| JojoD818 |
Hi everyone,
Just want to ask which bridge+parallel connection should I try.
The one from AN-1192... |
|
|
| JojoD818 |
| Or this one which uses a DRV134 at the input? |
|
|
| bigparsnip |
| I would go for the second one, just so that all of the power amp chips are opperating int eh same mode, as with all the discussin on here about the varying performances of (non)inverting clones you may end up with a mildly unbalenced bridge setup. |
|
|
| JojoD818 |
Hi,
I have a working gc but it is in a non-inverting mode, can I still use the same topology as in the 2nd schematic?
Thanks,
JojoD |
|
|
| bigparsnip |
| yes, you can do this with n/i amps, just change over the necesary parts, making sure you keep all of the critical components in the circuit. |
|
|
| supereri |
| What size and how many capacitors should be used on the rails? I currently have 1000 on each rail on my single channel IGC. |
|
|
| bigparsnip |
there are I believe two main schools of thought on htis one:
First: Place the same capacitance as you would for each single channel where you normaly would as though each chip was acting independantly of the others.
Second: Have one large bank of caps (as with a conventional PSU) and then use smaller decoupling caps close to each individual chip as before (but using much smaller caps than you normaly would).
I have noe experience of either of these layout (yet), but if this is a first attempt at this, I would use the one which is easier to implement for you at first, and if you are not happy you can easily change it later on (unless you use a PCB).
Andrew. |
|
|
| supereri |
The 4 10K resistors and the 4 220K what tolerance should they be? Also who makes the .1ohm 5w 1% resistor?
Also can the 10k variable resistor ( I assume thats what it is on the signal input) be left out in favor for a preamp?
18v is the most the drv134 can handle, correct? |
|
|
| bigparsnip |
The resistor tolerences for the feedback resistors should be as good as you can get them, idealy 0.1% or better.
Also, I would consider slightly larger output resistors (national recomend 0.22 as a minimum if I recall correctly, or else they may not work correctly) but I'm not too sure where you might find them (I know Dale produce high power and close tolerance wire wound type, and I think Vishay produce precision resisotrs in a T220 package which would run at 5W with a small heatsink).
As far as the DRV134 I'm not too sure, I would just go to naationals website and look it up in the data sheet. |
|
|
| JojoD818 |
bigparsnip,
What you said is what I have in mind! I'll build 4 usual channels, with the caps and decoupling caps and then make a bank of caps in the main psu.
The only problem I have is finding the 0.1ohms 5W 1%, we don't have such resistors here, just 5% 5W. Can I use that?
JojoD |
|
|
| theChris |
0.1 ohm 5 watt resisitors? hmm. i think you could almost make that with a little bit of wire!
5% would be no good. were looking for matching here, not so much accuracy on each, but matching. it creates a network so each opamp will work equally as hard. |
|
|
| JojoD818 |
Hi,
AN1192 states that a 385VA trafo is needed per channel. One channel consists of 4 chipamps right?
JojoD |
|
|
| JojoD818 |
Aside from the transformer question, do I still need to put zobel networks on each of the 4 chipamps?
Thanks,
JojoD |
|
|
| bigparsnip |
Hi, the tollerence of the output resistors is not too critical when it comes down to it, as they are only there to make sure the chip amps share the current equaly between them, and if you can get this down so that there is no more than 5% difference between the chips that should be good enough (although, if you do want better, just buy more resistors than you need and match them up by yourself with a good multi meter).
As far as the Transformer goes, you will only need one for each speaker you use, so yes, one for each set of chip amps.
And if you do use a zobel, I think it wouild be ok to just yse one, placed directly at the output after all of current matching resistors. |
|
|
| JojoD818 |
bigparsnip,
Thanks for the info. I already have a 2 channel gc that works. My plan is to parallel my existing 2 channel and build another 2 channel and parallel it with the 0.1 ohms of course. Then I will bridge them with a drv134 (on it's way :D ). The reason I ask about the zobel is because my pcb has provisions for them and my existing 2 channels have zobel at the output. Thanks again!
If you have other pointers please post them.
JojoD |
|
|
| JojoD818 |
I have successfully bridge/parallel the OPA549. Now a I want to try the OPA541.
There is already a 0.1ohms resistor from the output going to the speaker which is also used for current sense pin.
Do i still need to series another 0.1ohms to parallel 2 OPA541s? If not, for a 4 cell amp (bridge/parallel), the amps in parallel will have their current sense pin 8 tied together, is this ok?
Please help.
