I can see that you are using coherent averaging. I am sure that you know that this method makes FFT plots look better than they really are, regarding noise floor. It still measures (THD and THD+N numbers) properly, however an inexperienced reader may be fascinated by the low noise floor, which is both false regarding noise density and mains spuriae.
Yes, I know and that’s why amplifier's noise floor was measured by the proper method (post #4). I implore anyone to point out an amplifier with lower output noise. Noise floor is about 25 nV/rtHz.
Not to mention that considerable part of measured noise is not form the amplifier output, rather by EMI incursion into non professional level measurement gear.
Coherent averaging was used to reveal low high order distortion harmonics. Check post #423. 🙂
Not to mention that considerable part of measured noise is not form the amplifier output, rather by EMI incursion into non professional level measurement gear.
Coherent averaging was used to reveal low high order distortion harmonics. Check post #423. 🙂
Did you measure the THD on the real amp?
When I looked into BJT vs Lateral Mosfet, I always found contradictory information. Is Lateral Mosfet actually better than BJT?
When I looked into BJT vs Lateral Mosfet, I always found contradictory information. Is Lateral Mosfet actually better than BJT?
Of course.
I’m sure you know that there is no general answer to that, rather it depends on application and design.
For this specific design, lateral MOSFETs were superior to BJTs. The same design, only with a CFP output was actually built, listened to and measured.
What makes them better is not all in that table. It’s about Ic/Vce curves.
Did you read the first post (or any) of this thread?
Hi tombo56, I might need your help here. I finally finished the pcbs for this amp and am testing each channel at a time on the bench. The supply is regulated +/-17.00VDC. I am trying to adjust the Iq bias to your recommended 1A.
With pot RV1 set totally clockwise (max R) I get 43mVDC across both R16/R17 = 536mA bias current in output stage. I am measuring the bias current with a DMM on DC current range in series with the pos and neg supply - so that is also measuring the current for the input etc on top of the Q8 or Q9 OS current.
With pot RV1 turned totally anticlockwise (min R) I get 157mV across both R16/R17 = 823mA bias current in output stage.
Both channels give virtually the same results.
What do I need to do to get a wider adjustment range for Iq in RV1? Does the value of R10 (150R) need to change.
Your suggestions would be very much appreciated, thanks.
I have the input shorted and the output open circuit (no load)
Gary.
With pot RV1 set totally clockwise (max R) I get 43mVDC across both R16/R17 = 536mA bias current in output stage. I am measuring the bias current with a DMM on DC current range in series with the pos and neg supply - so that is also measuring the current for the input etc on top of the Q8 or Q9 OS current.
With pot RV1 turned totally anticlockwise (min R) I get 157mV across both R16/R17 = 823mA bias current in output stage.
Both channels give virtually the same results.
What do I need to do to get a wider adjustment range for Iq in RV1? Does the value of R10 (150R) need to change.
Your suggestions would be very much appreciated, thanks.
I have the input shorted and the output open circuit (no load)
Gary.
Hi gary s, lower Iq range is caused by wide J113 Idss variation from one production batch to another.
As you said, solution is to reduce R10 value. Easiest way would be to solder in parallel another 150 Ω resistor or change it to 47 – 100 Ω. Another resistor can be placed on the PCB B side, so it won’t spoil how assembled PCB looks.
For any new build, I would recommend to replace R10 with proposed lower value.
As you said, solution is to reduce R10 value. Easiest way would be to solder in parallel another 150 Ω resistor or change it to 47 – 100 Ω. Another resistor can be placed on the PCB B side, so it won’t spoil how assembled PCB looks.
For any new build, I would recommend to replace R10 with proposed lower value.
Thanks for the quick reply tombo, I will try your suggestion (parallel another 150R for R10) tomorrow and report back - many thanks.
I had to parallel a 68R resistor with the 150R for R10 = 47R to get the Iq up to where I needed it. Even then the RV1 bias adjust pot was near the end of its travel - so maybe R10 could be even lower in value. But all working now and sitting at 1A Iq for the output stage. Now to bolt it all up and finish wiring in the chassis.
