Reducing voltage of TL494 based supply?

[theAnonymous1]

I have a donor supply from a car amp that I would like to use with some amplifier modules that take +-35v max. The supply gives dual +-40v output @ 13.8v input. I would like to reduce to closer to +-30v @ 13.8v input.

I would like to know if the voltage can be adjusted without rewinding the toroids? The toroids are a PITA to remove because I only have a low wattage iron.

The chip is actually a KIA and not TL, but I doubt it matters. There is also a KIA4558S dual op-amp in part of the circuit.

 

[luka]

Hi

Yes you can, you will have to change feedback. I assume it is basic resistor devider. It is probably sensed from + to GND, via two resistors, one big, other is smaller(value). Smaller stays the same, you will have to change only upper one.

You could have feedback done too with those op-amps, but I think they are over and under voltage settings. You may need to thange those too.

Could you take one big/high resolution pic of whole board from top and bottom and send them to my mail. I would like to see how they have connected toroids... Thanks..

 

[theAnonymous1]

Sure, I'll send some high res pics.

If I can't manage to lower the voltage the correct way do to my lack of skills, would say 8-9 high current diodes in series on each rail work? I know, that's a completely lazy and noobish approach to the problem.

 

[theAnonymous1]

Luka, I uploaded the files to a host instead of e-mailing them do to the size.

http://s000.tinyupload.com/?file_id=11303397806307010986

 

[luka]

hi

We will do this the right way, no diods in series.
First thimg for you is to find feedback, which probably will be connected to pin1, pin2 will constant voltage, made also by resistor devider.

You could also draw how is 494 around pin1, pin2...

 

[theAnonymous1]

I am no good at drawing a schematic.

I hope this is understandable.....

Pin 1 (+IN) ---> 33k (R501) ---> +Vs
Pin 1 (+IN) ---> 1k (R503) ---> 6.8k (R505) ---> Pin 4 (Dead Time)
Pin 1 (+IN) ---> 1k (R503) ---> box cap? (C508) ---> Pin 5 (Ct)
Pin 1 (+IN) ---> 1k (R503) ---> 3.9k (R506) ---> Pin 6 (Rt)
Pin 1 (+IN) ---> 1k (R503) ---> box cap? (C526) ---> Ground
Pin 1 (+IN) ---> 1k (R503) ---> Large 1k (R512) ---> Ground
Pin 1 (+IN) ---> 1k (R503) ---> Transistor A1023 (Q503 & Q502)
Pin 1 (+IN) ---> 1k (R503) ---> 1k (R509) ---> Pin 10 (E2)
Pin 1 (+IN) ---> 1k (R503) ---> 1k (R510) ---> Pin 9 (E1)
Pin 1 (+IN) ---> 1k (R503) ---> Jumper to op-amp circuit

Pin 2 (-IN) ---> 22k (R504) ---> box cap (C507) ---> Pin 3 (FB)
Pin 2 (-IN) ---> 33k (R502) ---> -Vs

 

[luka]

Hi

I will try to decode this and get back at you tomorrow. I will draw shema for myself for better understanding. So until tomorrow...

 

[djQUAN]

most amps are unregulated. that might be the case in this one.

try running the supply with a variable supply and vary the input voltage from 11-15V and watch the output voltage (use some load of sort like bulbs etc)

if it is indeed unregulated, adding a resistor from the deadtime pin to the +5Vref pin of the TL494 decreases PWM modulation so that should lower the output voltage.

from looking at the pic, try measuring the voltages across R507 (1M) it looks like this component in parallel with C509 appears to be the soft start circuit. if the voltage across this is around 5V, you replace R507 wiith a 1Mohm pot and decrease resistance until you get to the desired output voltage.

 

[theAnonymous1]

Thanks for the input djQUAN.

The supply is indeed unregulated. At 12.5v input the output is down to around +-35v. Yes, R507 is in parallel with the 220uf 16v cap. I will measure the voltage across it in a bit.

Is this the best way to lower the voltage, or is it better to change the feedback?

Thank you both for your input and time.

