panson_hk, I am a bit curious, what did you set the bias to? and what is the output noise level?
I have an old HP334A and a heathkit IG18. The best I can get with those 2 pieces of equipment is around .016%. When I tested the boards I had I was able to measure around .018 - .02% which is close to my limit.
Your reasult seems similar to post #23 in the following forum but a little worse. These $50 boards perhaps with a better layout could do better.
http://www.diyaudio.com/forums/chip-amps/102973-lme49810-new-cousin-lm4702-3.html
I have an old HP334A and a heathkit IG18. The best I can get with those 2 pieces of equipment is around .016%. When I tested the boards I had I was able to measure around .018 - .02% which is close to my limit.
Your reasult seems similar to post #23 in the following forum but a little worse. These $50 boards perhaps with a better layout could do better.
http://www.diyaudio.com/forums/chip-amps/102973-lme49810-new-cousin-lm4702-3.html
I adjusted the bias level from zero (significant crossover spikes observed) while monitoring the THD value and distortion residual. I usually use an output power of about 25 W/8 Ohms in finding the optimum bias level. A sweep for THD vs power/freq will then be executed. Based on the results, the bias level might be further fine tuned.
The THD of this inexpensive board is worst compared to all my LME498xx amps. I have checked the driver chip (49810) alone. It performs as spec. I suspect there are layout issues. I will also check whether the transistors is the source of problem.
The THD of this inexpensive board is worst compared to all my LME498xx amps. I have checked the driver chip (49810) alone. It performs as spec. I suspect there are layout issues. I will also check whether the transistors is the source of problem.
Hello.
I have some questions regarding this design. I bought the PCB for this amp without components.
1.
I have PSU that gives me +- 81V. If I count in the losses at max operation than I get approx. +-75V. Will the amp operate at this voltage?
2.
Can I use 2SC5200/2SA1943 BJTs at this voltage or is better to use some other devices (wich one)? I will put in 5 pairs and drive the 8 Ohm speakers, so I'll get approx. 350W.
3.
Can I put 2 mono channels in bridge and drive 4 Ohm load with this voltage?
4.
I looked at datasheet for uPC1237. The chip is rated at maximum operating voltage of 60V. Well if I'm going to use my PSU I must change the values of some resistors. That is R8, R4 and resistor that goes on the relay (560 Ohm) if we look at datasheet. If I understand it correctly I must change the R8 to 27k, R4 to 47k, and the value of relay resistor depends on relay supply voltage and its coil resistance (rated current of relay).
Do I have to change anything alse for +-75V operation except driver voltage (for LME49810)?
I have some questions regarding this design. I bought the PCB for this amp without components.
1.
I have PSU that gives me +- 81V. If I count in the losses at max operation than I get approx. +-75V. Will the amp operate at this voltage?
2.
Can I use 2SC5200/2SA1943 BJTs at this voltage or is better to use some other devices (wich one)? I will put in 5 pairs and drive the 8 Ohm speakers, so I'll get approx. 350W.
3.
Can I put 2 mono channels in bridge and drive 4 Ohm load with this voltage?
4.
I looked at datasheet for uPC1237. The chip is rated at maximum operating voltage of 60V. Well if I'm going to use my PSU I must change the values of some resistors. That is R8, R4 and resistor that goes on the relay (560 Ohm) if we look at datasheet. If I understand it correctly I must change the R8 to 27k, R4 to 47k, and the value of relay resistor depends on relay supply voltage and its coil resistance (rated current of relay).
Do I have to change anything alse for +-75V operation except driver voltage (for LME49810)?
Hi!
I have some questions regarding this board.
I've already bought, 2 x 4.7kohm 5 watt resistors, for use in parallel, at r26.
And I bought the oscillation fix, components.
But what I don't get is, what value I should set the r30 at, if my supply voltage is 60volts +/-?
How and where did you guys measure the bias?
What do you use the AC-named input for? Is it for Speaker output, and is being used for detecting Clipping or DC?
Or is it for power supply positive input, so it can measures how close it is to clipping?
Is there any other mods that I have missed, in order to make these boards work properly with 60v+/- on on both power inputs. (the chip and the normal power input)?
I have some questions regarding this board.
I've already bought, 2 x 4.7kohm 5 watt resistors, for use in parallel, at r26.
And I bought the oscillation fix, components.
But what I don't get is, what value I should set the r30 at, if my supply voltage is 60volts +/-?
How and where did you guys measure the bias?
What do you use the AC-named input for? Is it for Speaker output, and is being used for detecting Clipping or DC?
Or is it for power supply positive input, so it can measures how close it is to clipping?
Is there any other mods that I have missed, in order to make these boards work properly with 60v+/- on on both power inputs. (the chip and the normal power input)?
1 - I left R30 as is
2 - measure bias across 1-0.22 ohm resistor or 2- 0.22ohm resistor
I set bias to an average of 30ma (.030 amps) ie 6.6mv across ONE resistor or 13.3mv across 2 resistors. that are beside each other. Use the E labeled pin and the speaker output connector for one resistor measurement. Use the E labeled pin on opposite transistors to measure across 2 resistors. Be carefull with your probes. If you have some distortion measuring equipment you can set the bias to an optimal value. At +/- 60 volts and 30 ma bias you amp will dissipate around 18 watts per channel. It will get warm.
3 - The ac input is for the upc1237 chip to detect that the power is on. I used an unused winding from the power transformer secondary.
