So I have been reading, reading, and more reading. Can’t focus on the coding because this is driving me loopy. I finally stumbled upon a small crumb that led to another small crumb. I was reading about power supplies and how capacitors are necessary for smoothing out the transition from AC to DC. I happened to remember a guy who built his own home-brew amplifiers and in his videos he talked about how frustrating it was with the current draw of the capacitors. Basically he built a circuit called a soft-start to solve this problem. What is happening is the transformer is putting out a tremendous amount of current to fill the big capacitors because it can handle such a large amount.
I then watched Great Scotts videos on power supplies for a switch mode build. However, I stumbled upon his video on a soft start video. Basically he explained why you need one and how to make a simple one. Then I stumbled upon this other guy with the exact same problem I had. Blew every single fuse he threw at his cabinet. He had a 100watt transformer. I clicked the link and found it to list 8.3 amps. Exact same problem I had. So, thoughts, ideas, suggestions? Anyone want to make this a sort of community build? Like we can all build it together and keep track of what problems we face. Hope its not a problem to post two URL’s
.
.
To be honest, I think your biggest problem is that you’re trying to feed regulators rated for max 1.5 amps each with a way over spec’ed supply. Unless you come up with an arrangement that can regulate more than about 2 amps in practice, feeding it with an 8.5 amp supply is at best pointless.
That is kind of the point of the soft-start. Its designed specifically to limit the power in amps with a resistor. When the caps are full the relay turns on which bypasses the resistor. Then power can flow normally. Basically providing a work around the amperage problem. This is just a guess. I don’t have the money for the relay at the moment. I have plenty of time to read up on all this.
edit;
Also amps does not exist unless the circuit is created. Ohms law states that amps equals the product of resistance in ohms and voltage. Also consider the equation
What shunt resister will we use?
12volts = 2amps * R
12volts / 2amps to isolate the resistance
R = 6 ohms
We will need power resistor 6 ohms to get 2amps. Of course something bigger like 10ohms will give us enough head room. After a small amount of time AKA capacitor “T,” can be calculated with a mosfet to turn the relay on to bypass the shunt resister when the in-rush current dissipates. If that makes sense
Are you using a toroid transformer? They need quite a bit of startup current for a few milliseconds to magnetize the toroid core, but 12×8.3 = 100 VA is below where soft-start circuitry is usually recommended. Can you post the schematics for the transformer side of your power supply, from mains to the rectifier, and with the fuse positions marked out? Also, what fuses are you using? (size and and type – i.e. fast/slow).
(But as others have pointed out, you’d save yourself a lot of headaches by starting out with a smaller transformer, closer to the capacity of the power supply you’re using it with. Designing robust high ampere stuff is hard and can be dangerous if not done right, and it’s not really needed for most synth gear anyway.)
Also amps does not exist unless the circuit is created.
Well, you are creating a circuit when you plug the transformer into mains.
I get the point of a soft start; just saying that your are creating the problem that requires it by spec’ing for massive current from the transformer and your regulators won’t be able to make use of that current anyway. If you do want to be able to eventually draw more than a couple of amps, you will need a different design anyway. In short, I’m saying if you want a simple circuit, power it appropriately or, if you want to build a more powerful supply, find a circuit made for that to build from.
Here is the transformer. You certainly are correct Fredrik about the primary being a complete circuit. I really need to stop looking at this like its a DC power source from cells. Old habits die hard I suppose. So, the current works differently due to the transformer. Surely there is a simple solution here though.
Also here is a picture of the transformer. It has a 10amp fuse. Then I add two fast burn blade fuses (2amp) to protect the power supply circuit. Could that be the problem? I will put my previous understanding of circuitry aside to learn something new. I really do appreciate the help.
I had a soft start module, the type you plug into a fuse box. I’ll find it tomorrow and see what its specs are. Got it to solve the problem of random breaker tripping on a big variac I have. Never got around to fitting it… I think it just switches in a resistor to limit the inrush current for a few cycles…
your are creating the problem that requires it by spec’ing for massive current from the transformer
I don’t agree here. The transformer does not create a problem. Maybe there is a peak current it draws when connected to the mains, so its primary fuse needs to be adapted to that. But the power the transformer can deliver has no effect on the following circuit here as long as it exceeds the power the circuit wants to draw at max. I agree that a transformer this big is a waste if the voltage regulators can only hold 2A but if this is the only transformer LP2Lily currently has or if in the future more power is needed, he can keep on using this transformer and add yet another 2A section for the new circuity.
The problem with the fuses blowing is more likely to be a short or a massive capacitance. @LP2Lily What is the current value of all capacitors combined at the moment?
That’s a nice beefy transformer, powerful enough to fry a lot of downstream circuitry without noticing , but doesn’t look like a toroid so while it will need extra current for a cycle or two (5×, maybe?), it’s not that extreme – the 10 A fuse should leave plenty of margin (you could also use smaller time-lag fuses, but I think you can trust the manufacturer here; a 10 A fuse seems reasonable).
