If you use the original low current setting 68R//68R and 12+12V mains transformer it can live without sinks as in many builds done so far like that. Individual small sinks that don't touch each other or any other metal part can be used even without insulation anyway. Its safer to use such if with a 15+15V transformer because it gets warmer, but still with the low current 68R//68R. No accidental short with wire, crock clips or probes must be done between them or grounded chassis if used so. For bigger ''hot rod'' currents you need a metal surface for L style Mosfets mounting, be it a thick panel or a chunky common sink. Insulation also is needed if on common metal between Mosfets as you know, like in Fran's example. TO-240 insulation pads you can find in any electronics store be it online or local.
Me neither have any hum, there must be a flaw in your implementation.
I'm glad we have a so valuable preamp for that ridicule cost. Less than 100€ for me, including trafo and case.
You don't need to insulate them, just be sure the heatsinks are not touching anything around them. I had little troubles specially with the ones closer to the CCS setting current resistors.
If you are not tweaking the CCS, just using the default resistors, you don't need heatsinks. The mosfets stay barely warm.
Hi
why do you prefer wima mkp 10 0,22 uF to bennic mkp 4,7 uF
tight space or sound
Hi samoloko,
the experiments started about page 106 or so (I think) but for my specific comments about the wima vs the bennics (which have performed very well for me before) see here:
http://www.diyaudio.com/forums/pass-labs/145201-building-symmetrical-psu-b1-buffer-113.html#post2118544
Fran
the experiments started about page 106 or so (I think) but for my specific comments about the wima vs the bennics (which have performed very well for me before) see here:
http://www.diyaudio.com/forums/pass-labs/145201-building-symmetrical-psu-b1-buffer-113.html#post2118544
Fran
Just a data point to report, I tried WIMA MKP 10 0.1uF 400VDC for about 3 weeks, tonight I thought the sound to be not as good as it used to be, so I put the original 100uF electros back in, which are Suntan CD286 105C low ESR from a local supplier, Jaycar in Australia. The sparkle and liveness of the music has returned. Maybe the 400VDC rating doesn't help on the WIMA's I used, and my system is modest, but the 100uF electros win over the 0.1uF in my case. Maybe larger value film caps are the way to go but I have none on hand at the moment. My DC B1 is using all garden variety components, so that may have something to do with it too. It's always good to try new possibilities because sometimes the improvement is well worth it, and if something is not tried and tested then we never know. Seems like maybe a larger value film cap is worth testing, as some people have been doing. The trouble for me is, the DC B1 buffer sounds so good as it is, and it is easy to hear a forward improvement in sound, but more difficult for me to hear if it goes backwards a little bit. That's why it takes 3 weeks for me to "wake up". Not unusual for me 
I am not sure which will work best.
member BLUES has matched set's for sale, and I would recommend him as a supplier for those. ampslab.com also sells the bl's still at a buck a pop unmatched.
Not from experience, never used GR or V for DCB1. Violet grade transcoductance is good and drops THD but also presents more input capacitance. In your shoes I would try tight V audio pairs (when Idss is strong slight mismatch manifests offset easier) and GRs elsewhere so not to overcook the Leds, they will get noisier. I would use a 10k Log if my source allowed when driving V Ciss.
Did anyone use the Fairchild equivalence of the IRF MOSFET?
I am using the Fairchild and having a hard time to set the B+'s current. I only got 0.7V across R1. The B- side is in a better shape, measured 2.12V across the "stock value" R1. However, the B- is only at -9.6X (don't quite remember the exact number).
The voltage of all the LED strings are on the mark of 5.4 and 9.0 +/- 0.01V. Is it possible that the Fairchild's may have a different bias point at this Vgs 5.4V? I couldn't tell from comparing the curves of the FQA12P20 and IRFP9240. Does anyone want to give me a hand?
Thanks!!
I am using the Fairchild and having a hard time to set the B+'s current. I only got 0.7V across R1. The B- side is in a better shape, measured 2.12V across the "stock value" R1. However, the B- is only at -9.6X (don't quite remember the exact number).
The voltage of all the LED strings are on the mark of 5.4 and 9.0 +/- 0.01V. Is it possible that the Fairchild's may have a different bias point at this Vgs 5.4V? I couldn't tell from comparing the curves of the FQA12P20 and IRFP9240. Does anyone want to give me a hand?
Thanks!!
Sorry to be a pain I've spend ages looking through this thread to find the most recent prints for a home etch and given up, could someone please link me to the post or pdf?
Also can anyone comment on how the onboard psu compares to the "teddypsu" or "superteddypsu" from Pink Fish forums, I was going to use it for the lm318 in myref amp but would it also be a better psu here?
OR
Can this psu be used to power the lm318 on myref to save building the PSU's I mentioned above if performance is on par?
Are there any sections that can be omitted if used as an internal buffer to myref?
I'd be really grateful for any replies as these are my last remaining questions before I finalise my design.
Thanks
Also can anyone comment on how the onboard psu compares to the "teddypsu" or "superteddypsu" from Pink Fish forums, I was going to use it for the lm318 in myref amp but would it also be a better psu here?
OR
Can this psu be used to power the lm318 on myref to save building the PSU's I mentioned above if performance is on par?
Are there any sections that can be omitted if used as an internal buffer to myref?
I'd be really grateful for any replies as these are my last remaining questions before I finalise my design.
Thanks
Last edited:
Posts 14,32,39. Then a GB was done (had its own ''heroic'' story in the GBs section) with an output delay included, plus an alternative multi input version.Those are exclusive Gerbers. If there is a capacitor input to drive, no delay/fast release is needed, no danger.
If you are interested in the reg only, this is not the best thread. This one is for those building on the GB pcbs full buffers.
Technically a shunt reg has much better and flatter Zo than a cap multplier.
If you are interested in the reg only, this is not the best thread. This one is for those building on the GB pcbs full buffers.
Technically a shunt reg has much better and flatter Zo than a cap multplier.
Thanks Salas, it is an excellent call to reduce the set resistors. After I spent the lunch time to compare the transfer characteristics curves, I also came to the same conclusion. It looks like he FQA12P20 need a little over 4.5V Vgs to turn on whereas the IR can do 0.5V sooner. I do have some IR MOSFETs, but they are matched in triple. So, I am trying to save those for other projects
I think I have a 50R 10W wirewound pot in the parts bin. I need to dig it out tonight so that I can experiment what R to use to // the 2 x 68R and don't need to solder and desolder too many times. The pessimistic side of me is telling me this is not the end of the story .....
I think I have a 50R 10W wirewound pot in the parts bin. I need to dig it out tonight so that I can experiment what R to use to // the 2 x 68R and don't need to solder and desolder too many times. The pessimistic side of me is telling me this is not the end of the story .....