| lumanauw |
I just made an experiment classD, and failed. It's quite different than building analog amps. I think I have to learn from basic again. The aim is good sounding offcourse :D
What is the difference (technically and sonically) between
1. ClassD made by full discrete (like UCD) and classD made by IC's (like IRFaudio)?
2. ClassD that takes feedback before LC filter and classD taking feedback after LC filter?
3. How bad is phase shifts made by LC filter? Why manufacturers like IRF and LT takes feedback before LC, while majority here likes to take feedback after LC filter? Pro's and Con's? |
|
|
| subwo1 |
Hi lumanauw, I am just going to ramble off some thoughts quickly as your questions do not have simple answers. Feedback after the filter tends to give lower distortion since filter nonlinearities are compensated for by feedback. A good discrete class D amp can give lower distortion if it is done so that it can almost operate cleanly in linear mode.
I think an amp made largely of IC chips can be done to give clean output as well, but it tends to be harder to eliminate discontinuities in the signal flow. Whenever the signal has to jump between on-off thresholds, distortion increases--the greater and/or slower the jump, the more distortion there tends to be. But, consistency or smoothness in signal flow may be more useful than pure propagation speed of the signal through the amp. Those are some quick thoughts.
I am still working on a P/N design. I am having trouble getting my pspice models for the IRF9540n and IRF540n to give me accurate simulations, though. |
|
|
| MikeB |
Hi subwo1 !
Welcome to the club, i also can't simulate with irf540n/9540n,
but wanted to use them for my P/N-ClassD (fully discrete).
The models from IRF don't really work in pspice...
Lumanauw, if you want to design dicrete ClassD, you will need
simulations. It's too hard to predict deadtimes/crossconductions,
rise/fall-times. I don't really trust sims at these speeds, but they
give a good idea.
Mike |
|
|
| lumanauw |
Hi, Subwo1,| quote: | | Feedback after the filter tends to give lower distortion since filter nonlinearities are compensated for by feedback |
How far is the phase shift made by LC filter? Will it making the front end/comparator/differential receiving the correct signal for NFB correction?
I reallly wonder why large manufacturer with deep knowledge in switching like IRF takes the feedback before the LC, assuming they know that they can take the feedback after LC.
Hi, Mike, | quote: | Lumanauw, if you want to design dicrete ClassD, you will need simulations. It's too hard to predict deadtimes/crossconductions, rise/fall-times. I don't really trust sims at these speeds, but they give a good idea.
|
Yes, in this case, SIM is a must. With analog amps, we can get without SIM, but not in this field, I think.:D |
|
|
| classd4sure |
Hi,
| quote: | | I just made an experiment classD, and failed. |
Naaaaah, sounds like you got something out of it, right? That's not a failure.
I would say start with the class d reference design thread, lots of good stuff in there for ya.
Sims are a total pain but if you get good with one you can do some neat things, and lots of experimentation rather cheaply.
IR just arent' in the amp design business so much as mosfets + drivers.
Regards,
Chris |
|
|
| subwo1 |
| quote: | Originally posted by lumanauw
How far is the phase shift made by LC filter? Will it making the front end/comparator/differential receiving the correct signal for NFB correction? |
Hi, it may appear that way at first, and to a point, if the delay is really great by using, say, a 100uH inductor, then maybe it will cause problems with filter resonance getting too close to the audio range.
From the opposite perspective, the filter tends to limit how many harmonics can rechurn through the amp. So the sound may actually gain smoothness and fidelity.
An advantage of feedback before the filter would seem to be that it is easier to get really high switching frequencies without producing more high frequency switching content at the output. I don't really see why it should not also benefit from fewer harmonics as well. I may be missing something with these thoughts on harmonics. Best Regards. |
|
|
| subwo1 |
| Hi again, indeed, I was overlooking at least one thing about after filter feedback. The phase lead (compensation) can keep the input and output pretty much in net phase over a full switching cycle. The input gets ahead for half a cycle and then behind for the other half. So I doubt with a switching frequency of a few hundred kilohertz the ear has any chance of perceiving the back and forth time difference. |
|
|
| lumanauw |
Hi, Subwo1,
I see majority of self-oscilating designs here are using after LC feedback point (with some exceptions, like IRF amp). From the IRF paper, fig 7 of iraudamp1.pdf, the red curve is peaking. They write that peaking is due to resonance of LC filter.
