Music volume and attenuators
Hi there,
I wonder why with the follwing amp setup :
a TA2024 under 12V, a 50K series attenuator
a TK2050 under 30V, a 10K ladder attenuator,
and the same active source and records I need to push the TK2050 up to 15 (3PM as for clock scale) when the TA2024 "shows" the same volume at 12? Of course the sound is cleaner at that volume on the TK, and I have no clues of the steps of that 10K ladder attenuator...
Hi there,
I wonder why with the follwing amp setup :
a TA2024 under 12V, a 50K series attenuator
a TK2050 under 30V, a 10K ladder attenuator,
and the same active source and records I need to push the TK2050 up to 15 (3PM as for clock scale) when the TA2024 "shows" the same volume at 12? Of course the sound is cleaner at that volume on the TK, and I have no clues of the steps of that 10K ladder attenuator...
Hi Krystan,
thanks for having pointed that, and I've looked on boards, both are on the 20/20K ratio, so same gain, and then even when I've set the Sure TK2050 on "High" 10/20K it's slightly less loud than the Sure's TA2024. I'll have a try with a PCDP and it's own volume control. That's not a real porblem to turn the pot far, but I want to be sure that's not a real technical problem, just the normal way it works with such attenuator.
Thanks, Matthieu
thanks for having pointed that, and I've looked on boards, both are on the 20/20K ratio, so same gain, and then even when I've set the Sure TK2050 on "High" 10/20K it's slightly less loud than the Sure's TA2024. I'll have a try with a PCDP and it's own volume control. That's not a real porblem to turn the pot far, but I want to be sure that's not a real technical problem, just the normal way it works with such attenuator.
Thanks, Matthieu
Last edited:
Main gain
The main gain is set in the feedback section between the output and the driver chip. There may be a difference there.
The main gain is set in the feedback section between the output and the driver chip. There may be a difference there.
If the TA2024 is also from Sure, the sensitivity is lower (0.66V) than the TK2050 board (0.77V), which means the TA2024 should be closer to full power than the TK2050 set on High for the same input. But, if your TK2050 is the single version (2x50W), the difference in electrical power (3 times) doesn't mean a large difference in subjective loudness. You need 10 times the electrical power for a perceived 2 times loudness. If this is combined with a different curve in the attenuator (your 10K ladder attenuator has an higher attenuation at 12PM than the 50K series attenuator at 12 PM), it could explain why the TK2050 requires an "higher" position of your attenuator.
I would try to bypass the attenuators on both amp and plug them in a source that has his own volume control (such as the headphone/line output of a PC, the quality of the source is not a problem to test the gain). Then, you'll know if this comes from a difference in the attenuators or the gain in the amplifier. Since, the output of a PC may be quite high if fed directly to the amps, start with the TK2050 and the volume set to zero, then rise it to normal listening volume, then switch to the lower power TA2024. This way you'll be sure you don't overdrive your speakers.
I would try to bypass the attenuators on both amp and plug them in a source that has his own volume control (such as the headphone/line output of a PC, the quality of the source is not a problem to test the gain). Then, you'll know if this comes from a difference in the attenuators or the gain in the amplifier. Since, the output of a PC may be quite high if fed directly to the amps, start with the TK2050 and the volume set to zero, then rise it to normal listening volume, then switch to the lower power TA2024. This way you'll be sure you don't overdrive your speakers.
Checked, it's that ultra cheap eBay attenuator wich has around half steps attenuated to almost max then the real deal on the last steps! Not to mention the hidden layer of resistors being odd and like used, the mismatch on one step...
Will choose something better, maybe even a SMD serie attenuator or a better quality ladder...
No problem in fact, thanks for clues fellows!
Will choose something better, maybe even a SMD serie attenuator or a better quality ladder...
No problem in fact, thanks for clues fellows!
Hi there,
it's in fact more simple than that, adding the poor attenuator's behavior, 10K sounds loud later than a 50K. Maybe it's because more signal is leaking via 10K to ground than 50K? That may sound so simple nobody dare to told me I'm not that smart to even think of it
So here is the question, beside having loud earlier on the pot, any reason to choose 20/50K than 10K? I've read the use here is to go lower than next impedance. And soundwise seems 10K does it better as experienced here, right?
