Large power supply with Leach Amp

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I plan on using 2 68000uF 50V caps in series on each side of my leach amp power supply connected to a 80VCT 800VA torodial transformer. I'm going to be driving some loads that dip into the 2.85ohm range and I'm wondering if the 200V 25A Bridge Rectifier (DigiKey 26MB20A-ND) will be able to handle the current or if I'd need a different one. Any other advice about this power supply design would be helpful as well.
 
You are better off using a bunch of 10000 at 80 volt caps and a larger bridge like 35 amp at 400 volt. With series caps that big you are not going to have the same actual capacitance on both the plus and minus sides as the tollerance of caps is 20% or more. You should at least balance the caps with 1% resistors so each gets the same voltage. Using them to save money is going to hurt the project rather than help by getting a big supply out of it, after all, 34000 isnt that big.
 
I have several 50v 68000uF caps. Let me know if interested. (My cap meter only reads till 20,000uF otherwise I would have matched them for you.)

If you plan on buying 3 pairs, I could match the closest (by series connecting and measuring.... but I guess 3X 68000uF per rail would be too much of a killer. :D
 
Interesting, I hadn't even considered the tolerances. What do you think about 3 80V 12,000uF caps wired in parallel on each side? Would it still need the resistor? If so, please explain how I would use a resistor to balance the voltage... Also, I was under the impression that 34,000uF per side was a VERY large amount of capacitance for this amp, is that not true?
 
Yes, 36,000 is big! Use the 3 x 12,000uF in parallel to get 36,000uF per side and you will have a very large power supply. In parallel you dont need to worry about the tolerance of the caps. I use 16 x 6,800 in one of my amps! Now that is big, but there are bigger.
 
Why is it that I don't have to worry about tolerances if the capacitors are wired in parallel, but I do if they are in series? Is there a way to solve the for the size of the transformer (in Amps), the voltage rating and the amperage rating of the bridge rectifier, and the amount of capacitance I will need in order to drive a load of a particular resistance? For instance:

I want to know how large of a transformer, what kind of bridge rectifier, and how much capacitance I will need in the power supply to be able to power a 2 channel Leach Amp, each channel driving a 2ohm load.
 
Hmm, I don't have a capacitance meter so there really isn't a good way to tell if they are "close'. What would happen if one side of the power supply has more capacitance than the other? Is there a way to solve the for the size of the transformer (in Amps), the voltage rating and the amperage rating of the bridge rectifier, and the amount of capacitance I will need in order to drive a load of a particular resistance? For instance:

I want to know how large of a transformer, what kind of bridge rectifier, and how much capacitance I will need in the power supply to be able to power a 2 channel Leach Amp, each channel driving a 2ohm load.
 
Many well respected people use the yardstick of 10,000 uF per rail for a 100 wpc amplifier (2 channels) so for 2 ohm loading, this amp would need 40,000 uF per side... this is more than enough.

The krell KSA-250 uses 80,000uF per rail and that can drive a 1 ohm load to 2000watts!

Some 375 watt carver used 13000 uF per rail. So as you can see there is a lot of variance and they all are proven commercial offerings.

BTW Motion, a 80VCT Trafo will give you rails of +/-55 volts (or loaded 53v) and that will probably fry your 50v caps.....
 
K-amps said:
Mark I am curious, why dont you need to worry about tolerances ina parallel situation? I guess with the law of averages, you are better off but you still could have a mis-matches set... right?
It's not about averages, it's about how they add up. In series they are added like resistors in parallel but in parallel they are just added up. In the series setup if they are unevenly matched the caps will not split the voltage between them evenly causing possible problems if the split is too one-sided. You could use 1% power resistors to even out the voltage between the two caps but why waste the power. There are also issues of wasting half of the capacitance and increased resistance. Keep it simple and you actually get a better supply with lower ESR if you just use multiple smaller uF caps with the right voltage rating in parallel.
 
I feel like I'm beating a dead horse here, but I want to make the best choice. If I use multiple smaller uF caps in parallel (say, 3 12,000uF 80V caps per side) as opposed to larger uF caps in series (say, 2 68,000uF 50V caps per side) won't I run into the problem of possible uneven capacitance? I guess I'm asking what the better solution is, which is going to result in the largest stable resevoir of power for transients. Also, ESR = electrostatic resistance? What exactly does that refer to?
 
m0tion said:
I feel like I'm beating a dead horse here, but I want to make the best choice. If I use multiple smaller uF caps in parallel (say, 3 12,000uF 80V caps per side) as opposed to larger uF caps in series (say, 2 68,000uF 50V caps per side) won't I run into the problem of possible uneven capacitance? I guess I'm asking what the better solution is, which is going to result in the largest stable resevoir of power for transients. Also, ESR = electrostatic resistance? What exactly does that refer to?
ESR= equivolent series resistance, lower is better.
It really isn't about being uneven if they are in parallel, it becomes a problem when they are in series as one works harder than the other if they are mismatched. In parallel they each just do what they can, one does not rely on the other to work as they do in series. With a large number of smaller caps the law of averages will work in your favor and electrolytics are usually on the plus side of error.
 
I've been using a Leach Low TID amp for around 5 years now. Currently it drives a pair of 2-way LS. The 20W75/4 bass-midrange units have a DC resistance of 3.2 ohms, which is not much higher than your anticipated 2.85 ohms.

Burst mode testing (8 cycles on, 24 cycles off) into 4 ohms, results in clipping at 180W. Slew rate measures 70v/uS, as per spec.

Power supply components are:

Txfmr 40v + 40v @ 200VA Toroidal

Rect pair of paralled MDA970-3 rects (200v 8A), one for each channel

Caps 4700uF + 4700uF (70v 10A ripple rating) per channel

Output devices 2 off, TIP35c + TIP36c per channel.

I'm not into home theatre, but my music ranges from opera and hefty organ music, to Genesis and the Eagles, with a lot of other stuff in between. And if I'm listeningto a piece from the organ at Notre Dame, I like it to sound like I'm there ........

My listening room size is about 16' x 10'.

So ........ I would say the Rect you have in mind would be more than adaquate, and the Caps are way over what you really need - even when connected in series - I'd say 2 x 10,000uF per channel would be more than enough. If your thinking of using the
amp to drive the heater for your central heating .... well .... an 800VA Txfmr might be needed, and the radiation from the output device heat sinks would probably raise your listening room temperature by a few degrees before they melt into a pool of loose silicon! And if you really think you're going to need 800VA, you will need some real big heat sinks, and a big case to put it all in, too!

But, whatever way you go, I doubt you'll be disappointed with the Leach Amp.

Regards, and good luck

JohnT
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.