Will the amp which take feedback after LC do not have this?
How about sonics and stability comparison of feedback after LC and before LC. Any member here have made both?
This is my experimental classD. The feedback is taken after LC. I use IR2111 driver, I have clean square drive now (but in higher frequency, the square have oscilation/ringing in the edges)
Without R1,C2,R3,C3, the oscilation frequency is about 24khz. When I put 10k for R1, the oscilation frequency rises alot, but the square is ringing in edges, so I lowered the frequency by putting 1k for R3 and 220pf for C3.
I cannot put R3 and C3 in parrarel with feedback resistor (C2 place) like UCD patent. If I put it there (or just a single C2=100pf), all the mosfets blown.
There are things that I dont understand about L1 and C1 value.
If I put L1=200uH and C1=220uF (nonpolar Elko), the amp looks normal, but cannot drive high volumes. The sound is wrecked. In low volume, seems fine.
If I put L1=30uH and C1=2.2uF (WIMA MKP), the mosfets blows instantly (upper and down mosfets, all blown)
If I put L1=30uH and C1=47uF (nonpolar elko), or C1=220uF(elko), the elkos and L are heating up very quickly, and the current draw at idle rises alot.
If I put L1=200uH and C1=47uF (nonpolar elko), both the L and C heating up.
I wonder how come the designs here uses L only 40uH and C only 1uF without problem for self oscilating classD?
Seems I cannot have that low values of L and C. How to do that? Is this value of 40uH and 1uF valid only on switching frequency >300khz? My current is only 40khz, none of these low value can work. I've tried to raise the frequency by not using R3 and C3, but the idle current is very high, I shut down the arrangement imediately.
I've blown many-many IRF640 today. |
|
|
| subwo1 |
Hi lumanauw, | quote: | | ...peaking. They write that peaking is due to resonance of LC filter. |
With suitable phase compensation and a sufficiently high switching frequency, and so long as the output does not clip and the input signal does not change too much too quickly, then post filter feedback can enable the amp to damp output filter resonance. There are many ifs; its true.
| quote: | | I have clean square drive now (but in higher frequency, the square have oscilation/ringing in the edges) |
I'd say the turn-on time is too fast, and there may also be a lot of gate ringing. The turn-on portion of each MOSFET driver buffer needs some resistance. I'd try separating the emitters, placing a 10 to 20 ohm or so resistor between them. Connect the intersecting point between the PNP emitter and the resistor to the gate.
| quote: | | There are things that I dont understand about L1 and C1 value |
I can only try to give some general guesses about the types of things that are happening since I can't tinker with the actual circuit. If you can, try to set up a simulation model of your circuit in LTspice so that you can compare the two and maybe get some ideas.
In some filter combinations, I suspect one or two things could be happening. One is that the filter could be going into heavy resonance, and the other is that the inductor is saturating. It would seem to be most likely with the 30uH unless it is air core, but especially then it could be inducing interference into your circuit. But a saturating core will overload and overheat your MOSFETs and cause them to blow.
| quote: | I wonder how come the designs here uses L only 40uH and C only 1uF without problem for self oscilating classD?
Seems I cannot have that low values of L and C. How to do that? Is this value of 40uH and 1uF valid only on switching frequency >300khz? My current is only 40khz, none of these low value can work. I've tried to raise the frequency by not using R3 and C3, but the idle current is very high, I shut down the arrangement imediately. |
One reason is that a high switching frequency permits the inductor impedance to remain high enough. The general answer to excessive power consumption during higher switching frequencies is switching losses. These can be causes by switching on one MOSFET too quickly after the other has turned off. Maybe the added gate resistor will help that problem. Sorry about the string of dead MOSFETs.
If you are going to get a functional amp, the IR2111 IC is a relatively easy path to that goal. I am confident you can get it to work! Best regards. |
|
|
| lumanauw |
Hi, Subwo1,
Yes, it already produces sound (in low level), I guess this is a good start.
Anything weird in my schematic (besides the buffers you mentioned)? I've got no simulator, I make experiment by W2W connection.
