Generic snubber values: fixed-/variable-voltage regs

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
LED sub for R2 or R4 for pretracker

Thx for your reply ...
martin clark said:
Yes, the voltage-set mechanism is not affected by the curent drawn from the regulator.

With the SAA7220 a tracking-pregulator setup using two LM317 works very well IME.
I'm glad that you noted the tracking-pregulator specifically above because that's what I'm using. Your acoustica article mentions the tracking-pregulator after the note on LEDs, so I didn't know the recommendation still applied: in the datasheet pre-tracker, R4 is in the same position as non-pretracker's (std LM117 schematic) R2. Or did you mean R4?

Just to be clear, when using the two in-series LEDs for 5VDC out ...

- do the other resistor (R1, R3-4) values remain as given in the datasheet?
- does the 10uF Tantalum cap across R2 tweak you noted here still apply for the two LEDs -- i.e., across the two green in-series LEDs?

And while I've got you you here, one more question about the tracking pre-reg:

Any recommendations for total rectifier capacitance**?

**For a 5VDC-output PS+reg, the rectifier math works out to ~5000uF (Horowitz, p 330). But ... (a) that sounds like overkill (esp. with modern components and topological techniques and "shortcuts"; the Horowitz text is 20 yrs old!); and (b) I don't have that much room in my PS box! Snubbers may help as might soft+fast-recovery diodes and damper resistors.
 
Minimal Reactance Power Supply

Hi Eric,

I really like the idea of your Shared Current Power Supply. I totally agree that reducing the capacitor size sounds better and I developed such a supply for a tube preamp that I call the Minimal Reactance Power Supply (MRPS).

You mentioned in your article that you have a problem with noise. You might want to experiment with replacing the series regulator with a shunt regulator. I have designed several low-voltage MRPS using an LM317 as a constant current source followed by a TL431 shunt. All of the designs were silent.

Here is the original article http://www.raleighaudio.com/chapter_5.htm which doesn’t need a shunt following the CCS because the load is also constant current. Here is an example of a high-voltage MRPS with a shunt http://www.raleighaudio.com/cps_schematic.htm The input capacitor for both designs was 3.6uF.

Dave
 
Hi David,

Many thanks four you very interesting article. Absorbing 100V undulation without rumble is very interesting. I will try your MRPS. What capacity do you use to smooth tension (2uF?).

On manufactured product you can find :
- 100uF on Audio Research SP11,
- 60uF on Conrad Johnson Premier 10,
- 40uF on Conrad Johnson PV11, PV12...

A first (and simple) application of the shared Current Power Supply is entry point of rectifier. I catch this tip on this DIY forum about power supply. One person say that He connect rectifier near amplifier and not on the filtering capacitors. This point don't seems to find any interest!
This wiring greatly improve sound : more details (less foggy sound), sweet and softness, depth bass. I implement in my Model 4 preamplifier.

I give comprehensive schematic on attached picture. To understand differences, draw current during charging and discharging capacitors.
 

Attachments

  • wiring_power_supply.gif
    wiring_power_supply.gif
    3.5 KB · Views: 430
Hi Eric,

The size of the capacitor is determined by four things: First is the power line frequency, which determines how often the capacitor gets charged. Second is the input voltage level, which determines how high the capacitor is charged. Third is the amount of current drawn by the load, which determines how much the capacitor is discharged. And forth is the desired output voltage, which determines how low the input peak voltage can be.

The schematics shown in the earlier post have a 250V center-tapped transformer so the input voltage is 350V. The load for that preamplifier is 30mA plus a little more for the shunt, say a total of about 40mA. The output voltage is 214VDC. The charging frequency is either 100 cps or 120 cps depending on your power source.

With that, 1.8uF was fine for a 60Hz power line but not enough for a 50Hz power line. I did not have a 50Hz power source to test so I tried adding a second 1.8uF capacitor in parallel and that worked fine. The actual value needed is probably about 2.5uF for 50Hz. You could put together a formula to calculate the size of the capacitor but I find it easier just to make a few measurements.

If the capacitor is not big enough, you will see a notch in the output voltage at the charging frequency. The notch is caused by the negative peak of the input ripple dropping below the level of the output. The capacitor needs to be only big enough to keep the negative peak of the ripple from dropping below the level of the output. It doesn’t help at all to make the capacitor any larger than it needs to be – in fact it hurts because the charging pulse is shorter.

Your comment on the importance of the entry point is interesting. I know that the resistance of a few inches of wire can cause a problem if you get the routing of the ground wrong. I had not thought about the importance of the routing of the voltage, but of course it is just as important as the ground.

Dave
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.