1. Home
  2. Electronics Design
  3. a few electronic fuse circuits
  4. Page 2

a few electronic fuse circuits

Ok, while walking the dogs a couple more ideas came to mind, a bit unorthodox though:

a. Find a cheap gate driver or PWM bridge driver chip that has UVLO. Hardwire the one for +15V to output high, the one on -15V to output low. Of course, that only works if it won't engage either of its output devices when in UVLO, which isn't always the case for sync buck chips. Some motor drivers have a coast function, might be useful here as well.

b. An optocoupler feed back whether or not both output voltages come up. Requires little on the output side since that most likely doesn't have to be precise. uC on the other side pulses on (via a PMOS or something) but lets go if the output voltage won't come up in due course. An interrupt or fast polling routine then keep monitoring the opto output.

Version b, in the extreme, might only require a zener and a resistor so the opto trips off if the grand total on the output side goes below 25V or whatever you deem ok. Current sense there is feasible as well because I guess you won't care about a few hundred mV of drop at that end.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg
Loading thread data ...

One fix (I used to do it in CD ignition circuits) is to use a much bigger bias resistor, just enough to get it started. Then add a lower value resistor in series with a diode, from the feedback winding CT to ground. As the thing starts to oscillate and that CT point starts to go negative, the lowside diode+resistor conduct and furnish the main base current path. It can probably be made to work with just an added schottly diode if you get all the numbers right.

If you do that, you can probably eliminate the feedback winding cap to ground, so the whole thing devolves to replacing that one cap with a small schottky diode.

I've also seen the feedback winding CT bypass cap replaced with a resistor. That will work, but it's dicey as regards starting up when it's cold.

John

Reply to
John Larkin

Use "pre-glow" like on the old Mercedes-Benz Diesels :-)

--
SCNR, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

I guess if you are making your own coils, you can add a winding from the coil to generate a bias voltage for you that would supply the remainder needed. Just rectify it and pass the soft start bias R through it. When it it starts to oscillate, that winding will supply additional voltage to increase the current on the bases via the bridge you have placed on that added winding. unless you regulate it, it will be a little unstable since a little load is going to influence this output unless you plan on over saturating the coil a bit to compensate for this.

The old buck boost topology..

I still think a resetable time delay shut down is cheaper, you may need the initial full current to get the attached load in happy land..

Reply to
Jamie

Place a 10 ohm R on the LM317 output to common. Modulate the Adjust pin via an Op-amp. Decouple the output via a cap to the speaker.. Supply the LM317 with ~12 volts..

How's that for hack! No, I've never done this before, but I can see where it may actually work! ;)

Jamie..

Reply to
Jamie

I believe (I'd have to dig out a notebook that isn't available today) that I tried that configuration and it clipped the negative half cycle. Maybe not.

Jim

Reply to
RST Engineering

Isn't that going to suck an ungodly amount of current (as opposed to discrete class AB design)?

Jim

Reply to
RST Engineering

The simplest circuits are probably class A with a constant-current thing in the other leg. Of course it will be inefficient and have bad response.

With signal-path transformers, you could drive signal into both legs of a totem pole of 317s, and get class AB. You could get a lot of power gain that way, so the whole amp could be just the output pair.

I'd rather make a parametric amplifier with ceramic caps as the gain elements.

John

Reply to
John Larkin

That is what the cap is for with the 10R to give you that - output with a class A in mind. The 317 has a source only output, no sink. You could maybe improve it by using a series 10R to the other shunt 10R and the cap then would connect at the node..

of course, you need to adjust the levels from the Op-amp to the adjust pin via a R network with the op-amp in single rail mode.

Reply to
Jamie

Having obtained that level of product info, I'd have felt justified in looking further, for something else.

RL

Reply to
legg

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.