UD2.7C and news coil 150khz 70cm

[JCF]

Hello all,
what do you think of oscilloscope signals: full bridge and primary current output.
the frequency of the primary and secondary are close, is it normal to have such a strong resonance on the primary?

what about end of scope trace of primary signal  ?





Specs  SavedCoil PVC 8 cm DIA 0.20mm FIL 2800 TOURS 18cm 37cm tore V0 AFFINE ok.txt

[Hydron]

Looks fine to me. The odd waveform at the end of the trace is due to the energy in the tank circuit being transferred back into the DC bus capacitors via the anti-parallel diodes of the IGBTs.

Probably the more important question now is if it's making sparks

[JCF]

Thank's for the message.
Yes big spark, but only try with 200V (I must go outside with the tesla for testing with 350V/400VDC

[Mads Barnkob]

You are running some 600us on-time?

Seems like you are quite efficiently transferring all energy to the secondary coil, as the current envelope peak and slowly goes down, but your DC bus voltage also seems to be sagging!

Reminds me of my small DRSSTC, I could not get it to hit OCD limit, it would just eat and transfer all the power to the secondary circuit. You see this when you have a coil in equilibrium of power from primary to secondary to spark.

Maybe try to add some more DC bus capacitance

[JCF]

Mads Merci !

on time is 750µS.  I was wondering why the level of impulses in the primary was going down? is it a lack of capacity?
I have in all 390µFx4 = 1560 µF. it's not a lot..

I can add more in the big box, this is an ATX PC power supply box, space is limited.

How much ? I will not exceed 400V/450v DC

Where buy them at a correct price? I usually use RS Components, so Ebay (never news)?

https://www.arrow.com/en/products/als30a472np450/kemet-corporation 4700µF 450V 64€ (x2 ?)

[Mads Barnkob]

Its not the purple primary current that is going down, that is just a result. The root cause can be seen on your yellow inverter output voltage trace. For each cycle the voltage gets lower and lower, you are draining your DC bus capacitance.

You can calculate how much you need, according to a desired maximum voltage drop in a burst, I wrote about that here: https://kaizerpowerelectronics.dk/tesla-coils/drsstc-design-guide/dc-bus-capacitor/

I can sell you some used large 3300-4700uF/400-450V'ish caps from old UPS if you want? That can with shipping hopefully be less than half price of what you found

[JCF]

Mads,

Is it possible to parallel 4x390µF plus 4700µF V capacitors (electrical risk?) with balancing resistors?

if there is any doubt I will only connect the 4700µF capacitor only

[Mads Barnkob]

Mads,

Is it possible to parallel 4x390µF plus 4700µF V capacitors (electrical risk?) with balancing resistors?

if there is any doubt I will only connect the 4700µF capacitor only

The current would be shared according to capacitance ratio. So the 4x390= 1540uF would only conduct 25% of the current and the 4700uF the 75%. So you would most likely overload the single large capacitor.

[JCF]

well, the result is correct at 200Vdc (maxi 350VDC) with this new coil.

I would like before going to higher voltages to adjust the OCD to the driver UD2.7C

the full bridge is composed of 4 IGBTCM200DU-12F (600V 200A). The current sense for OCD on the primary is 1: 2000 also the same 1:2000 to feedback

At which voltage to adjust TP1, how to calculate it ?

is somebody can help me ?

[Mads Barnkob]

well, the result is correct at 200Vdc (maxi 350VDC) with this new coil.

I would like before going to higher voltages to adjust the OCD to the driver UD2.7C

the full bridge is composed of 4 IGBTCM200DU-12F (600V 200A). The current sense for OCD on the primary is 1: 2000 also the same 1:2000 to feedback

At which voltage to adjust TP1, how to calculate it ?

is somebody can help me ?

A 1:2000 ratio current transformer corresponds to 2000A in the wire going through the CT, gives 1A in the CT output. So the voltage output depends on the burden resistor across the CT output. Too little resistance might give too low voltage, too high resistance gives unwanted phase shift. A normally used value is 5-50 Ohm depending on number of windings. So if you need 5V for 2000A primary coil current (1A output from CT), that is just Ohms law to get the required resistor. R = 5V / 1A = 5 Ohm. So if you want to set the OCD to 500A, 25% of 2000A, set it to 25% of 5V ~ 1,25V. Be sure to check what the OCD range of the driver really is, I seem to recall that the effective input range of UD1.3 was only something like 1-3.75V. But that might be different for UD2.7C

[JCF]

Merci !
I made measurements at C14 to find out what was the input voltage of LM311. with a signal generator sinus 150Khz at 20 volt input OCD. Current sense 1:2000.
on the oscillo trace we can see the rectified signal average value 300mv on 2.5 Ohms = 120mA
i.e. 120 * 2000 = 240 A
is it correct  ?

[davekni]

Current sense 1:2000.
on the oscillo trace we can see the rectified signal average value 300mv on 2.5 Ohms = 120mA
i.e. 120 * 2000 = 240 A
is it correct  ?

Yes, your calculations look correct.  However, to reduce noise sensitivity, I suggest increasing resistance above 2.5 ohms in order to increase OCD trip voltage.  Around 25 or 50 ohms would be great (3 or 6V trip threshold).  As a minimum, remove R28 (or R17) to get to 5.1 ohms.