Marantz amp question (Transformer).

Thanks for the drawing,

but I am finding that just angling the LED's works good enough. Ill post pics when I am done.

If I ever desire more light then I will use your idea :)
 
Oh yeah it looks like I am finally finished.

I cleaned up all of my work and mounted the breadboard up to the casing.

Overall I am very pleased with the results. :)

If anyone is wondering heres the parts I used:

Radioshack:
1 X 6" breadboard (Cut in half to get 8.25")
1 X 2.5" Breadboard
1 X 1000uF Cap
1 X Bridge Rectifier
3 X 100ohm 10watt resistors
and
8 LED's from Hong Hong (Ebay it)
(Supply voltage directly off transformer is ~23V. Multiply times 1.4V for peak wave and you get 31V. Use the parts above to achieve the same effect I produced)

And the final photos :)

final_a_amp.JPG


final_a_leds.JPG



Thanks to everyone who has contributed to this project. It was a huge success :D
 
Very cool looking!

Quick question. Did you have the circuit diagram for that amp, or were you just doing guess and test with a multimeter pretty much?

You've inspired me to become more creative, perhaps someday I'll start fooling around with electronics and see if i can make some goodies.

By the way, how do you like the Marantz receiver anyways? I see them on E-bay all the time for like 50$us, and they really look nice.

Oh, and those are some funky looking 100ohm resistors....... never seen some like that.
 
Thanks :)

Answers to your questions:

1. I used a plain old multimeter. No diagrams. I just poked around til I found a suitable voltage.

2. I LOVE it! It really workd well for its age. Its ashame to see they are so cheap :p (shippings gonna be high though on a ebay one :D) Its really a solid amp. Hell, everything down to the front is made of like 1/8" metal! It sure does crank though (I use it mostly for headphones (it REALLY helps boost the highs and bass!)).

I believe Mr. X has some experience with the marantz's. Maybe he can offer some more insight on them!

Hope this helps!
 
Mister X said:
ewww.....if those where all radioshack LEDs that was an expensive lesson.
You got questions?
They got silly prices. :D
Click to expand...

i work for radioshack ( corporate office doing support for stores )... about the silly prices.. you have no idea. ( im probably not allowed to mention actual cost of items vs selling cost) Getting a 30% discount on all parts / electronics does help somewhat.
 
nice!... off topic but i had that amp too... still might not sure... it kicked ass.. still kinda does... i couldnt turn it past like 1/2 way with out fear of blowing my speakers
 
Same here :p

I think I have to check my LED's something caused 2 of them to go really dim. Probably all that voltage I pumped into them earlier really shortened their life.
 
UPDATE!:

So its been a few weeks since I last posted here- and put the LED's in the amp. Over time I noticed they got dimmer, and dimmer, and dimmer, then came the day that the LED's did not turn on.

Instead of leaving it that way- my usual solution, I took the amp apart again. Ohhed out the ressistors, all good there, and the cap was still good. Took a look at all the LED's and they were burnt on the inside.

So, I kind of knew that they were getting a bit too much voltage.

I later broke out the Weller and unsoldered the dead LED's, then I thought what I could do inorder to such up some more voltage. So I went with the logical solution and stuck some more LED's in my array- 3 more to be exact. That knocked the voltage down just enough to keep the LED's intensity at their fullest (without exceeding their rated voltage).

Now that I had 3 LED's I got those signal meters finished.

I must say it came out AWESOME this time!!!

The panels are now COMPLETELY EVENLY lit.

I tried to be the best I could with the pics, so they should give you a good idea of what the real deal looks like:

Amp with the lights off:
ampdarkrev1.JPG


Signal Meters:
ampdarksignal1.JPG


Amp with a flash:
ampflashrev1.JPG


Im really impressed how well this turned out this time. Lets hope it STAYS this way :)

Oh and I was digging through the file cabnet and found the ORIGINAL owners manual, and a specs sheet- its really cool to find stuff like this :p

Heres the specs sheet, for the hell of it:

ampspecfront.jpg

ampspecback.jpg


Now thats what I call one HELL of a spec sheet! EVERYTHING you need to know :p
 
Wow. I saw this just at the beginning and haven't seen it again until now. I gotta hand it to you, it looks very very nice. Good work!

 
Thanks alot :)

For some reason the 3rd LED from the end is flickering a little, like it will be bright for 5 seconds then it will go dim for a second. I dont know if the LED is defective or what, its not a bad solder joint because the last 2 LED's arent affected.

