OK, I just bought a prototype amp off of ebay from a company called Quad Audio just because I was curious. Well, I pulled the back panel off & to my untrained eyes, it appears to be a decent amp with more potential. I found that it had 4 spaces on the output side with unused FETs. Would it really be as simple as throwing in more matching FETs & resistors/diodes to upgrade this thing, or would a power supply upgrade be necessary. I know just about enough to be dangerous which is why I'm asking.
Here's some pics:
Here's some pics:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
If you have a working amp, I'd suggest simply using it as it is.
Some class D amps are very sensitive to changes. Adding the FETs and the related components may cause reliability issues.
From the layout, it's amazing that they produced a working prototype. The application notes recommends keeping the IC as close to the output transistor as possible. This amp has the IC far from the outputs.
Some class D amps are very sensitive to changes. Adding the FETs and the related components may cause reliability issues.
From the layout, it's amazing that they produced a working prototype. The application notes recommends keeping the IC as close to the output transistor as possible. This amp has the IC far from the outputs.
Yeah, I'm not sure if that's the cause, but I hear some odd higher frequency noise coming out of it (hooked to a power supply). It's funny, because there doesn't appear to be any reason that they needed to put the IC so far away either, because there doesn't seem to be anything immovable in the way.
I'm not too worried about possibly messing the amp up if there's a possibility of increasing the power. Like I said, I was just interested in playing around.
I'm not too worried about possibly messing the amp up if there's a possibility of increasing the power. Like I said, I was just interested in playing around.
More FETs
It appears they were using groups of three FETs in parallel and decided later to cut that down to 2 FETs. Adding the deleted FETs will possibly damage the amp unless you know exactly what criteria was used to match them.
Regardless, they will do nothing to give you the least bit more output power. Maximum power output is determined by the power supply. All that additional devices to do is to allow you to operate to a lower load impedance. However, doing this will meed beefing up of the power supply as well and recalibration of the protection circuits.
If anything the transformer looks a little light. I use 2 transformers that size for a 125W x 4 Class AB car amp I designed for Tru Technology. And while Class-D amps are more efficient, they are still not that much more efficient as I also design Class-D amps.
Looking at the main power filter capacitors, the pair of 1500uF caps to the right of the transformer, I have to tell you that this amp is incredibly low for the amount of capacitance there. These capacitors determine how tight the bass will be and this amp is light in that department. Paralleling these caps with a lot more capacitance is the best way to improve this amp without blowing it up. Just remember that the longer the wires to extra caps, the less well they will help.
Dan
It appears they were using groups of three FETs in parallel and decided later to cut that down to 2 FETs. Adding the deleted FETs will possibly damage the amp unless you know exactly what criteria was used to match them.
Regardless, they will do nothing to give you the least bit more output power. Maximum power output is determined by the power supply. All that additional devices to do is to allow you to operate to a lower load impedance. However, doing this will meed beefing up of the power supply as well and recalibration of the protection circuits.
If anything the transformer looks a little light. I use 2 transformers that size for a 125W x 4 Class AB car amp I designed for Tru Technology. And while Class-D amps are more efficient, they are still not that much more efficient as I also design Class-D amps.
Looking at the main power filter capacitors, the pair of 1500uF caps to the right of the transformer, I have to tell you that this amp is incredibly low for the amount of capacitance there. These capacitors determine how tight the bass will be and this amp is light in that department. Paralleling these caps with a lot more capacitance is the best way to improve this amp without blowing it up. Just remember that the longer the wires to extra caps, the less well they will help.
Dan
Well, darn, why does everyone have to ruin my fun. Though a lower impedence is kinda nice, I'd pass if that's all I'd get. I was hoping that it wouldn't be power supply limited, but I guess the more I look at it, the more it looks lacking.
I was already assuming that the amp was far below the 1000w rating. I was thinking in the 500w neighborhood when driven at 4 ohms mono. So, no suprise there. The lack of capacitance did suprise me as well compared to some of the better amps that I have.
I do have a bunch of axial 1500uf capacitors sitting around. I wonder how many I could stuff in there without it being dangerous/hideous? Looks like maybe 6 before I started to get pretty far from the traces.
I really do appreciate the help, I've been wanting to learn more over the years. Actually, about 2 years ago, I bought your CDs Perry. I was suprised to find you on this forum (not that I should have been). I can't say enough about how much I learned & am still learning from your stuff. Was worth every penny.
I was already assuming that the amp was far below the 1000w rating. I was thinking in the 500w neighborhood when driven at 4 ohms mono. So, no suprise there. The lack of capacitance did suprise me as well compared to some of the better amps that I have.
I do have a bunch of axial 1500uf capacitors sitting around. I wonder how many I could stuff in there without it being dangerous/hideous? Looks like maybe 6 before I started to get pretty far from the traces.
I really do appreciate the help, I've been wanting to learn more over the years. Actually, about 2 years ago, I bought your CDs Perry. I was suprised to find you on this forum (not that I should have been). I can't say enough about how much I learned & am still learning from your stuff. Was worth every penny.
I'd be pretty tempted to try and upgrade that thing if I had it as well. Wouldn't it be less sensitive to adding the extra output drivers than an A/B style amp, for example? Since the output signal is a switching signal instead of varying voltage/current signal? There also may be higher current power supply devices available in the same package size, as well. I mean, of course you'd want to try and get matched date outputs and power supply transistors, but wouldn't it be more critical in an "analog" style amp than it would be in this switched type? Just asking, seems like I've read it somewhere before..
