Are big wirewound resistors acceptable for snubbers

[Ignoramus28489]

Two questions:

Question 1.

I suspect that the answer is no, but want to doublecheck.

My tig inverter's snubber specifications call for a 0.16 Ohm snubber
resistor.

I have a large, about 4" long, wirewound (better called metal tape
wound), 0.2 ohm resistor. It is rated for 100 watts and is about 1"
thick and has maybe 20 turns of metal ribbon perhaps 3/16" wide.

Can I use this monster as a snubber resistor, or would it not be
suitable due to its inductance?

If not, what resistor would you suggest.

Question 2.

Can I simply skip the resistor in the RC snubber, in my application.
Especially since my bridge is shorting and there are no times when it
is open.

I experimented with a nice big cornell dubilier capacitor and found
that it reduces ripple to next to nothing, without a resistor. Adding
a 0.05 ohm resistor made things much worse for some reason.

I understand that resistor is there to suppress oscillations, but,
ultimately, why should I care too much about those oscillations.

i

 

[Terry Given]

Two questions:

Question 1.

I suspect that the answer is no, but want to doublecheck.

My tig inverter's snubber specifications call for a 0.16 Ohm snubber
resistor.

I have a large, about 4" long, wirewound (better called metal tape
wound), 0.2 ohm resistor. It is rated for 100 watts and is about 1"
thick and has maybe 20 turns of metal ribbon perhaps 3/16" wide.

Can I use this monster as a snubber resistor, or would it not be
suitable due to its inductance?

bingo. if you dig up the solenoid equation, or the modified version
(google nagaoka) you can figure out roughly what inductance it is. XL =
2*pi*Fring*L must be small c.f. Rsnubber, where Fring is the frequency
of the ringy thing you want to snub.

If not, what resistor would you suggest.

one with less inductance. several in parallel would work better. are you
sure about the value? it sounds awful low.

Question 2.

Can I simply skip the resistor in the RC snubber, in my application.
Especially since my bridge is shorting and there are no times when it
is open.

without R, all cap energy gets dumped into your IGBTs when they switch
(I think, cant recall your sch). this may be a problem. The R is what
provides the damping - converting the oscillatory energy to heat.
without it, all you do is move the resonant frequency around. but you
may be able to alter Fring such that the ESR of your main inductor
provides the damping....

I experimented with a nice big cornell dubilier capacitor and found
that it reduces ripple to next to nothing, without a resistor. Adding
a 0.05 ohm resistor made things much worse for some reason.

the "L" of the resistor no doubt.

I understand that resistor is there to suppress oscillations, but,
ultimately, why should I care too much about those oscillations.

i

if it gets up your gatedrive, you will care.

Cheers
Terry

 

[Ignoramus28489]

bingo. if you dig up the solenoid equation, or the modified version
(google nagaoka) you can figure out roughly what inductance it is. XL =
2*pi*Fring*L must be small c.f. Rsnubber, where Fring is the frequency
of the ringy thing you want to snub.

Yep. Too bad.

one with less inductance. several in parallel would work better.

So, what kind should I look for, carbon film, or what?

are you
sure about the value? it sounds awful low.

I took some formulas from a Fuji IGBT guide. I will appreciate some
clarification.

As you know, I already have a working bridge:

http://igor.chudov.com/projects/Homemade-TIG-DC-to-AC-Inverter/01-Prototype-1/

but without snubber.

Feel free to disagree with what I will say: Essentially, in a shorting
H bridge (one that is never fully open), the snubber is not used for
much other than to suppress internal ringing on power rails going to
half bridges that open and close. Which requires relatively small caps
on each IGBT (which I will provide).

So, the job of the big snubber is to handle wrong situations like
switching the bridge off while welding (ideally should not happen),
wires going loose etc.

What I want is to never have voltage on the DC supply rails that
exceeds 150 volts. That is so because my welder specifies max voltage
as being 150 volts, and I do not want to ruin it.

That means that the resistor should allow all 200A of welding current
(let's say 400A for safety sake) to go into the capacitor at 150 V,
that means

400A = 150V / R

or R = 150/400 = 0.375 ohm.

without R, all cap energy gets dumped into your IGBTs when they switch
(I think, cant recall your sch). this may be a problem. The R is what
provides the damping - converting the oscillatory energy to heat.
without it, all you do is move the resonant frequency around. but you
may be able to alter Fring such that the ESR of your main inductor
provides the damping....

Got it. You are 100% right.

the "L" of the resistor no doubt.

Well, I was not clear. The resistor that I tried was not that
wirewound resistor. I tried a small, golden colored, heatsink
mountable ribbed resistor. I am not sure what's inside it, but it does
not seem to be wirewound.

