Panson,
If I use this circuit to switch the lme49811 to mute, will that offer full protection? I'm thinking that it won't. Your thoughts.
Ken
Without driver?Hi Panson!
Yes, I'm using the LME49830 with IRFP240/9240.
You don't need the 180R if driver is not used. Try to run the amp with low bias current (50mA) at the beginning for a while. If the amp is stable, you can increase the bias gradually. Note the current measured in this way includes the 49830 operation current.
Make sure you short the input when doing bias and output DC measurement.
You can verify if the amp is operating by using your PC sound card to generate a sine tone to drive the input and measure the output with a meter in AC Volt. Or, you can with the input open and touch the +input with you finger. The meter in AC Volt should show you some readouts. The amp amplifies a hum signal.
What is the value of the MOSFET gate resistors? Since you don't have a scope, you should be more conservative. Try to use 30 pF and put ~100R for gate resistance. The wires from PSU to the amp board should not be too long too.
These are fine. The value won't cause any problem. Just to make sure no soldering issue.
Pls assemble it and use it.I bought this exact kit from eBay some time ago! It´s the exact same circuit but I forgot it and never assembled it!!! Should've known better...
You can hear it since it is usually in kHz or Mhz. It could damage a speaker driver if the level is high. Can you make the envelop detector as mentioned earlier to check the output of oscillation?Can an oscilating amplifier, not heardable on speakers, cause them to fail or the only thing I should be concern about is DC at the output?
Thank you Panson! Your help is very important!
Regards,
Paulo.
LME49830 does not have a pin for a clip LED. The LED is only for 49810.
Dear Panson,
I guess it is due a PCB design error from mine. Most likely the sink and source PCB lines to close to each other.
Already advanced happy new year!
With kind regards,
Bas
Hi Bas,
I tested the amp with 1k and 10k square wave input. The load was 8, 4 ohm. No oscillation found.
Here is waveform of 10 k input. Scope is in full BW, 500 MHz. The average gives us background noise removed for better view.
Attachments
Yes, I'm using the LME49830 with IRFP240/9240.
Without driver?/QUOTE]
Yes, without driver. Do I need a driver?
What is the value of the MOSFET gate resistors?/QUOTE]
I'm using 100R. Is it OK?
Try to use 30 pF/QUOTE]
But that value will low the slew rate...
These are fine. The value won't cause any problem./QUOTE]
Great!
Pls assemble it and use it./QUOTE]
You're absolutely right. Do I need a dedicated transformer or can I use the amp's transformer with a voltage divider?
Can you make the envelop detector as mentioned earlier to check the output of oscillation?/QUOTE]
Thanks, I will do it!
LME49830 does not have a pin for a clip LED. The LED is only for 49810./QUOTE]
Then I don´t need it... In this case R3, R8 and D11 are needed?
Thank you for all your help!
No driver needed. 100R gate resistor is okay.
Lower slew rate ... but you don't have a scope to check oscillation. I suggest you use larger cap to avoid problem. When you can get a scope, you can than try lower cap. Play safe!
Sure you need R3, R8 and D11 which are needed for the mute pin.
Use a separate transformer for the speaker protector if you main transfomer's voltage is too high.
Lower slew rate ... but you don't have a scope to check oscillation. I suggest you use larger cap to avoid problem. When you can get a scope, you can than try lower cap. Play safe!
Sure you need R3, R8 and D11 which are needed for the mute pin.
Use a separate transformer for the speaker protector if you main transfomer's voltage is too high.
lme49811 circuit
Hello,
http://www.unisonus.com/pdf/unisonus_lme49811_amp.pdf
Could someone please explain the purpose of D1,D2,D3,R9,R10, C6,C7.In my point of view i would omit these parts...because i see no benefit from them.Or maybe my knowledge to little here...
Hello,
http://www.unisonus.com/pdf/unisonus_lme49811_amp.pdf
Could someone please explain the purpose of D1,D2,D3,R9,R10, C6,C7.In my point of view i would omit these parts...because i see no benefit from them.Or maybe my knowledge to little here...
D2 and D3 are the so-called catching doides which protect T1 and T2 from voltage transients induced at the output by inductive loads.
Dear All,
I am glad those diodes come into question here... I was always wondering, do those Diodes affect the bass quality? My thinking was this. If a cone resonates from it's own mass it sends EMK back to the amplifier. The amplifier sees this small induced voltage as an error and the feedback loop send a opposite signal in opposite phase to correct this error. When catching diodes are from output to voltage rail, this EMK will send to the power rails and don't get corrected by the feedback loop. Any thoughts?
