First Gainclone..with a 'switcher' and 'XO'

[mpmarino]

Hi Folks,

I've been a member for a short time and have bugged you guys a few times recently. Thanks for anyones input in the past. Anyway, my ultimate goal is to build a 4 channel 'powerhouse' Pass amp but that task is a bit daunting considering I have never built an amp before. ..

I have decided to go with a gainclone as a first electronics project to get my feet dirty. I am not new to electronics, just new to building amps. I am way more experienced with loudspeakers.

I have two problems: I hate to keep things simple ...and I am cheap (or maybe my wife keeps me cheap!) Therefore I am using anything I have on hand to complete the job.

My amp will incorporate a 'humpty' sort of. It will be a switching power supply at 56v capable of 15a robbed from a small surplused Lorain telecom power supply. The unit has virtually no ripple and will see extra filtering. It has a full 0-56v output so I will be splitting that w/ 2 caps and resistors. The power supply is behind the amp in the pic. It will be installed in a matching chassis to the amp.

The amp will be a non inverted LM3875 4 channel p-p with a 2 way active crossover internally. The crossover is an Audio Control 2 way automotive unit which will be removed from its chassis and incorporated into the amp. It is a super quiet unit and will probably be a great 'buffer'. Most importantly I had it kicking around and it was doomed to collect dust forever unless I did something with it.

The chassis' are made completely of 'junk' I had kicking around in my garage and my company truck. Sorry for the lengthly post. See pic. Whadya think?

Marc

[mpmarino]

"fist' gainclone?

I mean FIRST gainclone!

[li_gangyi]

from what I gather...you will be trying to split the 56V in to 2 post and neg rails?? I dun think it will be that easy...perhaps a GC with single rail supply will fit you more...or modify the SMPSU to give +/- voltages?

[mpmarino]

I have heard of splitting single supply to split supply. I have seen this done by running 2 caps in series (15000 uf in my case) across the + and - rails. Also paralleling a 1k or so resistor w/ each cap to 'lock down' the voltages.

I do have my doubts though: if there were an unbalanced draw, wouldn't the voltages get skewed? Would it throw the resistor ladder out of balance and cause lower voltage on the high draw rail? I think it will. Whadya you guys think...

Is this a viable option?

Thanks,
Marc

[li_gangyi]

exactly...the imbalance could throw the thingy off wack...which is why it should be better to obtain the slpilt supply anyother way...

[djQUAN]

if you bridge the outputs, splitting the PSU rails that way wouldn't be much of a problem. the center "ground" would just be a reference point and there is no risk of uneven loading of the rails.......

[jackinnj]

Originally posted by mpmarino
I have heard of splitting single supply to split supply. I have seen this done by running 2 caps in series (15000 uf in my case) across the + and - rails. Also paralleling a 1k or so resistor w/ each cap to 'lock down' the voltages.

I do have my doubts though: if there were an unbalanced draw, wouldn't the voltages get skewed? Would it throw the resistor ladder out of balance and cause lower voltage on the high draw rail? I think it will. Whadya you guys think...

Is this a viable option?

Thanks,
Marc

It's easier to run the GC in the "Single Supply" mode than to try to split the rails --

but if you must, run a pair of diodes, one from K on the "+" rail, A on the ground, the other A on the "-" rail, K on the ground -- this will help.

I have a half-dozen 48VDC Sorensen switchers made for the Telecom industry -- they are almost completely devoid of switching noise and ripple -- haven't run GC's with them yet., but will use the single supply option.

[mpmarino]

Thanks Guys,

I am convinced to go 'single supply'. That brings about more questions as I am just looking at it. I can see that the big disadvantage is having such large caps in series with the output to kill the DC. The speaks I am using I expect will never see (or want to see) below 60Hz or so. Can I get away with using a smaller cap?

Are the output caps supposed to be so large so they will not interfere with bandwidth, or are there other reasons as well?

I did plan already to use a series cap anyway on 2 of the channels as they will be direct coupled to my tweeters...so the single supply option doesn't change THAT much for me. I am buildg a small set of line arrays with 4.5" and .75 inchers run with this clone,....they will hand off to a sub run with a Adcom 555.

My real question is in regards to the size of the output cap, I guess. Can I get away with tailoring it to the impedance and freq. response of the speakers, somewhat.

Thanku
Marc

[mpmarino]

Anyone?

The obvoius answer to me is 'yes' a smaller cap can be used..but I may be missing something...

Also, anyone have any info on a single supply schematic other than the National datasheet?

Marc

[pinkmouse]

Originally posted by mpmarino
"fist' gainclone?

I mean FIRST gainclone!

Sorted!