Re: Re-adjust the Bias of Power Amp
Hi everyone
I bought a China made solid state power amp which rated 90w RMS output per channel and have been using it for 2 weeks. Today, I use a multi-tester to measure the quiescent current at the emitter resistor (0.22 Ohm) and the voltage reading is 0.5mv which I think the current is too low (0.5mv*0.22Ohm = 2.27ma). Therefore I have tried to increase the voltage across the emitter resistor to 22mv which is approximate 100ma. However after every 15 minutes run in, the voltage keep dropping to 20.5mv. Would anyone please tell me what caused the quiescent current unstable and what action can be taken to fix this issue?
Hi everyone
I bought a China made solid state power amp which rated 90w RMS output per channel and have been using it for 2 weeks. Today, I use a multi-tester to measure the quiescent current at the emitter resistor (0.22 Ohm) and the voltage reading is 0.5mv which I think the current is too low (0.5mv*0.22Ohm = 2.27ma). Therefore I have tried to increase the voltage across the emitter resistor to 22mv which is approximate 100ma. However after every 15 minutes run in, the voltage keep dropping to 20.5mv. Would anyone please tell me what caused the quiescent current unstable and what action can be taken to fix this issue?
Attachments
What is the correct quiescent current for that type of output stage? It depends on the circuit used. 2.2mA sounds a bit low, but for a CFP output 10mA could be correct. A Darlington output might need 100-200mA. The reduction you are now seeing may be due to thermal overcompensation.
Hi DF96
Thank you for your inputs.
I do not know the part number of the output transistors but it has 2 pairs per each channel. To my observation, the transformer and the heat sinks are big enough to run the output at 100ma (22mv across each 0.22 Ohm emitter resistor).
2pair output stage with each output device passing 100mA is a lot of heat to dissipate.
90W into 8r0 indicates supply rails @ ~+-48Vdc.
Each of those heatsinks will require to dissipate ~20W, if you stay at 22mVre.
Let me guess the Rth s-a is ~1C/W @ 70Cdegrees Delta T s-a
20W * 1 * 1.5 (de-rating factor for 20W), gives Delta T ~30C.
If ambient is 28degC and Delta T is 30Cdegrees then the devices will be running with a Tc ~ 62degC. That is quite hot for quiescent temperature.
90W into 8r0 indicates supply rails @ ~+-48Vdc.
Each of those heatsinks will require to dissipate ~20W, if you stay at 22mVre.
Let me guess the Rth s-a is ~1C/W @ 70Cdegrees Delta T s-a
20W * 1 * 1.5 (de-rating factor for 20W), gives Delta T ~30C.
If ambient is 28degC and Delta T is 30Cdegrees then the devices will be running with a Tc ~ 62degC. That is quite hot for quiescent temperature.
The 'instability' in bias you see is not abnormal in my experience.
It takes a while to heat up your heatsinks from cold,
and the thermal compensation, when mounted on the main heatsink,
starts compensating after 15 mins, when the heatsinks have warmed up.
This is most probably the reason that some manufacturers (for example AKSA)
mount the vbe-multiplier to the back of one of the output-transistors,
to make it react faster.
Best regards,
Klaas
It takes a while to heat up your heatsinks from cold,
and the thermal compensation, when mounted on the main heatsink,
starts compensating after 15 mins, when the heatsinks have warmed up.
This is most probably the reason that some manufacturers (for example AKSA)
mount the vbe-multiplier to the back of one of the output-transistors,
to make it react faster.
Best regards,
Klaas
Hi everyone
Thank you for all your inputs. The un stable issue has improved after I connected the input to the ground of the chassis. I set the bias current @ 100ma per each transistor and I had listened to the amp for several hours last night. The amp has remarkable sonic improvement. The heat sinks are capable to dissipate the heat from the transistors even after few hours operation. The transformer is a liitle too hot but within acceptable margin.
Thank you for all your inputs. The un stable issue has improved after I connected the input to the ground of the chassis. I set the bias current @ 100ma per each transistor and I had listened to the amp for several hours last night. The amp has remarkable sonic improvement. The heat sinks are capable to dissipate the heat from the transistors even after few hours operation. The transformer is a liitle too hot but within acceptable margin.