4 - depending on the relay used on the board you may have to change the value of the 4.7k resistors. When testing the amplifier put a meter across the 560ohm R26 resistor. wait until you hear the relay click, about 4-6 seconds.
Now calculate the voltage across the relay (Vcc+) - (V across R26).
Look at the relay coil voltage rating. Is it 12, 24, 36, 48 etc. Write this down somewhere.
If the relay coil is rated at 24 vdc, then you want 36 volts across R26 and 24V across the relay coil, for a total of 60Volts, your Vcc value, a little less across the relay is ok. ie 20V
Since you know the voltage across R26, you can now calculate the resistance of the relay coil. Current through the relay is the same as R26 so I=Volts across R26 divided by 560. Take the voltage across the relay and divide by the current and you will get the relay resistance. R relay.
Now you can calculate the ideal relay current and you can now calculate the ideal value for R26.
If you haven't installed the boards you can measure the reisistance of the relay and this will save you a few calculations. I had installed and removed the boards so many times I did my mods and measurements from the top with the boards mounted on heat sinks as the screws for the power transistors were getting loose and I didn't want to drill and tap 13 more holes.
Going back to the bias measurements the boards must be on heat sinks. Adjust the bias after the amp warms up a bit. If the bias current keeps climbing and climbing the bias transistor is not mounted properly. Given the distance between the bias transistor and the power transistors there will be a delay as the transistors warm up and the heat get disipated through the heatsink to the bias transistor.
for a picture of my nearly completed project try this link post #210, its missing a case
http://www.diyaudio.com/forums/clubs-events/199025-diy-ottawa-winter-2012-a-21.html
2 - measure bias across 1-0.22 ohm resistor or 2- 0.22ohm resistor
I set bias to an average of 30ma (.030 amps) ie 6.6mv across ONE resistor or 13.3mv across 2 resistors. that are beside each other. Use the E labeled pin and the speaker output connector for one resistor measurement. Use the E labeled pin on opposite transistors to measure across 2 resistors. Be carefull with your probes. If you have some distortion measuring equipment you can set the bias to an optimal value. At +/- 60 volts and 30 ma bias you amp will dissipate around 18 watts per channel. It will get warm.
3 - The ac input is for the upc1237 chip to detect that the power is on. I used an unused winding from the power transformer secondary.
4 - depending on the relay used on the board you may have to change the value of the 4.7k resistors. When testing the amplifier put a meter across the 560ohm R26 resistor. wait until you hear the relay click, about 4-6 seconds.
Now calculate the voltage across the relay (Vcc+) - (V across R26).
Look at the relay coil voltage rating. Is it 12, 24, 36, 48 etc. Write this down somewhere.
If the relay coil is rated at 24 vdc, then you want 36 volts across R26 and 24V across the relay coil, for a total of 60Volts, your Vcc value, a little less across the relay is ok. ie 20V
Since you know the voltage across R26, you can now calculate the resistance of the relay coil. Current through the relay is the same as R26 so I=Volts across R26 divided by 560. Take the voltage across the relay and divide by the current and you will get the relay resistance. R relay.
Now you can calculate the ideal relay current and you can now calculate the ideal value for R26.
If you haven't installed the boards you can measure the reisistance of the relay and this will save you a few calculations. I had installed and removed the boards so many times I did my mods and measurements from the top with the boards mounted on heat sinks as the screws for the power transistors were getting loose and I didn't want to drill and tap 13 more holes.
Going back to the bias measurements the boards must be on heat sinks. Adjust the bias after the amp warms up a bit. If the bias current keeps climbing and climbing the bias transistor is not mounted properly. Given the distance between the bias transistor and the power transistors there will be a delay as the transistors warm up and the heat get disipated through the heatsink to the bias transistor.
for a picture of my nearly completed project try this link post #210, its missing a case
http://www.diyaudio.com/forums/clubs-events/199025-diy-ottawa-winter-2012-a-21.html
Thank you multisync, for your detailed reply.
I ended up using Cristi's smps2000R 45volt supply instead of my original idea of using a linear power-supply.
I kept the bias at the same bias-voltage drop as you described, but maybe it wouldn't hurt turning it up a bit more. Maybe there won't be any point in it, since the "transistors" are already open, and it would just defeat the purpose of a class B (A/B) setup?
As an A/C source for detection of power on/off, I used a positive DC source instead, since the circuit on the board has a rectifier on the as input anyway.
I also modified the board with the 75 Pf and the resistor mentioned in this thread.
Do you guys have any suggestions to improvements on this setup?
I don't need a lot of power in 8 or 4 ohms loads, but it should be able to handle loads down to two ohms, hence the lower input voltage.
I ended up using Cristi's smps2000R 45volt supply instead of my original idea of using a linear power-supply.
I kept the bias at the same bias-voltage drop as you described, but maybe it wouldn't hurt turning it up a bit more. Maybe there won't be any point in it, since the "transistors" are already open, and it would just defeat the purpose of a class B (A/B) setup?
As an A/C source for detection of power on/off, I used a positive DC source instead, since the circuit on the board has a rectifier on the as input anyway.
I also modified the board with the 75 Pf and the resistor mentioned in this thread.
Do you guys have any suggestions to improvements on this setup?
I don't need a lot of power in 8 or 4 ohms loads, but it should be able to handle loads down to two ohms, hence the lower input voltage.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.