I am not using STD15P/N pair instead using STD03P/N with external resistors. I am using 7Watts resistors from Vishay Sfernice. These below
RWM0622R220JA15E1 Vishay Sfernice | RWMB-.22CT-ND | DigiKey
My only adjustment trim pot is the 100Ohm trim pot as mentioned above which is between two temperature compensation diode terminals of the Sanken devices. And there is only one adjustment per channel. So you mean to say I when I adjust for 8.5mv across one of the 0.22Ohm resistor, other also will be 8.5-9mv range and that will set the bias?
No, you misunderstand me. I meant you are no longer using SAP15 transistors. Now that you have STD03 pairs, you don't need to set the current in the same complicated way.
You could not do this with SAP15 pairs because the resistors were internal and that was why the special procedure to set the bias current. You have already shown that the bias setting procedure was for the old (SAP15) transistors - not the new ones.
The 0.22 resistors are in series, correct? So you can measure across one or both to calculate the current that flows through both transistors and resistors. Sure, you can take just one resistor and measure the voltage across it because that is easy to understand. However, I suggested measuring the voltage across both resistors because it is more accurate at such low voltages . Still, it doesn't matter to me if you don't understand why it calculates the same for 1 or 2 or 100 resistors in series but you certainly know the total resistance so why not use it.
As long as you can understand ohms law, you can calculate your required voltage across the resistor(s) to determine the current according to my previous post. You should be able to adjust the bias current in a similar manner to the previous transistors, if the internal diodes are similar.
As I said in the past, I haven't tried this substitution but since you have decided to do it this way now, I hope you remain successful and keep a very close watch on the current and pull the power to the amplifier board the split-second you see it rise by itself. One thermal runaway and your parts will be history.
You could not do this with SAP15 pairs because the resistors were internal and that was why the special procedure to set the bias current. You have already shown that the bias setting procedure was for the old (SAP15) transistors - not the new ones.
The 0.22 resistors are in series, correct? So you can measure across one or both to calculate the current that flows through both transistors and resistors. Sure, you can take just one resistor and measure the voltage across it because that is easy to understand. However, I suggested measuring the voltage across both resistors because it is more accurate at such low voltages . Still, it doesn't matter to me if you don't understand why it calculates the same for 1 or 2 or 100 resistors in series but you certainly know the total resistance so why not use it.
As long as you can understand ohms law, you can calculate your required voltage across the resistor(s) to determine the current according to my previous post. You should be able to adjust the bias current in a similar manner to the previous transistors, if the internal diodes are similar.
As I said in the past, I haven't tried this substitution but since you have decided to do it this way now, I hope you remain successful and keep a very close watch on the current and pull the power to the amplifier board the split-second you see it rise by itself. One thermal runaway and your parts will be history.
I have assumed here, that you have fitted both emitter resistors as this borrowed sketch by moderator tomchr here shows. Otherwise, I could understand why you may have trouble following the suggestions.
Here is the thread link, for interest: http://www.diyaudio.com/forums/chip-amps/161337-yet-another-lme49811-std03-build.html
Here is the thread link, for interest: http://www.diyaudio.com/forums/chip-amps/161337-yet-another-lme49811-std03-build.html
Sorry for the confusion, that's not what I meant. I was not saying I did not want to follow your suggestion. Yes your suggestion was clear enough, it's just that I was not sure, how to set that so called 2.5mA forward current for those thermal diodes.
Yes, seems like the circuit's output stage is similar to what you posted from the other thread. Except for, the MF amp has a 4.7Ohm+cap zobel at the output also the output comes through an parallel inductor+resistor stabilization.
I do not have the exact circuit, what I did was, I substituted the STD03 instead of SAP15 and connected the Emitter pin to where originally was the sense pin of the SAP15 device. It was very easy because of the reduced form factor of the newer STD03 device. Later I added the 7W 0.22Ohm resistor between the old sense and Emitter pads. Everything else in the circuit is retained, except for faulty trimmer replaced with exact same value/brand.
OK, you have the required parts in place so all should be fine.
I will repeat this again because you are still referring to the bias setting procedure for SAP15 pairs. Is that because you can't find any other procedure for the STD03 pairs?