I finished doing some preliminary mods on my amp board and here are my findings :
- Don't use the rotary encoder if you want the best sound, I tried it and wasn't happy with it.
- Do the input cap mod, really big improvement. I installed 2.2 uF obbligatos that I had lying around and it sounded like a different amp. Much more depth, truer sound. Stock I wasn't impressed.
- I installed a connexelectric 300R smps power supply set to 30 VDC. Nice quality, great form factor. It allowed me to use a nice small enclosure. It did improve the sound quality also.
- I picked up a gigawork DACT type 21 stepped 20K pot for around 10 bucks on the net. I am very impressed with the sound quality, sounds at least as good, probably better than the preamp I was using.
- The board is very cheap and easily damaged and the solder that Sure uses is hard to remove so you need to be very careful. I ended up lifting a trace when doing the cap mod, but luckily was able to work around the damage. It gives me pause before doing any more mods like cap replacements.
I auditioned it against two chip amps, a chip amp.com LM3886 based amp and a twisted Pear Rev C. At first it was no contest, both of the chip amps sounded significantly better. After the modifications the Sure amp was much closer, I still need to live with the Sure amp for a while, but for now I'd have to score the RevC first, the Chipamp.com amp second and the Sure amp a close third. I'd also have to say that the noise floor on the chip amps seems lower than the Sure board. All in all this has been a relatively cheap and fun experiment with class D, I'll be continuing my tweaking.
PJN
- Don't use the rotary encoder if you want the best sound, I tried it and wasn't happy with it.
- Do the input cap mod, really big improvement. I installed 2.2 uF obbligatos that I had lying around and it sounded like a different amp. Much more depth, truer sound. Stock I wasn't impressed.
- I installed a connexelectric 300R smps power supply set to 30 VDC. Nice quality, great form factor. It allowed me to use a nice small enclosure. It did improve the sound quality also.
- I picked up a gigawork DACT type 21 stepped 20K pot for around 10 bucks on the net. I am very impressed with the sound quality, sounds at least as good, probably better than the preamp I was using.
- The board is very cheap and easily damaged and the solder that Sure uses is hard to remove so you need to be very careful. I ended up lifting a trace when doing the cap mod, but luckily was able to work around the damage. It gives me pause before doing any more mods like cap replacements.
I auditioned it against two chip amps, a chip amp.com LM3886 based amp and a twisted Pear Rev C. At first it was no contest, both of the chip amps sounded significantly better. After the modifications the Sure amp was much closer, I still need to live with the Sure amp for a while, but for now I'd have to score the RevC first, the Chipamp.com amp second and the Sure amp a close third. I'd also have to say that the noise floor on the chip amps seems lower than the Sure board. All in all this has been a relatively cheap and fun experiment with class D, I'll be continuing my tweaking.
PJN
PJN,
I suspect that you would get a significant improvement from adding a buffer between the 20K stepped attenuator and the Sure amplifier. The Sure amplifier input impedance is 11-22Kohm according to the manual, which would suggest that a passive attenuator should be less than 5Kohm (which your source may not like).
You probably don't want to go down the route of using a lower value attenuator (as this may cause as many problems as it solves) so the best idea is a buffer after the attenuator that has an high input impedance (100Kohm or more) and low output impedance (below 1Kohm).
The Nelson Pass B1 buffer would be an obvious and excellent solution. I am currently building an integrated amp with the Sure board and (original single ended) B1 buffer both running from a 24V Meanwell SMPS. My plan is to have the SMPS and B1 elements powered up constantly with the front panel 'power switch' in line with the 24V to the Sure amp only (this way there will be no power on noises as the B1 warms up). I will connect the output coupling caps on the B1 directly to the Sure amp downstream of the onboard crappy coupling caps (no need to have two coupling caps in series).
I've used the B1 with the Helder TK2050 monos and loved the result. Warmer, richer and more natural than a passive attenuator.
Go for it!
I suspect that you would get a significant improvement from adding a buffer between the 20K stepped attenuator and the Sure amplifier. The Sure amplifier input impedance is 11-22Kohm according to the manual, which would suggest that a passive attenuator should be less than 5Kohm (which your source may not like).
