Hoping to start my build within next few weeks as my gout fingers are now not far from pain free!
The matter of resistor sound has always fascinated me and I have a few favourites.
In the USSA power and pre amps which resistors do you think have the greatest influence on sound quality?
Many thanks
The matter of resistor sound has always fascinated me and I have a few favourites.
In the USSA power and pre amps which resistors do you think have the greatest influence on sound quality?
Many thanks
Hoping to start my build within next few weeks as my gout fingers are now not far from pain free!
The matter of resistor sound has always fascinated me and I have a few favourites.
In the USSA power and pre amps which resistors do you think have the greatest influence on sound quality?
Many thanks
We did a fewof usspa
Resistor.. The feedback R. Input output and impedance R. I use caddock when able. Feedback 22 ohm and output R.. Can use carbon as well. Better musicality and better bass
Mosfet gate R .. Use takman carbon. Much better
Otherwise... Dale rn cmf. Takman metal caddock. All fine.
Low ohm.high watt.. We prefer mox or cement
Many thanks for that AnthonyA. I will be building the pre as well as the power versions and am now getting the BoM in line. I have a good number of Tantalum resistors and also quite a few of the 'BE' series made by Sfernice to Be Yamamura's requirements...they sound very good indeed - to my ears.
Brianco,
Be series? What's its like?
Sounds like fun. Just for u to know DC coupled. And relatively short chain.. The resistor doesn't exactly impart as much color as normal.. After burn in.. Ehh it's almost the same.. Give or take..
Cheers
Had a bit of a scare this morning.! I decided to run a few audio tests on the second completed amp module that I had biased and installed Rv1,2 in recently.
I was powering it up using 2 identical bench supplies. I reached about +- 24 volts and everything looked good with a cold bias current of about 1.25 amps. I was fine tuning the power supply voltage on the positive supply and when I looked at the voltmeter on the - supply it was reading 41 volts. I shut everything down and there was no apparent damage or hot parts.
I was most concerned about the 4700uf 35 volt capacitor on the board - supply.
It measured OK at 5kuf. Luckily the current limit, which I had set to 1.5 amps, was still working.
I took apart the cheap Ebay variable voltage supply to see what happened. There are 2 2n3773 on a flat piece of aluminum near the rear panel. The thermostat controlled fan draws air from the front sides of the chassis and out the back panel past the heatsink. One of the 2n3773 was shorted. I happened to have some used ones in stock. The root cause of the problem seemed to be that the mounting bolts on these TO3 type transistors were loose. The thermal insulator could not do it’s job of transferring heat from the transistor to the heatsink. I replaced the part, tightened both transistor’s mounting bolts and tested the supply. It ran fine but the transistors got very hot before the snap action fan thermostat activated. I shorted the thermostat so the fans are on all the time. I’ll put up with the noise.
Anyway, checked the second supply and tightened it’s loose mounting bolts and shorted the thermostat. Ready to try the amp module again and hope for no damage.
Long story short- It worked fine, settling to 1.05 amps and 1.4mv of offset after warmup.
I tested at 1khz - gain=10.75, THD at 2.83v 8 ohms -<.01%, THD at 10v rms -8 ohms -12.5watts = .048% Damping factor -done by adding 10ohms in parallel with 8 ohm load and measuring output voltage change = 51 - output Z calculated to .155 ohms
Maximum output before clipping with +- 27.5 v supplies, 1khz sine wave 20% duty cycle and an 8 ohm load was just over 17v rms or 36w.
I am impressed with the offset and bias stability of this design! The thermal performance of my heatsinks is very good and with a small fan it barely gets above 40c. The output mosfets are not even hot to the touch. So I think I am going to try for +-30v supplies and 1.1 amps of bias current as an operating point.
It will be a while before I get the chassis finished but I am now anxious to get it done and have a listen!
I was powering it up using 2 identical bench supplies. I reached about +- 24 volts and everything looked good with a cold bias current of about 1.25 amps. I was fine tuning the power supply voltage on the positive supply and when I looked at the voltmeter on the - supply it was reading 41 volts. I shut everything down and there was no apparent damage or hot parts.
I was most concerned about the 4700uf 35 volt capacitor on the board - supply.
