Resistors getting so so...Hot :o

Thread Starter

RIKRIK

Joined Oct 11, 2019
129
Okay update, thanks for all your advice , it's not exactly 50hz , but Its close enough. More worried about a large changes in frequency, the 4047, drifts quite a bit from 11 to 13v's, this circuit less than 1hz. Okay the total of the 4 resistors before @12.7v was 30w, now it's 6.8w . A 77% improvement. Also using a more available mlcc ,1uf, it's a bit cheaper too .

upgraded.png
 

SteveSh

Joined Nov 5, 2019
105
"Too hot to touch". That means the resistors is probably at 140 deg F, or 60 deg C. In the grand scheme of things, that's not all that bad. That's well under the max temperature rating for most resistor.

Don't forget that the maximum operating temperature for a resistor depends on what type it is. Max temperature is lower for carbon than it is for wire wound, for instance.
 
Last edited:

Thread Starter

RIKRIK

Joined Oct 11, 2019
129
Hi Steve, I used all 1% wire wound resistors, can put my finger on it for about 3 seconds before it gets hot enough for me to feel that the heat coming out of it is a bigger risk than the high voltage .

One other question I have a electrolytic cap, I still feel like its one of the parts to most fail over time, IV looked into Tantalum, 1uf 35v. However from what I can sus being a 90's kid, something went down in the 80s involving exploding capacitors and the 2000s I have read a ethical article on why not to use the stuff as it's comes from war torn countries. However I only care about the how long it will last.
 

SteveSh

Joined Nov 5, 2019
105
Hi Steve, I used all 1% wire wound resistors, can put my finger on it for about 3 seconds before it gets hot enough for me to feel that the heat coming out of it is a bigger risk than the high voltage .

One other question I have a electrolytic cap, I still feel like its one of the parts to most fail over time, IV looked into Tantalum, 1uf 35v. However from what I can sus being a 90's kid, something went down in the 80s involving exploding capacitors and the 2000s I have read a ethical article on why not to use the stuff as it's comes from war torn countries. However I only care about the how long it will last.
IMHO - the resistor temperature is not a lot to worry about. From what I said earlier, 60 deg C (if that's what the resistor it at) is well within the temperature rating of that type of part to ensure long life.

As for the tantalums, we use them all the time. But they have to come from qualified suppliers, have been tested/qualified every which way, which usually leaves out war-torn countries.
 

MisterBill2

Joined Jan 23, 2018
7,964
The 270R resistors set the hz's of the ossilator, changing them would mean I would need a higher ceramic cap value, and there hard to find, I did see some where about using Darlington transistors, to resolve the issue, let me just find it
You could keep a similar R/C ratio increasing the resistors to 1200 ohms and reducing the capacitors to 1 Mfd from 4.7 Mfd. It may take a small amount of adjusting, but if you are feeding lights they are not very picky about frequency. What is the frequency of the circuit that you have now? Also, the circuit in post #16 is reasonable and more efficient because no power is wasted in the driver transistors. so there are alternatives that will be much better, But be careful about handling the CD4047 because it is more static sensitive. You could also use a 555 timer IC driving a 4013 flipflop to get a perfectly symetrical square wave. That would be cheap, easy, and use fewer parts.
 
Last edited:

Thread Starter

RIKRIK

Joined Oct 11, 2019
129
MisterBill2 it seems you have red my mind.i have a twin brother, and he's not even that good at knowing what I'm thinking, this was the updated version I posted recently . If you scroll up a little , but thank you for you input. upgraded.png
 

Thread Starter

RIKRIK

Joined Oct 11, 2019
129
Okay heres some more data via a simulation, also changed my low battery circuit so it's less of a power hog
f1.pngf2.png
 

Wolframore

Joined Jan 21, 2019
2,233
Simulation and what happens reliably in actual transistors are two different things. Especially if you don’t know the h-parameters and thermal characteristics of your models
 

Thread Starter

RIKRIK

Joined Oct 11, 2019
129
And wolf you were right, did do the same circuit with a 11v zener, worked over a wider hfe, but will definitely have to have a pot I think. But currently designing a board , the inverter prototype is finished :D
 
Top