Some questions :
1. Is it OK to use Elkos (elektrolit capacitor) for LC filter? Or MUST use non-elektrolit type (like WIMA or Mylar capacitor)?
2. I cannot use low valued L and low valued C (below 100uH and below 100uF). Where can I learn more about the design of LC filter of classD power amp?
My elkos and cores are heating very quickly in some values (low values). Something must be wrong in my experiment, since others here can use L<50uH + C<2.2uF without heat or current draw or other complain at all.:D |
|
|
| subwo1 |
Hi lumanauw, I am going to guess that with low loss output filter capacitors like polypropylene, the filter Q is higher than the noise tolerance of your circuit can handle. The electrolytic ones are lossy and provide damping to the filter. Electrolytic capacitors commonly get hot at high frequencies.
Your cores heating at low inductance values indicates to me they are being subjected to too much magnetic flux and maybe even saturating. Lower switching frequencies let the flux on the core build up longer. Try the modified gate driver buffer first to see if it will let you raise the switching frequency so that you can try a lower valued inductor. You are right that your goal will be something on the order of a 40uH choke and a 1uF capacitor.
You may also try to adjust the positioning of your parts and incorporate anti-interference techniques like twisting together parallel runs of wires, especially if they constitute a closed circuit path. So, a "hot" and a return wire should be twisted when possible. Keep high frequency circuit paths short and tight. Place a poly .22uF or so decoupling capacitor directly from the drain of the upper MOSFET to the source of the lower. They should be very close to each other. Keep all component lead lengths as short as possible. I recommend using 1/8w resistors as the optimal size for low voltage p2p. You may like to try to to keep the circuits in tight sections like your own versions of integrated circuits. Component bodies may be touching when voltages present permit. |
|
|
| classd4sure |
Here is your new simulator:
http://ltspice.linear.com/software/swcadiii.exe
It's free, full version, perfectly legal, is a decent program, sooner you get comfy with it the better, it will allow you to do all kinds of experimentation quickly and easily. |
|
|
| pat allen |
:bigeyes: :bigeyes: :bigeyes:
Funny, i am working with LT step-up regulators for a specific application here....thaaaank you for this link !! |
|
|
| phase_accurate |
Hi David
At least the combinations with a large cap make for a low corner frequency and therefore a low switching frequency as well. But if we have a combination of 20 uH and more than 1 uF of capacity we do not only have a low cutoff frequency - we also have a highish Q and a very low inductance for an amp switching at low frequencies, giving quite a high idle current through the coil.
Do you have a scope to see how fast it actually switches or do you have access to one ?
Regards
Charles |
|
|
| lumanauw |
Hi, Subwo1,
| quote: | | I'd say the turn-on time is too fast, and there may also be a lot of gate ringing. The turn-on portion of each MOSFET driver buffer needs some resistance. I'd try separating the emitters, placing a 10 to 20 ohm or so resistor between them. Connect the intersecting point between the PNP emitter and the resistor to the gate. |
Your suggestions seems improving my experiment amp. I put R=47ohm in the totem pole (after emitor of NPN), and it improves the max level that I can play this amp. Before putting this R, in medium level it already wrecked sound, but after putting this R, in the same volume it doesn't wrecked.
I've been using basing totem pole (without R) for years. I don't know if putting R in totem pole emitors is a must :D
The gate pulse are changing shape a bit. Before R it is full square, but after putting R, the left top edge is rounding.
Subwo1, what is the proper totem pole for classD should looks like? (driving parrareled mosfets with small capacity chip like IR2111).
Hi, Charles,
The switching frequency of my current setup is only 24khz. I can raise it to more than 100khz, but the gate pulses are ringing (blurry in scope), and the edges are ringing.
My switching frequency is only 24khz. It works with L=200uH and C=220uF. I've put 12V zeners to protect the G-S. But when I put small inductance (50uH) or small capacitor (2.2uF), all upper and down mosfets are blown instantly. What is making all the mosfets blown? The zeners are OK after blown, it means the destructive thing is not from the gate, but from somewhere else. But where does it comes from? Too much Idrain? The mosfets are not burning. They just becomes shorted G-D-S. I'm confused what is ruining the mosfets.
