Yes indeed, everyone starts with zero knowledge. I still have plenty to learn on this forum from all the friendly folks. I give credit to them all for much of my knowledge. Too, welcome to the forum.🙂
A good point you have touched upon is that we try to see what we can do. We weigh the benefits against the risks. Decline comes with stagnation.
A good point you have touched upon is that we try to see what we can do. We weigh the benefits against the risks. Decline comes with stagnation.
I repaired a DX700 with a blown bridge. I used the HUF74639G3 from Fairchild and everything is fine. I think the sound was a bit fuller than before.
IMO, the design could use some work. I would remove the cap from the gates and add snubbing for the bridge, especially at the higher current of a 1 ohm load and because the HIP4080 is rather 'dirty'. I purchased a second set of HUF75639G3 which I can send you if you are interested. Email me or catch me on IM.
I also ran the DX700 through the standard battery of tests on an AP S1 up to the limit of my DC power supply. I can scan those in if you're interested also.
IMO, the design could use some work. I would remove the cap from the gates and add snubbing for the bridge, especially at the higher current of a 1 ohm load and because the HIP4080 is rather 'dirty'. I purchased a second set of HUF75639G3 which I can send you if you are interested. Email me or catch me on IM.
I also ran the DX700 through the standard battery of tests on an AP S1 up to the limit of my DC power supply. I can scan those in if you're interested also.
Evan Shultz said:
Absolutely, please send up everything you have info-wise. On the HUF75639G3 I am intending to get another set so I would be interested, what are you asking for them?
What do you mean by "add snubbing for the bridge"?
Last question; so you are saying that it would be safe to run this amp at 1 ohm with some minor mods, or should I add more mosfets just for the fun of it?
Damn, one more; the bridge gate resistors are 15 ohm, what about changing those to 5 ohm to improve efficiency?
Evan, is this what you are talking about?
CONTROLLING VOLTAGE TRANSIENTS IN FULL BRIDGE DRIVERS APPLICATIONS
To all: I got the amp up and running last night with a new HIP4080AIP, now I can begin making changes to it. 😀
CONTROLLING VOLTAGE TRANSIENTS IN FULL BRIDGE DRIVERS APPLICATIONS
To all: I got the amp up and running last night with a new HIP4080AIP, now I can begin making changes to it. 😀
Evan Shultz said:
Evan,
The information I am getting from the above link on controlling voltage transients seems to be at odds with your advice to remove the cap from the gates. Can you please clarify?
<from the above linked PDF file>
To insure reliable performance of a H-bridge drive circuit, the designer must insure that the
device operates within the maximum ratings of the device(s) used in the circuit. One of the critical
parameters to consider is the maximum voltage capability of the devices. To maintain the
reliability, the voltage transients due to switching inductive loads must be maintained within the
ratings of the device. Two techniques used to control the voltage transients in fast switching
applications are proper bypass filtering of the power supply and snubbing the outputs to control
voltage rise times. Using these two techniques the voltage transients in a DMOS bridge application
can be controlled to within safe levels.
Good link you've found there.
I see nothing at odds with what Evan's said, the advice coincides well.
The higher current flow brought on by heavier loads will promote ringing, therefore, snubbers.
The cap on the gate, I presume to be one going from gate to source? This does not qualify as a supply bypass cap or as a snubber. It is effective in slowing charging of the gate, thereby providing a slower overall switch, and much higher losses in the device. At the higher currents of a 1ohm load, this could quickly lead to failure, so removing that cap will help provide a faster and more efficient switch.
A cap from gate to source is also a technique you'll see some using (even some app notes) to help limite gate step (CdV/dT induced turn on causing cross conduction) but I don't think that'll work because you have the series gate resistance, inductance, and source pin inductance in there as well. So the voltage may look smaller by the time it gets to the outside cap but it's already had it's effect of stepping up the gate voltage. I think it's just a haywire kind of half fix that probably does more harm than good.