..... The picture of the amplifier you have posted is the same with alesis RA 500...
to my understanding a bad designed machine
----heat sinks are not capable of handling the power if the request is 100%
----seems that your model is also degraded by having only 4 outputs per Chanel ( ra 500 is six )
----design wise the piggy boards often have soldering issues
----quality of make is low ,...
----expect problems in the secondary power supply
---- to protection circuits
-----and blown outputs if driven hard
that is the picture of RA 500 other than that beefing up the idle on such a machine will eventually end up to stability issues unless machine is used for hifi service and not PA abuse ...
Eventhough service manual is very detailed there is no notes for any adjustments ...schematic is available if you think that you can use it
kind regards
sakis
to my understanding a bad designed machine
----heat sinks are not capable of handling the power if the request is 100%
----seems that your model is also degraded by having only 4 outputs per Chanel ( ra 500 is six )
----design wise the piggy boards often have soldering issues
----quality of make is low ,...
----expect problems in the secondary power supply
---- to protection circuits
-----and blown outputs if driven hard
that is the picture of RA 500 other than that beefing up the idle on such a machine will eventually end up to stability issues unless machine is used for hifi service and not PA abuse ...
Eventhough service manual is very detailed there is no notes for any adjustments ...schematic is available if you think that you can use it
kind regards
sakis
On rare occasions I will predict the future for people. They usually don't believe me, though my success rate is very good.
Increasing the bias will eventually destroy your amplifier.
This was not a wise path to follow, as no matter what the quality level, the manufacturer probably knew what was stable and safe. If the transformer is hot, the internal temperature is probably way higher than it should be. The insulation will break down and fail. Though the outputs may be OK in casual testing, there's a good chance that you can now send the unit into thermal runaway or find some other way to exceed the SOA. Any capacitors that are now running with higher ripple current, or living in a warmer environment will fail faster.
Of course I am occasionally wrong. 😉
Increasing the bias will eventually destroy your amplifier.
This was not a wise path to follow, as no matter what the quality level, the manufacturer probably knew what was stable and safe. If the transformer is hot, the internal temperature is probably way higher than it should be. The insulation will break down and fail. Though the outputs may be OK in casual testing, there's a good chance that you can now send the unit into thermal runaway or find some other way to exceed the SOA. Any capacitors that are now running with higher ripple current, or living in a warmer environment will fail faster.
Of course I am occasionally wrong. 😉
hi sakis
Thank you for sending the information about RA500 amp.
The rear panels of both the amps are same design but do not know if they are both using the same circuit design. Would you please post the schematic of the RA500 here for me to check? Thanks
Some more information:
I am currently using this power amp in my home stereo system to drive a 8" Fostex full range speaker with open baffle design and it sounds better than before after the modification of increasing the bias current.
The heat sinks are working very well, at 25*c home temperature they are just a little warm, I have no doubt that they can capable to handle the heat generated from the output transistors.
The transformer (1 transformer for two channel) is a little bit hot after few hours operation but as I said it is still within acceptable margin. My Pioneer Class A power amplifier M22 the power transformers (one for each channel total 2) are even hotter than this one.
As to the sonic performance between this amp and the M22. This amp can beat the M22 in general except the bass frequency the M22 is thicker.
e mail me for the schematic ...
i can only think of your results as wrong and one may say that "i like the RA500" better than the Pioneer but the truth is that this is wrong .
if you like it presumably is fine for me .
if you keep this amp for HIFI use .... it will remain safe i think
kind regards
sakis
i can only think of your results as wrong and one may say that "i like the RA500" better than the Pioneer but the truth is that this is wrong .
if you like it presumably is fine for me .
if you keep this amp for HIFI use .... it will remain safe i think
kind regards
sakis
Hi there
I set the bias current of Channel A of the amp @ 22mv (100ma) and trying to set Channel 2 @ the same current however the maximum I can get is 14mv (63.63ma) with original trim pot 330 Ohm on the pcb.