I dare not raising the frequency because I still feel something is wrong right now. The value of LC seems to have small connection to the switching frequency. It is determined more by feedback arrangements, global feedback+local feedback around opamp in my experiment amp. |
|
|
| lumanauw |
In the iraudamp1.pdf, in page 6, in the "Self Oscilating PWM Modulator" topic, it is writen that IRF amp uses "Sigma Delta" modulation that makes all error in audible frequency is shifted to higher upper frequency.
How come the errors are converted to upper higher frequency, doesn't this means higher order distortion (in analog amp, higher order distortion are not wanted, because they make not pleasant sound)?
Pat Allen, you have built this IRF amp (feedback before LC). By any chance, have you compared it to feedback after LC (like UCD)? Or, how does IRF classD sounds?
What is "Sigma Delta" modulation? I've read this term in DAC. Is it a way to convert square wave to sinusoidal? Is this making the IRF amp taking feedback before LC? But in that chapter it writes "2nd order integrator with U1, C17, C18 converts rectangular wave form from classD switching stage and outputs a quadratic oscilatory waveform as a carrier signal". It is not converting square to sinusoidal, but square to what form?
If it is converting square to sinusoidal, then for me taking the feedback before the LC makes more sense than after LC, cause the square is already converted to analog (to be feedback to front end and compared with analog input signal) before LC.
If it works like that, then in IRF classD amp, even without LC at all the amp should work without destroying the mosfets? (unlike selfoscilating classD with feedback after LC, like my experiment amp. Very sensitive to LC value even to destruct the mosfets).
Destroying the tweeters maybe? :D
LC filter in selfoscilating classD that takes feedback before LC (like IRF) and after LC (UCD) has differen't purpose? In feedback before LC, LC just becomes a low pass filter, but in feedback after LC, the LC becomes energy storage? |
|
|
| subwo1 |
Hi lumanauw,
I am glad about the improvement from the modified gate drivers. :) I cannot say what a totem pole should look like for paralleled MOSFETs since it varies depending on the circuit. But the MOSFETs turning off need enough time to get "off" before the other ones turn on, and it is common practice for the sake of efficiency to use the energy stored in the choke (or transformer) to carry the voltage away from the source of the device turning off to the other rail before the other device turns on. Best Regards.
PS. It is a general no-no in high frequency circuits to connect the cathode of a zener diode directly N-channel MOSFET gates. It definitely can cause ringing on the gate terminal, and MOSFET destruction. Reliance on the power supply voltage as the limiting factor in setting the maximum gate drive voltage is the accepted method. |
|
|
| analogspiceman |
Hi lumanauw,
Could you repost your schematic or point me to it and any information about your layout.
Thanks -- analogspiceman |
|
|
| phase_accurate |
Have you measured how fast your "hand-sewn" comparator and level-shifter works ? I don't know how the IR 2110's input logic reacts to slowly changing input signals. Maybe you should increase the current and decrease the resistors in said part of the circuit.
Regards
Charles |
|
|
| lumanauw |
The schematic is in post #9. http://www.diyaudio.com/forums/atta...tamp=1111291589
It is an initial trial, after reading IRF classD amp, I wanted to try feedback before LC, seems more understandable to me (than self oscilating after LC-feedback like UCD).
I think for heavy loads (like 1ohm parrareled speaker played at full power), feedback before LC is more suitable.
I haven't make PCB yet, everything is P2P connection :D
This is the photo of the driver (IR2111). The emitors of MPSA06 of driver buffer is cut, and I insert 47ohm like Subwo1 suggest.
Analogspiceman, your Guestimate of UCD, is it working in real CCT or you just SIM it only? |
|
|
| MikeB |
| quote: | Originally posted by phase_accurate
Have you measured how fast your "hand-sewn" comparator and level-shifter works ? I don't know how the IR 2110's input logic reacts to slowly changing input signals. Maybe you should increase the current and decrease the resistors in said part of the circuit.
Regards
Charles |
Hmm, according to the datasheet the ir2110 has schmitttriggers
at it's inputs, so unless the signals are not heavily shifted the
quality of these signals should not matter much ?
Mike |
|
|
| phase_accurate |
In this case the amp can be made hysteresis-controlled self-oscillating which wouldn't be a bad start. In this case also the 10 k can stay. Maybe the gain of the diff-stage could be reduced as well in order to make it behave less like a comparator (like Ivan's circuit).