My point was that you don't follow the old procedure because STD03 pairs are different, as you know. You do not need to set up a 2.5 current first and then slip some dummy emitter resistors into the circuit for the final 40 mA setting. You no longer need to do that. Just set the DVM to lowest DC Volt scale (It must be capable of reading down to 1 mV reliably). Clip with clip leads, IC clips or solder, (don't just hold the meter probes in place) across the emitter resistor(s). Then measure the voltage drop all the way from zero up to the required mV reading - slowly adjusting the control and waiting for the heating effect to stabilize as the current increases.
Obviously, if you read some high voltage above the required mV at the start, you need to turn that control down, fast. I'm sure that as long as you have made secure meter connections, you will quickly find the correct direction in that event.
I have no idea where the position of adjustment lies now, so it could be set to minimum, maximum or anywhere between. That is your problem to make a safe starting position. You could guess with some confidence that CCW means minimum bias but unfortunately, there are no guarantees that is the case. With many MF products, you come to expect the unusual rather than normal so I would be cautious, as I have said.
You could look at the circuit components and connections to deduce in which direction the wiper becomes more positive or negative and that would help your decision, as the wiper closest to positive rail usually means maximum bias current.
I will repeat this again because you are still referring to the bias setting procedure for SAP15 pairs. Is that because you can't find any other procedure for the STD03 pairs?
My point was that you don't follow the old procedure because STD03 pairs are different, as you know. You do not need to set up a 2.5 current first and then slip some dummy emitter resistors into the circuit for the final 40 mA setting. You no longer need to do that. Just set the DVM to lowest DC Volt scale (It must be capable of reading down to 1 mV reliably). Clip with clip leads, IC clips or solder, (don't just hold the meter probes in place) across the emitter resistor(s). Then measure the voltage drop all the way from zero up to the required mV reading - slowly adjusting the control and waiting for the heating effect to stabilize as the current increases.
Obviously, if you read some high voltage above the required mV at the start, you need to turn that control down, fast. I'm sure that as long as you have made secure meter connections, you will quickly find the correct direction in that event.
I have no idea where the position of adjustment lies now, so it could be set to minimum, maximum or anywhere between. That is your problem to make a safe starting position. You could guess with some confidence that CCW means minimum bias but unfortunately, there are no guarantees that is the case. With many MF products, you come to expect the unusual rather than normal so I would be cautious, as I have said.
You could look at the circuit components and connections to deduce in which direction the wiper becomes more positive or negative and that would help your decision, as the wiper closest to positive rail usually means maximum bias current.
Thanks Ian. I did not just blindly replaced the trimmer before I posted my questions. Remember I had one working channel? So I took the trimmer out of that channel and pre-adjusted both my trimmers to what I read on that old trimmer before installing the new trimmers on to the board. It was around 60Ohms to start with. So yeah I did not just start at an extreme at the first place..
Then when I powered and measured as you suggested it was hovering around 26mV. For both the channels across 0.22Ohm. Now I have reduced both to 18mV range. Hope that is fine?
I used crocodile clips on the DMM to measure. I also checked the DC offset on both channel.
One reads -11mV and the other reads -15mV. Guess that is fine too... Thanks a lot for all your inputs...
Good that you were prepared. However, let's be clear. I specified 9 mV for a single 0.22R. 18 mV was for two, in series as they appear on the sketch and as I explained. If you have measured 18mV across only 1 resistor, the current must be 80 mA - twice Sanken's recommended bias current.
Trust and use Ohm's law, I=V/R 0.04A=0.0088V/0.22R
..............................................0.04A=0.0176V/0.44R
On the other hand, MF may have used higher current according to an estimate based on your measurement of the bias setting with SAP15 transistors. The voltage drop of 26 mV across only 1 resistor suggests a current of roughly 120 mA, 3 times Sanken's recommended current.
I was afraid that introducing a measurement of more that one object will be hard to understand. It seems to have confused you.
Still, you would have some idea of how hot the working channel became before replacing parts. A gross difference like this would be easy to tell from the temperature of the heatsink. Meantime, I would monitor temperature and the bias voltage over several hours and again, after playing some music loud, to get the temperature up and see if it stabilizes and falls back to the original level afterward. Take care measuring the temperature rise regularly with the cover on but just resting on the test leads. Listen for smoothness of the crossover, such that you don't hear harshness.
Good work so far but watch and preferably measure temperatures for your own assurance.
