I thought this was going to be an easy circuit. It probably is. But I’m making a stupid mistake, somewhere. My head feels like it isn’t able to solve the most simple problem. I think I broke my soldering iron out of frustration.
I’ve already blown 3 78l05’s from my power supply. I don’t have any left. But next week 5 new ones should show up. (even though I need 4 for another project! So I’m feeling a bit… Euhm… Like I can’t make too many mistakes for now) (So I’m assumig there’s an overcurrent/short somewhere)
rant over
So a simple digital 5v chip.
Datasheet:
Pictures the messed up board.
I’m assuming my mistake is in the switch connected to the board. Also the resistance between ground and vcc is 2.5k, in the isolated section of knobs and switches. Which I don’t understand. Switches are off. Pots are 10k. Turning pots does 0 to this resistance.
Also worth noting. The switch was connecting yellow and red when I smelled my voltage regulator burning up…
You probably need more info… But I’ve been staring at the thing for too long to give decent information.
Where should I look?
Thanks for any help. In the meantime. I’ll work on something else. A front panel or something.
I don’t see anything obvious here, but have you tried using it without the IC installed? Also, do the 78l05’s work fine outside of your circuit?
Oh and yeah, sometimes copper plated perf or stripboard hasn’t been etched well and has some pads that actually have small shorts. I’ve had a couple of projects really take longer than they should because of that.
What’s the exact circuit you’re building? The one from the datasheet, minus the 3360, or something else? What do the switches do (they’re not in these schematics).
(the 78L05 is supposed to have thermal and overload protection so seems you’re pushing it beyond what those mechanisms can handle).
The caps on the pots look pretty big for a 100nF, btw. Are they 100uF? (that’s 1000× too big, and just charging up 400 uF to 5 V could stress the regulator, even if there’s probably enough resistance between +5V and the caps to prevent that).
The schematic is what’s there minus 3360 like you said. edit: also no voltage divider on the output.
+Mode cv switch (the one I don’t know if I connected correctly. Voltage divider for pin 3. Using 100k resistors instead of 10k. Because voltage divider would still work the same? And I replaced them thinking this might solve the problem. I also tested the voltage divider, which seemed to work perfectly with the switch.)
+switch for pin 2, which I connected to ground instead of 5v, so it’s not working, but shouldn’t do anything either as it’s just grounding the pin like the schematic right now.
+voltage overload protection on the gate input
-trigger input (didn’t have space on my panel, that’s why there’s 1 input protection too many. So it’s hanging there doing nothing.)
Didn’t have 0.1uf so put in 0.22uf… Thought it was close enough. So it’s not x1000, but x2.2. Also put in a 100ohm and 220ohm capacitor in series on pin 12 for the level CV. Because I didn’t have 330ohm. Every capacitor is 0.22uf instead of 0.1…
Thanks for taking a look. I’ll take a fresh look again tomorrow.
Edit: For a second I could get it to work in loop mode. But at that time my regulator was already malfunctioning and giving 7 volts instead of 5. So it did kind of work for a second.
Edit: edit: In despair of trying to fix things, I also omitted the resistor on the gate input.
Huh, haven’t seen an electrolytic below 1 uF in ages, but a quick visit to Mouser tells me they’re still being made (but 0.22 uF is about as low value as you can get). TIL
You may want to replace them with ceramics once you get things working, though – those are decoupling capacitors that are there to deal with high-frequency noise, and electrolytics suck at that. Also, 100nF is everywhere, since that’s the standard value for decoupling, so you should get a bunch next time you buy components (they’re literally 1 cent each from Tayda, or if you want brand names Mouser will happily sell you 500 for ~$20).
But that’s right, isn’t it? There are four 10k pots with Vcc at one end and ground at the other. That makes 2.5k. Changing the wiper position doesn’t change the total resistance between Vcc and ground.
So sending 5V to pin 3? It’s not entirely clear from the photos.
Is there any way the switch could be miswired to connect +5V to ground? It doesn’t look it but again it’s not that clear. Or could the bare ground wire be touching the middle terminal? Could the switch be defective? What happens if you remove the switch and leave pin 3 grounded?
Mmm… Well I do ceramic once, but thought the symbol in the schematic was asking for electrolytic. If I understand correctly, all of them can be swapped wit ceramics, because all of them are decoupling caps, right?
Going to breadboard some things this evening, to see where it went amiss. Basic schematic, without switches first. I thought I learned from my mistakes and just make everything without breadboarding first…
@analogoutput Yes. 5v to pin 3. I understand it’s not that clear from the pictures. I’ll try it out this evening!
Mmm… For some reason I missed Sam’s schematic. Maybe I’ll just put a regulator on the pcb instead of using the 5v from my power supply. Saves some hassle if it breaks again. Going to breadboard now…
That’s just weird US notation – some people use an asymmetrical symbol also for non-polarized caps, while the ones that are polarized are marked with a + (e.g. C3 in that schematic). Other hints here are where they are in the circuit, and their value & unit prefix.
(TBH I’m not entirely sure why he bothers with caps here, I at first thought it was to deal with potentially noisy external CV but there’s no external CV here. If it’s to avoid the wiring picking up noise you want them closer the chip, not on the front panel.)
Voltage divider for pin 3. Using 100k resistors instead of 10k. Because voltage divider would still work the same?
The analog PIC/AVR inputs are a bit special; while they measure voltage, they need a certain amount of current to charge an internal capacitor, and datasheets recommend staying at 10k or below:
But none of this explains why you’re frying regulators.
Of course it is! I heared something interesting yesterday. That when you’re emotional, your IQ goes down.
I also did this:
I’m assuming it’s not a big deal… But noticed it when making my breadboard. It’s all layed out, but waiting for the baby to wake up, so I can grab a power supply.
Ok breadboard update! No smoke, nothing got burned… I got loop-mode to work, but the gate gave me nothing… So it’s progress. I’m sure I’m getting 5v in… Just nothing comes out.
Yet another update. Breadboard complete and everything working fine… (note I’m using a different power supply… I’m going to test the panel with this power supply after changing all the tiny mistakes I made…)
Final update (for today I mean*)… Tried my panel again… It’s kind of working… Nothing blew up. But it’s acting really weird. Negative voltage on the output constantly… When I put a gate in, it goes positive. How, what?
Since my breadboard is working perfectly… I think I’m going to start fresh tomorrow. It will be less hassle redoing everything, than to find the problem. But can I live with not knowing what’s going on? I think I can… I consider today. A win.