Oh and BTW, by the end of the array there is exactly 3.3V, I had about 3.6 before I added the 3 extra LED's. That should help the LED's by not putting so much voltage through them (which I believe is the reason they died in the first place).
 
DaRkF0g said:
For some reason the 3rd LED from the end is flickering a little, like it will be bright for 5 seconds then it will go dim for a second. I dont know if the LED is defective or what, its not a bad solder joint because the last 2 LED's arent affected.
.
Click to expand...
LEDs that go bad will do this, they will work for a while, then overheat, then cool down and repeat. just replace the bad LED, it should be a 1 time thing. if you overvolted them in the past it may have contributed to the failure.
 
Another LED was also doing this just for a minute or so when I turned the unit on. It seems that the voltage just simply may be stabilizing, I mean the transformer voltage. It stops once the caps and such are charged. I think its just the voltage drop that I am seeing in the LED's, I mean I AM running 11 LED's off of a 30V power source. At this point I dont think its anything to worry about, if they start to get dim I may try parallel or something instead.
 
Are you running all the LEDs in one series string? If so, the one resistor may not be limiting the current sufficiently. As LEDs warm up, they drop less voltage and sink more current. When this happens, the resistor normally compensates--the increased current means that the resistor drops more voltage. (Follow closely here, it gets a bit hairy) The problem comes when the increase in the voltage drop (across the resistor) due to increased current is smaller than the decrease in voltage drop across the LEDs. It's called thermal runaway. The resistor isn't big enough to compensate.

It's a bit like a nuclear reactor with under-sized Boron control rods. If the reaction stays at a low enough level, the rods can control it. Once it gets past a certain level, they're too small to stop it.

I guess what I'm trying to say is that you should split your LEDs into two parallel strings, both running off the same 30VDC (which used to be 24VACrms), each with its own current-limiting resistor (you'll have to recalculate the values--they'll be bigger than before). How does that solve the problem? It means that if the LED current goes up, the voltage dropped across the resistor will increase more than the voltage across the LEDs will decrease, and all will be well. No more burnt LEDs. You'll be drawing twice as much current as before, but it will still be well within the limits of your components.

 
I was thinking of doing that, but I am afraid I will have the same problem as I did before. Heh, did you see how many resistors it took to get my current setup- it took 3 100ohm resistors that can handle up to 10 watts per resistor.

If these burn out however I will be forced to go with the parallel idea, which is not much of a problem, I judt dont want ot run into the overheating resistor problem I had before.

And are you talking about splitting me LED's up into to rows of say, 10, and putting 1 resistor in the beginning of each orw, or have a resistor before each LED?

Thanks :)
 
Mohonri said:
The resistor isn't big enough to compensate.
Click to expand...
Well, he is using 10watt resistors... I think this would be a problem if he was using 1/2 watt, or even 1 watt, but 10 watt POWER resistors? I don't think so
 
Yeah, the 1 and .5 watters got hot, like enough to burn your hand, so I decided to go to 10 watt (next best thing I could find at RadioShack). Even they get a little warm, but nowhere near hot. The only reason I used 3 100ohmers was because 100ohm was the biggest value radioshack sells in 10watt.
 
Do you know how to measure current with your DMM?
The question after that is:
Does your meter have the ability hold a peak reading?
 
Why the hell do you ask me if I know how to use a MultiMeter?! I post what it tells me, its as simple as that!

I mean hell it only can do AC/DC and Ohms- it doesnt take an electronics genius to figure that much out :p

Do you really think that my RadioShack POS has the ability to hold data:p- I may do some bidding on ebay tonight for a half-decent fluke!

Now, with that said why it was 24VDC before and 30VDC now- I have NO idea?!
 
:eek: smartass :p

Ok, in that case no, I cannot. My dad has a multimeter that can however. I will borrow it from him when I get a chace.

Once I do get it what am I looking to measure?
 
DaRkF0g said:
I was thinking of doing that, but I am afraid I will have the same problem as I did before. Heh, did you see how many resistors it took to get my current setup- it took 3 100ohm resistors that can handle up to 10 watts per resistor.

If these burn out however I will be forced to go with the parallel idea, which is not much of a problem, I judt dont want ot run into the overheating resistor problem I had before.

