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

These should correct the errors to post# 1439:
Cs = .47uf
Cx = .047uf
Rs = 154 ohms

When I went to verify it on a different day, the middle pic from Post #1439 was different. That image showed a vertical front with an almost 90 degree bend going across the screen. It was over damped, not good.

I went through the process again and this time the capacitor values were different:
Cs = .047uf
Cx = .33uf
Rs = 154 ohms

Why it likes the larger value of Cx I don't know. If anyone cares to suggest answers I'm listening again.
 
Member
Joined 2011
Paid Member
I've never tried to Quasimodo a transformer that was installed inside the final electronic equipment and connected to the rest of the circuitry. My intuition says this arrangement would probably screw up the measurements but my intuition has been wrong many times in the past.

If it were me I would either remove the transformer from the gear and Quasimodo it all by itself on the lab bench, or else, figure out how to modify the circuit design of Quasimodo itself so that its stimulus pulse is less than 0.5V (a diode drop). Then lash the new circuit together on a solderless breadboard, debug it on the bench, and finally try low-amplitude Quasimodo stimuli on the transformer-inside-equipment. No guarantee this would work, but it may be worth a try.
 
I've never tried to Quasimodo a transformer that was installed inside the final electronic equipment and connected to the rest of the circuitry. My intuition says this arrangement would probably screw up the measurements but my intuition has been wrong many times in the past.

If you open-circuit the winding under test and short all the others, I don't see how it can possibly not work.
 
It's a new day in Q-Modo- Land, with fun and games on the ARC SP3-A1. :)

What you must might beable to see in post# 1439, pic#5 if you look closely at the outlet wiring between brown #2, and green #3, you will see a tacked on piece of wire that shorts the primary 120V AC line.

Now that I tried the easy way out...(and exhausted those possibilities) I'm grateful that MJ stopped by and made a few suggestions.

Aggie Sync finally got smart and removed ALL three secondaries. It didn't kill me either. I should have done it with the first post of the ARC SP3-A1.

There are still questions that I have regarding the oscillations after the bell toles. The part I understand is damping the ringing. Do we go to the point of a 90 degree right angle? or to the point where the damping just flattens out.

Unlike MJs images he's shared with us in the modo guide and other threads...there is a point where 1. the damping starts 2. the ringing starts to flatten 3. the flattening continues 4. the ringing gets lumpy 5. the ringing gets more distortion-y?

6. the ringing increases again 7. until it looks like a smoother ring (only 3 to 6 minor mounds).

From what I have gleaned from doing this for too many hours

and 5 pages of lab notes, I believe the target image lies somewhere between 3. and 4. that I just listed.

Target is an almost smooth line, though it may contain small humps but if kept dialing in the Rs...the ringing or lack there of gets lumpy and the ringing starts to increase again.

I don't thing we want lumpy...I think we want nice smooth humps.

Also noted when Rs is dialed too far, it almost looks like we want to add more capacitance to the Q-Modo because it is no longer smooth. at the start of the ring curve. I've got too many pics but I'll try and post a few later that show what I'm describing.

Also note that in my case my old analog scope is at it's limits. This is puzzling because where the probe's BNC plug in it states, "1M ohm 20pV <- 400Vpk. My scope starts whining to me somewhere between 10Vpk - 14Vpk. I would really like to zoom in a bit more cut don't want to damage it when "ouch" comes on.

cheers
 
Board going quick, only 4 left.



I’d love to have one of the Quasimodo boards as well!

Also, a friend of mine loaned me his SMD Quasimodo board, but I’ve been having some problems getting it to work (at all).
It is hooked up per the guide. I am using a dual channel Tek 465b, a bench supply set to 12v to power it. Both transformers I’ve tried it on are dual secondary; one is dual primary. Primaries shorted, other secondary shorted. The trigger waveform looks fine but I’m not getting any output. If I disconnect the secondary from the Quasimodo, the output matches the input. Connected I get nothing. I removed the pot, still no change. I’m thinking it is the STS chip (don’t recall the number) which drives the output.
Any thoughts?
 
Last edited:
Remove the trimmer (giving Rs = Infinity) and try it on an inductor whose value is known; anything greater than 50 microhenries. You should see a waveform similar to Figure 12 in the Quasimodo design note. If you don't, apologize to your friend for destroying his Quasimodo V3.
Already tried that (minus the inductor, don't have any). My friend is already aware of it, didn't work right off the bar.
I'll replace the STS12N3LLH5 and possibly the TC4427ACOA and go from there.
Question: when you guys discuss Cs and Cx, do these equate to C2 (series) and C3 in the schematic? I've read some of the thread but not all of it.
 
Hi Pars,

Please read the design document that Mark wrote. Think about what the terms Cs means? Would it go together with the Rs?

But then again I have the Q-Modo V4 through hole.

I haven't breadboarded my Q-Modo for .5Volts, but I do have a question.

