Hartke HA3500 with partially clipped output

[Nyqusit5]

Hi,

I am working on an HA3500 with a clean output up to more or less 50W output. After that, the top side of the signal clips (under load). I have checked the output transistors, the ballast resistors and even tried to change the driver but no change. the voltage rails are steady, even under clipping condition. Without a load, the output is clean all the way. I should mention that the input signal is a clean sine wave. here is a pic of the output on a scope:


The HA3500 is available in many posts on this forum.

Any ideas?

[J M Fahey]

Hi,

I am working on an HA3500 with a clean output up to more or less 50W output. After that, the top side of the signal clips (under load). I have checked the output transistors, the ballast resistors and even tried to change the driver but no change. the voltage rails are steady, even under clipping condition. Without a load, the output is clean all the way. I should mention that the input signal is a clean sine wave. here is a pic of the output on a scope:
[ATTACH=CONFIG]46786[/ATTACH]

From dim memory, 3500 has 2 or 3 pairs output transiustors.
If only 1 out of 3 is working on the V+ side, it will work fine up to a point.
Clip your *floating* scope across each of the 3 emitter resistors to see voltage across each, which will also show current through each ... if they are not open that is.
We want to check that all 3 transistors equally share load.

You might also have an open collector top driver, so it just works as a bypass without adding its own current gain.
In that case power transistors will be underdriven and unable to drive a speaker load, no problem if unloaded.

You may also have a defective short protection circuit, which triggers too soon.

And on and on and on ... so check these 3 hypotesis first.

The HA3500 is available in many posts on this forum.

Close but no cigar, *you* post one and we all follow the same.

[Nyqusit5]

Thank you for your input JM.

Th Ha3500 has 4 pairs. I have tested and replaced all of the 2SC5200 (or 2SD2155) and Q309 (voltages on its pin are ok) with no change. While in the process, I tested the 0.47Ohm ballast resistors and all of them measure good. All of these transistors voltages are coherent at idle or in action.

The schematic is here:
http://music-electronics-forum.com/t45625/

[Nyqusit5]

After more testing, I can confir that all power transistors and their associated ballast resistor work fine: successively taking them out of the circuit ends up in the reduction of the positive (portion of the sine wave before clipping.
Again, Q309 has been replaced...
Still looking but can't figure it out!

[g1]

How about Q311 and associated circuit?

[Nyqusit5]

Bingo!
I just received some 2sc3198 this morning and although the original Q311 measured ok in situ, it was faulty.
Thank you all for your help!