Tri-amp with Brian's board

[guitvinny]

Hi,

I'm about to build a tri-amp system for my speak from chipamp.com boards and I got few questions :

-I want to know if the trace of one power supply board from Brian will be able to support 3 x lm3886 ? More info : I use a 22-0-22V transformer 300 VA and the load will be 2 x 8ohm and 1 x 4ohm (2x Vifa MG18WK and 1 x DX25SG05-04)

-Also, I'm tempted to trade the 100uF supply capacitor on the amp board for 1000uF low ESR. Again, from a power supply point of view, will it be too much ? I will have 2x 10000uF on the power board and 2x1000uF on each 3 amplifier board.

-Except from what is explained in chipamp build guide, is there any other considerations I should observe to ground everything ?

Thanks for your time, I will post some pic when completed

(see attachment for my speaker)

[pacificblue]

Did you try to mail the question about the power supply traces to chipamp.com?

1000 µF on the amp board is exactly, what the datasheet recommends. Nevertheless some people seem to like the smaller capacitance better due to its effect on the mids and highs. For the woofers it is better to use as much capacitance as you can afford, while the tweeter will probably be fine with 100 µF. On the other hand you are effectively connecting all those capacitors in parallel, so what you do on one PCB will affect the others as well.

The question about grounding implies that everybody knows, what the chipamp assembly manual explains and what is missing. The best way to go about it is to build the amps according to the manual and ask first chipamp.com and then the Forum for help, if it does not work out.

[guitvinny]

I know Brian is somewhere around in this forum .. if you read this, please give me a small hint .

Else, if some one who has already done tri-amping with chipamp board (successfully) .. please raise your hand

Thanks in advance

[guitvinny]

Hi again,

I got some news from Brian : should be fine without "beefing-up" the board by soldering some extra wires underneath. So, I have solder all 6 amplifiers and 2 power boards.. everything is tested and it's working fine, my test set-up : Cd line-out >> simple active crossover (12dB/oct : 500Hz low-pass, 2300Hz low-pass, 2300Hz high-pass) >> 3 x amplifier module (all of them got 30dB gain = 20.log(1+22000/680)) >> to speakers (mg18wk09-08,mg18wk09-08 and dx25tg05-04, it's a 2.5 way). Please note that all crossover values have been arbitrarily chosen for testing.

Now, I need to adjust the power coming to the mid/sub and to the tweeter, since tweeter got a higher sensitivity and different impedance (93.5dB/m vs 86.5dB/m and 4ohm nom. vs 8ohm nom). I've done some math in order to adjust the gain and the power coming to tweeter. I'm not 100% sure of what I've done, so if I missed something or if I'm going in the wrong direction, please give me some indications.

So :

-Mid/sub amplifier can deliver 45W max (lm3886 datasheet @ ±30V in 8ohm)
-Vout_max_sub = sqrt(45W * 8ohm) =18.97V, this is the voltage max coming from the amp to deliver 45W into 8ohm
-The amplifier got a gain of 1+22000/680 = 33.35 = 30.46dB
-So the max input it can take without clipping is 18.97V/33.35 = 0.5689V
-Now, the tweeter have a sensitivity of 93.5dB and the sub got 86.5db, the difference = 7dB
- So I need a amplifier with 23dB of gain (30.46dB - 7dB = 23.46db). In order to do so, I change the ground feedback resistor form 680ohm to 1787ohm (R3 on Brian's board).
- To match the power of the sub, I need an amplifier of 8.98W (= 45W / (antilog(7/10))

Still with me ? Next part is the one I'm not really sure..

