okay, i'm working on a memphis pr-500.1:
i installed 8 irf3205's in place of the irfz44n's. i scoured the data sheets of both, and they are nearly identical, save for the 3205's having more current capability, a little bit higher gate charge, and slower response time. should work fine on paper. they are driven by a pair c945's for the 8. it powers up and plays just fine, and stays perfectly cool. i ran it for over 10min, over 10a draw, which i am assuming is 100w+ for now.
what concerns me is the how and the range of the regulation. my fluke seemed to be having an intermittent problem on dc with latching/ picking up dc, v/hz/%, since i was wiring up a bunch of lights on my house outside this evening. had to switch it to ac, and then back and then it seemed to work just fine. so, i took several repeated measurements to be sure.
here is what i found at the gate resistors:
at full duty and initial power-up, it would reach 3.497v @40khz, 49 to 50% duty cycle. this is fine by me, and it runs perfectly stable.
at maximum regulation and lowest duty, (13.6+v supply @ idle) it would drop down to .18xv @8.6khz and 2.9% duty cycle
in between i was getting different readings, and the voltage seemed to ramp in a hysteresis at 12.88, and 12.38 on my power supply readout/adjustment. the readings are as follows:
gate charge voltage:
increase to 12.88= +2v ramp down to +0.512, constantly variable matched with supply down to the .19 minimum.
decrease to 12.38minus= +1.2v ramp to +3.5v fixed
switching frequency hz:
increase to 12.88= 39khz ramp down to 12.5khz constantly variable matched.
decrease to 12,38= 18khz ramp up to 39khz fixed.
duty cycle:
increase to 12.88= 49% ramp slowly to 6.5 to 7% minor variation.
decrease to 12.38= 13% ramp moderate to 49% fixed.
..... so my main concern, is if it is okay for them to be in the ranges of +.2 to +1.2 on the gate charge and as low as 2%duty? i did measure for voltage drop across the 47ohm gate resistors, and there really isn't one. that is where my first thought was headed, as i was taking my readings on the gate side of the resistors. also, both the 44's and the 3205's have 2v listed as the minimum gate charge in the characteristics.
other question i have out of curiosity...
i know frequency changes inductance some, but does it really reduce power transfer that much? it is in the low power modes 50w and under, or with low supply voltage, and it appears that the frequency is the main regulating tool.
I would like to put it to work on a full load, but i'm being a little tentative as not to send it up in flames in my tahoe. i enjoy the smell of leather over burned amp
am i just being paranoid?
thanks in advance,
-Heath H-
i installed 8 irf3205's in place of the irfz44n's. i scoured the data sheets of both, and they are nearly identical, save for the 3205's having more current capability, a little bit higher gate charge, and slower response time. should work fine on paper. they are driven by a pair c945's for the 8. it powers up and plays just fine, and stays perfectly cool. i ran it for over 10min, over 10a draw, which i am assuming is 100w+ for now.
what concerns me is the how and the range of the regulation. my fluke seemed to be having an intermittent problem on dc with latching/ picking up dc, v/hz/%, since i was wiring up a bunch of lights on my house outside this evening. had to switch it to ac, and then back and then it seemed to work just fine. so, i took several repeated measurements to be sure.
here is what i found at the gate resistors:
at full duty and initial power-up, it would reach 3.497v @40khz, 49 to 50% duty cycle. this is fine by me, and it runs perfectly stable.
at maximum regulation and lowest duty, (13.6+v supply @ idle) it would drop down to .18xv @8.6khz and 2.9% duty cycle
in between i was getting different readings, and the voltage seemed to ramp in a hysteresis at 12.88, and 12.38 on my power supply readout/adjustment. the readings are as follows:
gate charge voltage:
increase to 12.88= +2v ramp down to +0.512, constantly variable matched with supply down to the .19 minimum.
decrease to 12.38minus= +1.2v ramp to +3.5v fixed
switching frequency hz:
increase to 12.88= 39khz ramp down to 12.5khz constantly variable matched.
decrease to 12,38= 18khz ramp up to 39khz fixed.
duty cycle:
increase to 12.88= 49% ramp slowly to 6.5 to 7% minor variation.
decrease to 12.38= 13% ramp moderate to 49% fixed.
..... so my main concern, is if it is okay for them to be in the ranges of +.2 to +1.2 on the gate charge and as low as 2%duty? i did measure for voltage drop across the 47ohm gate resistors, and there really isn't one. that is where my first thought was headed, as i was taking my readings on the gate side of the resistors. also, both the 44's and the 3205's have 2v listed as the minimum gate charge in the characteristics.
other question i have out of curiosity...
i know frequency changes inductance some, but does it really reduce power transfer that much? it is in the low power modes 50w and under, or with low supply voltage, and it appears that the frequency is the main regulating tool.