Why not try it with open loop first, using an external "clock" in order to check the behaviour of the switching stage first. In a second step try it with pre-filter NFB. If everything is O.K. the loop can be rearranged for post-filter NFB.
If everything is working fine - but you are still not content because the circuit looks much too simple - you can add a double NFB-loop ! ;) :cool:
Regards
Charles |
|
|
| analogspiceman |
Hi lumanauw,
Here are a few observations about your circuit:
Small electrolytic capacitors are generally not suitable for use as part of a class d amp's output filter due to their being unable to handle the power lost in their series resistance. Their lifetime is loosely inversely proportional to internal temperature, so as long as the temperature is not too high, they will still work, but maybe for as short as hours or minutes. The series resistance does more than just cook the capacitor, however. It also serves to dampen the resonance of the output filter (which may be partially why you got immediate failures when using a 2.2uF film capacitor). You could check this by adding a small resistor (3 ohms or so) in series with the 2.2uF capacitor when using it with the 200uH inductor. Your circuit has other problems, however, that could lead to failures regardless of the damper resistor.
Your schematic does not show the details of your housekeeping power supplies. Exactly how are they implemented? With resistor-Zener strings? Or lab bench supplies? (Bad hookup here could also lead to failures.) By the way, I assume your NE5532 also has been provided with a -15 volt supply, correct?
Note that the IR2111 driver IC is contributing a lot of delay to the forward signal path. Its deadtime, its output transition times, plus the slewing time on the input required to overcome the hysteresis may add up to as much as 1us. Due to the equivalent phase shift this represents, your maximum possible operating frequency will be limited to well under 250kHz (your practical maximum operating frequency will probably be closer to 100kHz).
I have several concerns about your opamp circuit. It appears that nothing prevents in from over driving and severely Zenering the BE junctions of the following differential PNP pair. You really should consider grounding its positive input and running it in inverting mode with the minus input as a summing node at virtual ground (sum the input here as well). Then you could connect a couple of anti parallel diode strings (say two diodes in each string) from the output back to the minus input. This would ensure that the following PNP stage is never overdriven and would also keep the opamp from saturating and getting momentarily stuck to a rail. Once you have made these changes (and if you have hooked up your housekeeping supplies properly), then you should be able to (in fact should need to) add in an RC lead network in the "C2" position (something between 330pF and 1nF in series with 2.2k).
What is the configuration of the mosfet output power circuit traces and how do you connect in the drive signal?
There are more possible issues, but I don't want to overburden you and would rather wait for a little feedback before running off down a possible dead end, anyway. |
|
|
| lumanauw |
Hi, Charles,
| quote: | | (like Ivan's circuit). |
In the Amploid schematic by IVX, I don't understand the level shifter section. Output of opamp is connected to emitor of BC856 while it's base is tied to -15V. How much VBE will this transistor experience? I think more of 0.6V will happen in that BC856.
| quote: | | In a second step try it with pre-filter NFB. If everything is O.K. the loop can be rearranged for post-filter NFB |
For NFB point taken before the LC filter, don't we need something like U1+C17+C18+R23+R26 in the iraudamp1.pdf? They call it Delta Sigma 2nd order integrator. What does this integrator do here?
I don't saw this on after LC feedback like UCD. Only need R+C parrarel with feedback resistor?
I feel that feedback before and after LC needs different CCT (this is from my current knowledge about classD). Or do you think they are the same, so with the same cct, we can take feedback before and after LC?
| quote: | | looks much too simple |
Charles, because I wanted to learn about classD, I feel that this cct is too complex. Too many non-basic parameter that makes the amp fails. I will post my idea about very basic classD.