Trust and use Ohm's law, I=V/R 0.04A=0.0088V/0.22R
..............................................0.04A=0.0176V/0.44R
On the other hand, MF may have used higher current according to an estimate based on your measurement of the bias setting with SAP15 transistors. The voltage drop of 26 mV across only 1 resistor suggests a current of roughly 120 mA, 3 times Sanken's recommended current.
I was afraid that introducing a measurement of more that one object will be hard to understand. It seems to have confused you.
Still, you would have some idea of how hot the working channel became before replacing parts. A gross difference like this would be easy to tell from the temperature of the heatsink. Meantime, I would monitor temperature and the bias voltage over several hours and again, after playing some music loud, to get the temperature up and see if it stabilizes and falls back to the original level afterward. Take care measuring the temperature rise regularly with the cover on but just resting on the test leads. Listen for smoothness of the crossover, such that you don't hear harshness.
Good work so far but watch and preferably measure temperatures for your own assurance.
Last edited:
Sorry that was a typo i measured 18mv across 0.44ohm(0.22+0.22), i got that part Clearly. I measured the voltages betweem emitter to emitter from the beginning.
26mv was also between emitter to emitter (with 0.22+0.22 in series between the emitters) with no load connected
I read in another thread that 50mA-60mA idling current makes the STD03 sounds much better. And unlike the SAP15 devices, as it does not have the emitter resistors, which was the weakest link, it was comparatively safer to do so... So may be I should indeed adjust for 22mV to 26mv.. Which will be roughly there? Or may be to play it safe and keep it at 20mV which is like 45+mA is better?
So far the heatsink is only slightly warm while playing at around 11'O clock position on the volume dial. And it is quite loud with a speaker that has sensitivity of 87db/W/m.
So far the heatsink is only slightly warm while playing at around 11'O clock position on the volume dial. And it is quite loud with a speaker that has sensitivity of 87db/W/m.
Yes, from the thread I linked, guys were using around 100 mA bias current which probably sounded better for some ears. Some folks are very focused on class A sound and this often leads them to crank bias up as far as they dare within their power requirement - dangerous for the next owner who may want it up louder and causes it to overheat and blow.
The big questions are: How does a higher bias relate to heatsink size and safe full power operation of the MF A3.2 amplifier?
What is bias is required to match the original sound quality of the amplifier?
As you are only running recommended current, I expect it will be quite cool, so let's try testing fully with the 26 mV/0.44R setting you found. It may well be just fine with that 60mA bias current. As before, keep a close eye on it for an extended period. Muffle the speakers if you don't have dummy loads for testing. I have used cushions, bedding, foam matreses etc. around the speakers when I need to reduce noise and I have nothing else.
If you find no improvement, sure, set it lower but I have my doubts about higher current with the standard amplifier. Heatsinking must be adequate at full power, which you must account for as you are responsible for the safety of the amplifier, even if your friend doesn't consider it.
The big questions are: How does a higher bias relate to heatsink size and safe full power operation of the MF A3.2 amplifier?
What is bias is required to match the original sound quality of the amplifier?
As you are only running recommended current, I expect it will be quite cool, so let's try testing fully with the 26 mV/0.44R setting you found. It may well be just fine with that 60mA bias current. As before, keep a close eye on it for an extended period. Muffle the speakers if you don't have dummy loads for testing. I have used cushions, bedding, foam matreses etc. around the speakers when I need to reduce noise and I have nothing else.
If you find no improvement, sure, set it lower but I have my doubts about higher current with the standard amplifier. Heatsinking must be adequate at full power, which you must account for as you are responsible for the safety of the amplifier, even if your friend doesn't consider it.
I settled for 45mA. And played it for 3-5hours with top covers on. seems to be fine.. I now reckon that, I should have taken some more clear pictures of the STD03 substitution, for future reference for someone in need... May be tomorrow I should re-open it...
This was a simple job, but then not having schematics was the biggest challenge...
This was a simple job, but then not having schematics was the biggest challenge...
guys any luck with manuals form what i understand by federal law they have to provide repair and service manuual even for a nominal fee..
i have a n a5cr that i'dl liekt o do a few mods too but seeing the issues with the MF products may sell it and get sumin else till i can build me 3 new amps.
i have a n a5cr that i'dl liekt o do a few mods too but seeing the issues with the MF products may sell it and get sumin else till i can build me 3 new amps.
- Status
- Not open for further replies.