M-audio AV30 speakers recap and repair

[mariushm]

Re: M-audio AV30 speakers recap and repair

Well, I went with the maximum written on the picture, 14v ac @ 0.8A.

Actually, i forgot something. The peak DC current will be about 0.62 x 0.8A = 0.5A , because you can't magically make 19v @ 0.8A from 14v AC @ 0.8A without any losses. So if you assume that 0.8A transformer current is correct, redo the math assuming the maximum output DC current to about 0.5-0.6A, not 0.8A.

You can always put a multimeter in series with the regulator and push the volume to the max and measure the peak current, then put a suitable heatsink.

ps. I went back and edited the previous posts to take account of 0.5A peak DC current, if we assume a 0.8A AC current.

[lexwalker]

Re: M-audio AV30 speakers recap and repair

If the connector says 14v AC @ 0.8A that means there's going to be about 20v after rectification, minus about 1v from the voltage drop on the diodes.

edit: The peak DC can be estimated using the formula Idc = 0.62 x I ac = 0.62 x 0.8A = ~ 0.5 A.

So 19v @ 0.5A... But keep in mind at very low currents the voltage drop on the rectifier diodes will be smaller, and the transformer may output 5-10% more AC voltage, so the DC voltage may reach a peak close to 25v, so it's safer to use 35v rated capacitors.

Usually I always use 1.41xVacRMS and ignoring the voltage drop across the diodes/rectifiers, thus my result would be 19.74Vdc peak voltage. Not forgetting that this is a linear power supply using a transformer which the voltages on the secondary side would correspond to the voltage on the primary side. There are also variations in voltage in different places, anywhere from 220VacRMS to 250VacRMS. The designers have to account for that. Worst case scenario, I would assume the transformer to generate up to 17.5VacRMS on mains voltage surges (of up to 275VacRMS), thus giving 24.675Vdc peak voltage which is close to 25Vdc. Possibly the main reason they originally chose 25Vdc rating for the capacitor. Having 35VDC does give some headroom (about 10Vdc) in case peak voltage ever exceeds that. Quite often designers would use 2X the VacRMS for the capacitor voltage. Usually I would prefer around 2X the peak DC voltage (assume the expected input voltage to the transformer is around 220V to 230V for the specified secondary VacRMS)...

Let's assume there may be some voltage drop on the transformer at full 0.5A and say 17v DC minimum at any time after rectification.

With 0.5A at the output, normally I'd recommend 1117 regulators with 1-1.1v maximum dropout voltage, but these only support up to 15v at the input.
So you'd have to use 7812 and 7912 or LM317, LM337.. the classics, which have about 2v dropout voltage.

Therefore, if you really want +/-12v in the worst conditions, you want to have at least 14v at the inputs so then you can use the formula to determine capacitance.

C = Current / ( 2 x mains frequency x Vripple) = 0.5A / (2x60Hz x (17v - 14v) ) = 0.5 / 120x3 = 0.5 / 360 = 0.0013888[88] F = ~ 13888uF

So in theory, with minimum 17v DC after rectifier, with a 1500uF capacitor you'd have at any time 14v or more at the input of the regulator, even when 0.5A is pulled through the regulator. Going with higher capacitance, you will just get the minimum DC voltage higher, closer to 17-19v, which in turn will cause more power to be wasted in the regulator and therefore heat it up more, and you'd need a larger heatsink.
The amplifier won't use 12v @ 0.5A constantly, it will only use as much as needed, the average power draw will depend on the volume and how the music is.

But to be on the safe side, I'd still recommend going with a 2200uF capacitor or why not, even 2700-3300uF.

Even if you use a small capacitor, in the worst case scenario (when the volume is pushed very high and power consumption peaks), the input voltage will simply drop a bit below 14v, which means the regulators in turn will output a little bit below 12v. The amplifier chip is designed to work with a wide range of input voltages, so it won't mind a bit less than 12v for a few ms, but you may get a bit of distorted sound. Considering you won't use the speakers at maximum (at least I hope), this won't be a problem.

Most of Europe, especially the British isle including Scotland uses 50Hz mains frequency. Your calculation is based on USA or Japan's 60Hz mains frequency. Since the original design uses zener diodes, my guess is they would assume worst case voltage drop down to 12V, with 14V (the RMS voltage) minimum on full current draw/load. If going by your formula, then the result would be 2500uF. Thus with capacitance value of 2500 and above, the revamped 3300uF value would fall in line with those assumptions...

