NO!!!!
most of the resistors will dissipate <<500mW.
you should check the current and/or voltage for each resistor. Check it's dissipation for worst case loading, not always the highest output current.
Then select the very few resistors that need to be >250mW and use 125mW or 250mW for all the low power resistors.
eg.
the 33r for setting the CCS current. If it passes 150mA, as you want 100mA of output, will dissipate 0.15*0.15*33 watts = 742mW. This could be a 1W single 33r resistor or two 500mW 68r resistors each dissipating 382mW. For 150mA I would probably use three parallel 100r resistors each dissipating 250mW. I would choose any of 400mW or 500mW or 600mW for that duty.
only if the voltage drop across the 33r is 1.98V to 2.145V.
Changing the FET and/or changing the LEDs will change the CCS current.
As an example of the variable nature of our devices.
I have 1.050V and 1.430V across the CCS resistors in a pair of DCB1 pcbs.
Last edited:
Hi Salas,
yes I have three LEDs in each of those strings.
I chose to buy the DCB1 pcbs. The least I can do is compare the standard DCB1 with any modifications I choose to try/adopt.
As I said way back, the voltage across the 5LED string will be more dependant on the diode current than on the selection of Vf. and so it has turned out. 8.6V & 10.2V, a difference of >300mV across each LED. Rather than <5mV when selected by Vf.
yes I have three LEDs in each of those strings.
I chose to buy the DCB1 pcbs. The least I can do is compare the standard DCB1 with any modifications I choose to try/adopt.
As I said way back, the voltage across the 5LED string will be more dependant on the diode current than on the selection of Vf. and so it has turned out. 8.6V & 10.2V, a difference of >300mV across each LED. Rather than <5mV when selected by Vf.
not done a comparison yet.
The standard B1 goes into enormous 2nd harmonic at raised output levels when fed with equal +-supplies.
Juma recognised this and designed his bf862 version to optimise the supply voltages to suit the maximum output levels. You missed that and went for an odd arrangement around that k170. I cannot see how misusing a jFET that way is right, but as above I still have to listen. Maybe this is the better sounding.
Yes I will use the Hypnotize board so I will follow your advice thks.
Can I use the same heatsink listed in the BOM for IRFP9140?
287-1ABE Wakefield Heatsinks
Can I use the same heatsink listed in the BOM for IRFP9140?
287-1ABE Wakefield Heatsinks
This is to take low excess heat for those who would have 15+15VAC Tx for the DCB1 instead of 12+12VAC when still set at about 60mA CCS.
You say you want to power a +/- 100mA application, that means to set at least about +/- 130mA with 33//33R 1W per side. If your application really draws +/-100mA all the time then 0.72W is the dissipation per shunt Mosfet, and about 1W per CCS Mosfet with the Tx you described. Those sinks look like taking it. If its less on the load in reality, no. Consider the option of mounting the Mosfets underneath and fixing them on chassis with insulation pads. That way there will be no problem if there is more heat than we think, and you could go 22R//22R 1W for a nice 200mA setting that lets it behave a bit better subjectively as some members have found in the DCB1's case.
You say you want to power a +/- 100mA application, that means to set at least about +/- 130mA with 33//33R 1W per side. If your application really draws +/-100mA all the time then 0.72W is the dissipation per shunt Mosfet, and about 1W per CCS Mosfet with the Tx you described. Those sinks look like taking it. If its less on the load in reality, no. Consider the option of mounting the Mosfets underneath and fixing them on chassis with insulation pads. That way there will be no problem if there is more heat than we think, and you could go 22R//22R 1W for a nice 200mA setting that lets it behave a bit better subjectively as some members have found in the DCB1's case.
new to the thread
Hello Salas,
Hello all,
I've been industriously reading on your work, regarding the phono stage AND the shunt reg. Awesome stuff! Haven't been posting this far, but as I decided to jump on the wagon, I do have some questions.
I intend to build a shunt reg. for my DDDAC1543MKII, which is now using 8 chips, but will jump in final version to 16. Current draw on this is about 390mA per 8 chips. I assume a value of 850mA, just to be safe. Voltage will need to be 7.2V. For reference, right now I am using a variation of Jung with a blue LED and an LT1034/2.5 as Vref. The op amp is an LM49710. I am extremely interested in a direct comparison between the series supplying and the shunt, as I heard very good things about the latter.
My first question is - is there a PCB for this, hopefully including relay delayed start, that is non-symmetrical? I do understand that I could use just half of the board, but saving on space inside my DAC's enclosure is an issue I need to consider.
I intend to fully separate everything that is "digital" (such as clock, USB receiver, etc.) from the actual "analog" DAC stage in terms of power supplying, and this shunt reg will feed the DAC. The DDDAC PCB design etc. lends itself easily to doing this.
Please let me know if single-rail PCBs exist (I couldn't find any in the threads), and I will follow up with further questions, if I may.
Thank you much in advance for your help!
Cheers,
Radu.
Hello Salas,
Hello all,
I've been industriously reading on your work, regarding the phono stage AND the shunt reg. Awesome stuff! Haven't been posting this far, but as I decided to jump on the wagon, I do have some questions.
I intend to build a shunt reg. for my DDDAC1543MKII, which is now using 8 chips, but will jump in final version to 16. Current draw on this is about 390mA per 8 chips. I assume a value of 850mA, just to be safe. Voltage will need to be 7.2V. For reference, right now I am using a variation of Jung with a blue LED and an LT1034/2.5 as Vref. The op amp is an LM49710. I am extremely interested in a direct comparison between the series supplying and the shunt, as I heard very good things about the latter.
My first question is - is there a PCB for this, hopefully including relay delayed start, that is non-symmetrical? I do understand that I could use just half of the board, but saving on space inside my DAC's enclosure is an issue I need to consider.
I intend to fully separate everything that is "digital" (such as clock, USB receiver, etc.) from the actual "analog" DAC stage in terms of power supplying, and this shunt reg will feed the DAC. The DDDAC PCB design etc. lends itself easily to doing this.
Please let me know if single-rail PCBs exist (I couldn't find any in the threads), and I will follow up with further questions, if I may.
Thank you much in advance for your help!
Cheers,
Radu.
Nice.
(thank you for the welcome
Per my numbers in my initial post, my Rset will need to be in the vicinity of 1.2ohm (2.2V/1.8A), and say, 10W, is that correct? I assume soe hefty HS would be called for for such heavy draw.
I think I need to talk to Tea-bag about buying some boards in the US, right?...
Thx,
aR.
(thank you for the welcome
Per my numbers in my initial post, my Rset will need to be in the vicinity of 1.2ohm (2.2V/1.8A), and say, 10W, is that correct? I assume soe hefty HS would be called for for such heavy draw.
I think I need to talk to Tea-bag about buying some boards in the US, right?...
Thx,
aR.
Not really a big problem is sinking for you, due to the low voltage you need. If you run it at 850ma there is 460mA left to burn, times 7.2V = 3.3W on the shunt Mosfet. If with 15V input, then your CCS Mosfet will be double hotter bcs it has to burn all mA times drop. Use 1C/W sink and fasten both Mosfets on it with insulation pads.
P.S. Yes, talk to Tea Bag, I have nothing to do with supplying boards.
P.S. Yes, talk to Tea Bag, I have nothing to do with supplying boards.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.