Hi, Analogspiceman,
| quote: | | Small electrolytic capacitors are generally not suitable for use as part of a class d amp's output filter |
Ahh, that explains why they heat up so quickly (in seconds)
| quote: | | Your circuit has other problems, however, that could lead to failures regardless of the damper resistor. |
Yes, I know something(s) are not right, I just dont know what they are. | quote: | | Your schematic does not show the details of your housekeeping power supplies. Exactly how are they implemented? With resistor-Zener strings? Or lab bench supplies? (Bad hookup here could also lead to failures.) By the way, I assume your NE5532 also has been provided with a -15 volt supply, correct? |
NE5532 have -15V, I forgot to draw it. The 12V (relative to -50V) is generated by LM317 (from 0 to -50V) for powering the IR2111. | quote: | | (your practical maximum operating frequency will probably be closer to 100kHz). |
It's OK with me for learning. IR2111 has automatic dead time ( no need to adjust) about 650nS. For F=100khz, what value of LC filter do you suggest?
If I have understand about how classD works, I dare to make 400khz cct. | quote: | | I have several concerns about your opamp circuit |
You're right. I feel that opamp add much complexity for learing about classD. I the next post, I will post a very simple classD, just to learn how classD works. | quote: | | then you should be able to (in fact should need to) add in an RC lead network in the "C2" position (something between 330pF and 1nF in series with 2.2k). |
This is my big question. I cannot put anything in C2 position that not-making the mosfets destroyed. From 100pf, 1k+220pf, they all makes my experimental amp failure. I can only put R1 or R3+C3 to adjust the switching frequency without destroying the mosfets. This is very contradictive to UCD papers. |
|
|
| analogspiceman |
Hi lumanauw,
I now understand the 12 volt supply on the negative rail, but you neglected to mention where the +/-15 volt opamp supplies come from or how they are hooked up. - a.s. |
|
|
| lumanauw |
NE5532 (opamp) in my first experiment classD seems to add complex bananas. I should be able to do self-oscilating without using any opamps. This is the very basic classD I draw to learn how a classD works. Will figure A works as self-oscilating classD?
NFB is taken after LC, so it might have to use R3-C3 in the feedback resistor.
I couldn't resist (even I have not knowing wheter A will work or not), but I'm intrigued by the possibility of this design. It can make NelsonPass's concept of Susy, like I draw in B. This option works only if A works. :D
SuSy can eliminate absolute distortion that the speaker sense, because the speaker outputs are taken floating from both outputs, if both outputs have the same distortion, the speaker wouldn't sense it. |
|
|
| lumanauw |
Hi, AS,
Sorry I forgot. +/-15V comes from emitor followers with NPN/PNP transistor which their bases is hold at 15V with zeners, relative to 0/ground. (the exact voltage is not +/-15V, but 14V4 more precisely). The transistors are TIP31/32C, to hold 14V4 from +/-50V only for powering the opamps.
After reading about BUS-Pumping in iraudamp1.pdf, I'm afraid this method of generating +/-15V from main rail is not good. Or is it OK to use this way to power the opamps? |
|
|
| classd4sure |
Hi Lumanaux,
Try something like this, it's about as simple as UCD gets, not best, but simple. It works. Sorry it's a bit of a mess, I think R40=1K. I'm just using an air coil, I don't really care about EMI right now, it's low power, doesn't heat up, doesn't saturate.. does the job.
I have the output stage working in P2P right now, not with the exact values shown but it's also not the exact mosfets shown.
The comparator portion is still on a breadboard, will be doing that in P2P tonight as well.
I find it's nice to start with something that you know works, and then go from there.
Regards
Chris |
|
|
| lumanauw |
Hi, Chris,
Thanks for the schematic. I have to study it further.
There are some components I don't understand.
-R78+D80, what are they for? Helping to fill C34 besides the work of D84?
-R71 and its lower 100k, what are they for?
-R75 = 40k, what is it for?
If it is to be used for higher rail (+/-50V), what values to be changed?
I tried IRF640N in my experiment cct, and they blow up far more easily than IRF640 (without N). For classD, IRF (without N) is more appropriate than the ones with N suffix?
A question. Is it a must to construct classD (any cct with fosc >250khz) with SMD component? Or classD can works with ordinary component too, like 1/4W resistors, TO-92/TO-220 transistors at the drivers? |
|
|
| classd4sure |
| quote: | Originally posted by lumanauw
Hi, Chris,
Thanks for the schematic. I have to study it further.
There are some components I don't understand.
-R78+D80, what are they for? Helping to fill C34 besides the work of D84?
-R71 and its lower 100k, what are they for?
-R75 = 40k, what is it for?