But if I’m understanding you correctly you’ve been blowing the 2 A fuses, not the mains one? That could indeed be the capacitors (they’ll effectively act as a short for some small multiple of R×C seconds, where R is the source impedance; no idea what that may be for a 100 VAC bobbin transformer, though). You may have to use slower fuses.
EDIT: Some googling indicates that the resistance of the primary winding of a 120 V/100 VA transformer is somewhere around 5 ohm, which (using crude estimates) gives a secondary resistance at around 5/(120/12)^2 = 0.05 ohm and every 10,000 uF corresponds to a short in 5×0.05×10,000u = 2.5 ms. Select your fuses accordingly.
Sorry for the late response. (reason in side note). I am using a 1 uf (micro-farad) and 3’300 uf for the big ones. In theory I assume this is the same as DC you just add the values. Please correct me if I am wrong.
Total capacitance
+
6 of 3300uf = 19’800
(+ 2) of the 1uf
That would be 19’802
Minus the 2 big caps in the trouble shoot
13’202 before the input pins total
Output pins has 2uf total
+
Side note
Sorry, I crashed yesterday. I woke up thinking it was still Saturday. I slept in till 8am and got to looking at my cell to think it was 8pm. I would not have known it was Sunday unless it said it in the corner. I did an all night cleaning of my kitchen. I had a big boost of positive energy the other night. A big plus to facing depression like I have been. Depression does really weird and aggrivating stuff to me. Like I will loose a good amount of weight only because I have very little motivation to do anything. My kitchen was in really bad shape before my boost because of it.
So have you thought about the suggestion to split the problem in 2 and work on one side of the power supply until that works and following that on the other half?
Also, using slow fuses as fredrik suggested which are also of a higher value would possibly prevent them from blowing when the capacitors are charging up when powering on.
I managed to get it to some-what work with two nine volt batteries. Basically no connection to mains lines. The leds I have on the output lit up. However, I only got a little less than 9v. I expected this to be honest. I just wanted to see if the fuse will blow with all the parts in their proper places.Surprisingly, aside from one of the voltage regulators getting hot from reverse voltage I got it to work. Had to move all my alligator clips around. So, yes its the transformer I have causing the problems. I want to try a few more things before I give up. The transformer I wall adapter thing I have cost me 50usd. Not exactly something I want to give up on. It will stay on the bread board until I have the motivation to experement with it again.
I will buy one 500mA wall wart and then use that till a I can get with a friend of mine who is going to school for being an electrician. Right now I have to pay things off. We used to be class mates before I lost who I was from being yelled at by the teacher. In all fairness I began yelling back because I did lose who I was when I was in tech school. My friend used to be an engineer of sorts. Basically got fed up with all the hateful men. She was like a mentor to me for a little while. With any luck I will be able to find reasonably sized transformers at something like 2500mA and then I can wire them all on a 8amp or 9amp line. I really do have a lot to learn for assembly language. I wrote the code for converting a MIDI message that will output onto a serial buss.
Basically my goal right now is to learn how to program EEPROM’s and build several embeded systems to work in a cluster. After I see that the arduino successfully outputs the 16 bit number onto two 595 shift registers things will be picking up pace. It will be a while before I can connect a DAC to convert the 16 bits into two control voltages (Pitch and velocity). That will be when I am ready to really build and solder my first synth module. I will be using a DC bench power supply until I know the cluster computer works exactly the way I want it to and the power requirements. It will have a basic OS that will use 2 8-bit LCD’s and a basic keyboard. One LCD to keep track of incoming MIDI messages and the other for displaying perimeters for the MIDI to CV cluster computer.
I totally get that you’d want to salvage your investment, so how about a different regulator set and higher current slow blow fuses, maybe? Something like the LM138, which is specifically designed to handle high peak currents. Not sure what the negative counterpart would be, but it seems doable. A bit of a hassle, but doable.
Unless you need all the amps in the same circuit, building several small supplies and hooking them up to the same 12 VAC line is easier than building one big supply. Linear supplies don’t scale nicely.
That said, if you insist on building a big dual supply, the LM138 (or the commercial grade LM338 more likely) would take care of the positive rail but I don’t think there’s any negative regulators in the same league. On the other hand, once you go above 1-2 amps you might as well go for regulator + power transistors anyway, using the regulator to drive the transistors instead of the output. Basically this example from the 78xx datasheets:
(and similar for the negative rail, but with a 7912 and NPN transistors)
The transistor in the example is no longer manufactured, but there’s plenty of replacements; TIP35/36 is a quite common choice iirc, and is widely available. You can stack a bunch of transistors in parallel to handle more current, with small balancing resistors in series with the emitters. And you need beefy bridge rectifiers and big heatsinks…
Went looking to see if Rod Elliott had covered these kinds of supplies, got a bit distracted by a 4-part article series with more information about transformers than I even knew existed, but finally found an article that shows how to wire up multiple output transistors:
, but doesn’t look like a toroid so while it will need extra current for a cycle or two (5×, maybe?), it’s not that extreme – the 10 A fuse should leave plenty of margin (you could also use smaller time-lag fuses, but I think you can trust the manufacturer here; a 10 A fuse seems reasonable).
