rc4558 vs OPA4227PA

[crippledchicken]

hi, i'm looking for opinions as far as being compatible in the same circuit, and sound wise in a 250 watt rythmik audio plate sub amp. i realize, the 4558 is dual sil pack and the 4227PA is quad pack DIP but i can get around that part. i mainly wanted some thoughts about circuit, compatibility and whether the OPA4227PA chips would be an upgrade to the RC4558 chips soundwise since i have some 4227 chips on hand any way? thanks!

[crippledchicken]

i removed preamp board and the actual op amps are BA4558N SIP PACK also, has max. power dissipation rating of 900 mw. and i dought the OPA4227 rating is that high wasn't listed in its data sheet? any ideas on any other op amps in the 8 pin sip pack which would be an upgrade to these originals, would be appreciated! Thanks very much

[macboy]

You are lucky actually, since it is easy to replace a dual 8-SIP with a quad 14-DIP. If you look at the pinouts, you will notice that one side (7 pins) of the DIP match exactly with 7 pins of the SIP. The only thing missing is the other power line. So you just have to install a small jumper wire from that pin to the 8th SIP hole. You should also connect both input of both unused opamps of the quad to ground rather than leaving them 'floating'.

[crippledchicken]

hi macboy, thanks for the reply. i installed the opa4227 quad chips like you suggested the only problem i found was, with amp unhooked except for power cord the op amp closest to the amp inputs got really hot to the touch so i added .1uf stacked film caps fro power pins to ground but still ran hot? the other four chips did not get hot nor the original ba4558n chip. the ba4558n amps are rated 900mw. max, could the 4227 chips maybe not be beefy enough for replacements? i never hooked it up and ran it since that one op amp got so hot. thanks again!

[macboy]

hi macboy, thanks for the reply. i installed the opa4227 quad chips like you suggested the only problem i found was, with amp unhooked except for power cord the op amp closest to the amp inputs got really hot to the touch so i added .1uf stacked film caps fro power pins to ground but still ran hot? the other four chips did not get hot nor the original ba4558n chip. the ba4558n amps are rated 900mw. max, could the 4227 chips maybe not be beefy enough for replacements? i never hooked it up and ran it since that one op amp got so hot. thanks again!

Good idea to to add power supply decoupling. There is no such thing as a supply that too well decoupled. Try a small cap across the two supply pins as well (mind the voltage rating!).

The opamp is probably either unstable (oscillating), or connected wrong. Double check that it is connected correctly, not backwards, or with +/- supplies mixed up. The 4227 should be unity-gain stable, so that isn't it. But it is a very fast opamp, much faster than the 4558. There are a few things you could try. It is possible that there is some RF noise pickup, which the slower 4558 didn't respond to, but the fast 4227 is. Try adding a very small cap (<100 pF) across the feedback resistors; this creates a low-pass filter well above the audio band, to help reject RF. You should do a quick calculation to make sure that the reactance of this cap at 20000 Hz is greater than 10x of the value of the feedback resistor to avoid messing with the audio frequency response. ( Xc = 1 / ( 2*Pi*F*C ) where F is in Hz and C in Farads). A small cap, around 100 pF, connected from the input jack signal pin directly to ground, can help reject RF pickup. You could also try a small cap (guessing 10 to 33 pF) across the opamp's two input pins.

Also you could change the connection of the unused opamps. Short the output to inverting input (to create a voltage-follower) and short the non-inverting input to ground. These unused pins should not be left floating, as that is an invitation for oscillations especially with FET input opamps.

Another possiblity is simply a bad chip. Try swapping it for another.

[crippledchicken]

that's alot of great info Thanks! should keep me buisy for awhile lol! yeah, i added 2 1k. 50v caps in the middle of the four op amps that are close together for the +/- 12v. which feeds all five chips but, i'll try more work at the problem chip and see what i can come up with. i thought about ordering some of the brown dog adapters to adapt from sip to dip since i also, have some opa 2134/2604/2132 chips i could try also as suggested by djk. you think one of those chips might be a better choice maybe the 2604? thanks very much for your help!

[kevinkr]

hi Crippled Chicken, just wondering what prompted you to upgrade the op-amps in this plate amp? Does it have a line level x-over to your other speakers?

In most applications op-amp upgrades can make a very large difference, in this case if this is just a sub plate amp with internal line level x-over I wonder how much if any of a difference you will hear.

It's true the 4558 is a fairly mediocre op-amp for high performance audio applications, but if its operating over a limited low frequency range it is a pretty reasonable choice in terms of slew rate, noise performance, and stability. The 4558 is pretty forgiving in poorly decoupled designs where a higher speed device might oscillate.

Hope that none of those op-amps are inside the power amplifier loop, because substituting a faster part may very well require changing the loop compensation to keep the amplifier from oscillating and destroying its output devices.

Do you have a schematic you could post here? Also do you have a scope, meter, generator and the like?

You may be able to glue (thermal epoxy) a small heat sink on the hot running op-amp once you ascertain that it is actually running within its design limits. Small heatsinks for this purpose can often be purchased surplus on ePay or new from Mouser or Digikey or you can make one.

[crippledchicken]

Originally posted by kevinkr
hi Crippled Chicken, just wondering what prompted you to upgrade the op-amps in this plate amp? Does it have a line level x-over to your other speakers?

In most applications op-amp upgrades can make a very large difference, in this case if this is just a sub plate amp with internal line level x-over I wonder how much if any of a difference you will hear.

It's true the 4558 is a fairly mediocre op-amp for high performance audio applications, but if its operating over a limited low frequency range it is a pretty reasonable choice in terms of slew rate, noise performance, and stability. The 4558 is pretty forgiving in poorly decoupled designs where a higher speed device might oscillate.

Hope that none of those op-amps are inside the power amplifier loop, because substituting a faster part may very well require changing the loop compensation to keep the amplifier from oscillating and destroying its output devices.

Do you have a schematic you could post here? Also do you have a scope, meter, generator and the like?

You may be able to glue (thermal epoxy) a small heat sink on the hot running op-amp once you ascertain that it is actually running within its design limits. Small heatsinks for this purpose can often be purchased surplus on ePay or new from Mouser or Digikey or you can make one.

hi kevinkr, sorry i don't have schematic for the rythmik audio basic 250 amp and checked with them and they had none available for the general public. something i noticed is that from the main board containing ps/rect/outputs etc. is a bundle of wires about 6 or 8 tiny, i'd guess maybe 24 guage or so. it seems the +/-12v. op amp power feed is being fed from those as well, as the ground for the preamp board? could this be considered poor circuit layout just asking, for oscillation problems? should i consider running larger gauge wires especially, ground wire to reduce chances for oscillation? i have no special equipment just a simple Fluke meter to test with. thanks very much! also, amp is run using speaker level input, and full signal to mains.

[kevinkr]

Hi crippledchicken,
It all depends on how long those wires actually are, if they are only 6" or less inches long there is little benefit to changing them to a heavier gauge, and always something of a risk of mis-wiring.. (ask me how I know that.. LOL)

There is always a certain amount of risk in modifying something if you don't have the schematic and haven't figured out the topology and/or the reasons behind the design choices made.

One area where you might derive some benefit would be improving the power supplies. Better/Larger electrolytics in the supply, less lossy rectifier bridges, and in many cases beefing up of higher current traces with heavy wire will buy you something. Usually a little more power and a little less distortion.

Have fun!