I see nothing at odds with what Evan's said, the advice coincides well.
The higher current flow brought on by heavier loads will promote ringing, therefore, snubbers.
The cap on the gate, I presume to be one going from gate to source? This does not qualify as a supply bypass cap or as a snubber. It is effective in slowing charging of the gate, thereby providing a slower overall switch, and much higher losses in the device. At the higher currents of a 1ohm load, this could quickly lead to failure, so removing that cap will help provide a faster and more efficient switch.
A cap from gate to source is also a technique you'll see some using (even some app notes) to help limite gate step (CdV/dT induced turn on causing cross conduction) but I don't think that'll work because you have the series gate resistance, inductance, and source pin inductance in there as well. So the voltage may look smaller by the time it gets to the outside cap but it's already had it's effect of stepping up the gate voltage. I think it's just a haywire kind of half fix that probably does more harm than good.
classd4sure said:
Good thinking, I experimented with such external gate capacitors a bit, and they did not turn out to be worthwhile as I saw things. They also cause the driver IC to dissipate more power, leading to potential failure.
I had have the exact same thoughts about the internal MOSFET voltages being higher than is seen on the outside. The gate step turn-on is attributed to the Miller Effect I mentioned earlier, just to tie things together.
classd4sure said:
Ah! OK, thanks guys. That makes sense now.
I wasn't questioning the validity of your advice Evan, I just didn't quite understand. I thought that cap was part of a snubber network so it seemed like taking it out would defeat the purpose.
I did gain a better understanding of how a snubber works after talking to my boss about it, so now it all makes sense to me.
ClassD: I have been studying the link you posted about gate ringing, and then looking over the amp design. On the power supply, the gates have 15 ohm resistors, and the outputs have 10 ohms. Is this part of the problem with this amp as well?
The output side gates have a small circuit (I think to prevent overshoots?) connected to them; From the HIP4081 it branches to two 1/4w surface resistors, a 10 ohm res. goes to the cathode of a diode US1B
the 100 ohm res. goes to the anode/gate-source cap/and gate of the MOSFETS.
I read that a Schottkey diode can be better matched to the needs of the circuit but are not in common use due to the cost, and the amount of space they take up.
I am not that concerned with either of those problems, so how about this for a plan; remove the gate-source caps, change all gate resistors to 5 ohm (except for the 100 ohm ones), replace the diode with a faster/lower foward bias Schottkey type, add snubbers to the power supply gates, and double up on all MOSFETS?
Or am I totally off track here?
Thanks Sub.
I hadn't even considered the extra dissipation of the driver.
It seems there's no substitute for careful device selection.
Here's a decent link about snubbers:
http://www.ridleyengineering.com/snubber.htm
It would appear as though implementing a snubber (if required) is more than a matter of taking a best guess, and must be carefully designed.
Here is a more in depth look at them, also covering the RCD version of it:
http://www.cornell-dubilier.com/tech/design.pdf
SHOdown, it's cool to question advice. Good advice is easily defended.
Is there a problem with this amp?? I don't know the first thing about it 😀 Switching supply right? Switching supplies often require a minimum load maybe that's what the 10 Ohm is for, or to decouple it from the output a bit and provide a feedback point? I don't know. I'd advise against messing with the supply in any way, they're carefully designed to meet specific goals and there likely isn't anything that can be easily changed on it.
Also, looking at the amp, is it possible to parallel the MOSFETs while keeping the pins almost on top of each other to minimize loop inductances, and still have them heatsinked?
Is there overcurrent protection on the supply? If so trying to drive a one ohm load might not work well at all if it keeps tripping protection circuitry, maybe you'll want to disable or modify it, maybe not. Do you have schematics for this amp?
I take it the cathode points away from the gate and towards the driver output, with the ten ohm in series, and the 100 in parallel to all that?
Hmm, I'd remove the 10 ohm on the output driver, gone.
Faster it turns off the better.