How fast, or what is the rise time of the square wave?

Pulse? Maybe I can use it as a Pulse Generator, PG? Now I'm wondering if I can build the PG cheaper than buying the old TEK PG506 for example.

I'm learning one way to adjust the square wave hight is simply dialing (adding) more volts to the board.

Post #1439 shows a measly 9V output of my power supply, but that is cranking out 500mA.

I"ve also gotten sloppy with the dialing up the volts to 30 something. The scope didn't like it much but Q-Modo didn't seem to mind. Of course I didn't leave it there for long.

I hope that helps, also let me know if any of the scope pics help.

Cheers
 
Member
Joined 2011
Paid Member
... try it on an inductor whose value is known; anything greater than 50 microhenries

For those who might be intimidated by the idea of choosing a number greater than 50, I recommend re-reading post #487 in this thread.

For those who are using an older, low bandwidth oscilloscope: If the bandwidth of your oscilloscope is "B" Hertz, select an inductor that gives fosc less than (B / 10) Hertz. This guarantees that the waveform you are trying to display, will be comfortably slower than the bandwidth of your scope.
 
Yes, MJ, that is easy for you to say.

But for us mere mortals...the appeal, and even for you omnipotent type, the appeal of no-math snubbing is worst than the ancient sin lust...it is overwhelming! Perhaps I've succomed [(sp) unable for spell check to find] to the sin of pride even.

So be it, I shall continue.

My old general purpose scope is a measley 100MHz. If I recally my recent EET courses that would be 100 e6 or 10 * ^7 (the extra zero must be accounted for).

Plugging the values into MJs equation we want less than B/10 Hertz.

where B = 100MHz, 10 still equals 10.

100,000,000/10 = 10,000,000 Hertz 10MHz.

Lets find fosc.

a. Google says butt splice for fiber. NOT.

from MicroChip: fosc = Frequency of the OSCillator. and also tsoc = Time of the OSCilator. = 1/fosc (inverse function on calculator).

Recall we want a number greater than 50uH. I'll go try and find on, I"m sure I have some laying around somewhere. I wonder if it might be too large though...we'll see.

We also have to find the formula for calculating xHenries to Hertz. The search continues...I'm sure it's just around the corner somewhere in reference.
 
Member
Joined 2011
Paid Member
I recommend post #487 in this thread. It was mentioned this morning, in #1451 above.

If scrolling to a certain post# is inconvenient, perhaps scrolling to a certain date might be easier. The date on post #487 is 16th November 2014.

The attachments are illuminating.

_
 
Last edited:
Hi Pars,

No worries. It's funny that when we write to ask a question etc, and post it, the answer will at times just pop into our heads. It happens with me too and I'm sure many people.

We call this constructivist learning, where we experience something then cogitate about it; constructing our understanding and knowledge of the the subject etc. Asking questions helps us whittle away the gray areas that we are unsure about. To that end we can delve right into Socratic Teaching or Learning.

Now, if I can find those pics I was going to post here....

Cheers,
 
Hi Mark,

Can you please remind me what brand/model of affordable trany or coil can be used as accuracy reference with classicCx and Cs (100nF and 150nF) caps? I need one of these just for my sanity check with known-tested Rs to repeat it on my tester. Thank you in advance.
My apology. I just realized that ended up with an extra “0” for Cx. To remind you, I’m looking for your known cal/sanity-check coil or transformer with Cx-10nF and Cs-150nF Rs. Transfers from manual are obsolete parts now and hard to find. Thank you.
 
I ordered my boards last month, I built Quasimodo 2 weeks ago, this weekend I've finally gotten around to setting up and testing the jig on one of the coils on my transformer. What a neat little project!

With Rs = infinity (i.e. no resistor), my waveform looked like the following:
Undamped.png

I wound down the trimpot (it took ages to see any result), and eventually snubbed the ringing like so:
Critically damped.png

I measured the trimpot: 19.8 ohms! Seems a bit low to me, though the trace is hard to argue with. Should I perhaps try other values for Cs? I've yet to try it at other frequencies; my DIP switches haven't even had the plastic film pulled off. This transformer has another secondary to snub, and I have another identical transformer I'll be using in an impending amplifier build.

As you can see in the first attached image, I ordered snazzy white boards from Seeed, so I'd like to sell off at least 8 of the others to other Australian users (at this stage I think I'll reserve another one for myself). I'd give them away, except the postage here was expensive, and posting them out will also be expensive. Without having worked it out that much, I'd like to sell them @ $2 each + postage (I reckon one will fit happily in a standard letter envelope). To sweeten the deal, I'll optionally toss in one of the NTD4906 MOSFETs for only another dollar, which I ordered in a pack of 100. PM me if this appeals.

Cheers, Andrew
 

Attachments

  • quasimodo.jpg
    quasimodo.jpg
    93.4 KB · Views: 455