If it is to be used for higher rail (+/-50V), what values to be changed?
I tried IRF640N in my experiment cct, and they blow up far more easily than IRF640 (without N). For classD, IRF (without N) is more appropriate than the ones with N suffix?
A question. Is it a must to construct classD (any cct with fosc >250khz) with SMD component? Or classD can works with ordinary component too, like 1/4W resistors, TO-92/TO-220 transistors at the drivers? |
Hi,
Yes, R78+D80 are part of the precharge circuit of the bootstrap capacitor, some delay is required before enabling output stage to make sure that cap is fully charged. During operation though, D84 does all the work.
The two 100k's only purpose are to increase PSRR.
R75 is another part of the bootstrap circuit, it ensures a path for the pre-charge circuit previously discussed, even under no load conditions.
For higher voltage.... hmmmmmm..... off the top of my head, R78 would obviously have to be increased accordingly.
Q142 and Q143 should be changed to ones with higher breakdown voltage.
I wouldn't recommend starting off trying to get one working with +-50V.
Start lower, work up.
Ideally you'd have a good PCB and SMD components all nicely layed out. At lower power levels like this circuit, you can easily get away with P2P, sound quality will likely still suffer somewhat, but it can sound very good.
I don't think anyone would want to do P2P with SMD but if they ever do I want to see it!
My circuit as it currently is using very ordinary parts, but my Mosfets are very high end.
I've had similar experience with this circuit in the past regarding mosfets. I really don't think newer generation types are worse, but they do burn out easier. I believe the older types take alot more abuse because of their greater parasitics, slowing everything down. They're harder to switch so it would take more noise to turn them on spurriously for example.
Hope that helps. Yeah, don't try doing 50 Volts with this, not at first.
Regards,
Chris |
|
|
| lumanauw |
Hi, Chris,
Is it right that feedback before LC is more suitable for heavyworking classD (like for parrareled 1ohm subwoofer)?
Do you know where I can see a good design of selfoscilating with feedback before LC (like IRF amp) but with discretes like your UCD design? |
|
|
| classd4sure |
| quote: | Originally posted by lumanauw
Hi, Chris,
Is it right that feedback before LC is more suitable for heavyworking classD (like for parrareled 1ohm subwoofer)?
Do you know where I can see a good design of selfoscilating with feedback before LC (like IRF amp) but with discretes like your UCD design? |
Hi,
Yeah I forgot to mention R73 could also be scaled for higher voltage... that's not to say I didn't overlook something else as well.
I'm convinced pre filter feedback is inherently inferior so it's not something I've ever tried.
If you really want to drive 1ohm loads I'd say I'd say use a fully differential full bridge. It's easier to start small and work up though.
Sorry, you won't see many fully discrete class d's. I think you can learn alot from that example I gave you though. Try downloading LTspice for free and go to town with it. It will let you learn about how each part of the circuit interacts with the whole of it. Being fully discrete, every small change affects everything else to a certain extent, and a simulator is a great way to learn about it.
Cheers,
Chris |
|
|
| subwo1 |
| quote: | Originally posted by analogspiceman
I have several concerns about your opamp circuit. It appears that nothing prevents in from over driving and severely Zenering the BE junctions of the following differential PNP pair. You really should consider grounding its positive input and running it in inverting mode with the minus input as a summing node at virtual ground (sum the input here as well). Then you could connect a couple of anti parallel diode strings (say two diodes in each string) from the output back to the minus input. This would ensure that the following PNP stage is never overdriven and would also keep the opamp from saturating and getting momentarily stuck to a rail. Once you have made these changes (and if you have hooked up your housekeeping supplies properly), then you should be able to (in fact should need to) add in an RC lead network in the "C2" position (something between 330pF and 1nF in series with 2.2k).
|
Hi analogspiceman,
the idea about the anti parallel diodes is cool. :yes: |
|
|
| lumanauw |
| quote: | | You really should consider grounding its positive input and running it in inverting mode with the minus input as a summing node at virtual ground (sum the input here as well). Then you could connect a couple of anti parallel diode strings (say two diodes in each string) from the output back to the minus input. |
I dont get it. From output to minus input is the feedback resistor place? What is the drawing looks like for the anti parrarel diode and its place? |
|
|
|