Good call on the diode, the one you have has a 1volt fwd drop and I think is party the reason why they used that cap on the gate. Removing that cap sounds like a good plan.
See how it works after that.
If it passes all tests with flying colors... and you want to squeeze more out of it, see how it likes lowering the 100 Ohm resistor, 50 might be a good goal but go in small increments (10). If it starts to run hotter while idle you're running into trouble and need to back off.
That'll help lower THD.
Get all that worked out before you ever consider attempting to parallel mosfets, you'll have less of them to replace this way (once they smoke).
Only add snubbers if required.
Make this your first stop:
http://focus.ti.com/lit/ml/slup169/slup169.pdf
I hadn't even considered the extra dissipation of the driver.
It seems there's no substitute for careful device selection.
Here's a decent link about snubbers:
http://www.ridleyengineering.com/snubber.htm
It would appear as though implementing a snubber (if required) is more than a matter of taking a best guess, and must be carefully designed.
Here is a more in depth look at them, also covering the RCD version of it:
http://www.cornell-dubilier.com/tech/design.pdf
SHOdown, it's cool to question advice. Good advice is easily defended.
Is there a problem with this amp?? I don't know the first thing about it 😀 Switching supply right? Switching supplies often require a minimum load maybe that's what the 10 Ohm is for, or to decouple it from the output a bit and provide a feedback point? I don't know. I'd advise against messing with the supply in any way, they're carefully designed to meet specific goals and there likely isn't anything that can be easily changed on it.
Also, looking at the amp, is it possible to parallel the MOSFETs while keeping the pins almost on top of each other to minimize loop inductances, and still have them heatsinked?
Is there overcurrent protection on the supply? If so trying to drive a one ohm load might not work well at all if it keeps tripping protection circuitry, maybe you'll want to disable or modify it, maybe not. Do you have schematics for this amp?
I take it the cathode points away from the gate and towards the driver output, with the ten ohm in series, and the 100 in parallel to all that?
Hmm, I'd remove the 10 ohm on the output driver, gone.
Faster it turns off the better.
Good call on the diode, the one you have has a 1volt fwd drop and I think is party the reason why they used that cap on the gate. Removing that cap sounds like a good plan.
See how it works after that.
If it passes all tests with flying colors... and you want to squeeze more out of it, see how it likes lowering the 100 Ohm resistor, 50 might be a good goal but go in small increments (10). If it starts to run hotter while idle you're running into trouble and need to back off.
That'll help lower THD.
Get all that worked out before you ever consider attempting to parallel mosfets, you'll have less of them to replace this way (once they smoke).
Only add snubbers if required.
Make this your first stop:
http://focus.ti.com/lit/ml/slup169/slup169.pdf
Probably slightly off topic.
Guys, i think that looking specs of the amp would be useful way to saving efforts for the right direction. If i remember correct, this amp have THD about 1% (or 2% even!) at rated power, and damping factor 100@4ohm i.e. 25@1ohm! Not so much precision, isn't it? BTW, modern caraudio subwoofers for the small box ability have a tend: little Vas(by hardness and big moving mass) and Qes(double coil and ~5kg magnet). On the most powerfull caraudio russian forum I saw many comments about the such amps sounding (MTX, etc have similar products), and the comments was pretty ruthless- fatty&slowly bass, input- bass solo, output- hum chaos. And many people consider it shame like the inherent class D sound in generally!
Well.. Let's replace the poor kicker modulator to the something cool one? UcD with second feedback loop, like Hypex subamps, would be nice solution.
Guys, i think that looking specs of the amp would be useful way to saving efforts for the right direction. If i remember correct, this amp have THD about 1% (or 2% even!) at rated power, and damping factor 100@4ohm i.e. 25@1ohm! Not so much precision, isn't it? BTW, modern caraudio subwoofers for the small box ability have a tend: little Vas(by hardness and big moving mass) and Qes(double coil and ~5kg magnet). On the most powerfull caraudio russian forum I saw many comments about the such amps sounding (MTX, etc have similar products), and the comments was pretty ruthless- fatty&slowly bass, input- bass solo, output- hum chaos. And many people consider it shame like the inherent class D sound in generally!