[mariushm]

Re: M-audio AV30 speakers recap and repair

I used 60 Hz because I saw he lives in US, under his name.

Let me expand a bit on a few things, which matter when designing such a linear power supply.

Most small transformers (such as the one in these speakers, probably rated for 30VA or something like that) are usually designed in such a way that at low current consumption, the output of the transformer can be up to 5-10% more than the voltage written on the label.

So, at 0.1A this transformer may output 14v AC rms + 10% or about 15.5v ACrms.

Then, you also have to consider that the mains voltage may be higher or lower than the standard 110v (in US)..

As the transformer has a 7.85:1 ratio (110v/14) under load and 7.1:1 at idle (110v/15.5v), with 120v you may find that the transformer outputs 120v/7.1 = 16.9v AC rms.
On the other side, worst case scenario, you could have 100v/7.85 = 12.73v AC rms (under load, with undervoltage at mains).
That's why it's important to leave room for safety when choosing capacitors... depending on conditions, you can get much higher voltages.


In Europe, you have to be even more careful, because the mains voltage is standardized at 230v but some countries have the default still at 220v, others have it set at 240v.... with the +/-5% of tolerance. So one should normally design for 210v-250v

When using a bridge rectifier and capacitor to get DC voltage you have:

V peak DC = 1.414 x V ac rms - 2 x Vdrop rectifier
I peak DC = ~ 0.62 x I ac

The voltage drop on each diode varies with the load. Some diodes are better than others... they may be 0.4v @ 0.1A and 0.8v @ 1A.

In this example, the KBL406G is kind of lame... the drop on each diode is 0.7v @ 0.1A and 0.8v @ 1A (see datasheet here: page 2, right).


So I'm going to assume 110v AC rms and 60 Hz as he lives in US, and I'm going to assume the amplifier will typically use a few hundred mA, therefore the transformer will not output 10% more than its rated voltage.

Vdc = 1.414 x 14 - 2 x 0.8v = 18.2v
Idc = 0.62 x 0.8 = 0.5A

Now that's a peak dc voltage of 18.2v... again, formula is C = I/(2x Fac x Vripple) = 0.5A/ [2x60x(18.2-14)] = 0.5 / 504 = ~ 990 uF.

In the previous comment, I was even more conservative and assumed the 14v value was an approximation and used 17v for the AC voltage instead of 18.2v. So 2200uF should definitely be enough, for running this in US.


I doubt he bought those things from Goodwill for $5 only to fix them and sell them to someone in Europe (probably the mains transformer doesn't even have a 220-230v selection) so I don't see the point of making the power supply overkill.

[spiff59]

Re: M-audio AV30 speakers recap and repair

Thank you, everyone, for all the useful information.
I am in the US and our utility provides a consistent 118vac @ 60hz.
A tall pair of 18mm nichicon VZ caps rated at 3300uf/50v/105c fit nicely as replacements for the primary filter caps. The higher voltage was probably unnecessary, but I replaced the caps on the output side that had literally melted down with nichicon PJ's at 220uf/35v/105c (was a snug fit).

I looked in my junk box and found I had salvaged a pair of these heatsinks off of some dead critter: http://www.photonage.com.au/heatsink...321a1e79140540
I found I had just enough room to nicely affix the heatsinks to the metal back panel of the speaker between the circuit board and the speaker's cylindrical air port. Then I looked up the 7812 and 7912 parts, and their pinouts and am left saying "what?". The tab is ground on the 78xx, but hot (Vin) on the 79xx? Grumble. Sure would have been convenient had they both had the same pinout. It seems to me the 7812 with just some Arctic Silver between it and the heatsink will be running considerably cooler than the 7912 whose contact with the heatsink is separated by a mica or mylar insulator

[ben7]

Re: M-audio AV30 speakers recap and repair

Thank you, everyone, for all the useful information.
I am in the US and our utility provides a consistent 118vac @ 60hz.
A tall pair of 18mm nichicon VZ caps rated at 3300uf/50v/105c fit nicely as replacements for the primary filter caps. The higher voltage was probably unnecessary, but I replaced the caps on the output side that had literally melted down with nichicon PJ's at 220uf/35v/105c (was a snug fit).