Thanks Vunce - I remember you finished yours quite a while ago now, can you remind me of your impressions of its sound.
Very good impressions indeed, Gary.
This amp just sounds right in my system and room. Great detail retrieval without ever fatiguing to listen to for long periods of time. Percussion work has that fast snappy hit sound, especially when it's El Estepario Siberiano behind the kit!!
It's all subjective of course, and you're close to being able to form your own impressions. I'm not very good at describing to others what sounds good to me, if you don't need gobs of power Bliss is an amplifier that won't disappoint.
Wishing you good luck on a drama free power-up and first sound, Rock on Gary!
This amp just sounds right in my system and room. Great detail retrieval without ever fatiguing to listen to for long periods of time. Percussion work has that fast snappy hit sound, especially when it's El Estepario Siberiano behind the kit!!
It's all subjective of course, and you're close to being able to form your own impressions. I'm not very good at describing to others what sounds good to me, if you don't need gobs of power Bliss is an amplifier that won't disappoint.
Wishing you good luck on a drama free power-up and first sound, Rock on Gary!
Thanks again Vunce, both channels bias well on the bench - so will take a day or so to get it all together in the chassis. Then another bias check for good measure. Then some listening - will report back.
Vunce, I think it's time for you to join anonymous amp builders group for 12 steps program... 😀Very good impressions indeed, Gary.
This amp just sounds right in my system and room. Great detail retrieval without ever fatiguing to listen to for long periods of time. Percussion work has that fast snappy hit sound, especially when it's El Estepario Siberiano behind the kit!!
It's all subjective of course, and you're close to being able to form your own impressions. I'm not very good at describing to others what sounds good to me, if you don't need gobs of power Bliss is an amplifier that won't disappoint.
Wishing you good luck on a drama free power-up and first sound, Rock on Gary!
While connecting amp to the supply, take care to connect signal ground to the ground connector at the center of GND connectors group and decoupling ground to the edge of that group.
Also, connect loudspeaker ground next to the signal ground at the power supply board.
Not doing so will not spoil the sound, only changes that 0.0000x distortion at the last digit.
Completed the build and it is singing!! Wahoo..
Some tech notes:
Regulated +/-17VDC supplies and the bias settles at 1.05Amps DC, rock steady. DC offset is less than 1mV - so negligible. Heatsinks sit at a comfortable 45 deg C.
Sound:
First of all the amp is dead quiet, no hiss or hum at all. First impressions with music through my Yamaha NS10M speakers. Very revealing, fast and dynamic, for only 10 watts or so, it has great authority. Very much on a par with my Stasis 50wpc amp and Zen Mods Babelfish F8.
It is certainly right up there with the best I have built, and it is also the smallest output amp, so tombo has done a remarkable job with this amp.
Highly recommended for any who want superb sound with an easy build and a good price point.
Oh, by the way, I forgot to mention it is powered by dual Meanwell SMPS LRS-50-15. The SMPS output is adjusted to give the +/- 17VDC supply at the amp terminals, with AmyAlice filters for plus / neg supplies. No background noise at all. Benefits - cheap and light weight, takes up less chassis space.
Overall - a big winner for me.
Some tech notes:
Regulated +/-17VDC supplies and the bias settles at 1.05Amps DC, rock steady. DC offset is less than 1mV - so negligible. Heatsinks sit at a comfortable 45 deg C.
Sound:
First of all the amp is dead quiet, no hiss or hum at all. First impressions with music through my Yamaha NS10M speakers. Very revealing, fast and dynamic, for only 10 watts or so, it has great authority. Very much on a par with my Stasis 50wpc amp and Zen Mods Babelfish F8.
It is certainly right up there with the best I have built, and it is also the smallest output amp, so tombo has done a remarkable job with this amp.
Highly recommended for any who want superb sound with an easy build and a good price point.
Oh, by the way, I forgot to mention it is powered by dual Meanwell SMPS LRS-50-15. The SMPS output is adjusted to give the +/- 17VDC supply at the amp terminals, with AmyAlice filters for plus / neg supplies. No background noise at all. Benefits - cheap and light weight, takes up less chassis space.
Overall - a big winner for me.