JojoD |
|
|
| GregGC |
| quote: | Originally posted by JojoD818
I have successfully bridge/parallel the OPA549. Now a I want to try the OPA541.
There is already a 0.1ohms resistor from the output going to the speaker which is also used for current sense pin.
Do i still need to series another 0.1ohms to parallel 2 OPA541s? If not, for a 4 cell amp (bridge/parallel), the amps in parallel will have their current sense pin 8 tied together, is this ok?
Please help.
JojoD |
It could be OK. Not 1000% sure but it should work fine. If anything you will experience current limmiting earlier than the calculated value. Other than that I don't see any danger of doing it. Get a second appinion though. |
|
|
| theChris |
| yes, it should work, but the current limiting feature will be affected by DC-offset in the amplifers, which i guess is actually still what you want. |
|
|
| JojoD818 |
Guys,
Please be clear, do I need another 0.1ohms resistor or not?
JojoD |
|
|
| GregGC |
| quote: | Originally posted by JojoD818
Guys,
Please be clear, do I need another 0.1ohms resistor or not?
JojoD |
No more resistors.
Go for it and tell us how it works. |
|
|
| JojoD818 |
I forgot to mention that I used the OPA549 for my bridge parallel amp.
Regards,
JojoD |
|
|
| GregGC |
| VERRY IMPRESSIVE! :bigeyes: |
|
|
| JojoD818 |
GreGC,
Thanks for the kind words, the amp is easy to implement and up to this time, a trouble free amp. :)
JojoD |
|
|
| avenger |
| HI , How hard was it to match the components? Its seems here that this matching is what causes the most headaches when building this type of configuration...almost nobody here has accomplished this without the servos in case of the LM3886......please helps us out |
|
|
| JojoD818 |
avenger,
Please bear with me on this but here it goes.
When I was building my bridge parallel, I was too lazy to include the dc servo but not that lazy in matching my components. The resistors for example are 5% tolerance! So I have to measure them carefully with a DMM, that is the reason why I bought a lot of them but they are very cheap. Even the 0.1ohms 5 watt resistors are 5% tolerance so I bought a lot. I then pick the ones with the closest values and used them in my non-inverted pcb for the 549 (see my website). I wired it up, and couple it to a DRV134 for the input. Guess what? the whole amp heats up with a load! :hot:
I then said to myself, maybe I need the servos after all:smash: , but there was this basic circuit in AN1192 that uses dc blocking caps instead of servos. I desoldered one leg of the 1Kohm resistor (going to ground) and insert a small 22uf electrolytic and my problems were solved! :cool: No more heat dissipation without load, even with a load without volume.
I still can't believe that a single cap is all I need to clear that output. I tried it in a single 549, without the cap I measure 0.005V dc offset and with the cap I can barely measure a 0.001V. At a 20mV setting I can measure a 0.01mV dc offset which in my book is a very outstanding offset. :D
I urge you to try it, I never thought a chipamp in such a configuration can handle a lot of power before but now I have proven to myself that sometimes simple things can be good, and in a diyer's hand, can be better.
Best regards,
JojoD |
|
|
| Optical |
IM not sure i follow..
Which is the 1k resistor going to ground?
Which figure in AN-1192 were you looking at there? |
|
|
| Optical |
also did you use the exact schematic as in your second post? with the balanced input? how does it sound?
im planning on building an identical one this weekend as i just got the transformers :) so im just trying to iron out details so the project goes as smoothly as possible!
Thanks,
Matt |
|
|
| JojoD818 |
I use the drv134 for the input. Now, consider the schematic below, it's a non-inverting version from the datasheet. I build 2 pairs of this wherein each pair is in parallel. Now each pair gets their input from one of the drv134's output. Simple isn't it?
The 22uf cap that saved me a lot of headache was inserted between R1 and ground.
HTH,
JojoD |
|
|
| Optical |
interesting! thanks for sharing that
how does it sound?
ive not heard anyone comment of the sound of one of these higher wattage amps i dont think! :bigeyes: |
|
|
| JojoD818 |
Optical,
I cannot comment on the sound since I use it to power a sub. But power is definitely there!
JojoD |
|
|
| Optical |
| I dont suppose you want to give it a try out on some full range speakers and give us a run down? :angel: |
|
|
| JojoD818 |
JDeV,
Hi, my bridge/parallel gc is still working well. In fact, I never built any other amp for my sub since the b/p gc already has sufficient power to feed my sub.
I donot really have any idea regarding the LM4780, I haven't had the chance to read it's datasheet yet. :(
JojoD |
|
|
|