Well.. Let's replace the poor kicker modulator to the something cool one? UcD with second feedback loop, like Hypex subamps, would be nice solution.
SHO, no problem. I should have explained what I meant more in my original post. Email me about the FETs!
Unfortunately I can't find this great document I have for easily choosing snubbing components. It will get you performance indistinguishable from days and days of work on a calculator and Spice. Do some searching online and find perhaps you can find it. It goes something like choosing a resistor and then you only need a moderate scope to select all the rest of the components.
Removing the gate cap will greatly reduce losses. Keep in mind that this amplifier was made in 1999 [or before] and under apparently a strict budget. When/if you remove that cap you will most likely have to change gate resistors [greater resistance] and certainly add snubbing because the caps slow down the rate of charge moving to the gates. Removing them will add lots more overshoot and ringing. Be sure that you have fast reverse diodes across the gate resistors.
I don't remember too much about the extra circuit you are reffering to hanging off the bridge, but there's no reason I can think of that you woulnd't help things by switching to a Schottkey. Just make sure that circuit isn't somehow controlling the feedback to the HIP.
Sorry it was been so long since I have looked inside that amp - I am starting to forget some of it. I think with a little tweaking you can make the amplifier perform much better and that it will be stable at 1ohm as long as you don't get silly with it. Send me an email and we can work something out with the FETs.
EDIT [from rereading earlier posts]: I favor adding snubbing regardless. IMO, there's no reason not too have it. The gate cap was obviously a band-aid to cover something up, and more than likely it's because a snubber would have cost more.
I was unable to get any information on this amplifier from Stillwater Designs when I was working on it before. However, I after I messed with the amp I gained a contact at Kicker - give me a few days to check with him 🙂
Unfortunately I can't find this great document I have for easily choosing snubbing components. It will get you performance indistinguishable from days and days of work on a calculator and Spice. Do some searching online and find perhaps you can find it. It goes something like choosing a resistor and then you only need a moderate scope to select all the rest of the components.
Removing the gate cap will greatly reduce losses. Keep in mind that this amplifier was made in 1999 [or before] and under apparently a strict budget. When/if you remove that cap you will most likely have to change gate resistors [greater resistance] and certainly add snubbing because the caps slow down the rate of charge moving to the gates. Removing them will add lots more overshoot and ringing. Be sure that you have fast reverse diodes across the gate resistors.
I don't remember too much about the extra circuit you are reffering to hanging off the bridge, but there's no reason I can think of that you woulnd't help things by switching to a Schottkey. Just make sure that circuit isn't somehow controlling the feedback to the HIP.
Sorry it was been so long since I have looked inside that amp - I am starting to forget some of it. I think with a little tweaking you can make the amplifier perform much better and that it will be stable at 1ohm as long as you don't get silly with it. Send me an email and we can work something out with the FETs.
EDIT [from rereading earlier posts]: I favor adding snubbing regardless. IMO, there's no reason not too have it. The gate cap was obviously a band-aid to cover something up, and more than likely it's because a snubber would have cost more.
I was unable to get any information on this amplifier from Stillwater Designs when I was working on it before. However, I after I messed with the amp I gained a contact at Kicker - give me a few days to check with him 🙂
Hi,
The gate-source cap would help hold the gate lower than what the big 1 volt FWD drop of the cheap diode they used in place of a schottky. That's the band-aide as I see it. So it would serve to help return some measure of noise immunity. If it was of sufficiently low value it wouldn't have delayed turn on by a great margin, and with a 100ohm gate resistor, it likely wasn't very high, what's the switching frequency of this amp anyway? For all I know it could be 10Khz.