Thanks Terry. Since I have the bridge basically working, having a
super safe, idiot proof snubber is now my main focus.

i

 

[Ignoramus28489]

Forgot to say. My Mom, whose job 20 years ago was to process stolen US
technology for Soviet industrial and military applications, just gave
me a big lecture about how dangerous it is for amateurs to create such
big power projects.

So, that reinforces the safety message and I want the snubber circuit
to be the pillar of safety of this whole thing. To save my ass in case
if everything goes wrong and all my mistakes somehow create a domino
effect.

I also installed a thermal normally closed switch that would turn the
circuit off if the heatsink temperature exceeds 70 degrees C, for the
same reasons.

i

 

[JeffM]

I tried a small, golden colored, heatsink mountable ribbed resistor.

Ignoramus28489

http://www.google.com/images?q=Dale+5-watt+alltronics ??

I am not sure what's inside it, but it does not seem to be wirewound.

Don't put any money on that.

 

[Ignoramus14135]

http://www.google.com/images?q=Dale+5-watt+alltronics ??

Yes, quite similar.

Don't put any money on that.

I think that you are right.

So, again: what kind of resistors should I use for snubbers...

I read in some document (cornell-dubilier snubbing guide) that I
should use "carbon composition resistors".

I need about 0.2-0.3 ohms. (I have a huge inductor in front of the circuit)

I could parallel 20 of these resistors:

http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=108350&Row=418718&Site=US

It's kind of pricey at $10 for 20 of them, but it would do what I
want. Am I correct?

i

--

 

[JeffM]

So, again: what kind of resistors should I use for snubbers...

I read in some document (cornell-dubilier snubbing guide)
that I should use "carbon composition resistors".
Ignoramus14135

Yeah. That's the stuff. Carbon comps can take a lot of abuse.
Problem is they don't make them any more.
A Russian physicist I worked with called these *bulk resistors*.
Here's the modern equivalent:
http://www.google.com/search?q=resistor-in-a-TO-*-package
..
..

http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail?Ref=108350&Row=418718&Site=US

Your dataset has expired. Please search again.
$DEITY, I hate sites so heavily (and poorly) scripted
that you can't deep-link to them.

 

[Ignoramus14135]

Yeah. That's the stuff. Carbon comps can take a lot of abuse.
Problem is they don't make them any more.
A Russian physicist I worked with called these *bulk resistors*.
Here's the modern equivalent:
http://www.google.com/search?q=resistor-in-a-TO-*-package
.
.
Your dataset has expired. Please search again.
$DEITY, I hate sites so heavily (and poorly) scripted
that you can't deep-link to them.

Try digikey item such as

TDH35PR250J-ND

TDH35PR150J-ND

i

 

[JeffM]

http://www.google.com/search?q=resistor-in-a-TO-*-package

Try digikey item such as
TDH35PR250J-ND
TDH35PR150J-ND
Ignoramus14135

Yup. We're on the same wavelength.

 

[Ignoramus14135]

Yup. We're on the same wavelength.

sounds good. I will check my trash pile tonight also, I could have
some such resistors, I recall seeing something in a TO package on a
little heatsink, coming from a UPS.

i

 

[JeffM]

http://www.google.com/search?q=resistor-in-a-TO-*-package

sounds good. I will check my trash pile
Ignoramus14135

I forgot you're working from cast-offs.
..
..
:::::I read in some document (cornell-dubilier snubbing guide)
:::::that I should use "carbon composition resistors".
::::: Ignoramus14135
:::::
You may find some of the old stuff in your stash.
http://www.google.com/images?q=carb...globalsources+-ave-nl+-sourceresearch+-taiohm
Notice the very flat ends.
The body is phenolic (typically medium-dark brown)
and only the bands are painted.

 

[Ignoramus14135]

I forgot you're working from cast-offs.

Some things are castoffs, the stuff that I can get the fastest s what
I already have...

i

.
:::::I read in some document (cornell-dubilier snubbing guide)
:::::that I should use "carbon composition resistors".
::::: Ignoramus14135
:::::
You may find some of the old stuff in your stash.
http://www.google.com/images?q=carb...globalsources+-ave-nl+-sourceresearch+-taiohm
Notice the very flat ends.
The body is phenolic (typically medium-dark brown)
and only the bands are painted.

--

 

[[email protected]]

I dont know the details of your project, but I would generally be
inclined to prefer easily available parts. A standard power wire wound
R can have a smaller powered higher R non inductive paralleled with it,
in many cases this works well. Whether its good for your project I dont
know, as I dont even know what youre making.

NT

 

[Terry Given]

I have pulse-tested these resistors before. The peak pulse power rating
is not much more than the average rating - its a thin film.

carbon composition resistors *are* still made. check out IRC, HVR to
name a couple. We bought them from a japanese company, but I forget who.