I would like to put it to work on a full load, but i'm being a little tentative as not to send it up in flames in my tahoe. i enjoy the smell of leather over burned amp
am i just being paranoid?thanks in advance,
-Heath H-
Re-check them. Place the black probe on pin 7 of the 494. Place the red probe on pin 9 or 10.
If you want the duty cycle at the transistor, place the black probe on the source leg and the red probe on the gate leg.
If this is a self-oscillating amp, I doubt that the class D circuit is synchronized with the power supply. If it's not, the frequency should not be changing on the power supply.
If you want the duty cycle at the transistor, place the black probe on the source leg and the red probe on the gate leg.
If this is a self-oscillating amp, I doubt that the class D circuit is synchronized with the power supply. If it's not, the frequency should not be changing on the power supply.
the only change in readings is the voltage goes up to +3.935 at full power across pin 7, and 9, where it would be +3.497 at the gate pin. i re-checked all readings, including the duty cycle across the source/gate.
the frequency/etc seems to be set dependent on rail charge, and mostly variable at low power states, otherwise, it locks in at 40khz/3.5vgate charge, and 49%d.c.
i am using voltage adjustment at the power supply, because it was changing too much with music. the output db fluctuated too much for my liking over 100w, but i'm sure that has more to do with my power supply voltage drop than anything.
i want to push it in the truck, but it doesn't seem right to me either.
oh, yah. there is also a 2904d ic nearby, which is tied into the thermal resistor, and i'm not sure what else. i'm just assuming this is the protection ic, and don't think it handles these adjustments, but i could be wrong. i would have to pull the board again to trace what it is tied into, and check the p-sheet.
thanks perry,
-Heath H-
the frequency/etc seems to be set dependent on rail charge, and mostly variable at low power states, otherwise, it locks in at 40khz/3.5vgate charge, and 49%d.c.
i am using voltage adjustment at the power supply, because it was changing too much with music. the output db fluctuated too much for my liking over 100w, but i'm sure that has more to do with my power supply voltage drop than anything.
i want to push it in the truck, but it doesn't seem right to me either.
oh, yah. there is also a 2904d ic nearby, which is tied into the thermal resistor, and i'm not sure what else. i'm just assuming this is the protection ic, and don't think it handles these adjustments, but i could be wrong. i would have to pull the board again to trace what it is tied into, and check the p-sheet.
thanks perry,
-Heath H-
Check the frequency pin 5 to pin 7. Does the frequency change?
Set the voltage so that you get the maximum duty cycle at idle. If the FETs remain cool (clamped to the heatsink) and the current draw doesn't increase as when it reaches the maximum duty cycle, the drive circuit is likely OK.
You mentioned that the regulation had hysteresis. This is unlikely. It's more likely that the current draw by the output stage wasn't enough to draw the rail voltage down. If there was significant hysteresis, the rail voltage would vary greatly.
Set the voltage so that you get the maximum duty cycle at idle. If the FETs remain cool (clamped to the heatsink) and the current draw doesn't increase as when it reaches the maximum duty cycle, the drive circuit is likely OK.
You mentioned that the regulation had hysteresis. This is unlikely. It's more likely that the current draw by the output stage wasn't enough to draw the rail voltage down. If there was significant hysteresis, the rail voltage would vary greatly.
across pins 5&7, the frequency is 75.12khz on the high side of the hysteresis, and 75.06 on the low side of the hysteresis. this the only point it changes, otherwise remains constant at all times. current and temps have remained cool and pretty constant, other than the typical fluctuations of charging/discharging my bench cap with the voltage adjustment, since the meter is before that at the moment.
i assure you the hysteresis is present, but i only really see it at idle with the 1/2volt supply range. rails charge to 92v across and stay there. i suppose i can play a tone on it and adjust the output with the sub hooked up tomorrow (2ohm load) and see how it reacts. as mentioned, i did notice the output changes. i can also re-check the frequency variations i was seeing tomorrow.
i assure you the hysteresis is present, but i only really see it at idle with the 1/2volt supply range. rails charge to 92v across and stay there. i suppose i can play a tone on it and adjust the output with the sub hooked up tomorrow (2ohm load) and see how it reacts. as mentioned, i did notice the output changes. i can also re-check the frequency variations i was seeing tomorrow.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.