I looked in my junk box and found I had salvaged a pair of these heatsinks off of some dead critter: http://www.photonage.com.au/heatsink...321a1e79140540
I found I had just enough room to nicely affix the heatsinks to the metal back panel of the speaker between the circuit board and the speaker's cylindrical air port. Then I looked up the 7812 and 7912 parts, and their pinouts and am left saying "what?". The tab is ground on the 78xx, but hot (Vin) on the 79xx? Grumble. Sure would have been convenient had they both had the same pinout. It seems to me the 7812 with just some Arctic Silver between it and the heatsink will be running considerably cooler than the 7912 whose contact with the heatsink is separated by a mica or mylar insulator

How are you mounting the heatsinks onto the back panel?

[spiff59]

Re: M-audio AV30 speakers recap and repair

I haven't decided upon any solid course of action yet, I'm still brainstorming.
The panel is metal and grounded, I might have tack-soldered the heatsinks on, had both devices had ground on the mounting tab. Epoxy is an option. I could leave a "live" heatsink inside the speaker, it's not like -14v or -20v at 1A is going to curl the hair of someone shocked by it (I'll probably be the last person inside this speaker). But, that doesn't seem a clean solution. I'm tempted to search for a couple insulators with decent thermal properties and, just to be consistent, insulate both the 7812 and the 7912 from their heatsinks.

[stj]

Re: M-audio AV30 speakers recap and repair

a pair of gray or pink thermal pads and a pair of insulating bushes for the screws.
salvage both from any pc psu.

[spiff59]

Re: M-audio AV30 speakers recap and repair

I finally got back to this project. I plugged the Vin and Vout from the regulators into the vacant holes where the (removed) 2W resistors had been and drilled a couple other holes on a nearby beefy area of the ground plane. I put the regulators (insulated) on a single oversized heatsink that snugly tucks into a gap between the transformer and the heatsink for the output transistors and after double-checking things, and crossing my fingers, I fired it up. Am reading almost 34Vac across the transformer, making me wonder how the zeners ever held up. I tested a solid + and - 12v out of each regulator and after assembly the speakers work great and the heatsink stays relatively cool. Thanks, guys

[jgiebler]

Re: M-audio AV30 speakers recap and repair

I know this is an old Post, but I'm hoping to find some help with a Pair of AV30s.

This set suffered from 2 fried Zenner Diodes (Z1 & Z2) I also Replaced the 2 220uf 25v Caps as well as the 2 Ceramic 104 caps beside them and 2 22uf 25v Caps on the edge of the board.

The speakers are now putting out sound but it is extremely soft. Source volume is all the way up.

Even the headphone jack on the speakers is very soft.

What would you suggest I replace next? This is a Pro-Bono fix for a friend and I've already got a couple hours into it. Hoping the next part will be the last.

[ben7]

Re: M-audio AV30 speakers recap and repair

I know this is an old Post, but I'm hoping to find some help with a Pair of AV30s.

This set suffered from 2 fried Zenner Diodes (Z1 & Z2) I also Replaced the 2 220uf 25v Caps as well as the 2 Ceramic 104 caps beside them and 2 22uf 25v Caps on the edge of the board.

The speakers are now putting out sound but it is extremely soft. Source volume is all the way up.

Even the headphone jack on the speakers is very soft.

What would you suggest I replace next? This is a Pro-Bono fix for a friend and I've already got a couple hours into it. Hoping the next part will be the last.

Check R57 and R58 (those two power resistors ... 100 ohms each)
If the zeners shorted then those resistor might have blown because of that.
Also, if the zeners went open circuit, then a higher than intended voltage might have been supplied to other circuitry, and damaged something.

[jgiebler]

Re: M-audio AV30 speakers recap and repair

Check R57 and R58 (those two power resistors ... 100 ohms each)
If the zeners shorted then those resistor might have blown because of that.
Also, if the zeners went open circuit, then a higher than intended voltage might have been supplied to other circuitry, and damaged something.

Resistors Test Good... any other possible fixes for low volume? I'm concerned you are correct and voltage bled across the damaged Zenners. I don't have the time or the knowledge to go through piece by piece and determined the failed components on my own, unfortunately.