Anyone want to take a guess on how this circuit works?

Thread Starter

Quintilis_Telescope

Joined Oct 13, 2009
10
Ok. Im having trouble figuring out how this circuit works. Its essentially a transistor oscillator with a transformer on the end to generate a high voltage AC. There are two transistors.Q1 is PNP and Q2 is NPN. When Vcc = 5V, point A, which is the output that feeds the transformer gives a mildly distorted square wave at that goes from 0V to 3.4V at ~50Hz. Point B gives a capacitor charging/discharging transient waveform that alternates between 1.5V-3.7V. Point C gives a high voltage AC of 350V at ~50Hz. I would really appreciate if someone or multiple people could explain anything they can so I can understand how this circuit oscillates. My guess is that it has to do with the capacitor. It should charge up to a certain voltage, turn on a transistor, discharge, turn off the transistor and the other turns on.Also, how can a transformer generate such a high voltage? Maybe its similar to an inductor with the collapsing magnetic field? Anyway. Thanks for your time.





Also note that point A is attached to the center tap of the transformer. It can be attached to pin 3 of the transformer and work the same.
 

Attachments

mossman

Joined Aug 26, 2010
131
I've seen similar circuits. The term "switched mode" comes to mind. Could be mistaken. I found an article at one time that explained in detail how it works. I'll try to find it and get back to you. Transformers can create larger voltages (step-up transformer) because of the magenetically coupled windings. Current through the primary winding generates a magnetic field which couples to the secondary winding and induces a voltage proportional to the "turns ratio" (number of windings in coils). Note that as the voltage goes up, the current goes down proportionally because a transformer is a passive device (power in equals power out, minus losses of course).
 

Kermit2

Joined Feb 5, 2010
4,162
It is acting more like a spark coil in this circuit than a transformer. You are building up a magnetic field and at some point the voltage is quickly changing. I suspect it is a 'spike' that generates the high voltage. It appears the secondary is DC connected to the primary,which is not like a transformer and more akin to the old time sparking coils.
 

R!f@@

Joined Apr 2, 2009
9,918
The circuit itself is an oscillator. Frequency is determined by the RCL of the circuit.
The HV is generated at the output when the field in the transformer collapses.

In this circuit I doubt it is HV. Since the transformer ratio is 1:1. Out put equals the input and the diagram says it's audio tx.

I think it's just an audio frequency oscillator circuit.

Not a HV one
 

Thread Starter

Quintilis_Telescope

Joined Oct 13, 2009
10
The circuit itself is an oscillator. Frequency is determined by the RCL of the circuit.
The HV is generated at the output when the field in the transformer collapses.

In this circuit I doubt it is HV. Since the transformer ratio is 1:1. Out put equals the input and the diagram says it's audio tx.

I think it's just an audio frequency oscillator circuit.

Not a HV one
It is definitely HV. I have it hooked up here. The transformer can be attached to a voltage multiplier to get 450-600V DC but the current is very low. My multimeter drops a lot of the voltage when I connect it. But ignoring the transformer, can anyone explain how the circuit biases itself to oscillate?
 

Kermit2

Joined Feb 5, 2010
4,162
You could try modeling it with LTspice and look at the current in the cap. There is really nothing in the circuit that could cause oscillation other than the cap and inductor. I suspect the cap charges through the induction of the transformer, and when it reaches a high enough voltage the transistors are biased into conduction which brings the voltage down quickly. If you turn the diode around does everything stop working? this would mean it definately has something to do with the induction of the xformer and cap charging. Also dis-connecting the jumper from primary to secondary on the transformer should stop the high voltage spikes and give you a standard 1:1 transform ratio for the waveforms.
 

R!f@@

Joined Apr 2, 2009
9,918
It is definitely HV. I have it hooked up here. The transformer can be attached to a voltage multiplier to get 450-600V DC but the current is very low. My multimeter drops a lot of the voltage when I connect it. But ignoring the transformer, can anyone explain how the circuit biases itself to oscillate?
You admit it's not HV, but you did not know that.

U said u can attach it to multiplier, then this makes it HV.
To get Higher voltage you have to use higher Vcc, since the RX is 1:1. The transformer itself does not step up the voltage.
It's a poor design since I do not see any isolation.

A voltage multiplier cannot handle noticeable current. It' just multiplies voltage not power thus very low current. Even a DMM will heavily load a voltage multiplier if HV probe is not used.
 

SgtWookie

Joined Jul 17, 2007
22,230
It's a simple flyback converter with no output voltage regulation.

I whipped up an approximate model in LTSpice, and ran the simulation for you.

Feel free to download LTSpice (google that phrase) and try it out yourself.

See the attached files.

If you change where Q2's collector connects to the transformer, you'll increase the output current and decrease the output voltage.

Rifaa, sorry but you are incorrect about the output voltage being the same as Vcc.

If I had done a better job modeling an actual 600:600 audio xfmr, I'd probably see a lot higher voltage out. This one is just a rough guess.
 

