Hey All,
I am working on integrating a 4780 parallel amp (using Brian's PCB art) into a plate amp for a subwoofer. In the process of designing the woofer box, I have been using the 4780 boards I currently use for a 2-channel amp.
The woofer I'm using is a TangBand W8-740C and it is rated at 4 ohms. When trying to feed the woofer a full load the GC seems to be crapping out and muting the channel every second or so.
Could the problem be related to having too much supply voltage on the rails? That doesn't make much sense since I'm only pushing around +/- 32V. The other possibility is that the chip is overheating.
The trouble is that I can't tell which of the two issues are occurring.
Right now, I have two 4780's hooked up on a 1/2" thick piece of 9"-long aluminum in my chassis, which I would think provides enough heat protection. Also, as I mentioned before, the voltage is well within range.
Any help would be greatly appreciated, as I would hate to have to start fresh and design/build some 3886 boards or something for this purpose...
Thanks,
Chris
I am working on integrating a 4780 parallel amp (using Brian's PCB art) into a plate amp for a subwoofer. In the process of designing the woofer box, I have been using the 4780 boards I currently use for a 2-channel amp.
The woofer I'm using is a TangBand W8-740C and it is rated at 4 ohms. When trying to feed the woofer a full load the GC seems to be crapping out and muting the channel every second or so.
Could the problem be related to having too much supply voltage on the rails? That doesn't make much sense since I'm only pushing around +/- 32V. The other possibility is that the chip is overheating.
The trouble is that I can't tell which of the two issues are occurring.
Right now, I have two 4780's hooked up on a 1/2" thick piece of 9"-long aluminum in my chassis, which I would think provides enough heat protection. Also, as I mentioned before, the voltage is well within range.
Any help would be greatly appreciated, as I would hate to have to start fresh and design/build some 3886 boards or something for this purpose...
Thanks,
Chris
CrisTUFR said:
These are not really 2 issues, but same issue, a cause and its symptom. But that's not important right now, since your voltage and heatsink size are at least reasonable, although not conservative.
Paralleling chip amps is not trivial. What component values did you populate your board with, and how did you deal with issues of equalizing gain between the amps and also DC offset?
The details of the design can be found in this thread:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=32551&highlight=
I used the values listed in the schematics found in that thread. Here is the actual post by BrianGT with the schematic diagram attached:
http://www.diyaudio.com/forums/showthread.php?postid=384993#post384993
I used 1% resistors for everything except the two sandbars on the output. I also have a zobel at the speaker terminal.
This design works perfectly fine with an 8 ohm woofer in the same cabinet, so I'm sure I'm just pushing things beyond their limits. I am only using a 90VA transformer with 18V secondaries for this plate amp's power supply, and was wondering mainly if that would potentially solve my problem.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=32551&highlight=
I used the values listed in the schematics found in that thread. Here is the actual post by BrianGT with the schematic diagram attached:
http://www.diyaudio.com/forums/showthread.php?postid=384993#post384993
I used 1% resistors for everything except the two sandbars on the output. I also have a zobel at the speaker terminal.
This design works perfectly fine with an 8 ohm woofer in the same cabinet, so I'm sure I'm just pushing things beyond their limits. I am only using a 90VA transformer with 18V secondaries for this plate amp's power supply, and was wondering mainly if that would potentially solve my problem.
Sorry to be unclear. The existing amplifier, which is having troubles, is on a 225VA toroidal transformer. It has 22V secondaries.
The new amp that I intend to build will only be 90VA, and 18V secondaries. It's only going to be hooked to a single channel for the sub. I want to avoid this problem on my new amp and was hoping that perhaps I was just outside the voltage that would be safe to use with a 4 ohm load.
The chip gets pretty hot after "loud volume" playback for a long period of time. Once I hook the 4 ohm driver in, It gets hotter than when the 8 ohm driver is hooked up.
The new amp that I intend to build will only be 90VA, and 18V secondaries. It's only going to be hooked to a single channel for the sub. I want to avoid this problem on my new amp and was hoping that perhaps I was just outside the voltage that would be safe to use with a 4 ohm load.
The chip gets pretty hot after "loud volume" playback for a long period of time. Once I hook the 4 ohm driver in, It gets hotter than when the 8 ohm driver is hooked up.
Wow. Talk about interesting issues this past weekend...
So I hacked away and built my sub amp out of a BrianGT LM4780 board I had left over. I just used el cheapo 10% 'peanut' (both in shape and colour) resistors and changed the values a little bit from the original schematic to match what I had laying around. I think that by increasing the gain a little bit from the original, that helped give a little more oomph.
Anyway, I used my 90VA transformer to build this, and wasn't pleased. The speaker tended to pop a lot during use. When I plugged my scope into the GC's output, I noticed some nasty things showing up in the output. I correlated those pops with a dip in the voltage on the rails. Turns out that my transformer was 18VCT, and hence I only had 9V rails before the AC-DC transformation. Once there was enough draw from the 4 ohm speaker, it would dip *just* below 9V on the DC side, which threw the LM4780 into its protection state quite often. Boo-urns!
So in keeping with my el cheapo approach, I decided to grab a Radio Shack 25.2VCT transformer (I think it's only 50VA, but who cares) for less than half what I paid for the nicer Hammond unit. I wasn't expecting much, but after plugging it in with 12V per side (18, I think, after conversion) it worked quite well.
This amp, combined with a P3 CPU heatsink is running strong (and loud) while producing the same amount of heat (quite a bit) as my other GC. The difference is that it NEVER shuts off when under high stress, even though there's a lower-quality transformer powering the thing, AND I have a regulated PSU running the other amp!
I am starting to believe that perhaps my success is tied to using the different values of resistors for a higher gain... Does this make sense?
Just thought I'd document my success so far. I will try to post pics when it's fully built. The 4780 does a great job of driving a very nice amount of bass out of my woofer. Certainly not worth spending the cashola on a 200W+ plate amp!
Cheers,
Chris
So I hacked away and built my sub amp out of a BrianGT LM4780 board I had left over. I just used el cheapo 10% 'peanut' (both in shape and colour) resistors and changed the values a little bit from the original schematic to match what I had laying around. I think that by increasing the gain a little bit from the original, that helped give a little more oomph.
Anyway, I used my 90VA transformer to build this, and wasn't pleased. The speaker tended to pop a lot during use. When I plugged my scope into the GC's output, I noticed some nasty things showing up in the output. I correlated those pops with a dip in the voltage on the rails. Turns out that my transformer was 18VCT, and hence I only had 9V rails before the AC-DC transformation. Once there was enough draw from the 4 ohm speaker, it would dip *just* below 9V on the DC side, which threw the LM4780 into its protection state quite often. Boo-urns!
So in keeping with my el cheapo approach, I decided to grab a Radio Shack 25.2VCT transformer (I think it's only 50VA, but who cares) for less than half what I paid for the nicer Hammond unit. I wasn't expecting much, but after plugging it in with 12V per side (18, I think, after conversion) it worked quite well.
This amp, combined with a P3 CPU heatsink is running strong (and loud) while producing the same amount of heat (quite a bit) as my other GC. The difference is that it NEVER shuts off when under high stress, even though there's a lower-quality transformer powering the thing, AND I have a regulated PSU running the other amp!
I am starting to believe that perhaps my success is tied to using the different values of resistors for a higher gain... Does this make sense?
Just thought I'd document my success so far. I will try to post pics when it's fully built. The 4780 does a great job of driving a very nice amount of bass out of my woofer. Certainly not worth spending the cashola on a 200W+ plate amp!
Cheers,
Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.