Holton's AV400

[bigpanda]

I've finished wiring up one channel of AV400 and done a pre-flight test. I had test pt voltages taken and come in safe range as require by HOLTON expect R12, R13 I had 400 instead of 500mV. I did check my components and wiring. Any idea what might had gone crazy? If it is way off, I would go for mis-wiring but.....

Tks

BTW- I do my own pcb but I check & recheck the drawing before I get the parts. There is thin chance of a mis-lay circuit.

[naula]

Hello bigpanda!

I think 400 mV is acceptable. It was right direction? Plus on the emitter of the transistors?

Did you have 70-0-70 Vdc PSU in your "pre-flight-test"? If the voltage was smaller it causes the smaller values.

- naula

[bigpanda]

Hi

I pulled the amp. out again this morning and connected the power(still in the pre-freight stage). This time I got 380mV across R12 and 420mV across R13. Should I change Q7 & Q9 with a matched pair?

I also find that Q6, Q8 to be quite hot. I have to pulled my fingers off when touching them for 10 sec. Could that be a potential problem? The 2 MJE350 I had was totally plastic case with no metal on the back.

I used 4 BC546 with close hfe and got a 50mV at the output.

naula,
I had +-70.3V after connected to the amp.. That shouldn't be the cause of 400 instead of 500, right?.


Could somebody please kindly give me some advice.


Tks

[thylantyr]

I'm reading the AV400 document. It says R12 and R13 ~500mV.
~ = approximate. It's probably close enough to operate well.
Have you measured the resistor values with a meter ? Maybe
match resistors first.

also find that Q6, Q8 to be quite hot.
Attach a small heatsink

The document specifies +/-70V rails so I wouln't worry about
that.

If pre-flight is close in parameters to the default numbers
then proceed. Worse case, smoke test and you start again.
Part of the learning curve.

[Kilentra]

Hi Peter,

I'm sorry I haven't been able to respond to you, because I was on vacation. The 400mV readings are just fine, mine were a little low too. It is perfectly acceptable for the small signal transistors to be hot to the touch, but if they are painfully/burning hot then something may be wrong. Does the amp work?

[bigpanda]

Is that you Mike,


Sorry to distract you from your vacation. There is no quiet and isolated place on this planet anymore. Tks to technology....

I've just finished bolting the O/P Fets to the angle blacket with the Vbe mutiplier mounted on one of them. Much concentration and eye sight spent. I was about to throw the switch and I just check here if anybody said 'DON'T....'.

I'll switch the power about an hour from now. If I am lucky enough not to get killed, I'll post the outcome later in the afternoon. Wish me good luck, I need it.

[Kilentra]

Yep, it's me. And I certainly wasn't distracted, given that I was away from computers and all for two weeks...

Good luck, and please continue to be paranoid. In the process of getting my amps to work, though I didn't have all the information you did, I had one resistor catch fire, one transistor explode in my face, and at least three thermal runaway failures.

Basically, if you don't know what you're doing it's dangerous, expensive and time consuming. But since you do know what you're doing, if you're careful then nothing will go wrong. Have fun.

[bigpanda]

OK, I had it connected and did what HOLTON said and measured the source resistor voltage. (I was wondering what is going on with the math of this part : V=IR if I had a voltage of 18mV across a 0.22ohm resistor, I should have a currnet of .018/.22 ~= 80mA? but we should had a 100mA????)

I had the N-channels voltage ranging from 15 --> 18mV and the P-channel ranging from 16 --> 17.8mV. I think they should be OK, right? When I choose the N-channels, I choose from a Vgs range of within .02V but I got a 3mV variation when put toghther. Just wondering why it is so.

I am wondering how hot the amp. can get (or how long it takes to get hot) with no input signal and no output load. (Or shouldn't I even try to do this?)

[Kilentra]

That's a close enough setting, though the amp might sound slightly better with more bias (I tried a few between 60 to 200mA per device and I settled around 100mA). The Mosfets Vgs is variable depending on current and temperature so your initial measurements are never quite perfect. Don't worry because your current sharing is pretty close. I would start getting worried if one conducts 50% more than the others.

The important thing is to verify thermal stability by playing music and then seeing what happened to the bias. I have pretty large heatsinks and the Mosfets are warm (40 degrees or so) at idle. It doesn't matter if there is a load.

[bigpanda]

I tried inserting the speaker/DC protection circuit (I've raised some question on this in another thread) between the circuit and the speakers. I've tried the protection circuit by taking a longer delay (about 4-5 sec) and applied a 2.4V DC to the output. It works fine on its own. Now I inserted it in between the AV400 and the speaker (1 only for I have only one channel built) : I join the GRD of the protection circuit to the GRD of the amp. and applied a 12V DC with an adaptor (which I've done in testing the protection circuit). The contacts do not close this time. Even if its does, it closes with a much longer delay. Moreover, I notice the LED on the relay doesn't glow as bright as before. Sometimes the contacts trip after close for a while. All sorts of funny things happen.

I tried pulling out wires 1 by 1 from the amp and power up the protection. It seems that it is behaving with no predictable pattern.

After trying for almost 2 hrs., I really get frustrated and decide to run the amp. without the protection. I hook up a moving pointer type of multimeter to the O/P, seeing if it swings while switching on the amp. No noticeable movement of the pointer so I decide to take my chance. I skip the protection and hook my speaker directly to the amp. Feeding the analog output of my computer CD ROM to the input of the amp. and play a CD, I got it working!

I've been playing music continuously for 2 hrs already but the temp. seems to be stable. (I check the temp. every 15min.) The heat sink is a bit uncomfortably hot but did not get hotter. Maybe the heat stays around the FETs and the angle bracket. The source resistors are OK., Not much heating in the PSU part except the bleeding resistors. I've also add 2 1"x1"x1" U-heat sink to Q6, (not easy to do so) and the heating problem of these two have gone.

How does it sound? Of course better than the one I had before Hey I have to tell myself it "must" sound better after all that effect I have paid (just kidding ). But there is one thing I am quite surprise : it seems that I have to turn the volume of the CD ROM quite high to get the same loudness as before. I was expecting a very low input should give a lound output but its wasn't.

Tomorrow I would burn it in for longer period (its almost 1am. in the morning here) and see if it can stand it.