Cheers
Terry

[snip]

 

[Terry Given]

Forgot to say. My Mom, whose job 20 years ago was to process stolen US
technology for Soviet industrial and military applications, just gave
me a big lecture about how dangerous it is for amateurs to create such
big power projects.

cool!

I was doing some browing a while back, and found a whole bunch of white
papers on nonlinear control and Liapunov's method - in the 50s/60s the
US got very interested in russian control systems, and their analytical
techniques.

So, that reinforces the safety message and I want the snubber circuit
to be the pillar of safety of this whole thing. To save my ass in case
if everything goes wrong and all my mistakes somehow create a domino
effect.

I also installed a thermal normally closed switch that would turn the
circuit off if the heatsink temperature exceeds 70 degrees C, for the
same reasons.

smart move.

i

Cheers
Terry

 

[Terry Given]

Yep. Too bad.




So, what kind should I look for, carbon film, or what?

carbon composition.

I took some formulas from a Fuji IGBT guide. I will appreciate some
clarification.

As you know, I already have a working bridge:

http://igor.chudov.com/projects/Homemade-TIG-DC-to-AC-Inverter/01-Prototype-1/

but without snubber.

Feel free to disagree with what I will say: Essentially, in a shorting
H bridge (one that is never fully open), the snubber is not used for
much other than to suppress internal ringing on power rails going to
half bridges that open and close. Which requires relatively small caps
on each IGBT (which I will provide).

yep.

So, the job of the big snubber is to handle wrong situations like
switching the bridge off while welding (ideally should not happen),
wires going loose etc.

OK

What I want is to never have voltage on the DC supply rails that
exceeds 150 volts. That is so because my welder specifies max voltage
as being 150 volts, and I do not want to ruin it.

OK

That means that the resistor should allow all 200A of welding current
(let's say 400A for safety sake) to go into the capacitor at 150 V,
that means

400A = 150V / R

or R = 150/400 = 0.375 ohm.

max. Fair enough, lots of power so low impedance.

Got it. You are 100% right.




Well, I was not clear. The resistor that I tried was not that
wirewound resistor. I tried a small, golden colored, heatsink
mountable ribbed resistor. I am not sure what's inside it, but it does
not seem to be wirewound.

it will be. smash it to bits and have a look. you can (theoretically)
get non-inductively wound wirewound resistors, where the wire zig-zags
rather than spirals. I've never seen one though....

Thanks Terry. Since I have the bridge basically working, having a
super safe, idiot proof snubber is now my main focus.

i

No worries mate

Cheers
Terry

 

[Ignoramus1487]

I have pulse-tested these resistors before. The peak pulse power rating
is not much more than the average rating - its a thin film.

carbon composition resistors *are* still made. check out IRC, HVR to
name a couple. We bought them from a japanese company, but I forget who.

I thought that the above ones (digikey) were carbon comp ones. Am I
selecting wrong ones? If so, which ones would you suggest.

 

[Ignoramus1487]

carbon composition.

Thanks. I ordered a 35 W carbon composition resistor in a TO-220
package. I have some TO-220 heatsinks lying around.

OK

I am glad that you agree.

max. Fair enough, lots of power so low impedance.

Yes. Exactly. I want it to pass the current along and then dampen it.

it will be. smash it to bits and have a look. you can (theoretically)
get non-inductively wound wirewound resistors, where the wire zig-zags
rather than spirals. I've never seen one though....

You are 100% right, it is definitely a wirewound POS.

No worries mate

I will play with my working bridge more over thanksgiving. I have a
bigger DC power supply an will try to make an inductor out of low
voltage winding of a transformer.

i

 

[Terry Given]

I thought that the above ones (digikey) were carbon comp ones. Am I
selecting wrong ones? If so, which ones would you suggest.

http://www.irctt.com

the website doesnt work well with firefox. but they have peak pulse
power/energy ratings in their datasheets. eg a CHP-2 resistor is rated
at 2W, 25C and 1.33W 70C, but can eat a repetitive 800W pulse thats
100us wide....


Cheers
Terry

 

[Ignoramus1487]

http://www.irctt.com

the website doesnt work well with firefox. but they have peak pulse
power/energy ratings in their datasheets. eg a CHP-2 resistor is rated
at 2W, 25C and 1.33W 70C, but can eat a repetitive 800W pulse thats
100us wide....

Thanks Terry. I bought this TO-220 resistor:

TDH35PR150J

Digikey part number TDH35PR150J-ND.

If it does not work, I will experiment a little more, I have my eyes
set on something else as well. I appreciate your help. I am very
eager to test that stuff under a little higher load and worse conditions.

i