TC4420 MOSFET drive question

[ka1tdq]

I have about 10 ideas about where to go with this project, but this is what I have so far...

I put two crystals in this rig (3880 and 3885 with a switch for an externa crystal).  The oscillator goes to two 2N3866 trasistor stages so that I could build up to at least 5 volts pk/pk for a MOSFET driver input. 

I'm using a TC4420 MOSFET driver and I'm putting about 8 volts pk/pk into it but the output stays high at 5 volts dc.  The scope shows that the waveform actually goes about 1.5 volts negative ad peaks around +7 volts. 

I'm not sure why the TC4420 output doesn't change.  I've attached a schematic.

Jon
KA1TDQ

[KA2DZT]

This is a question for Steve QIX, but I would try another MosFet, seems maybe the one you're using is bad.

Just a guess.

Fred

[ka1tdq]

I replaced the IC, but I'm still getting the same result.  I'm thinking that the TC4420 maybe wants a good solid 5 volts for up and zero for down rather than some freaky fast analog signal.  Hell, it's not Jimmy John's. 

I could throw it into an AND/NAND gate to do the conversion first and then put it into the MOSFET driver.

Jon
KA1TDQ

[DMOD]

Unless a positive pulse comes in at over 0.7V the 2N3866 transistor won't switch.


I would also put a 4.7k to ground at the base. How much voltage is feeding the base?

[ka1tdq]

Actually, it switches ok... this is the 2nd transistor stage with 2.5 volts coming through the capacitor. It switches and i get over 10 volts peak and 0 volts on the bottom of the wave.

Im going to get a JK dual edge triggered flip flop to clean it up (im guessing dual edged are made). That should clean things up.
Sent via my phone so please for give errors in grammar.

Jon
KA1TDQ

[DMOD]

OK, misunderstood you, I thought the 2N3866 was not switching.

So you are using the 8-Pin dip version.

Acording to the specs, you need at least 2.4V above ground to switch it.

And there is no load on the output?

The only other thing may be that your switching times are shorter than 80 ns?

[ka1tdq]

I ended up putting a NAND chip after the 2N3866 and had to experiment to get the output large enough.  I replaced the resistor in the collector with a 10-turn FT37-43.

I borrowed the NAND configuration from Steve's VFO design, and it seemed to work pretty good.  The output of the NAND chip is a 5 volt peak/peak signal that seems to ride about two volts above ground.  I'm thinking that this is because the input to the NAND gate never really reaches zero volts due to the PN junction in the 2N3866.  A fix for this might be to place a diode in the ground lead of the NAND chip to reference it about .7 volts above ground.

I am getting a duplicate waveform out of the TC4420 MOSFET driver though.  I've attached a screen shot of the oscilloscope and it looks like a bunch of Motorola symbols.  The 2N3866 transistor that I'm using was made by Motorola, so this might be part of their design.

Anyway, I'm thinking that the TC4420 waveform might go rail-to-rail with 12 volts applied.  If that's the case, I just need to throw that baby into a MOSFET gate and belch away!

Jon
KA1TDQ

[ka1tdq]

...forgot to put the waveform in...

[Tom WA3KLR]

The TC4420 input is a n-channel MOSFET with TTL thresholds and diode clamps to gnd and Vdd.  

With the ac drive voltage you have to the 2N3866, I think that you could do away with the 2N3866 stage and capacitively couple right into the 4420 input along with adding a high resistance bias voltage divider to the input pin providing 1.55 volts dc.

Confirm with an ohmmeter that the GND pins of the i.c. leadframe have continuity to the board groundplane.

Latest scope photo - you have a higher rf frequency filling in on the upper half of the waveform, bad.

[ka1tdq]

Just stopped at Radio Shack and got a 220k/77k combination. I will also supply the 4420 with 12 volts. Out and about for the evening so I will probably try it out in the morning. Thanks!

[ka1tdq]

I pulled out the last 2N3866 stage and messed around with biasing the input to the 4420, but I couldn't seem to get it to work.  It got late tonight and I was getting frustrated, so I put the 2N3866 stage back in with a scroteful NPN for a final and made it Class C.  Scope shows a clean sine wave about 22 volts peak/peak with 33 volts DC.  My calculations show that it's 15 watts output. 

I could modulate it or use it as a driver.  The website for the 1 fet rig also shows a schematic for an analog driver to a 2-buss fet rig...

...or there's a Class D schematic out there for 10 watts drive. 

...or morse code key the carrier.  Unfortunately, It's crystal controlled so I can't swish the VFO around though.

Jon
KA1TDQ

[ka1tdq]

I finished wrapping things up this morning.  Attached are the output waveform, schematic and finished product. 

I do have room to squeeze in a modulator off to the right, or I could use a modified H-bridge as a modulation transformer.  Another option would be just to leave it as-is and keep it for an exciter.

My sister is sending me all my really large variable capacitors that she was storing for me.  I'm thinking about maybe a large tuner, another amplifier (I miss my old one) or an E-RF deck. 

Jon
KA1TDQ