Long story short, the whole goal here is to speed up switching, get it as close as possible to the point where it goes up in smoke, yet manages to behave itself. This is where best performance is achieved. So adding a bigger gate resistor to negate the benefits of the other modifications seems fruitless, and if it actually did come to that, it's time to look for a better mosfet, because 100 ohms is on the high side already, and there are certainly better mosfets out there to be had these days.
With faster switching at heavy currents it is very likely ringing will occure, and so snubbering would be beneficial, but _if_ there is little to no ringing, or the period of ringing wasn't a few orders of magnitude above the switching frequency, there is certainly no point in having it, as the only thing it will be dampening is efficiency. It also has to be designed according to the nature of the ringing which is there, how could you design it if there wasn't any?
Anyway at this stage it isn't ruled out, or in, do what can be done to improve the drive signals and then see where it stands for ringing.
BTW that circuit, if as I imagined it, would be in place to provide a faster turn off and a slow turn on. This isn't necessarily to do with ringing, though it could have an affect to lessen it by slowing rate of change, you can look at that as a side affect.
IVX, changing the modulator would be neat, may as well smoke this one first though and learn a few things 🙂
😉
The gate-source cap would help hold the gate lower than what the big 1 volt FWD drop of the cheap diode they used in place of a schottky. That's the band-aide as I see it. So it would serve to help return some measure of noise immunity. If it was of sufficiently low value it wouldn't have delayed turn on by a great margin, and with a 100ohm gate resistor, it likely wasn't very high, what's the switching frequency of this amp anyway? For all I know it could be 10Khz.
Long story short, the whole goal here is to speed up switching, get it as close as possible to the point where it goes up in smoke, yet manages to behave itself. This is where best performance is achieved. So adding a bigger gate resistor to negate the benefits of the other modifications seems fruitless, and if it actually did come to that, it's time to look for a better mosfet, because 100 ohms is on the high side already, and there are certainly better mosfets out there to be had these days.
With faster switching at heavy currents it is very likely ringing will occure, and so snubbering would be beneficial, but _if_ there is little to no ringing, or the period of ringing wasn't a few orders of magnitude above the switching frequency, there is certainly no point in having it, as the only thing it will be dampening is efficiency. It also has to be designed according to the nature of the ringing which is there, how could you design it if there wasn't any?
Anyway at this stage it isn't ruled out, or in, do what can be done to improve the drive signals and then see where it stands for ringing.
BTW that circuit, if as I imagined it, would be in place to provide a faster turn off and a slow turn on. This isn't necessarily to do with ringing, though it could have an affect to lessen it by slowing rate of change, you can look at that as a side affect.
IVX, changing the modulator would be neat, may as well smoke this one first though and learn a few things 🙂
😉
😉 No pain - no gain, Chris. I still remember how much irf540n smoked by me in the my first class D amp/experiments (1999)..
Switching freq is 125kHz.
Sorry about gate resistors, I missed the 100ohm deal. Yeah, I'd remove the gate cap and snub it. While doing that try to get down to a lower gate resistor, perhaps 50ohm or less.
The slow on/fast off circuit is there to set dead time. If you can measure dead time, once the above is done, try to get back to the same dead time as before modification. This will mostly affect THD and that will more than likely be a good overall balance already.
I will have access to a scanner on Mon and I will get the documents scanned and posted then.
When you do any of these measurements make sure you're using a good probe. IMO a strand of RG-58 with bare wire and wrapped shield is the best probe under the sun. Solder those 2 'tips' directly to what you want to measure. Also, to aviod capacitively loading the gates when you measure them, 1kohm in series with the probe should help to isolate the gate and give you a better reading. Best of luck!
Sorry about gate resistors, I missed the 100ohm deal. Yeah, I'd remove the gate cap and snub it. While doing that try to get down to a lower gate resistor, perhaps 50ohm or less.
The slow on/fast off circuit is there to set dead time. If you can measure dead time, once the above is done, try to get back to the same dead time as before modification. This will mostly affect THD and that will more than likely be a good overall balance already.