- To max the tweeter amplifier with the same signal amplitude as the sub amp : Vout_max_tweet= sqrt(8.98W / 4 ohm) = 6V
-The tweeter amp got a gain of 13.31 (23dB), so Vout_max_tweet happen when there is an input signal of 6V / 13.31 = 0.45V
-But, the sub amp deliver his max when he got 0.5698V at his enter. So, to adjust the tweeter amplifier max power I put a resistor divider in front of it with a ratio of 0.45V/0.5689V = 0.79 (a 5.6k with a 22k)

As you see it's the first time I made those kind of calculation and I need a bit of help. Do I change the ground feedback resistor from a 68ohm to a 1787ohm, in order to reduce the tweeter amp gain ? And, do I add a resistor divider in front of the tweeter amp to adjust his max power ?

Sorry for this long post, many many thanks for the time taken and happy holiday everyone !

References :

Tweeter : DX25TG05-04 XT/DX 1″ Tweeter Vifa Datasheet
Sub/Mid : MG18WK09-08 | Tymphany
ESP biamp : BiAmp (Bi-Amplification - Not Quite Magic, But Close) - Part 2

[AndrewT]

Hi,
you have a 4ohm driver.
This means you must select your PSU voltage and VA rating to suit the 4ohm load.

Once you have done that, check what power the other amp/s is capable of sending to the 8ohm driver/s.

If you are looking at -7dB of attenuation, I would do that with an adjustable pot or cheap switched pot at the input of the amp that has too much gain.

But I am not sure your 7dB is correct.

Let's start with 60W into 4r0. That gives +17.8dB +93.5dB = 111.3dB spl.
That's your maximum SPL from the tweeter, not continuous but on a very short transient.

An 8r0 load will draw a maximum of ~35W from the amp using the same PSU as the tweeter amp.
that gives +15.4dB + 86.5dB = 101.9dB spl.
Adding a second driver and second amplifier will lift the max SPL, over a very restricted frequency range, by 6dB, giving 107.9dB spl. That's -3.4dB bass to treble.
But, you now have a hole in the midrange of about -9.4dB.
Active speakers are much more difficult than simply placing active filters in front of power amplifiers.

[pacificblue]

Originally Posted by guitvinny

- To match the power of the sub, I need an amplifier of 8.98W (= 45W / (antilog(7/10))

If you don't want to use several power supplies, you can skip that step.

Originally Posted by guitvinny

- To max the tweeter amplifier with the same signal amplitude as the sub amp : Vout_max_tweet= sqrt(8.98W / 4 ohm) = 6V
-The tweeter amp got a gain of 13.31 (23dB), so Vout_max_tweet happen when there is an input signal of 6V / 13.31 = 0.45V
-But, the sub amp deliver his max when he got 0.5698V at his enter. So, to adjust the tweeter amplifier max power I put a resistor divider in front of it with a ratio of 0.45V/0.5689V = 0.79 (a 5.6k with a 22k)

All correct in theory. In practice you should measure the sensitivity of the drivers in the final box. Datasheets show the sensitivity measured in a very big enclosure and on a very big baffle. The enclosure type, volume and dimensions have a significant influence on sensitivity and frequency response. And the theoretical amplifier gain is not the real gain.

Originally Posted by guitvinny

Do I change the ground feedback resistor from a 68ohm to a 1787ohm, in order to reduce the tweeter amp gain ? And, do I add a resistor divider in front of the tweeter amp to adjust his max power ?

Use a potentiometer/trimmer. That allows for precise adjustment. Once you have the correct setting, you can replace the potentiometer with fixed resistors according to the potentiometer setting. Or you can leave the potentiometer in for re-adjusting at a later time, when you had enough time to listen and find the faults your first listening session will not reveal due to the excitement of a finished project.

Changing the resistor value from the inverting input to ground influences DC offset and the filter frequency at the inverting input. If you do that, you should change the blocking resistor at the non-inverting input accordingly.

AndrewT is right about the 4 Ohm driver. In your case you may get away with ±30 V, because the 4 Ohm driver is the tweeter, which has to cope with less energy and will be reduced in volume.

[jleaman]

Brian's a pretty busy guy as audio is not his main hobby any more, altho he still does some stuff ( SOME ) I have forwarded this link to him for you.


Jase