It measured OK at 5kuf. Luckily the current limit, which I had set to 1.5 amps, was still working.
I took apart the cheap Ebay variable voltage supply to see what happened. There are 2 2n3773 on a flat piece of aluminum near the rear panel. The thermostat controlled fan draws air from the front sides of the chassis and out the back panel past the heatsink. One of the 2n3773 was shorted. I happened to have some used ones in stock. The root cause of the problem seemed to be that the mounting bolts on these TO3 type transistors were loose. The thermal insulator could not do it’s job of transferring heat from the transistor to the heatsink. I replaced the part, tightened both transistor’s mounting bolts and tested the supply. It ran fine but the transistors got very hot before the snap action fan thermostat activated. I shorted the thermostat so the fans are on all the time. I’ll put up with the noise.
Anyway, checked the second supply and tightened it’s loose mounting bolts and shorted the thermostat. Ready to try the amp module again and hope for no damage.
Long story short- It worked fine, settling to 1.05 amps and 1.4mv of offset after warmup.
I tested at 1khz - gain=10.75, THD at 2.83v 8 ohms -<.01%, THD at 10v rms -8 ohms -12.5watts = .048% Damping factor -done by adding 10ohms in parallel with 8 ohm load and measuring output voltage change = 51 - output Z calculated to .155 ohms
Maximum output before clipping with +- 27.5 v supplies, 1khz sine wave 20% duty cycle and an 8 ohm load was just over 17v rms or 36w.
I am impressed with the offset and bias stability of this design! The thermal performance of my heatsinks is very good and with a small fan it barely gets above 40c. The output mosfets are not even hot to the touch. So I think I am going to try for +-30v supplies and 1.1 amps of bias current as an operating point.
It will be a while before I get the chassis finished but I am now anxious to get it done and have a listen!
Last edited:
Unfortunately now unobtainable! They are a zero inductance 1 watt 5% made for Be Yamamura by Sfernice in France. The prototypes were ordered at a meeting of the then top man for Vishay (UK) at my kitchen table!!! Be lived in London for many years and was a good friend. He has returned to Japan and have a look at this link re his latest work!
Mono and Stereo High-End Audio Magazine: Jean Hiraga meets ALE and Be Yamamura: one of the VERY best systems in the world?
I have a soft spot for those resistors!
Hi Bfpca
Glad you found your bench power supply issue and not damaged the USSA5 board
You get the expected results for THD and damping factor. Thanks for providing this data.
You can crank up the bias with such obtained low temperature. For 4 ohms speakers you need more bias current for more class A. For 8 ohms speakers you can have more voltage for more class AB.
Fab
Glad you found your bench power supply issue and not damaged the USSA5 board
You get the expected results for THD and damping factor. Thanks for providing this data.
You can crank up the bias with such obtained low temperature. For 4 ohms speakers you need more bias current for more class A. For 8 ohms speakers you can have more voltage for more class AB.
Fab
Hi Fab
Some more progress, I put in a jumper on the RV2A and to make the 27R resistor I did (22.8 + (10R+10R) in parallel) so I could get to 26.8R
Did the 8.17.1 again with P1/P2 to adjust the bias to 50mV and below are the values:
VGSP = 1545, VGSN = 986 which makes it like
R13=47R
RV1A = 26.8
So im3pre = 1545 / (47+26.8) = 20.93mA
R14=47R
RV2A = 0 (jumper)
So im3pre = 986 / (47+0) = 20.97mA
After this I let it cook for 30 mins then biased up slowly to 65mV by making sure the DC offset is <5mV. Heat sink was nice hot but warm to touch. So with this I have completed the first amp board successfully. Next up is the 2nd board with the same procedure and revert back. Once these are done I will try to get 1% tolerance of 27R resistors for 1st amp board and based on the calculations on the second board and replace the same.
Thanks
Some more progress, I put in a jumper on the RV2A and to make the 27R resistor I did (22.8 + (10R+10R) in parallel) so I could get to 26.8R
Did the 8.17.1 again with P1/P2 to adjust the bias to 50mV and below are the values:
VGSP = 1545, VGSN = 986 which makes it like
R13=47R
RV1A = 26.8
So im3pre = 1545 / (47+26.8) = 20.93mA
R14=47R
RV2A = 0 (jumper)
So im3pre = 986 / (47+0) = 20.97mA
After this I let it cook for 30 mins then biased up slowly to 65mV by making sure the DC offset is <5mV. Heat sink was nice hot but warm to touch. So with this I have completed the first amp board successfully. Next up is the 2nd board with the same procedure and revert back. Once these are done I will try to get 1% tolerance of 27R resistors for 1st amp board and based on the calculations on the second board and replace the same.