 

[luka]

Hi

Ok now is it regulated or not? If not then there is no feedback to change and only way to change output voltage would be with rewinding trafos.

 

[djQUAN]

nope. the best way to lower the voltage is rewind the transformer.

the one I posted earlier is the easiest.

there is no feedback as you already know, the supply is unregulated. unregulated supplies don't have feedback since they don't need to control the output voltage.

 

[theAnonymous1]

nope. the best way to lower the voltage is rewind the transformer.

the one I posted earlier is the easiest.

there is no feedback as you already know, the supply is unregulated. unregulated supplies don't have feedback since they don't need to control the output voltage.

Well color me red.

I measured 4.97v across R507. Should I attempt to lower the voltage as you described, or is this not a good idead after all? What are the downsides of this? I REALLY do not want to rewind the transformers.

Again, thank you both. Some people are just hopeless and have to face the humility of needing others brain power. I try to keep my begging to a minimum here.

 

[luka]

Hi

I am at lost here and you have nothing to lose so try what djQUAN said.

 

[djQUAN]

if you reduce the voltage as I described, it should be fine as there are output inductors.

as far as downsides, I think the mosfets will see a higher peak current flowing through them but I'm not a guru of SMPS's so I really can't tell. but it should atleast work without blowing up.

 

[theAnonymous1]

I'm not as worried about blowing it up as I am about touching the 80v on the output again.

I just tried a 2Mohm resistor across R507 and it didn't change anything. Is this value too high? I don't have any trim pots in that range.

EDIT: OK, I've tried 500k in parallel and still no change.

Yeah yeah, I know.... "Just rewind the stupid toroids you lazy noob".

 

[luka]

Hi

So you will rewind it, no big deal. I have rewinded my trafo for offline smps, well close to ten times. You will have to do only once...

 

[theAnonymous1]

OK, before I do this let me ask a few questions so I can do it correctly the first time...

1) Does it matter which side is rewound; the primary or secondary? I ask because the primary side is on top of the secondary and It would be a lot easier to just rewind the primary.

2) I assume I can't just back off some turns; they need to be placed evenly over the core?

Now all I need is to go buy a higher wattage iron.

 

[theChris]

rewind the secondary. the calculations for switching loss and for magnetizing inductance are based upon primary turns. removing a turn on the primary can increase the current, as well as the peak flux seen by the core. depending on design this will either increase power losses OR allow the core to saturate which would be worse.

 

[theAnonymous1]

Well, I don't think I will have to rewind now. I was playing around some more and managed to lower the voltage by adjusting the resistance of R501. This was the 33k from Pin 1 to +V. I just removed the resister and threw in a 50k trim pot. I had tried this earlier, but I guess the value I used was too high.

Tomorrow I will put it back on the heatsink and connect some high power 20 ohm resistors on each rail and see what happens. Hopefully no smoke.

So does this mean the supply IS regulated and I'm just a retard? Maybe my test load (1k) wasn't drawing enough current for it to regulate?

EDIT: I guess it is regulated. Now I get the same voltage on the output at 12.5v and 13.8v input. One strange thing though; one of the supplies is about 2v higher than the other. One is at +-30v and the other +-32v. No big deal as long as it doesn't blow up.

EDIT EDIT: OK, very strange. The supply that is 2v higher completely collapses under load. The other supply only drops about 0.5v on the + rail and 1v on the - rail with 20R across them.

SUPER EDIT::::::: The supply that collapses only does it with the feedback(?) adjusted. If I set it back to 33k the rails don't collapse, they just drop from +-40v to +-35v with a 20R load. Why would only one of the supplies do this and not the other?

 

[theAnonymous1]

SUPER MEGA NOOB EDIT:::::::::::: At the risk of setting my house on fire, and on a whim of insight; I paralleled the supplies. No fire, no explosions, and both the supplies settled to +-34v; down from the +-40 when both separated.

I don't know if you noticed from the picture, but I chopped of the other half of the board that had the power amp section. I never checked to see if the two separate supplies came together on that half of the board before I did the cut and threw it out.

Man, talk about a steep learning curve.