http://www.linkwitzlab.com/STK4050V_1_3.pdf
can i use this with 45-50V rails? or is the maximum supply voltage the difference between the rails?
can i use this with 45-50V rails? or is the maximum supply voltage the difference between the rails?
The spec-sheet is pretty clear.
With +/-66V it will make 200 watts in 8Ω. 200W in 8Ω is 56 volts peak, so they are allowing 10V drop across the chip at full rated power, not unreasonable at the rated distortion. It clearly must have more than 112V DC across the chip to make that power; actually they want 132V rail-to-rail to cover losses and leave some bias on peaks for low THD.
Then they cite output noise and "neutral voltage" at +/-80V conditions. There is noise even when an amp is at full power, but we usually measure noise at zero signal. "Neutral voltage" seems to be DC offset, also most conveniently measured at zero signal. So they seem to be assuming a power supply that is +/-80V at no-signal, +/-66V at full-power, a fairly reasonable spec.
Then there is a maximum rating of +/-95V. The "+/-80V" will wander up and down with utility company voltage; they are allowing a reasonable 18% leeway from "normal idle voltage" to over-voltage blow-up rating. I suspect they really use 200V rated transistors and shave 5% safety factor when they rate the whole assembly. (BTW: this is not a monolithic "chip", it is a built-up hybrid.) I would expect 99 out of a hundred STK4050V to survive idling at 200V or even 250V (+/-125V) for days on end, though putting high signal through at that voltage might kill a bunch of them.
> i have to build 10 amps
This is a 200 watt amp. The power input when making 200W is at least 300W, and tends to be more like 400W at the wall-power side of the power transformer. Times 10 is 4,000 watts. Few homes have that much power on one circuit. Most home fuses will pop just tring to start-up the power supply you'd need for ten 200W channels.
> can i use this with 45-50V rails?
Yes.
If you mean +/-25V, this hybrid is way bigger than you need, and may not give its best THD at such low voltage, but it will certainly work. Power output is about 25W in 8Ω. The "GainClone" type chips are probably a much better plan at that power level.
If you mean +/-50V no-load sagging to +/-45V at full-load: power output is about 75 watts at low THD, touching 100W when clipping gets gross. Ten channels of that "can" be fed from common household outlets, but you probably need some soft-start system to keep from popping house fuses at switch-on.
> use 4 amplfiers instead of 2 for the 10" xls...
Note that these modules are rated 8Ω. At these power levels, a too-low impedance will kill an amp like this. You can not parallel these amplifiers; and you can't bridge them except in 16Ω load.
The Sanken (now Sanyo) modules are OK amplifiers, nothing special. Sound just like a good 1980 Japanese receiver. They do not have good short protection and do not have overheat protection. They tended to be pretty robust even under abuse, and I'm sure they would drive 8Ω speakers to 200W all day long if the heatsink temperature was not high enough to burn your finger.
Can you actually get these modules today? I used a LOT of their 30V and 80W modules back around 1980. I have never seen them used in large production. Being hybrid, they are stock parts assembled on a machine that can do many types of modules, including CATV amplifiers which do have a steady market. But be sure you can get these modules (and spares) before you plan a box around them.
With +/-66V it will make 200 watts in 8Ω. 200W in 8Ω is 56 volts peak, so they are allowing 10V drop across the chip at full rated power, not unreasonable at the rated distortion. It clearly must have more than 112V DC across the chip to make that power; actually they want 132V rail-to-rail to cover losses and leave some bias on peaks for low THD.
Then they cite output noise and "neutral voltage" at +/-80V conditions. There is noise even when an amp is at full power, but we usually measure noise at zero signal. "Neutral voltage" seems to be DC offset, also most conveniently measured at zero signal. So they seem to be assuming a power supply that is +/-80V at no-signal, +/-66V at full-power, a fairly reasonable spec.
Then there is a maximum rating of +/-95V. The "+/-80V" will wander up and down with utility company voltage; they are allowing a reasonable 18% leeway from "normal idle voltage" to over-voltage blow-up rating. I suspect they really use 200V rated transistors and shave 5% safety factor when they rate the whole assembly. (BTW: this is not a monolithic "chip", it is a built-up hybrid.) I would expect 99 out of a hundred STK4050V to survive idling at 200V or even 250V (+/-125V) for days on end, though putting high signal through at that voltage might kill a bunch of them.
> i have to build 10 amps
This is a 200 watt amp. The power input when making 200W is at least 300W, and tends to be more like 400W at the wall-power side of the power transformer. Times 10 is 4,000 watts. Few homes have that much power on one circuit. Most home fuses will pop just tring to start-up the power supply you'd need for ten 200W channels.
> can i use this with 45-50V rails?
Yes.
If you mean +/-25V, this hybrid is way bigger than you need, and may not give its best THD at such low voltage, but it will certainly work. Power output is about 25W in 8Ω. The "GainClone" type chips are probably a much better plan at that power level.
If you mean +/-50V no-load sagging to +/-45V at full-load: power output is about 75 watts at low THD, touching 100W when clipping gets gross. Ten channels of that "can" be fed from common household outlets, but you probably need some soft-start system to keep from popping house fuses at switch-on.
> use 4 amplfiers instead of 2 for the 10" xls...
Note that these modules are rated 8Ω. At these power levels, a too-low impedance will kill an amp like this. You can not parallel these amplifiers; and you can't bridge them except in 16Ω load.
The Sanken (now Sanyo) modules are OK amplifiers, nothing special. Sound just like a good 1980 Japanese receiver. They do not have good short protection and do not have overheat protection. They tended to be pretty robust even under abuse, and I'm sure they would drive 8Ω speakers to 200W all day long if the heatsink temperature was not high enough to burn your finger.
Can you actually get these modules today? I used a LOT of their 30V and 80W modules back around 1980. I have never seen them used in large production. Being hybrid, they are stock parts assembled on a machine that can do many types of modules, including CATV amplifiers which do have a steady market. But be sure you can get these modules (and spares) before you plan a box around them.
Well, over here it seems to be even more easy to find than an LM3875
http://www.bmm-electronics.com/Product.asp?Product_ID=341
It seems like a very cool chip for a subwoofer, no bridging necessary if you only need 200W or so
good luck
http://www.bmm-electronics.com/Product.asp?Product_ID=341
It seems like a very cool chip for a subwoofer, no bridging necessary if you only need 200W or so
good luck
I have six 1500VA 2*38V transformators, thats why i look for a chip that can handle higer voltage... (i know i need softstart or something)... As it is an active system most of the amps will not use much power... at most 50W or so for 4 channels, 100W for another 4 and around 200W for the last 2 
i didn't mean i would bridge the amps for the four xls 10" but to use one amp for each...
i didn't mean i would bridge the amps for the four xls 10" but to use one amp for each...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.