View Full Version : Amp7 problem (don't work)
Amp7 problem (don't work)
I've tested for the first time my amp7.
And don't woks.
1) the supply after the (first!) connection become +47v and -42v (and not +/-45).
2) the yellow led is always on,
but after the connection between the 'mute' and the 'agnd' it's switched off for a pair of second it seems that the ta3020 burn and switched on.
Nb. during the first test I've not shorted the mute.
3) the voltage on the speaker is 2.6v.
4) If we turn on the amp7 the ta3020 heats a little and also the four power transistor.
5) The two fuses on the supply (+/-45) are OK.
6) For the vn10 and vnn10 I'm using an handmade supply based on lm317, that I've largely tested.
7) The two off-set trimmer have no effect on the speaker voltage.
8) I've removed the ta3020 and the supply voltages on the board are OK (+/-45v)
9) I've discovered that pins 2 - 3 of Q6 (drain -> source of a power transistor) are shorted.
What's the possible problem?
It's possible that's I've received a defected ta3020?
Or, maybe, I've received a defected transistor?
Do you have some suggestion or another idea ....
Looks most like a defective power transistor. So it's consistently shorted? Measured both directions? Did you remove it and measure again?
Looks most like a defective power transistor. So it's consistently shorted? Measured both directions? Did you remove it and measure again?
Yes, measured in both direction and without ta3020.
What do you think about D13? That's another possibility.
Have you experienced something similar?
As a matter of fact I did once, but I don't exactly remember what the cause was...:(
D13 is a possibility as well, certainly. You can choose whichever is more convenient for you to take of the board. I know the fets can be a bugger to solder off....
Thanks
I'm going to try something.
But at this point I think that the problem is the transistor.
So, you completely exclude the ta3020?
Well, why did it do wrong? That's a good question! Are you 100% certain the VN10 voltage was connected the right way around? Usually the fets die when there's something wrong with the VN10 voltage. This also doesn't exclude the TA3020 from being defective....but that's hard to test as is...
Well, why did it do wrong? That's a good question! Are you 100% certain the VN10 voltage was connected the right way around? Usually the fets die when there's something wrong with the VN10 voltage. This also doesn't exclude the TA3020 from being defective....but that's hard to test as is...
I've removed D13, and as a conclusion the transistor is shorted.
I'm sure about the VN10 and my supply (2 x 10v based on lm317T) is working correctly.
I've connected the positive of the 10v supply to the 'vn10' and the negative to the VNN (Pwr N on the board).
That's correct? (I say yes)
I've also measured everything before mounting the ta3020 and the transistors.
I've also removed the transistor ...
What's your proposal ....
Change everythig ... transistor and ta3020.
or only the transistor.
I would personally order a couple of spare chips, and a few spare transistors. And I'd change both and be done with it :)
Do you have the TA3020 socketed? I suppose you do...?
I would think the fet is the only con. The TA3020 is worth another try, sure...
Very strange this, these fets usually don't just die...
Very strange this, these fets usually don't just die...
Some peeple say that an if an elecronic works for the fist two or three hour it works forever.
That's prbabaly a manufacturing defect, the probability is very low but not zero ...
That's true... and for VN10, that would probably have caused more damaged parts...
Good luck!
Haggis Basher
19-Aug-10, 20:26
Some peeple say that an if an elecronic works for the fist two or three hour it works forever.
That's prbabaly a manufacturing defect, the probability is very low but not zero ...
Dear Gigi,
There is a list somewhere on the 41Hz site, which shows the descending probability of failure to make things work: it starts with soldering defects and ends with defective components, with other nasties in between.
I have not built an Amp7, but it uses the same chip the TA3020, as the Truepath. You may find it helpful to follow Audiophool’s test system from his Maedos website, as I would think that the systems are similar. I call Audiophool's tests, the safe tests, and "my bacon has been saved" on at least two occasions, during my first two builds, using his methods.
I have managed to resurrect one partially destroyed Truepath system, and I am currently working to resurrect another Truepath system; but the bad news, is that it is slow going.
Ken.
Thanks
That's the link of the test procedure, that is useful for everybody.
http://www.maedos.com/Guideline.html#Power_Up_and_Test_Procedure
A soldering problem is always possible.
But burning a power transistor, without electrical load, after some seconds is really excessive.
Don't forgot that the this transistor are also used in the class A power amplifier with an enormous power dissipation. And they work for days and days.
Next time I'm going to do a more accurate test procedure, that's sure.
Amp7 problem (don't work)
I've tested for the first time my amp7.
And don't woks.
1) the supply after the (first!) connection become +47v and -42v (and not +/-45).
2) the yellow led is always on,
but after the connection between the 'mute' and the 'agnd' it's switched off for a pair of second it seems that the ta3020 burn and switched on.
Nb. during the first test I've not shorted the mute.
3) the voltage on the speaker is 2.6v.
4) If we turn on the amp7 the ta3020 heats a little and also the four power transistor.
5) The two fuses on the supply (+/-45) are OK.
6) For the vn10 and vnn10 I'm using an handmade supply based on lm317, that I've largely tested.
7) The two off-set trimmer have no effect on the speaker voltage.
8) I've removed the ta3020 and the supply voltages on the board are OK (+/-45v)
9) I've discovered that pins 2 - 3 of Q6 (drain -> source of a power transistor) are shorted.
What's the possible problem?
It's possible that's I've received a defected ta3020?
Or, maybe, I've received a defected transistor?
Do you have some suggestion or another idea ....
I have the same problem but the difference is:
-my TA3020 is always cold as the mosfet,
-supply +/- 55V;
-the Q6 mosfet has 8 ohm between gate drain;
- the yellow led is always on;
- the VN10 don't adsorb current;
- V10 has always 60mA;
- VPP/VNN 0.02A;
-if I put the load the current is always zero;
in the first test, I've wrongly connect both the AC supply at the two LM317 board, instead two separated AC
the TA3020 is good but the Q6 is broken ? (but I don't want influence your reply...)
ps:all the solder are good
As it appears AMP7 doesn't feature pull down resistors to make sure the fets are low when there's no voltage on the gates. This can damage the fet when the board is tested without the TA3020 installed...
A fet is by nature a latching kind of switch, so the state it was (last) in it will remain in until the voltage on the gate has changed state.
It is well possible that this has been the cause here and the TA3020 might be damaged too after installing...
I would start off with a new TA3020 and four new fets here.
I was wrong about mosfet Q6, it don't have 8ohm between gate drain; I read in all post that the mosfet are broken only when are hot, however I will buy them and TA3020,
thank for your patience Vincent.
Roberto
I've restarted with this amplifier:
I've replaced the damaged mosfet an the ta3020.
As a result the resistance between the G and S (pins 1 - 3) of D6 is about 27KOhm, this problem disappear we remove the chip.
If we measure the resistance between the pins 4 and 5 on the removed ta3020 is infinite.
The resistance between the left an right VNN is 8 Ohm and laeft and right VPP 15 Ohm
With the new ta3020 I've tested only the 5v ad described in the maedos web site:
http://www.maedos.com/Guideline.html#Test_V5
And I've not seen the step (I don't known if the truepath i really identical to the amp7).
Some idea?
Powered by vBulletin® Version 4.2.0 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.