OnSemi ThermalTrak

[Arius]

Anybody has any interest in using OnSemi's ThermalTrak devices as the output devices in a Leach-clone?

They are not necessary the best method for biasing an output stage as evidenced by D.Self's work (Chapter Output Stage I) but I'd like to experiment with them. Besides, OnSemi claims measurable improvement.

If there is interest, then we can discuss them here. If not, I can proceed with trying them out and reporting in later. Got the parts in hand, so it'll be wasteful to not try them out.

http://www.onsemi.com/pub/Collateral/NJL3281D.PDF
http://www.onsemi.com/pub/Collateral/NJL3281D.PDF
http://www.onsemi.com/pub/Collateral/AND8196-D.PDF

One immediate question I have is whether to use a modified Vbe bias circuit (as per Leach) or use the internal diodes directly and forgo adjustment capability.

Thanks for looking.

 

[dimitri]

where have you got them, lucky man?
i'm waiting for my samples for 6 months...

 

[Arius]

I was surprised myself at how fast OnSemi was. I ordered them last Friday and they came in today from Arizona.

I really didn't expect them to send any because I truthfully told them my sample request was for lab use, not production product.

Well I suppose it helps if you are currently an OnSemi customer, which I am. Hopefully you'll get yours soon.

 

[dimitri]

thank you, ordered them for the second time

or use the internal diodes directly and forget adjustment capability

pls, allow me a week and I'll compare diode and base-emitter junction current-voltage curves

 

[Arius]

Well, a quick exercise with the simulator shows a few possibilities. Think I'll use a modified VBE amplifier bias generator with the 6 diodes (3 output pairs) in the base-collector leg. Now all I need to do is try it out to see how well it tracks.

 

[panson_hk]

i saw these parts first time in CES2005. Really want to try them (available fom Digikey).

 

[telewatt]

I was surprised myself at how fast OnSemi was. I ordered them last Friday and they came in today from Arizona.

This is because the parts come from Digi-Key, next day air. Motorola has arrangements with Digi-Key to ship most of their samples from there.

I hope to God Motorola does very well financially, otherwise they may stop doing this.

 

[Mr Evil]

...I hope to God Motorola does very well financially, otherwise they may stop doing this.

You do subsequently buy their products if the samples prove to be good don't you? Of course if people were only ordering samples out of greed then they would probably have to stop eventually, but none of us are like that, are we?

 

[Arius]

Well, I use OnSemi parts at work and I can say that their sample service is never that fast (unless it's a part from their preferred 24hr program).

Many times, the parts come out of Asia. Even parts from within the US have about a one week turn.

Anyhoo, you're right. Digikey stocks them now and they're about the same as the regular MJL parts in price.

About using the thermal traks themselves, I was thinking of using them (their internal diodes that is) in a modified VBE multiplier, except for one device. That last device will be the temp sensor for the protection circuit. That means one output stage junction will be uncompensated. Assuming equal current sharing, I think that'll be ok. Comments, anyone?