Thanks
Unfortunately now unobtainable! They are a zero inductance 1 watt 5% made for Be Yamamura by Sfernice in France. The prototypes were ordered at a meeting of the then top man for Vishay (UK) at my kitchen table!!! Be lived in London for many years and was a good friend. He has returned to Japan and have a look at this link re his latest work!
Mono and Stereo High-End Audio Magazine: Jean Hiraga meets ALE and Be Yamamura: one of the VERY best systems in the world?
I have a soft spot for those resistors!
That's sounds awesome! Waiting for your reviews.
Fab have good taste in sounds. Your in good hands
Hi mannirajHi Fab
Some more progress, I put in a jumper on the RV2A and to make the 27R resistor I did (22.8 + (10R+10R) in parallel) so I could get to 26.8R
Did the 8.17.1 again with P1/P2 to adjust the bias to 50mV and below are the values:
VGSP = 1545, VGSN = 986 which makes it like
R13=47R
RV1A = 26.8
So im3pre = 1545 / (47+26.8) = 20.93mA
R14=47R
RV2A = 0 (jumper)
So im3pre = 986 / (47+0) = 20.97mA
After this I let it cook for 30 mins then biased up slowly to 65mV by making sure the DC offset is <5mV. Heat sink was nice hot but warm to touch. So with this I have completed the first amp board successfully. Next up is the 2nd board with the same procedure and revert back. Once these are done I will try to get 1% tolerance of 27R resistors for 1st amp board and based on the calculations on the second board and replace the same.
Thanks
That is excellent result. You can stay as-is for the first channel thus do not need exact matching of driver current between left and right amplifier channel. Out of curiosity can you give the VGSP and VGSN values at 65mv (1.3A bias) for my information. This would allow you to calculate the final driver current at 65mv.
Fab
Hi Fab,
I retested the first amp board today and with bias at 65mV the heat sink was too hot to the touch. So I had to reduce it to 61mV which was like bearable. With this bias my VGSP = 1710 and VGSN = 1140 with offset being 2mV. The current on positive is 1710 / (47+26.8) = 23.17mA and on the negative side is 1140 / (47+0) = 24.2mA. I have put RV1A = 26.8R with RV2A with a jumper. The current does not seem to match with a difference of 1mA, not sure if this mismatch is acceptable or do I need to tweak the RV1A value?
Now coming to the second channel, all went fine but the step 8.14 front end testing I am getting the negative side around 0.1Vdc while adjusting P1 and positive side around 0.576Vdc while adjusting P2. I am unable to get both the sides to 0.3Vdc as turning P2 makes the positive side very high but negative comes to less than half. I rechecked the P1 and P2 values and they are Bourns W103 which is 10K. I have installed the thermistors touching the heat sinks and also the 2sk2013/313 matched ones installed. Any idea what could be the issue with this board at this step?
Thanks
I retested the first amp board today and with bias at 65mV the heat sink was too hot to the touch. So I had to reduce it to 61mV which was like bearable. With this bias my VGSP = 1710 and VGSN = 1140 with offset being 2mV. The current on positive is 1710 / (47+26.8) = 23.17mA and on the negative side is 1140 / (47+0) = 24.2mA. I have put RV1A = 26.8R with RV2A with a jumper. The current does not seem to match with a difference of 1mA, not sure if this mismatch is acceptable or do I need to tweak the RV1A value?
Now coming to the second channel, all went fine but the step 8.14 front end testing I am getting the negative side around 0.1Vdc while adjusting P1 and positive side around 0.576Vdc while adjusting P2. I am unable to get both the sides to 0.3Vdc as turning P2 makes the positive side very high but negative comes to less than half. I rechecked the P1 and P2 values and they are Bourns W103 which is 10K. I have installed the thermistors touching the heat sinks and also the 2sk2013/313 matched ones installed. Any idea what could be the issue with this board at this step?