Would anyone please tell me should I increase the value of the trim pot to get 22mv and what value should I use?
Please advice.
Thanks
I set the bias current of Channel A of the amp @ 22mv (100ma) and trying to set Channel 2 @ the same current however the maximum I can get is 14mv (63.63ma) with original trim pot 330 Ohm on the pcb.
Would anyone please tell me should I increase the value of the trim pot to get 22mv and what value should I use?
Please advice.
Thanks
Unless you have a schematic for the amp and know what you are doing, I would not increase the current at all. Your output stage could be set up for class B operation and your mucking with it will probably destroy your amp.
Hi dirkwright
Yes I am sure the original setting of the amp was Class B which I measured 0.5mv (2.27ma) for channel 1 and 0.7mv (3.18ma) for channel 2.
After I increased the bias current for both channel 1 22mv (100ma) and channel 2 14mv (63.63ma) I can hear the sound different after the setting and I have been using the amp for more than a week's time and the amp is still working fine. Therefore I believe I can set the amp @ 100ma without destroying the amp.
look at a Vbe multiplier. Determine how it generates the bias voltage.
Determine how it is capable of making that bias voltage adjustable.
Now look at your amplifier and see if your amplifier's Vbe multiplier matches what you have just researched.
Only after that will you realise that changing the Variable Resistance is the wrong change to make.
Go and look and learn.
Determine how it is capable of making that bias voltage adjustable.
Now look at your amplifier and see if your amplifier's Vbe multiplier matches what you have just researched.
Only after that will you realise that changing the Variable Resistance is the wrong change to make.
Go and look and learn.
Hi Andrew
It takes time to found out the Vbe multiplier from the pcb. I will post the Vbe multiplier after I get it. But on the other hand, to make it simple, how about if I change the emitter resistors from 0.22 Ohm to 0.1 Ohm (11mv/0.11 Ohm = 110ma)?
wrong approach also ...before starting tweaks in one amplifier you need to understand how is working ...
since your amplifier is commercially made and the original intention was a small pa amplifier there is a million more things that can be done to alter the ratio between safety and quality of the amplifier before the bias scheme ....
if you change the emitter resistors the first thing that will not operate properly is the VI limiter and actually will introduce power limits and distortion in your signal regardless high or low bias .
Vi limiter is calculated to operate on a specific resistor and changing that will effect the operation of the system ...
small advice ... if you aim for quality high or higher bias is not the only solution .... removing totally the VI limiter will make miracles ....changing a few capacitors o play with the speed and bandwidth of the amplifier will also make miracles ...
Point is that any of these changes need to be verified and any of these changes will change the safety of your amplifier for worst ... choice is yours .....
since your amplifier is commercially made and the original intention was a small pa amplifier there is a million more things that can be done to alter the ratio between safety and quality of the amplifier before the bias scheme ....
if you change the emitter resistors the first thing that will not operate properly is the VI limiter and actually will introduce power limits and distortion in your signal regardless high or low bias .
Vi limiter is calculated to operate on a specific resistor and changing that will effect the operation of the system ...
small advice ... if you aim for quality high or higher bias is not the only solution .... removing totally the VI limiter will make miracles ....changing a few capacitors o play with the speed and bandwidth of the amplifier will also make miracles ...
Point is that any of these changes need to be verified and any of these changes will change the safety of your amplifier for worst ... choice is yours .....
Hi sakis
Sounds very interesting. I am reading the article: "VI Limiter In Amplifiers" from Rod Elliott, very helpful.
By the way, what capacitors you will suggest to change to improve the speed and bandwidth of the amp? These are what I want to achieve as well.
Another article: "Put A Tiger In Your Amp - Adjusting The Bias Of Your Amp - You'll Be Surprised...." by Dejan V. Veselinovic of tnt-audio is also very informative, I am impressed.
I could suggest a few things ...which as i have many times already said will be against amplifier's safety ,stability and high power long run ..... post a schematic though ..... if you don't have one this will not work .....
- Status
- Not open for further replies.