Ps Panson, Thanks for the square wave images. What can I say.. looks like a charm!
Pss. The Unison design with the STD03's. the 2.2K resistor between the Source and Sink pins seems to high to give the 2.5mA current for the diodes inside the STD03's.
With kind regards,
Bas
With kind regards,
Bas
Hi Bas,
I think your mentioned issue only occurs at nearly output clipping. If the output positive peak plus back EMF is less than the rail. The upper diode will not conduct. Same argument for the lower diode. The amp should then operate as without the diodes.
Panson
Lme49811 + std03
Hello All,
The 2.2K resistor is indeed way to high. Couldn't even get the voltage across the emitter resistor lower then 40mV! (With the potmeter set to 200r it was 200mV, explains why the darlingtons were getting very very hot)
The LME49811 gives a minimum output current of 7mA. It's typical 9mA. What we have now is a current divider with the 2.2K resister and the 47r + 47r + 200r potmeter. So with the potmeter set to 100r we need to replace the 2.2k resister with a +- 500ohm resistor. This worked very good. Very easy to get the bias current right now.
I also needed to replace the 10pF capacitor as the amp was oscillating at 4Ghz.
I will try to get some THD plots next month when I'm going to the audio lab at my uni.
Hello All,
The 2.2K resistor is indeed way to high. Couldn't even get the voltage across the emitter resistor lower then 40mV! (With the potmeter set to 200r it was 200mV, explains why the darlingtons were getting very very hot)
The LME49811 gives a minimum output current of 7mA. It's typical 9mA. What we have now is a current divider with the 2.2K resister and the 47r + 47r + 200r potmeter. So with the potmeter set to 100r we need to replace the 2.2k resister with a +- 500ohm resistor. This worked very good. Very easy to get the bias current right now.
I also needed to replace the 10pF capacitor as the amp was oscillating at 4Ghz.
I will try to get some THD plots next month when I'm going to the audio lab at my uni.
output pairs vs load condition
Some test results for LME49811 with one pair/two pairs ThermalTrak 4281/4302 vs load condition (4 Ohm, 8 Ohm). Supply rails is +/- 49 V. Measurement bandwidth is 80 kHz for the left and middle graph. Bias current per device is 50 mA. Output power is 18 W/8 Ohm, 36 W/4 Ohm.
4281 and 4302 are low beta droop devices. For one pair, the 10 kHz distortion is doubled from 0.003 % to 0.006 % (8 Ohm -> 4 Ohm). When there are two pairs, the distortion is reduced for the two load conditions. There is almost no addition degradation for 8 Ohm to 4 Ohm at 10 kHz. At 20 kHz, the THD+N are 0.002 % and 0.003 % for 8 Ohm and 4 OHm, respectively.
The measurement bandwidth is 22 k for the right graph.
Some test results for LME49811 with one pair/two pairs ThermalTrak 4281/4302 vs load condition (4 Ohm, 8 Ohm). Supply rails is +/- 49 V. Measurement bandwidth is 80 kHz for the left and middle graph. Bias current per device is 50 mA. Output power is 18 W/8 Ohm, 36 W/4 Ohm.
4281 and 4302 are low beta droop devices. For one pair, the 10 kHz distortion is doubled from 0.003 % to 0.006 % (8 Ohm -> 4 Ohm). When there are two pairs, the distortion is reduced for the two load conditions. There is almost no addition degradation for 8 Ohm to 4 Ohm at 10 kHz. At 20 kHz, the THD+N are 0.002 % and 0.003 % for 8 Ohm and 4 OHm, respectively.
The measurement bandwidth is 22 k for the right graph.
Attachments
Speed-up capacitor
The effect of speed-up capacitor is evaluated. A 0.1 uF is in parallel with the driver emitter resistor to speed-up output device switching speed.
The amp under test is still LME49811 with two-pair 4281/4302. There is no effect for 50 W into 8 Ohm load (lower two curves). The capacitor does reduce the high-freq THD+N considerably for 100 W into 4 Ohm load. We need the output stage working harder to observe the effect clearly.
The effect of speed-up capacitor is evaluated. A 0.1 uF is in parallel with the driver emitter resistor to speed-up output device switching speed.
The amp under test is still LME49811 with two-pair 4281/4302. There is no effect for 50 W into 8 Ohm load (lower two curves). The capacitor does reduce the high-freq THD+N considerably for 100 W into 4 Ohm load. We need the output stage working harder to observe the effect clearly.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.