Thanks
Just to add on top my earlier comments, I did the previous step of setting 2.5Vdc on both positive and negative sides without any issues. Then I installed the 2sk2013/313 driver transistors along with the thermistors slightly touching the heat sinks. I am stuck now at 8.14 step unable to set the 0.3Vdc on both the sides as one of the sides goes either too high or too low if I can get one side to 0.3Vdc.
Thanks
Thanks
Hi manniraj
For the first amp, the current difference between positive and negative is < 5% which is lower than 12% as indicated in the manual thus it is fine. When you adjust bias ensure to let it stabilize for at least 45 minutes as per the manual.
For the second channel, something happened when you installed the drivers. Some possibilities:
1) diodes D1 to D4 are mixed or shorted (zener and diodes signals ) which limits the voltage. You can try to lift D1 or D3 and D2 or D4 to remove the diodes from the circuit.
2) mosfet drivers are in problems ( polarity reversed, very bad match,...). With step 8.14 voltage set at 0.1V for both P1 and P2, measure voltage between TP5 and TP7. Also measure voltage between TP6 and TP8. Both voltages should be in the range of about 1.9 to 2.3v. The differences between the 2 voltages should be normally much less than 0.2V.
3) Have you put a load resistor between OUT and ground as per figure 18.
4) thermistor are damaged. Lifting of one thermistor away from heatsink should have your dc offset to change considerably if thermistor is fine.
5) proper GND should be connected. Connect both GND terminal to PSU GND to make sure.
Re-measure voltages between TP9-TP10 and TP11-TP12 to check if you still have 28mv for both.
Fab
For the first amp, the current difference between positive and negative is < 5% which is lower than 12% as indicated in the manual thus it is fine. When you adjust bias ensure to let it stabilize for at least 45 minutes as per the manual.
For the second channel, something happened when you installed the drivers. Some possibilities:
1) diodes D1 to D4 are mixed or shorted (zener and diodes signals ) which limits the voltage. You can try to lift D1 or D3 and D2 or D4 to remove the diodes from the circuit.
2) mosfet drivers are in problems ( polarity reversed, very bad match,...). With step 8.14 voltage set at 0.1V for both P1 and P2, measure voltage between TP5 and TP7. Also measure voltage between TP6 and TP8. Both voltages should be in the range of about 1.9 to 2.3v. The differences between the 2 voltages should be normally much less than 0.2V.
3) Have you put a load resistor between OUT and ground as per figure 18.
4) thermistor are damaged. Lifting of one thermistor away from heatsink should have your dc offset to change considerably if thermistor is fine.
5) proper GND should be connected. Connect both GND terminal to PSU GND to make sure.
Re-measure voltages between TP9-TP10 and TP11-TP12 to check if you still have 28mv for both.
Fab
Last edited:
Hi Fab
Here are my responses to check based on your suggestions.
1) diodes D1 to D4 are mixed or shorted (zener and diodes signals ) which limits the voltage. You can try to lift D1 or D3 and D2 or D4 to remove the diodes from the circuit. I have not removed any of these as of now to test. Based on the below test and your response will lift one leg of D1 and D4. If I remove the legs of these can I power it and recheck the steps all over again right from the initial bias and till step 8.14?
2) mosfet drivers are in problems ( polarity reversed, very bad match,...). With step 8.14 voltage set at 0.1V for both P1 and P2, measure voltage between TP5 and TP7. Also measure voltage between TP6 and TP8. Both voltages should be in the range of about 1.9 to 2.3v. The differences between the 2 voltages should be normally much less than 0.2V. Replaced the M4 with another matched one and before soldering checked. After replacing 2sk2013 the TP5-TP7 = 1.720 and TP6-TP8 = 1.906
3) Have you put a load resistor between OUT and ground as per figure 18. No I have not put any load resistor even for the good amp board for setup so far.
4) thermistor are damaged. Lifting of one thermistor away from heatsink should have your dc offset to change considerably if thermistor is fine.
5) proper GND should be connected. Connect both GND terminal to PSU GND to make sure. Yes I have used both the GND and PGND to connect to the PSU grounds
Re-measure voltages between TP9-TP10 and TP11-TP12 to check if you still have 28mv for both. Yes rechecked and both of them are at perfect 28mV
Thanks
Here are my responses to check based on your suggestions.
