Chris,
Don’t you ever sleep? If your in Canada then your worst then me…
Enter the DARKER side - why small, in this circuit topology I was thinking of say 470R or bigger!
John
Don’t you ever sleep? If your in Canada then your worst then me…
Enter the DARKER side - why small, in this circuit topology I was thinking of say 470R or bigger!
John
Sleep??.. they come in my dreams...
heh.
The higher Rb the higher the driver bias resistors need to be, which are 460ohm right now. I could possibly go higher but 460 Rb.. likely something like 2k for the driver resistors.
Bruno you can't be hinting to DC feedback can you?
Regards,
Chris
heh.
The higher Rb the higher the driver bias resistors need to be, which are 460ohm right now. I could possibly go higher but 460 Rb.. likely something like 2k for the driver resistors.
Bruno you can't be hinting to DC feedback can you?
Regards,
Chris
Goodness no, that would be a "Japanese" solution.classd4sure said:
I'm telling you it's simple. When you find it you'll recognise it straight away because you'll be banging your head why you hadn't thought of it before.
Come to think of it, wasn't that the reaction here about the whole UcD thing?😀
!!!!
I must've been sleeping while posting! I actually *do* use emitter resistors in the bottom transistors (the ones with their bases tied together)
The reason why I forgot that is simply that their function was not related to balancing and their presence is not required for proper operation. The balancing problem you're seeing is related to the other two which are seeing vastly different power dissipation.
The emitter resistors were added to prevent the bottom transistors from becoming a contributing factor to noise.
I must've been sleeping while posting! I actually *do* use emitter resistors in the bottom transistors (the ones with their bases tied together)
The reason why I forgot that is simply that their function was not related to balancing and their presence is not required for proper operation. The balancing problem you're seeing is related to the other two which are seeing vastly different power dissipation.
The emitter resistors were added to prevent the bottom transistors from becoming a contributing factor to noise.
Well, now I completely lost – guess that’s what happens when you try to second-guess a problem – where exactly is the problem with “Balance / current hogging”?
John
John
Hi,
I second that.
Balance problem? Perhaps the output pairs would differ in power dissipation but, from one side of the comparator to the other, that would be alternating would it not, therefor not much of a concern?
My current hogging concern is that of the quad outputs, having one transistor in each pair possibly hog the current from the other, resulting in one, or both, of the half bridges no longer being able to switch properly.
I think you're right John, it does only need two emitter resistors..
So with that in mind, why opt for a three part solution, which thus far has escaped me, and likely will continue to.
I thought before the monolithic IC would be enough to guard against such hogging, apparently it isn't, but with two resistors, it should be safe enough.
I'll likely never get to build this myself but hopefully some will and I would like it to work OK for them, otherwise just for a simulation I wouldn't worry about it at all.
I second that.
Balance problem? Perhaps the output pairs would differ in power dissipation but, from one side of the comparator to the other, that would be alternating would it not, therefor not much of a concern?
My current hogging concern is that of the quad outputs, having one transistor in each pair possibly hog the current from the other, resulting in one, or both, of the half bridges no longer being able to switch properly.
I think you're right John, it does only need two emitter resistors..
So with that in mind, why opt for a three part solution, which thus far has escaped me, and likely will continue to.
I thought before the monolithic IC would be enough to guard against such hogging, apparently it isn't, but with two resistors, it should be safe enough.
I'll likely never get to build this myself but hopefully some will and I would like it to work OK for them, otherwise just for a simulation I wouldn't worry about it at all.
Well you've got the two trannies driving the current balance by their emitters and the mosfet drivers by their collectors. The one driving the high side is getting hot, upsetting the balance.JohnW said:
Point is to keep that from happening. Using a monolithic matched pair (or quad in the case of a FB) is one solution, but not the most elegant one.
The emitter resistors I have are in the bottom pair (the "balance"), and are not related to the current hogging problem.
Well, I have a feeling I'll catch some hell for this .... so be it.
This is a slight modification to my earlier idea that I mistakingly thought worked (ZZzzZzZzz...). This does work.
