miller cap

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I made the changes the amp works ok , idle current on the output is around 19 mA + 19 mA which is just about fine I guess. The bias adjustment however doesn't seem to work at all earlier it used to have some effect on the idle current.

My transistors are currently directly mounted on the heatsink without any insulation and I accidentaly touched the heast sink ...disaster both fuses blew and before they could blow they toasted the bridge rectifier! Now why did that happen ?
 
Minor detail - 6.8k input resistance plus 2,2 uF coupling cap gives a low pass of about 10 hertz - This is okay but barelly acceptable for high quality amplifier - I would either increase resistance or raise capacitance or both. I Usually try for 2 HZ (Tenth of 20 HZ). Also - larger resistance allows less expensive film cap for coupling. YMMV good luck.
 
I found out some more points to correct, I think the heatsinks need to be earthed to the common, I was going through rod elliots troubleshooting guide and he's saying the heatsinks unless grounded act like antennas and may lead to hf oscillations. The PNP and NPN side are mounted on separate heatsinks I still need to isolate them. I also think that the BD139 vbe multiplier is too sensitive . Since my output topology is a CFP my VBe multiplier is monitoring the drivers, the drivers are barely warm but the bias still fluctuates a lot blowing some air over the drivers seems to change the bias :).
BTW for two transistors including the drivers my bias voltage should be around 1.4 volts not 3.0 right ?
 
One of the problems I figured out was there was too much bias if the servo voltage divider ratio was anything greater than one with MJE 340/350 drivers, the drop across the collector resistors of the drivers was around 6V ! I changed the bias servo resistors to 1.2k / 1k pot + 560 ohms, now at just above minimum bias I get around and 0.8 volts across the collector resistor of drivers and 8mV drop across the output resistors which translates to around 32mA + 32mA. This was after 20 mins of operation.

I made some more changes also, I change the output resistors to 0.25 ohms and all the current sharing ones from 0.1 to 0.25.

My earlier burn out was a result of not isolating the output transistors from heatsinks ( although each side was sitting on an isolated heatsink) and not grouding them. When I touched the heatsink to check the temperature I must have caused an oscillation of some sort leading to too much current in the ouput transistors.
Now all the outputs are on silpads and the heatsinks are grounded.

Haven't tested with an audio source till now :D just taking voltage readings to ensure it runs for a reasonable stretch of time without going up in smoke.
 
I am not too sure about the VI limiter that I have on my amp, I saw the scheme in Leach amps, although the working principle is the same they seem to have a voltage divider. My VI limiter was straight out of Douglas Self's book but I am not too sure of the values I am using. My output resistors are 0.25 ohms, my working voltages +-43, two pairs of output devices. I will be running the amp with a 6ohm nominal speaker.
I have a choice between NJW1302/3281 and MJL 21193/4 what should be my SOA limiting current which the VI limiter should monitor 0.25 ohm*5 amp ~ 1.25 ... I should reduce this through a voltage divider to make this valure around 0.5 ?
 
Last edited:
I got the VI limiter calculations from an old marshall leach article " Build a low TIM amplifier". I have changed the schematic a bit and designed a new pcb. Need some feedback on the new pcb, The old one was fine but was just designed to test the concept this one I wanted to refine a bit.
 

Attachments

  • LME49810amp_rev2.0.pdf
    57.7 KB · Views: 48
  • LME49810ampsch.pdf
    41.7 KB · Views: 40
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.