New TK2050 board

Status
Not open for further replies.
Jump to Latest
#961
Questions

I have a few newbie questions, if someone would be so kind to answer.
Can I put the v1 amp in a wood case? Is it possible to severe the on/off switch from the board and connect wires to the switch, once atached to my new casing? Same with volume pot? Internal wiring for the rca and speaker terminals: Can I use speaker cable for the speaker terminals? What sort of cable is ok for the rca terminals? I bought the HP C7296-60043 AC/DC sps for my power supply and would like to use the connected cable with a dc plug at the end. Does anyone know, which kide of dc jack would I need for it? I know I have asked a lot of questions and I have already scoured several threads to find the answers and haven't had any luck.

Any help would be really appreciated. Thanks
 
#966
benoar said:
OK, thanks for these informations. Nice to knom the relays resistor will be changed. But will you adapt the feedback resistors too, as I wondered earlier ? If I go for 36V Vcc, with 1kOhm Rfbb we must have 17kOhm Rfbc. Or maybe another combination since 17kOhm may not be easily available. This will also increase a bit the gain; should I also compensate for it by decreasing Rf ?

About the gain, I was always puzzled looking at the 1.5V input sensitivity of TC2000. But now, I may have found why: the datasheet says the output of TC2000 should be driven up to 4Vpp. With a unity gain on the input stage, which is biased at 2.5V this leaves a 1.5V headroom, our input sensitivity ! Am I wrong on that ?

This make me think : with common 2V line input voltage, shouldn't we decrease a bit the input gain ? Like making Rf 15kOhm ?

Maybe you already thought about that, I am just trying to understand better your schematics…
Click to expand...

Right, we will change the Rfbc as we do on T3. 1.5V is the maxium input sensitivity of TC2000, make sure not exceed this voltage will keep THD low. And we will get 1.5V*15=22.5V with Rfbc=14kohm and Rfbb=1kohm, that's enough for 100W@4ohm output. And we always suggest a potmeter or an attenuation before input to limit the input signal for TC2000 not too big, so it's not necessary to change feedback resistors.
 
#968
Hi

I hope that this is not a rude question for this thread:

Has anyone compared this amp module to the Hypex units? I am torn between the hifimediy TK2050, and the Hypex: 2x UCD180hg.

The prices are wildly different, but I have heard great this of the hifimediy unit.
 
#970
Hi Steinar and Hifimediy,

One bug on my V1.2 board: the power supply relay stay stuck even when the switch is open.
After knocking it a little bit, the contacts opened.
During the failure, I could hear the relay moving when switching on and off but the ampli was still playing.
After the capacitor that explode on V2.0, I'm starting to think there are some quality problem on the component choice. Is the relay correctly sized? It is given for 12A under 125 VAC but how much under 31V DC current?

Please give advise on what to do as this board is planned to be a christmas gift...

Eric
 
#972
First, my assumptions on the 1.5V sensitivity and 4Vpp output voltage were completly wrong. The input stage is made for 2V line level, with a unity gain. I read somewhere that some Tripath amp have a 1.5Vrms input sensitivity, which matches the 2V line level and 4Vpp voltage swing.

hifimediy said:
Right, we will change the Rfbc as we do on T3. 1.5V is the maxium input sensitivity of TC2000, make sure not exceed this voltage will keep THD low. And we will get 1.5V*15=22.5V with Rfbc=14kohm and Rfbb=1kohm, that's enough for 100W@4ohm output. And we always suggest a potmeter or an attenuation before input to limit the input signal for TC2000 not too big, so it's not necessary to change feedback resistors.
Click to expand...

You use the gain (15V/V) to prove that the resistor values are right, but the gain is actually deducted from these resistor values (in the datasheet), so of course it's a match ! What I am saying is that according to the datasheet calculations, Rfbc should be higher. If I choose Vcc=36V, the Rfbc/Rfbb voltage divider with 14k/1k Ohms will give a maximum 2.34V on the feedback circuit positive input, which gives a total voltage differential (4.68Vpp) exceeding the recommanded 4Vpp. We will saturate the feedback circuit. Of course, this only happen on maximum input, but in the the other (lower input) cases, even if we don't saturate, we give too much feedback and worsen the THD.

BTW, I wonder if we could also imagine a TC2002-like feedback topology with a 1k pull-up to 5V and different Rfb calculations, but this is another story (in the TC2002 it says it is pin-compatible with TC2000; I wonder if it means we can use the same topology for TC2000 or not, as I suppose the more recent TC2002 feedback circuit is better engineered).

Another question, not about the feedback circuit, but the dimensionning of other components : why only a 10 Ohms resistor in the Zobel network ? Datasheet says 15 Ohms. What's the reason ?
And why only 11mH coil in the output filter ? Is it only a price consideration or is there another reason ?

Thanks for those informations !
 
#973
benoar said:
First, my assumptions on the 1.5V sensitivity and 4Vpp output voltage were completly wrong. The input stage is made for 2V line level, with a unity gain. I read somewhere that some Tripath amp have a 1.5Vrms input sensitivity, which matches the 2V line level and 4Vpp voltage swing.



You use the gain (15V/V) to prove that the resistor values are right, but the gain is actually deducted from these resistor values (in the datasheet), so of course it's a match ! What I am saying is that according to the datasheet calculations, Rfbc should be higher. If I choose Vcc=36V, the Rfbc/Rfbb voltage divider with 14k/1k Ohms will give a maximum 2.34V on the feedback circuit positive input, which gives a total voltage differential (4.68Vpp) exceeding the recommanded 4Vpp. We will saturate the feedback circuit. Of course, this only happen on maximum input, but in the the other (lower input) cases, even if we don't saturate, we give too much feedback and worsen the THD.

BTW, I wonder if we could also imagine a TC2002-like feedback topology with a 1k pull-up to 5V and different Rfb calculations, but this is another story (in the TC2002 it says it is pin-compatible with TC2000; I wonder if it means we can use the same topology for TC2000 or not, as I suppose the more recent TC2002 feedback circuit is better engineered).

Another question, not about the feedback circuit, but the dimensionning of other components : why only a 10 Ohms resistor in the Zobel network ? Datasheet says 15 Ohms. What's the reason ?
And why only 11mH coil in the output filter ? Is it only a price consideration or is there another reason ?

Thanks for those informations !
Click to expand...

Hi, thanks for your valuable information to us.
And about the zolbel network, it's ok for 8.2~16ohm for our amplifiers, and I always use 10ohm on it. For coils, we can get single layer wounded ones if we choose 11uH (not bigger), it's better from this side.
 
#976
So will these issues be fixed in a new revision?
I have been contemplating 4 pcs for 7.1 surround of the amp v2 with STA510A, but will use the amps on 4 ohm speakers (so the zobel is incorrect for this use?).

Regarding the feedback circuit... will it work as intended if I feed the amps 30-32v?

Will the relay be upgraded on amps from now on?
 
Status
Not open for further replies.
Top Bottom