IMHO When it clips e.g. on positive output, the upper output device will saturate and the opamp output will rise to about +14V, causing the upper driver to dump about 0.5 Amp into the base of the output device.
Because it's heavily saturated it will be slow to turn off when the base current is removed, causing the output to "stick" to the + supply rail.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The easiest way to a happy future is probably to keep the power supply and the pre-amp circuitry, but dump the entire power-amp section and replace it with something else e.g. a chip-amp.
I don't know what the supply voltages and power output are supposed to be, but something like an LM3886 only costs about 10 dollars, could probably be bolted to the existing heatsink, has totally stupid-proof protection circuitry and offers better performance than the existing amp. A cheap and easy (and good) construction option would be point-to-point wiring on the chip.
Cheers - Godfrey
Last edited:
Not mine 😉 FWIW I noticed QSC adds free wheeling diodes across their OP devices on all their later models.
So this is for holding up the op-amp supply rails under hard clipping, So does this prevent the thing from blowing up or not? I'm not convinced that added KLH diodes will do anything as strong as QSC does here. Seems good to add these QSC diodes >
So dont adj for current on the OP emitters?
and the added 2K will keep the loop happy I guess
Does he need to decrease the drivers emitters resistors by 1/2 and add QSC C4 as well?
Seems sensible! Costs less than one Sanken OP device. LOL
So the LM3886 supply needs to be less than plus or minus 40 Volts. And speaker at 8 ohms. check
Should the OP increase the chip amp gain to match this one?
BTW Thanks for the "sticky" explanation.
Hi Brian
The amp has only 1 pair of output transistors and there are also no current limiting circuitry. Combined with a 1000W rms subwoofer, it is not difficult to blow the power transistors if you play loud.
Regards
Mike
10 inch car sub with super low sensitivity and cone area, begs the user to get the cone moving, so indeed keep off the gain knob is wise future advise. The stock amp design does have current limit but it's probably damaged or just too slow to save cross conduction faults. After the stock amp fixes are in-place.. It might be something to run one 4 ohm coil and leave the second open. This will test the current limit protection before hard voltage clipping.
Can i add two more output transistors to this set up off the existing drive stage this way i have two PNP & NPN and give it a little more power. Like This
An externally hosted image should be here but it was not working when we last tested it.
Yes, you can add parallel output devices.
It is important that the parallel devices match each other. That effectively means replacing all the output devices with matched pairs.
You must add emitter resistors to all the output devices. I suggest you also add base stopper resistors to all the outputs.
This will reduce the load on the driver stage.
It will reduce the junction temperature of the output devices.
It will run the output devices within their temperature de-rated SOAR if you deliver the same power to the load.
It will increase the current ability of the output stage.
If the power supply cannot deliver the current demanded by the amplifier you will gain nothing in extra current ability.
BF No that won't solve your problem. The only way you'll get more power is by lowering the subs impedance ie use one coil. You'll not get this one to drive lower than 4 ohms reliably.
You need to solve the mystery of why this design is failing first. You have an 8 ohm speaker right? Is this thing oscillating ie R52 burned? Does the SOA protect work? What are the rail voltages. What is the power transformer rated for? Is the bias adjusted correctly? Did KLH ever mod any units with the diodes as they show?
Study the QSC design for clues to get this one working properly. The transistors advised in post 3 will out class the QSC ones.
You need to solve the mystery of why this design is failing first. You have an 8 ohm speaker right? Is this thing oscillating ie R52 burned? Does the SOA protect work? What are the rail voltages. What is the power transformer rated for? Is the bias adjusted correctly? Did KLH ever mod any units with the diodes as they show?
Study the QSC design for clues to get this one working properly. The transistors advised in post 3 will out class the QSC ones.
Yes its a 8ohm and the woofer not oscillating normal play and R52 is in perfect condition now for the rail voltage is 37+ 37- and the transformer rating the sticker gone but it puts out 250ac each side and for the bias i have no idea this is a new field for me so i am still learning that why i asking you guys (The Pro's) to make sure i am doing it right. And for KLH as far as i know they did not add the diodes that changed the design and the outputs they now use TIP35C & TIP36C the driver stage is2SD1563 &2SA1249.P.S the SOA protect works.
It's up to you which way to proceed. Think it over as to which takes more technical ability, time, interest, and expense with possibly no guaranteed results.
1) You can take some suggestions from the above posts to mod the existing design and play with the bias adjust. You will get the most power here using a 4 ohms woofer.
2) Or mount up an LM3886 chip amp and build it up on some perfboard. You will only be able to run 8 ohms using the existing +/- 37 Vdc supply. You can get around 60-70 Watts output. Not as much power as above.
1) You can take some suggestions from the above posts to mod the existing design and play with the bias adjust. You will get the most power here using a 4 ohms woofer.
2) Or mount up an LM3886 chip amp and build it up on some perfboard. You will only be able to run 8 ohms using the existing +/- 37 Vdc supply. You can get around 60-70 Watts output. Not as much power as above.