And are you talking about splitting me LED's up into to rows of say, 10, and putting 1 resistor in the beginning of each orw, or have a resistor before each LED?
Click to expand...

I was suggesting that you split up the 11 (or 10) LEDs into two strings of 5 LEDs, and putting one resistor on each string. Wait, hang on, I (finally) went back to look at the specs for the LEDs. 3.5V, 20mA.

Ok, so if you have two strings of 5 LEDs, you have 5x3.5 = 17.5V taken by the LEDs, with somewhere between 24 and 31V total (more on that later). So between about 7 and 14V has to be soaked up by the resistor. 20mA * 14V = 280mW worst case. So a half-watt resistor should be sufficient for each string. 14v/20mA = about 700 Ohms for the value of the resistor. 680Ohms should be fine. I believe the reason you were blowing those 270Ohm resistors at the start (along with the LEDs) is that it was undersized and underspecced. It allowed the LEDs to draw too much current, and it couldn't handle the power it was dissipating.

Putting in a larger resistor will do more to help extend the life of your LEDs than putting in additional LEDs, since you still have to deal with the issue of thermal runaway.

If you wanted to get real fancy, you could put in a current regulator (consisting of a linear regulator, a MOSFET, and a few of resistors), but that's probably more than you want to do, especially since you've already got a pretty good result.

You can measure the current going through each string with the multimeter you have. Measure the voltage across the resistor and you can find the current (V=IR, and I'll leave the algebra to you).

The reason for the DC voltage dropping is this: the voltage coming out of the bridge rectifier looks like a series of humps (it's the absolute value of the AC signal, minus losses in the rectifier). The capacitor stores the charge when the voltage goes up, and supplies current to the LEDs when the voltage goes down. As the capacitor supplies current, it loses voltage. The more current the LEDs draw, the lower the voltage sags before the incoming 'hump' dumps charge back in the capacitor and the LEDs. Your voltmeter takes a sort of average. So when nothing is attached, the voltage stays right up at 30-31V, but when the LEDs start sucking juice, the average drops to about 24V. You can get a smaller sag by using a larger capacitor. (heh--this one would give you a nice small sag, but it's a bit big for the space :) )

 
Great, thanks for explaining what was going on to me. Once these LED's die (and from what it sounds like I am doing they eventually will (although they are holding up well for now)), I will divide them up into 2 strings. Its too much of a pain to take the damn thing apart, and besides LED's are cheap :p

As for a cap with more microfarad's I didnt realize that the cap had that much of an impact on the DC wave(sp?)!

If I find time I will redo them before they die :)


Its funny with the 1000mfd cap I put in there, the LED's stay dimly lit for like 8 minutes after I cut the power, dont want that baby to discharge into you :D
 
LEDs should run for years without problems, even ones that have been abused and still have some life left in them will usually keep going. That implies something connected wrong. This is how they should be connected.



If everything is connected like that you will only need the resistors Mohonri said. There is no way three 10 watt resistors should betting that hot, they aren't dissapating much power. If the cap can keep the leds lit for that long i's plenty big enough.
 
Cool, thanks alot for the schematic! :)

Seems easy enough, maybe ill do it today if I have time. I llready have a large assortment of resistors (1/2 watt) so I should be able to find something that is suitable. The only thing I need to get is a cap (sounds like the one that I have is a little overkill and is affecting my circuit too much.

Thanks again :)
 
Ok im glad to hear that I dont need a new cap, now I just have to divide the chains up into 2 rows of 5 (1 row of 5 and, 1 row of 6)
 
It looks like you are blowing LEDs because the current through them is too high. Maybe the voltage varies too much? The circuit below splits the LEDs into two groups, 5 and 6. The LM317 voltage regulators and 75 ohm resistors are wired as a constant current source. The current through the LEDs is 16 mA regardless of the voltage applied (within the range 19 to 31 volts applied). Current through the LEDs is determined by 1.2V / resistor (in ohms).
leds.jpg


I have used this circuit at work where I light several LEDs on a panel using 3.3V, 5V, 12V, and 15V sources and I want the LED brightness to be equal for all of them without calculating resistors. The same circuit works for all.
 
I think I many use your soution. I believe the voltage is indeed varying too much. One question, what exactly does the lm317 do? Like, does it just smooth out the voltage to a certian amount, or will only let so much voltage flow? Also, does it need to be tied to a heatsink?