I trust it should completely solve any current hogging issues whatsoever. That's 2 parts.
For a third.. and maybe a fourth, I may consider some sort of mirror/level shifter, to unload the full swing and balance issue from the upper drivers, haven't really put much thought into that one yet though.
I trust doing so would allow the upper and lower drivers to operate at the same temperature, thereby also helping with any hogging related issues.
So, have I barked up the wrong tree here or what?
Chris
This is a slight modification to my earlier idea that I mistakingly thought worked (ZZzzZzZzz...). This does work.
I trust it should completely solve any current hogging issues whatsoever. That's 2 parts.
For a third.. and maybe a fourth, I may consider some sort of mirror/level shifter, to unload the full swing and balance issue from the upper drivers, haven't really put much thought into that one yet though.
I trust doing so would allow the upper and lower drivers to operate at the same temperature, thereby also helping with any hogging related issues.
So, have I barked up the wrong tree here or what?
Chris
Attachments
You could start from the idea of post #410 and give "the second output pair" their own common emitter resistor (ie not tie the emitters to the balance).
Hi,
Yeah, that's likely where I'll finish with it too. I guess you don't like the outputs being removed from the mirror..
Yeah, that's likely where I'll finish with it too. I guess you don't like the outputs being removed from the mirror..
Hi,
My plastic protoboard version is ... downright amusing!
Hooked up to a 400W cerwin vega 15" (very efficient speaker)/4ohm load.. well.. I've had noise complaints. Two in fact.
1. It's too loud at 12am
2. It blanks out the TV upstairs heheh.. darn air core coils.
It doesn't affect the TV that's alot closer to it, but the other one has an antenna on it ..
With some heatsinks and a fan on it I can keep it going long enough to be annoying, then the upper FET gets hot and I have to shut it down for a minute or so.
I can't improve on it without a scope, so I admit there is switching noise comming through which I expected considering.., You dont' hear it at full power but it is noisy.
I've listened to it enough to say the following:
Being the first class d I've ever heard at anything over 1W..
What I can say is the control over the bass is like nothing else I've experienced. My old A/B pioneer was always a sore spot where the bass was concerned, I thought it lacked power but evidently it lacked the control. In comparison this thing does lack power, but still manages to take an iron grip over it, instead of just a rumble it's rich and detailed.
The rest of the range is nice but too noisy to bother commenting on more than that. It's given me an itch anyway.. where oh where are those 400w modules..
I post this for any students out there or anyone who wants to play around and considering making it up on a protoboard just for kicks with sub standard parts that they have laying around, it's worth it!
My plastic protoboard version is ... downright amusing!
Hooked up to a 400W cerwin vega 15" (very efficient speaker)/4ohm load.. well.. I've had noise complaints. Two in fact.
1. It's too loud at 12am
2. It blanks out the TV upstairs heheh.. darn air core coils.
It doesn't affect the TV that's alot closer to it, but the other one has an antenna on it ..
With some heatsinks and a fan on it I can keep it going long enough to be annoying, then the upper FET gets hot and I have to shut it down for a minute or so.
I can't improve on it without a scope, so I admit there is switching noise comming through which I expected considering.., You dont' hear it at full power but it is noisy.
I've listened to it enough to say the following:
Being the first class d I've ever heard at anything over 1W..
What I can say is the control over the bass is like nothing else I've experienced. My old A/B pioneer was always a sore spot where the bass was concerned, I thought it lacked power but evidently it lacked the control. In comparison this thing does lack power, but still manages to take an iron grip over it, instead of just a rumble it's rich and detailed.
The rest of the range is nice but too noisy to bother commenting on more than that. It's given me an itch anyway.. where oh where are those 400w modules..
I post this for any students out there or anyone who wants to play around and considering making it up on a protoboard just for kicks with sub standard parts that they have laying around, it's worth it!
The difference between building a switching amp on a prototype board or on a proper PCB (with groundplane !!) can be like day and night. Using the very same circuit of course !
You can try this for yourself if you have the time and patience.
Regards
Charles
- Status
- Not open for further replies.