, 2017 Report Share Mmmk so first, this is absolutely gorgeous! Second, can I ask where you get your boards fabbed and how much? Quote Link to comment
Recommended Comments
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.