Thanks for the solution :)
 
The LM317 is a variable voltage regulator. When wired as I showed above, it functions as a constant current source. Datsheet here >--- Link If the voltage applied is between about 19 to 31 volts, the current thorugh the LEDs will always be 16mA which is less than the limit for your LEDs. For 16mA of current you do not need a heatsink. They are fairly common so you can find LM317 at Radioshack or just about any decent electronic parts store.
 
Check the latest. I originally drew the "ADJ" pin to the wrong side of the resistor; latest is correct.
 
The fact that the voltage is not constant won't wear out your LEDs faster or damage them in any way. It just has the potential to give them a nearly-undetectable flicker at 120Hz. The reason they're burning out is because of overcurrent (caused by overvoltage). The LM317 method will work, but won't give you any additional benefit, except for maybe smoothing a power spike getting through the transformer.


 
Correction noted Frank4d, thanks.

Mohonri, so you are saying that the LM317 will not solve my problem? If not, will breaking the LED's up into chains of 5 and 6 fix my issue?

Thanks :)
 
DaRkF0g said:
Correction noted Frank4d, thanks.

Mohonri, so you are saying that the LM317 will not solve my problem? If not, will breaking the LED's up into chains of 5 and 6 fix my issue?

Thanks :)
Click to expand...

Breaking the LEDs up into chains of 5 and 6 will work if you use suitable series resistors, and if the voltage does not increase much which would cause the LED current to be too high.
 
DaRkF0g said:
Mohonri, so you are saying that the LM317 will not solve my problem? If not, will breaking the LED's up into chains of 5 and 6 fix my issue?
Click to expand...
The LM317 will work just fine. It's just a little more complicated, and isn't a great deal better than just doing the series-resistor thing.
Breaking the LEDs up into chains of 5 and 6 will work if you use suitable series resistors, and if the voltage does not increase much which would cause the LED current to be too high.
Click to expand...
Exactly. Just size the resistors for worst-case scenario, and you'll be fine. In this case, worst-case scenario is (31V - 5*3.5)/20mA ~= 680 Ohms (closest standard-sized resistor) on each leg.

 
Ok, so I finally got a chance to sit down and work on my amp. I ran out to radioshack and picked up 2 LM317 (well 317T) MOSFET's. I took all the LED's out, and took the huge 10W resistors out. After I got the MOSFET's put in I took a voltage measurement coming off of them, and my meter said 50V. So, I thought ok my meter finally went, so I tried another meter. Still had 50V.

So instead of hooking a chain of 5 LED's up to a 50VDC power soure, I decided to stop the project and seek some advice here.

Heres basically what my circuit now looks like:

LM31750V.jpg


So, why is it that I have 24V going in to my rectifier, and I have 50V coming out of it?

(Seesh, whenever I try and get rid of extra voltage I end up with more than double the voltage than what I started with!)

So as usual, any ideas?

Thanks in advance.

Update: I tried 5 LED's with the way I have it setup above. I didnt have SED's, but rather a few millaseconds of extremely bright LED's.

Need to go back and think about this one again for sure!
 
Oh and, I dont think this is good but, why is it when I toush the back of the MOSFET, and touch the casing of the amp I get this shocking feeling? Theres current traveling through me but its just a tingly feeling.

Also I disconnected the CAP and the Rectifier and got 23V, is it possible that the rectifier or cap is bad?
 
DaRkF0g said:
Oh and, I dont think this is good but, why is it when I toush the back of the MOSFET, and touch the casing of the amp I get this shocking feeling? Theres current traveling through me but its just a tingly feeling.

Also I disconnected the CAP and the Rectifier and got 23V, is it possible that the rectifier or cap is bad?
Click to expand...

This thread has been running now for almost two months and I have read the voltage is anywhere from 15v to (now) 50v.... depending on which post I am reading. I am thinking the voltage across the transformer you are connecting to is very sensitive to the load on it and is not suitable for driving LEDs.

You have probably dumped a lot of money into this mod so far without the results you desire. So... have you considered adding a low voltage power supply that will only drive the LEDs? You could buy a small supply for $20 that only powers LEDs, install it inside the case, and end your aggravation.
 
I was going to use that as a last resort, but I think I have no choice now.

The only thing I am worried about is the fact that DC transformers emit a ton of noise/interference. I dont want all of that line noise. (Try holding a speaker wire next to a DC transformer, you should know what I mean)
 
You must log in or register to reply here.
Back
Top