@JulesP-
If not already mentioned, you also need a snubber diode on your coil to prevent kickback against the transistor.

 

Merely the result of applying the 'forced gain of 10' strategy. A higher forced gain should work though.

Hi alec,
LTS shows a gain of approx 120, for a common emitter configuration it is line with the d/s.
Eric

 

@JulesP-
If not already mentioned, you also need a snubber diode on your coil to prevent kickback against the transistor.

My reason for building this circuit is to investigate the properties of the Back EMF pulses that appears at the MJE Collector when it is triggered by a square wave from the PWM. I had been measuring 1300V pulses of about 30us FWHM that do not appear to damage the MJE since they are directed away from the Collector to a storage capacitor. I had been using a FET for this role (IRP840) and which gets over the issue of the limited current output of my PWM but most FETs have a 'flyback diode' between the Drain and the Source which is supposed to quell the reverse pulse. (I'm not clear how as the diode seems to be the wrong way round to short the pulse to ground) For that reason I have been trying BJTs instead but did not reckon for the increased current demand on my PWM - hence the need for a small amplification to keep the current demand on it low.

From the figures suggested by others on the forum here, with a coil current of 2A the base current for the 2222A would be 20mA based on a 'limited' gain of 10 for each tranny and hence the 600R resistor with the 12V PWM output. If I took the gains to be 20 instead then R=2k4. So maybe I should use about 1k to compromise?

So far the HV pulses do not appear to be damaging the MJE so is the 'snubber' diode really necessary, especially since the pulses are what I am looking into. High dV/dt gradients produce very interesting effects and as a retired physicist they are of interest but my electronics is not as confident as I would like it

I hope that gives an accurate account of where I am at the moment.

 

Morning Jules,
The current gain of the common emitter configured 2N2222 is approx 120.
So for a emitter current of say 240mA, that would be 2mA, ie@ 10V [allowing for Vbe drops] 10V/2mA = 5K, you original 4k will be fine.
E

EDIT:
This 2nd image, where Rin =1R shows that the Base input current to the 2N2222 is self limiting due to the negative feedback action of its emitter resistor 39R.

 

Morning Jules,
The current gain of the common emitter configured 2N2222 is approx 120.
So for a emitter current of say 240mA, that would be 2mA, ie@ 10V [allowing for Vbe drops] 10V/2mA = 5K, you original 4k will be fine.
E

EDIT:
This 2nd image, where Rin =1R shows that the Base input current to the 2N2222 is self limiting due to the negative feedback action of its emitter resistor 39R.

Hi E,

Thanks for the elaboration. Actually yesterday I ordered a 47R 2W resistor so using that I hope will be fine instead of a 39R.

The negative feedback action you mention, will this allow the circuit to cope with an actual hfe of the 2222A of anywhere between 10 and say 200? If it's in the lower range towards 10 then won't the 4k be far too high and, via limiting the 2222A current, also limit the coil current to around 0.5A?

Jules

 

What I have at the moment is this:

 

hi Jules,
That should be OK.
E

 

Hi,

That does not look right. The 47R should go in the collector not the emitter.

 

hi,
As the PWM drive input is 12V, it makes little or no difference if the 47R is in the Collector or Emitter.
If the PWM was only 5V then the 47R would be in the Collector.
E

 

hi,
As the PWM drive input is 12V, it makes little or no difference if the 47R is in the Collector or Emitter.
If the PWM was only 5V then the 47R would be in the Collector.
E

Hi Eric,

Well, 200ma drops 9.4v across 47 ohms. 12v minus 9.4v minus 0.7v equals 3.3v across the 2N2222 base resistor.
3.3v divided by 4000 equals less than 1ma base drive.
3.3v divided by 2000 equals less than 2ma base drive.

Now with the 47 ohms in the collector, the 2N2222 base is around 1.4v, and 12v minus 1.4v equals 10.6v.
10.6 divided by 4000 equals around 3 times 1ma which is 3ma base drive.
10.6 divided by 2000 equals around 2 times that 3ma which is 6ma base drive.

So which of those approximations do you prefer?
1ma to 2ma base drive, or 3ma to 6ma base drive for the 2N2222.

[LATER]
Actually it is worse than that, because if we have 200ma through 47 ohms then we have 9.4v across 47 ohms, and then we have to subtract TWO base emitter drops so we get as overhead voltage:
12-9.4-0.7-0.7=1.2v across that 2k or 4k, so we get as base current:
1.2/4000=300ua, or
1.2/2000=600ua

and that's not very good. With the 47 ohms in the collector we have a lot more overhead voltage to work with. We could also talk about speed but i dont think that is too much of an issue because of the application.