I will have access to a scanner on Mon and I will get the documents scanned and posted then.
When you do any of these measurements make sure you're using a good probe. IMO a strand of RG-58 with bare wire and wrapped shield is the best probe under the sun. Solder those 2 'tips' directly to what you want to measure. Also, to aviod capacitively loading the gates when you measure them, 1kohm in series with the probe should help to isolate the gate and give you a better reading. Best of luck!
Hi,
You can tailor dead time on the HIP4081 independently by adding resistors to pins 8 and 9, or changing the ones they already added there.
The reason you want it to turn on slower than it turns off is just to give the body diode time to recover from the current that freewheeled through it during the dead time period and allow it go into blocking mode. It's a kind of a soft dead time. This being an "older" mosfet might not like being turned on very fast (slower recovery times higher output capacitance etc) and a more modern replacement would be a viable option in order to optimise how fast it can be turned on. I wouldn't be very shocked to find out it doesn't even like as "low" a gate resistor as 100 once that cap is removed. Ideally it could be minimized, but it might actually need increasing.
Making it turn on faster in order to gain efficiency will quickly reach a point of diminishing returns when the body diode starts allowing current spikes to shoot through, increasing dead time will likely only make this worse as it will have more time to become hard saturated and therefore take even longer to recover.
My only point here is, this is more like heart surgery than a minor operation, procede with caution.
I would think snubbing should come last, the faster you get it turning on the more ringing is likely to occur, once the limit is found that doesn't cause mass amounts of shoot through from the body diode, then an optimal snubber can be designed if required.
Would it be possible to limit current from the supply to facilitate (or make possible) this kind of tweaking? Would adding power resistors to the rails be the best way?
It's pretty cool of you to get the schematics Evan.
Ivan, I used to collect the parts I smoked but it was getting depressing, rows of TO-220 tombstones.
Best of Regards,
Chris
You can tailor dead time on the HIP4081 independently by adding resistors to pins 8 and 9, or changing the ones they already added there.
The reason you want it to turn on slower than it turns off is just to give the body diode time to recover from the current that freewheeled through it during the dead time period and allow it go into blocking mode. It's a kind of a soft dead time. This being an "older" mosfet might not like being turned on very fast (slower recovery times higher output capacitance etc) and a more modern replacement would be a viable option in order to optimise how fast it can be turned on. I wouldn't be very shocked to find out it doesn't even like as "low" a gate resistor as 100 once that cap is removed. Ideally it could be minimized, but it might actually need increasing.
Making it turn on faster in order to gain efficiency will quickly reach a point of diminishing returns when the body diode starts allowing current spikes to shoot through, increasing dead time will likely only make this worse as it will have more time to become hard saturated and therefore take even longer to recover.
My only point here is, this is more like heart surgery than a minor operation, procede with caution.
I would think snubbing should come last, the faster you get it turning on the more ringing is likely to occur, once the limit is found that doesn't cause mass amounts of shoot through from the body diode, then an optimal snubber can be designed if required.
Would it be possible to limit current from the supply to facilitate (or make possible) this kind of tweaking? Would adding power resistors to the rails be the best way?
It's pretty cool of you to get the schematics Evan.
Ivan, I used to collect the parts I smoked but it was getting depressing, rows of TO-220 tombstones.
Best of Regards,
Chris
It seems that the biggest danger from the body diode would be when the duty cycle is very high, or low, that is, for the conducting diode side.
Instead of using resistors to limit current, light bulbs might be a possibility since their glow is an easy visual indication of current draw. I blew oodles of MOSFETs, too, when I was first trying to make a power supply years ago.
Instead of using resistors to limit current, light bulbs might be a possibility since their glow is an easy visual indication of current draw. I blew oodles of MOSFETs, too, when I was first trying to make a power supply years ago.
Man I have some catching up to do. Thanks everyone for the great advice!
I had no time yesterday to log in and see what I missed this weekend.