You probably don't want to go down the route of using a lower value attenuator (as this may cause as many problems as it solves) so the best idea is a buffer after the attenuator that has an high input impedance (100Kohm or more) and low output impedance (below 1Kohm).
The Nelson Pass B1 buffer would be an obvious and excellent solution. I am currently building an integrated amp with the Sure board and (original single ended) B1 buffer both running from a 24V Meanwell SMPS. My plan is to have the SMPS and B1 elements powered up constantly with the front panel 'power switch' in line with the 24V to the Sure amp only (this way there will be no power on noises as the B1 warms up). I will connect the output coupling caps on the B1 directly to the Sure amp downstream of the onboard crappy coupling caps (no need to have two coupling caps in series).
I've used the B1 with the Helder TK2050 monos and loved the result. Warmer, richer and more natural than a passive attenuator.
Go for it!
Last edited:
Myself and my brother both have the latest revision which we've installed into our PC's and run off a 12v line.
At first we had a significant tone, quite high in pitch, which we all put eliminated through an additional ground lead.
We now have noise that seems to be enhanced by typing on the keyboard, or when the processor is working.
I have tried a different power supply, at 18v. I don't have anything higher right now. Also, we are running lines direct to the DAC on the Asus DS sound card with nichilon 3.3uF decoupling caps. When we remove the input lines the noise stops. These wires are very short, not enough to be picking up interference.
Same thing happens when we plug into the 3.5mm jack and put the Op-AMP back into the sound card.
At first we had a significant tone, quite high in pitch, which we all put eliminated through an additional ground lead.
We now have noise that seems to be enhanced by typing on the keyboard, or when the processor is working.
I have tried a different power supply, at 18v. I don't have anything higher right now. Also, we are running lines direct to the DAC on the Asus DS sound card with nichilon 3.3uF decoupling caps. When we remove the input lines the noise stops. These wires are very short, not enough to be picking up interference.
Same thing happens when we plug into the 3.5mm jack and put the Op-AMP back into the sound card.
Sounds like the age old problem of noisy PC grounds. Isolating transformers usually solve this. Read post 1822 to 1850 in this thread: http://www.diyaudio.com/forums/clas...path-board-tc2000-tp2050-183.html#post2308431
The two easy ways of getting clean (well, not obviously noisy) audio from a PC are (1) using isolating transformers on analogue outputs (2) connecting external dac to PC optical digital output. Option (2) is usually better and cheaper than (1), but you need a PC with an optical output and an external dac.
Yup, but essentially I'm doing #2 already as the output is coming straight from the DAC on the Asus DS. This is a HQ soundcard which does 116db snr max.
The noise isn't volume related, I stuck a 22k inline (so 44k total using one of the input resistors on the Sure board) and the noise abated to a decent level, but now the gain is too low.
Regards, Andrew
The noise isn't volume related, I stuck a 22k inline (so 44k total using one of the input resistors on the Sure board) and the noise abated to a decent level, but now the gain is too low.
Regards, Andrew
5V caps
Hi there,
latest PCB have the slots for Sure's volume control. 2 pins are on the 5V line prior to the TC2000 and TP2050, a good place for a cap, as the post reg cap is now far away from the ICs. You can choose or just try caps in the pins (plug it off !) or remove the pins and solder it through holes.
My speakers are nice but not the best, and I've added a 10V reg before the LM317 so can't tell what impact the most, but I feel there is an improvement in mids clarity. I've put what I have in hands, a 100µF FM close to pins (and in fact very close to the TC2000) and a 330µF FM post reg. Anyway, with the "in pins try", you can check and flavor to the best sound, free of solder!
Hi there,
latest PCB have the slots for Sure's volume control. 2 pins are on the 5V line prior to the TC2000 and TP2050, a good place for a cap, as the post reg cap is now far away from the ICs. You can choose or just try caps in the pins (plug it off !) or remove the pins and solder it through holes.
My speakers are nice but not the best, and I've added a 10V reg before the LM317 so can't tell what impact the most, but I feel there is an improvement in mids clarity. I've put what I have in hands, a 100µF FM close to pins (and in fact very close to the TC2000) and a 330µF FM post reg. Anyway, with the "in pins try", you can check and flavor to the best sound, free of solder!
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.