1) diodes D1 to D4 are mixed or shorted (zener and diodes signals ) which limits the voltage. You can try to lift D1 or D3 and D2 or D4 to remove the diodes from the circuit. I have not removed any of these as of now to test. Based on the below test and your response will lift one leg of D1 and D4. If I remove the legs of these can I power it and recheck the steps all over again right from the initial bias and till step 8.14?
2) mosfet drivers are in problems ( polarity reversed, very bad match,...). With step 8.14 voltage set at 0.1V for both P1 and P2, measure voltage between TP5 and TP7. Also measure voltage between TP6 and TP8. Both voltages should be in the range of about 1.9 to 2.3v. The differences between the 2 voltages should be normally much less than 0.2V. Replaced the M4 with another matched one and before soldering checked. After replacing 2sk2013 the TP5-TP7 = 1.720 and TP6-TP8 = 1.906
3) Have you put a load resistor between OUT and ground as per figure 18. No I have not put any load resistor even for the good amp board for setup so far.
4) thermistor are damaged. Lifting of one thermistor away from heatsink should have your dc offset to change considerably if thermistor is fine.
5) proper GND should be connected. Connect both GND terminal to PSU GND to make sure. Yes I have used both the GND and PGND to connect to the PSU grounds
Re-measure voltages between TP9-TP10 and TP11-TP12 to check if you still have 28mv for both. Yes rechecked and both of them are at perfect 28mV
Thanks
Last edited:
Hi manniraj
See above.
You can redo step 8.14 of manual but before restart from a more balanced adjustment point. With power OFF adjust P1 to measure a resistance of 715 ohms between TP1-TP7. Also adjust P2 to measure 785 ohms between TP2-TP8.
Afterwards Power ON the amp and them verify if close to 0.3v for step 8.14 and monitoring DC offset and adjust as necessary.
Fab
Last edited:
Thanks Fab, just to confirm my understanding -
1) diodes D1 to D4 are mixed or shorted (zener and diodes signals ) which limits the voltage. You can try to lift D1 or D3 and D2 or D4 to remove the diodes from the circuit. I have not removed any of these as of now to test F: easy to desolder only one lead of one of the diodes per polarity. Just ensure to have diodes working once the output mosfet are installed to ensure protection. Shorted diode is easy to check with a multimeter.
You meant to remove D1 and D4 from the PCB board and go through the steps till I install the output mosfets (Exicons) right? So for now I will remove both of these Diodes but before that I will check both the thermistors.
4) thermistor are damaged. Lifting of one thermistor away from heatsink should have your dc offset to change considerably if thermistor is fine. F: you should check that. If at one point a lead of the thermistor touched the heatsink crossing the paint it may be damaged. It happened a few times from builder feedback.
I will move each of the thermistors one after another to check if the offset drifts considerably. Your last suggestion of redoing step 8.14 I will do only after removing the diodes and setting 715R between TP1-TP7 and then 785R between TP2-TP8 while powered off. And then check for the 0.3Vdc for the same step.
thanks a lot again
1) diodes D1 to D4 are mixed or shorted (zener and diodes signals ) which limits the voltage. You can try to lift D1 or D3 and D2 or D4 to remove the diodes from the circuit. I have not removed any of these as of now to test F: easy to desolder only one lead of one of the diodes per polarity. Just ensure to have diodes working once the output mosfet are installed to ensure protection. Shorted diode is easy to check with a multimeter.
You meant to remove D1 and D4 from the PCB board and go through the steps till I install the output mosfets (Exicons) right? So for now I will remove both of these Diodes but before that I will check both the thermistors.
4) thermistor are damaged. Lifting of one thermistor away from heatsink should have your dc offset to change considerably if thermistor is fine. F: you should check that. If at one point a lead of the thermistor touched the heatsink crossing the paint it may be damaged. It happened a few times from builder feedback.
I will move each of the thermistors one after another to check if the offset drifts considerably. Your last suggestion of redoing step 8.14 I will do only after removing the diodes and setting 715R between TP1-TP7 and then 785R between TP2-TP8 while powered off. And then check for the 0.3Vdc for the same step.
thanks a lot again