I had no time yesterday to log in and see what I missed this weekend.
classd4sure said:
Ok I will leave it alone until I know what I am doing.
Oh, and thanks for the great links.
Yes to all of the above except schematics. I am still working on that but Evan just wrote that he may be able to help out there.
I have a CNC mill, so I can fabricate an addition to the factory heat sink that will work beautifully and still fit in the chassis.
Ugh... The thought of re-replacing all those SMDs'... Perhaps I can rig up a breadboard type of system to quickly try out new values without causing a problem? Path length seems to be a real issue in MOSFETS circuitry. It is hard to visualize the switching speeds we are discusing, and the complications inherent in that.
THD+N on this amp is < 0.6% at rated power into 2 ohms'
I can not complain about that for a sub amp.
I hope my brain doesn't start leaking...
SHOdown said:
Nice!
SHOdown said:
Hmm, probably not the best idea because you'll be splitting hairs between one value to the other just to squeeze that last bit out of it and alot of the performance will depend on the parasitics. If you add big jumper leads to a homemade PCB, the PCB traces wouldn't be representative of what's on the board itself, and the jumpers will add alot of inductance. Certainly don't use an actual breadboard, they're hell with parasitics. The other thing is if you do it so you can just try the values with easier to handle parts, like 1/4 watt carbon resistors for example, the parasitics of the parts themselves change. I think you'd only be making work for yourself.
Maybe you can just short the 10ohm? Depending on the packages you can solder one on top of the other to parallel them and test values that way as well..I don't know if that would be easier or not. Sounds like work alright, try to look at it as part of the fun 🙂
SHOdown said:
I was kind of wrong there, I think I did only mention lowering THD, which it will. But I think your main goal here is to increase efficiency, hence the faster switching. If you optimise the switching so it's as fast as can be it might run hotter than with slower switching, at least at some point it will, which is the point where the body diode comes into play causing shoot through spikes. I think you want to optimise it so it runs as cool as possible right? So THD might not be optimal in the end, but it should help keep it cooler with the higher current from the 1 ohm load.
SHOdown said:
I hope my brain doesn't start leaking...
Know what you mean. There is alot of information in that app note.
What works for me is to just read em real quick once and save them. Then you know what's in there and can always go back to it when needed, which is alot if it's any good. This one is good 🙂
I spoke with a guy in the Kicker tech dept, and we had a long conversation about the design of the amp. He said that because that was their first foray into Class D, the design was not as flexible in terms of loads, and are in fact very picky about impedance loads. He told me there is basically no way to make it run into 1 ohm short of rewinding the transformer, and going forwards from there.
Fine with me. I am going to buy a 1 ohm stable 1200w amp and get it playing in my system. This Kicker I am doing, because I want to learn something new but it wouldn't hurt to make a killer amp in the process. 😀
So whatever it takes I am willing to give it a go. Want to change modulators? (whatever those are) Lets do it. Put in monster MOSFETS? Heck yea.
The goal for me is to learn how to design electronics. That is why I got into repairing electronics for a living, and I have learned a lot in the three years since I first began, but I want to know more than just how to fix something that is broken. I have been doing that all of my life.
So what I am trying to get at, is please feel free to toss out any ideas you may have, I appreciate them all and have already learned a lot of great things from you guys.
Dean
Fine with me. I am going to buy a 1 ohm stable 1200w amp and get it playing in my system. This Kicker I am doing, because I want to learn something new but it wouldn't hurt to make a killer amp in the process. 😀
So whatever it takes I am willing to give it a go. Want to change modulators? (whatever those are) Lets do it. Put in monster MOSFETS? Heck yea.
The goal for me is to learn how to design electronics. That is why I got into repairing electronics for a living, and I have learned a lot in the three years since I first began, but I want to know more than just how to fix something that is broken. I have been doing that all of my life.
So what I am trying to get at, is please feel free to toss out any ideas you may have, I appreciate them all and have already learned a lot of great things from you guys.
Dean
- Status
- Not open for further replies.