Yeah, I have even considered upgrading the powersupply FETs, but don't want to go too far without the guru's here telling me if I'm wasting time or not. The small transformer might stop me from upgrading the power supply too much though. Still, I can't help but think that there was intention behind those extra output slots.
ppia600:
To have low distortion and good efficiency, the FETs must be switched on and off with very precise timing. If there is too much dead time (time that none of the FETs are 'on'), you will get high distortion. Too little dead time and you'll get shoot-through which can destroy the FETs. Changing the load on the drive circuit changes the dead time in many class D amps.
I've seen amplifiers that wouldn't work when using exactly the same part number replacement as the original part because the part was changed in production (they are generally changed to make it possible to get more FETs per wafer). The specs are very similar in the two versions but the difference is enough to prevent the amp from working properly.
While it's possible that the amp would work with additional FETs, it's also possible that the amp will fail and it could be difficult to repair.
normalicy:
If you want to experiment, do this...
Use the amp for about a week or so to confirm that it's in perfect working order and is reliable.
Find the part number of the outputs currently being used. Order replacements that have the same part number. Remove and keep the originals. Install the replacements and run it again for another week to confirm that the replacements are working perfectly.
If they work perfectly, add the new components. If it continues to work well, You can experiment with lower impedance loads or whatever you decide to try. If it fails, you know you have suitable replacements. If it doesn't work with the replacement parts, you have the originals to reinstall and you still have a usable amp.
To have low distortion and good efficiency, the FETs must be switched on and off with very precise timing. If there is too much dead time (time that none of the FETs are 'on'), you will get high distortion. Too little dead time and you'll get shoot-through which can destroy the FETs. Changing the load on the drive circuit changes the dead time in many class D amps.
I've seen amplifiers that wouldn't work when using exactly the same part number replacement as the original part because the part was changed in production (they are generally changed to make it possible to get more FETs per wafer). The specs are very similar in the two versions but the difference is enough to prevent the amp from working properly.
While it's possible that the amp would work with additional FETs, it's also possible that the amp will fail and it could be difficult to repair.
normalicy:
If you want to experiment, do this...
Use the amp for about a week or so to confirm that it's in perfect working order and is reliable.
Find the part number of the outputs currently being used. Order replacements that have the same part number. Remove and keep the originals. Install the replacements and run it again for another week to confirm that the replacements are working perfectly.
If they work perfectly, add the new components. If it continues to work well, You can experiment with lower impedance loads or whatever you decide to try. If it fails, you know you have suitable replacements. If it doesn't work with the replacement parts, you have the originals to reinstall and you still have a usable amp.
Hi
Transformer is not that small, lots of wires on it... if you would increase freq. of power supply, it would be able to put more power out, but wires on it are probably limitation, not core itself
But I wouldn't change this amp in any way, since it works, and if it works don't fix/change it...but do as you wish
Transformer is not that small, lots of wires on it... if you would increase freq. of power supply, it would be able to put more power out, but wires on it are probably limitation, not core itself
But I wouldn't change this amp in any way, since it works, and if it works don't fix/change it...but do as you wish
As crazy as this might sound, I have found it fairly easy to get in contact with the engineers of different products.
With a little research you could likely talk to the person that designed the amp and ask why they opted to not use all the FETs on the output. It was clear that they had either the "intent" OR "option" to add more fets. This is where it is a gamble, as if they had intent and didnt, there is a reason. This is likely not the only prototype and the ones with all the fets may have not worked properly. If they had the "option" then it might be OK but just not justify the cost in production VS a gain in performance.
This would make me think that by adding the fets you will make the amp work poorly if the engineers had intent or that it would be an unnoticeable gain if they had the option.
I say do it, I would like to know what happens……(its not my amp!)
BTW: I am somewhat new to doing this sort of thing but I could see a manufacture opt to use fewer fets of a higher value then the original design that had more fets of a lesser value. THis could be a cost cutting measure that equaled performance and lowered production times. This could also explain why there is fewer fets on the board. But take into consideration that I have no clue what I am talking about.
-JH
With a little research you could likely talk to the person that designed the amp and ask why they opted to not use all the FETs on the output. It was clear that they had either the "intent" OR "option" to add more fets. This is where it is a gamble, as if they had intent and didnt, there is a reason. This is likely not the only prototype and the ones with all the fets may have not worked properly. If they had the "option" then it might be OK but just not justify the cost in production VS a gain in performance.
This would make me think that by adding the fets you will make the amp work poorly if the engineers had intent or that it would be an unnoticeable gain if they had the option.
I say do it, I would like to know what happens……(its not my amp!)
BTW: I am somewhat new to doing this sort of thing but I could see a manufacture opt to use fewer fets of a higher value then the original design that had more fets of a lesser value. THis could be a cost cutting measure that equaled performance and lowered production times. This could also explain why there is fewer fets on the board. But take into consideration that I have no clue what I am talking about.
-JH
Well, I doubt I'll be able to find the engineer(s). Quad Audio as best as I can tell is an install shop that wanted their own brand. I'm betting that it's a somewhat generic design, but I can't find it anywhere on ampguts. Here's Quad Audio's rather sparse site:
http://www.quadaudio.net/index.html
I'm thinking I'm going to do it, because there's nothing to lose anyhow. I've got tons of working amps, so I won't be out an amp. I've also got tons of broken amps, so why not add to the stack.
http://www.quadaudio.net/index.html
I'm thinking I'm going to do it, because there's nothing to lose anyhow. I've got tons of working amps, so I won't be out an amp. I've also got tons of broken amps, so why not add to the stack.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.