Simple, no-math transformer snubber using Quasimodo test-jig

FA, the snubber tuning process applies a step load change to the identified PT secondary winding, and identifies the response of the winding's internal energy and related leakage inductance and capacitance when loaded by the tuned snubber. The jig makes the step load change as ideal as practicable, along with making the identified transformer winding appear like just its leakage inductance as much as practicable.

The operation of rectifier parts/circuitry is not as ideal as the jig's operation. How you set up the rectifier and filter parts/circuitry is up to you.

What noise are you referring to?
 
  • Thank You
  • Like
Reactions: 1 users
A perfect diode with no Vf, that transitioned from off to on, and on to off, in zero time and with zero resistance when on would effectively cause a step load change on the transformer secondary winding and so the winding leakage inductance would ring around through the winding shunt capacitance, and externally through the snubber. A diode is not perfect and subtle effects then come in to play.

A nicer diode is a valve diode in a high voltage power power supply for valve amps. The diode's on resistance increases as diode current falls to zero, which avoids a step change in winding current, and is helped by the winding having a substantial series resistance and shunt capacitance.

In between are ss diodes and lower voltage power supplies for ss amps that impose higher current loading.
 
Been using this for a while

the asc capacitors polypropylene at allied electronics are great

I always snubbed with the .01uf
I then tried the .047 asc instead and it eliminated some jaggedness in the trace --bad sounding higher order harmonics
Only the cap with low est and no resistor can clean up this

The cap across the secondary with no resistor cap

Then for the resistor cap I got away with a .1uf asc cap which violated mark rule that you wanted it at least 10 times larger worked great

Don't you want to use the smallest cap you can get away with here

The cap across the secondary with no resistor is not tuned so can rob the music of dynamics particularly in the highs so I would rather use the .01uf
Then I wonder if you might want to use a higher wattage resistor

The actual circuit may have much larger current pulses than the quasimotyo would generate?


I am no engineer so would like mark or someone smarter than me to respond
 
Been using this for a while

the asc capacitors polypropylene at allied electronics are great

I always snubbed with the .01uf
I then tried the .047 asc instead and it eliminated some jaggedness in the trace --bad sounding higher order harmonics
Only the cap with low est and no resistor can clean up this

The cap across the secondary with no resistor cap

Then for the resistor cap I got away with a .1uf asc cap which violated mark rule that you wanted it at least 10 times larger worked great

Don't you want to use the smallest cap you can get away with here

The cap across the secondary with no resistor is not tuned so can rob the music of dynamics particularly in the highs so I would rather use the .01uf
Then I wonder if you might want to use a higher wattage resistor

The actual circuit may have much larger current pulses than the quasimotyo would generate?


I am no engineer so would like mark or someone smarter than me to respond
 
I have a small additional question regarding the schematic of the THT Version:
There is written: "Vcc Power supply +2.6V to +18V"

but the LMC555 (from TI) has a absolute max. VCC of 15V
and the TC4420/TC4429 or MCP1406/07 (from TI) have a recommended min. Operating Voltage of 4.5V

and is the FETs gate capable of 18V ?
or the other way around: does the FET "fully conduct" with only 2.6V on its gate ? ( to give the wanted "exremely low Zout (less than 0.01 Ohms)"" )
 
Last edited:
Member
Joined 2011
Paid Member
Most people just use a 9V battery or a 5V power supply.

I enjoy powering QMs from a lab power supply and dialing the voltage up and down while observing the scope trace. The ones I've built work fine below 4.5V and above 15V. However, you get the exact same optimum snubber result at all supply voltages, so using what's most convenient (9V battery) is often the best idea.

Arrow sells the MOSFET part number shown in the Quasimodo V.4 schematic, for $0.26 (link to sales page). I've attached its engineering datasheet below.

_
 

Attachments

  • NTD4906N.PDF
    126.1 KB · Views: 107
I think it should be a 555 in CMOS technology to make use of the feature "output capable of swinging rail to rail".
from the Quasimodo docs:
The 120Hz oscillator uses a CMOS 555 with rail-to-rail output swing, for
the same reason: a non-CMOS 555 chip’s output only swings to VCC – 2VBE, which spoils low
voltage operation.
(and the NE555P is IMHO not in CMOS technology...)

I was in the same problem not to find a CMOS version as a THT part at the moment,
therefore I ordered a SMD [SOIC-8] part (https://eu.mouser.com/ProductDetail/595-LMC555IM/NOPB)
and an additional breakout board "Adafruit 1212" (https://eu.mouser.com/ProductDetail/485-1212) [https://www.adafruit.com/product/1212]
 
Last edited:
  • Like
Reactions: 1 user
Member
Joined 2011
Paid Member
Think of it as a teachable moment: different companies sometimes use slightly different part numbers for the same basic chip. So it's worth looking in DigiKey's Parametric Search for other chips with the same features (here are more than forty) , and now that you know the part numbers, check octopart.com for availability. You'll want to avoid the non-CMOS chips, of course.

And oh by the way, the thru hole BOM attached to post #1 in this thread, does suggest a substitute part number for CMOS 555. It's been attached there for more than eight years, anyone interested can read it whenever they wish. Fortunately, Newark has (610 of them on the shelf) . Now there are than one possibility: ICM7555, TLC555, et al.

_
 

Attachments

  • tlc555.png
    tlc555.png
    10.1 KB · Views: 79
  • Like
Reactions: 1 user