DIY Audio Projects Forum

hi, i noticed this 5 volt regulator which seems to deliver a fixed 5 volts rather than needing to adjust it. has anyone had any experience using this to heat the 300b’s or other such valves? Does this have the right specifications to do the job? if so are there any benefits/disadvantages of the lm323t over the lt1085 used in this thread? thanks, woodo
https://au.mouser.com/datasheet/2/389/C ... 470641.pdf
edit: although i see it says voltage is adjustable

_________________
John

Actually there is a big disadvantage; the dropout voltage.

The old LM323 design has a rather large dropout voltage which is exacerbated by increased junction temperature. There is no way a rectified 6.3VAC transformer, with rectification loss, and appropriate filtering will provide the necessary input voltage to maintain a stable 5.0v output using this regulator with appropriate design margin. Regulation margin would be severely compromised. I chose the LT1085 (new data sheet location: 108345fh.pdf) for this design due to it being first and foremost a low dropout regulator (>1.2v at Ia=1.2A) and it exhibits much better long term stability than the LM series regulators. Regulators like the LM323 require a higher input voltage, hence dissipate far more heat, and hence require a more comprehensive thermal design solution.

In post #127 to this thread (http://diyaudioprojects.com/Forum/viewtopic.php?f=9&t=2699&start=127) I discussed at length, all the design requirements that went into the 300B filament supply design process. I would strongly suggest that anyone wanting to change the filament supply design read this post so that they understand the various tradeoffs involved.

As always, questions and comments are more than welcome.

_________________
Matt
It's all about the Glass!
http://www.CascadeTubes.com
Cascade Tubes Blog

Great, thanks for the advice Matt.

_________________
John

Hi, has version of this amp been built? if so, how is the sound. thanks.

_________________
John

I purchased a Music Angel 300B in Hong Kong in 2010. When I tested it on the bench, it was not very impressive, which was not a surprise and I bought it with the intention of using it for a "test bed". The price was reasonable, so I bought it.

As it is today (per the schematic), it sounds very good. It is very quiet, dynamic and overall has a good tonal quality to it.

Some of the voltages and part values may look odd (i.e. choke values, etc), but that is because that is what was in the amp when I bought it. I threw out all the stock resistors, caps (except 2) and wire.

The original design was a 3 stage integrated amp design. I don't usually build integrated designs for myself, so I turned this into a 2 stage design for use with a preamp.

When new, the amp used 450V capacitors in the power supply for the CLC section. When the amp was turned on before the output tubes would start to conduct, the voltage on the CLC section would rise right up to 450V for a few seconds and then drop. The two 68 ohm rectifier plate resistors reduced this voltage at start-up. Down the road as tinkering progressed, I bought all new caps, but never took out the rectifier plate resistors. Since the first cap is 56uF, those plate resistors limit current just enough that the 56uF cap does not cause any issues.

In the filament circuit the two 6800uF caps are originals and looked like they were decent, so I continued to use them. If you are going to build this 300B design, I would use at least 10,000uF. They are only a little more than the 6800uF's. I used LM317K's (T0-3 case) since I have a good stash of them.

The SRPP stage easily drives the 300B to 7 Watts RMS output @ 5%THD. Depending on the output tubes, your mileage may vary. I am using standard Shuguang 300B's with brown base and I have a mix of Chinese, Russian and USA NOS 6SL7GT's. The 5Z3 is a USA NOS RCA. The amp came with an octal 5Z3P (Chinese 5U4G), but it shorted out within a short time so I swapped out the octal socket for a four pin socket.

The James transformers put the final touches on the amp... they are very nice transformers. Too bad they stopped making this series. If I had known they were going away, I would have bought a few more pairs.

Signal to Noise Ratio @ 1 Watt unweighted.

I went ahead and bought the components for this regulated supply for my pcb based amp which doesn’t include the lt1085 circuitry while ordering components for a future 12au7/6as7 amp of Matt’s design. I am planning on making some other mods to my 300b amp in line with the monoblock design in this thread (gulp). Would it be ok to use the perforated board in the prototype of @suncalc or some other means of putting together this circuit? It’s difficult to see exactly how the circuit was assembled by @mwhouston from the pics. May have been soldered to terminal strips? Any advice appreciated. Regards...

Edit: my apologies I found Mark's pics of underneath the silver dragon showing the detail

_________________
John