Transformer Type
GammaRay, Sun Mar 10 2019, 10:53PMHere's the scenario for which I seek friendly advice: A HV power supply periodically discharges through a spark gap at 1,400vdc at around 1Hz. I plan to use a 20:1 TX to step down the high voltage pulses to low voltage pulses of about 70vdc. I'll use a TX, rather than a buck configuration, for isolation. Question is.... at such a low frequency (1Hz +/-) what type TX is a good candidate? Flyback, pulse, toroid, other?
Re: Transformer Type
Sulaiman, Sun Mar 10 2019, 11:20PM
this is a high frequency problem rather than a low frequency one,
the spark period and waveshape are of key interest.
I think that Iron or even MnZn ferrites will be unsuitable, air or Ni-Zn ferrites are probably required.
Although 1400V is not particularly high, isolation (capacitive coupling) will be challenging due to the very high dv/dt expected.
I think I'd try a capacitively-coupled magnetic pulse transformer approach,
good detail during the event with 1 second for d.c. recovery.
Alternatively you could use a (high frequency) voltage divider with amplifier/buffer, isolated power supply (dc:dc converter) and a high speed linear opto- coupler.
_________________________________________ __
EDIT: on re-reading it seems that you are not trying to monitor the voltage but to use it ?
If so then a pulse transformer seems a good choice.
The size of the transformer will be largely determined by current (= wire diameter) and integral(voltage.time) (= #turns.core area) during the spark, but extra space will be required for pri:sec isolation.
Do you know the spark duration or energy ?
What type of spark gap are you using?
==================================
Most importantly, what is the power supply, and load ?
==================================
Sulaiman, Sun Mar 10 2019, 11:20PM
this is a high frequency problem rather than a low frequency one,
the spark period and waveshape are of key interest.
I think that Iron or even MnZn ferrites will be unsuitable, air or Ni-Zn ferrites are probably required.
Although 1400V is not particularly high, isolation (capacitive coupling) will be challenging due to the very high dv/dt expected.
I think I'd try a capacitively-coupled magnetic pulse transformer approach,
good detail during the event with 1 second for d.c. recovery.
Alternatively you could use a (high frequency) voltage divider with amplifier/buffer, isolated power supply (dc:dc converter) and a high speed linear opto- coupler.
_________________________________________ __
EDIT: on re-reading it seems that you are not trying to monitor the voltage but to use it ?
If so then a pulse transformer seems a good choice.
The size of the transformer will be largely determined by current (= wire diameter) and integral(voltage.time) (= #turns.core area) during the spark, but extra space will be required for pri:sec isolation.
Do you know the spark duration or energy ?
What type of spark gap are you using?
==================================
Most importantly, what is the power supply, and load ?
==================================
Re: Transformer Type
johnf, Mon Mar 11 2019, 08:55AM
What?
how are you going to use that as a flyback???
what has torroid got to do with it ( a core shape)
Pulse --so you only want the levels -no power
you probably mean a high frequency N:N ratio transformer
johnf, Mon Mar 11 2019, 08:55AM
What?
how are you going to use that as a flyback???
what has torroid got to do with it ( a core shape)
Pulse --so you only want the levels -no power
you probably mean a high frequency N:N ratio transformer
Re: Transformer Type
GammaRay, Mon Mar 11 2019, 05:59PM
The speak gap is an iridium tipped spark plug in proximity to a neodymium magnet for quenching. I've yet to measure the spark duration and energy, but needs to be done. The load is (2) 330,000uf caps in parallel.
GammaRay, Mon Mar 11 2019, 05:59PM
The speak gap is an iridium tipped spark plug in proximity to a neodymium magnet for quenching. I've yet to measure the spark duration and energy, but needs to be done. The load is (2) 330,000uf caps in parallel.
Re: Transformer Type
Sulaiman, Tue Mar 12 2019, 09:01AM
If your construction techniques are poor then your plan will work ... poorly.
If your construction technique is good then you should prepare for an explosion :)
In effect you are going to connect a charged capacitor (your power supply) to a not-charged capacitor (via a transformer)
Good construction technique would mean short thick wires/interconnects with little stray inductance (or transformer leakage inductance)
so the very low inductance of wiring + components is all that limits the peak current.
If your construction technique is poor then stray and leakage inductance, plus wiring resistances, will limit the current peak.
Or you could design-in some inductance (or transformer leakage inductance)
Sulaiman, Tue Mar 12 2019, 09:01AM
If your construction techniques are poor then your plan will work ... poorly.
If your construction technique is good then you should prepare for an explosion :)
In effect you are going to connect a charged capacitor (your power supply) to a not-charged capacitor (via a transformer)
Good construction technique would mean short thick wires/interconnects with little stray inductance (or transformer leakage inductance)
so the very low inductance of wiring + components is all that limits the peak current.
If your construction technique is poor then stray and leakage inductance, plus wiring resistances, will limit the current peak.
Or you could design-in some inductance (or transformer leakage inductance)
Re: Transformer Type
GammaRay, Wed Mar 13 2019, 09:07PM
Thanks Sulaiman, I'll try a capacitively-coupled magnetic pulse TX and also an old-style flyback TX observing good construction techniques all around. If either show notable results, I'll let you know.
GammaRay, Wed Mar 13 2019, 09:07PM
Thanks Sulaiman, I'll try a capacitively-coupled magnetic pulse TX and also an old-style flyback TX observing good construction techniques all around. If either show notable results, I'll let you know.
Print this page