Attachments

Last edited:

R!f@@

Joined Apr 2, 2009
9,918
jeeeeeeeeeeeeeez!! u are right ..it is a fly back.
I shud have kown.
I mean I can fix fly back stages of CRT
That was really stupid of me..
 

Bernard

Joined Aug 7, 2008
5,784
Try: http://en.wikipedia.org/wiki/Blocking_oscillator Explination attempt: Starting with C1 charged to Vcc, there is weak forward bias on Q1 base, causing Q1 & Q2 to start conducting. Current starts flowing in Xfomer & R1. Drop in R1 further increases forward bias on Q1 base. C1 starts discharging allowing V drop of R1 to shut Q1 & Q2 off. Rapid collaps of xfmr field causes V spike at C. With decreasing current thru R1, Q1& Q2 are drivev further into cut-off. Ci can now recharge back to Vcc. Timing about 2X R5X C1 or 50 Hz.
 

R!f@@

Joined Apr 2, 2009
9,918
Try: http://en.wikipedia.org/wiki/Blocking_oscillator Explination attempt: Starting with C1 charged to Vcc, there is weak forward bias on Q1 base, causing Q1 & Q2 to start conducting. Current starts flowing in Xfomer & R1. Drop in R1 further increases forward bias on Q1 base. C1 starts discharging allowing V drop of R1 to shut Q1 & Q2 off. Rapid collaps of xfmr field causes V spike at C. With decreasing current thru R1, Q1& Q2 are drivev further into cut-off. Ci can now recharge back to Vcc. Timing about 2X R5X C1 or 50 Hz.
beats me....

why the HV in the audio range ?
 

Thread Starter

Quintilis_Telescope

Joined Oct 13, 2009
10
Try: http://en.wikipedia.org/wiki/Blocking_oscillator Explination attempt: Starting with C1 charged to Vcc, there is weak forward bias on Q1 base, causing Q1 & Q2 to start conducting. Current starts flowing in Xfomer & R1. Drop in R1 further increases forward bias on Q1 base. C1 starts discharging allowing V drop of R1 to shut Q1 & Q2 off. Rapid collaps of xfmr field causes V spike at C. With decreasing current thru R1, Q1& Q2 are drivev further into cut-off. Ci can now recharge back to Vcc. Timing about 2X R5X C1 or 50 Hz.
I dont necessarily understand the bolded text. Isnt there always current flowing through R1?

and SgtWookie, thanks for the simulation but I get an error "unknown spice device N in "n=1"...." when I try to run it. Ill try to read more on flyback converters.
 

SgtWookie

Joined Jul 17, 2007
22,230
SgtWookie, thanks for the simulation but I get an error "unknown spice device N in "n=1"...." when I try to run it. Ill try to read more on flyback converters.
Rats - I made a bonehead mistake in the MPSA42 definition; the + goes on the beginning of the continuation line, not at the end of the line to be continued. Sorry about that.

I fixed the simulation file and re-uploaded it; try downloading it again.
 

t_n_k

Joined Mar 6, 2009
5,455
This was my attempt at simulation ....

Interestingly, the capacitor voltage is roughly sawtooth in character within a small range - not showing excursions to Vcc.
 

Attachments

SgtWookie

Joined Jul 17, 2007
22,230
t_n_k,
I recognized the circuit for what it is; a flyback converter.

You did not include the rectifier and cap on the output as I did. Without them, your output will appear quite strange. Also, you seem to be taking your output reading from the wrong place; you should have been reading at the top of your R6.
 

t_n_k

Joined Mar 6, 2009
5,455
Yes thanks SgtWookie - I drew the arrow to the wrong point which was indeed the top of R6. The probe on the lower resistor was not included in the view I show. My picture is a compilation of several bits.

I take your point about it being a converter but the original schematic from the OP doesn't seem to include an output filter - just an open circuit at C. I was attempting to replicate a reasonable approximation to what was originally drawn by the OP. Does the OP in fact have a filter on the output?

The operation of the circuit is quite difficult to explain - I guess that's why we've resorted to simulation to discover this for ourselves.
 

SgtWookie

Joined Jul 17, 2007
22,230
t_n_k,
Yep, but the way the xfmr was wired up was a dead giveaway.

Our OP didn't include the diode and cap on the output, because they didn't really understand how the circuit was supposed to work. I included them so they could see a somewhat meaningful output from my simulation.

Not shown in my simulation is parallel parasitic capacitance for each winding in the transformer of 1nF and 2nF; just another pair of wild guesses. But, my simulation did have an output that exceeded Vcc by a significant margin.

Rifaa,
An audio transformer was used merely for convenience' sake. The output certainly is not at an audio frequency. It's just that we can get audio transformers pretty cheaply here in the USA.
 

Bernard

Joined Aug 7, 2008
5,784
Always current in R1? Yes, about 4 mA when C2 is conducting & a saw-tooth current in operation. Eye-ball simulation has its limits.
 
Top