Slant Six Forum
https://slantsix.org/forum/

Experimental Active Shunt using one Si diode
https://slantsix.org/forum/viewtopic.php?t=21081		 Page 2 of 2
Author:  supton [ Fri Mar 16, 2007 12:31 pm ]
Post subject: 
How did this go?

Had a brain fart just now, so I dug up this thread. Instead of doing a diode shunt, has anyone thought of a resistor instead to shunt it? Make a resistance that is equal to the ammeter resistance, and only one half the current will flow into the ammeter. The downside being that the ammeter is half as sensitive.
Author:  dakight [ Fri Mar 16, 2007 2:58 pm ]
Post subject: 
You don't really want a resistance in series with your ammeter circuit because virtually everything in the car except the starter motor itself feeds from that circuit. I suppose you could put a shunt in parallel with the ammeter, having identical resistance ( very small ) as the meter, then half the curent would flow through each segment but I don't believe it works out so well inparctice. For one thing, the resistor would have to be able to handle a continuous current of several tens of amps; it would need to be rated at 500 watts or better. I think the active shunt using a diode is the most reasoned solution since it allows current in excess of a predetermined magnitude to be shunted around the Ammeter while not interfering with its normal function at more normal loads. Many folks just yank the ammeter and replace it with a voltmeter.
Author:  supton [ Fri Mar 16, 2007 3:20 pm ]
Post subject: 
Yeah, I was talking shunt. After thinking about it, you're right, it does seem to be not working out: power dissaption of a diode is approximately 0.7V*current; power dissapation of a resistor is R*current_squared. At 100A, the power dissapation of the diode solution is roughly 63W (assuming 0.06 ohm ammeter and around 10A into the ammeter); the parallel shunt would be upwards of 100W (thinking 0.02 ohm and 75A into shunt).

I was thinking of the shunt being much smaller (like 0.01 max), where only about 1/7th of the current drain is "seen" by the ammeter. But that doesn't seem to work out either. And, the ammeter would seem kinda lazy at that point.

So, nevermind--bad idea. Didn't think it out before opening my mouth (keyboard). The diode does make best sense.
Author:  dakight [ Fri Mar 16, 2007 4:56 pm ]
Post subject: 
When I was said 500W is was thinking of the P-IE formula but it escaped my senile brain that the E in this case would be the voltage drop ascross the resistor, not the overall voltage driveing the circuit. Still, 100W is a lot for a resistor to have to dissipate continuously. It's been a long time since I played with electrons, a little rusty and a lot of the technology has changed even though the underlying fundamentals haven't.
Author:  sixsignet [ Sun Apr 29, 2007 3:18 pm ]
Post subject: 
I am back to work on this and having trouble finding high-current schottkys in DO-5 or DO-9 (to fit my heatsink).

I have a digital meter that can measure 20 A for 30 seconds - 10 A continuous and I have an old analog meter with a 2.5v range to show voltage drop.

What non-schottky diodes will have a forward voltage drop of 0.60 to 0.70?

I am considering this 85HF100 diode.

It starts conducting at 0.68v, but how much forward current will it conduct at 0.75v?
At least 30 amps @ 0.75v would be good for my 65 amp alternator.
.75 volts at the diode terminals is just past the C line on my ammeter.

http://tinyurl.com/28d6xv
Author:  supton [ Mon Apr 30, 2007 6:48 am ]
Post subject: 
Hmm, that diode shows like 0.7V at 10A--and maybe 0.75V at 100A. Nice flat response. The graph doesn't go into detail for such low currents.
Author:  sixsignet [ Mon May 07, 2007 12:48 pm ]
Post subject: 
Well, the 85HF100 diode shunt barely works.
(actually, if the goal is to take the load off the ammeter circuit before it carries too much current, it arguably doesn't work)

The shunt barely starts working just before the ammeter's needle reaches the C line.
With the needle on the C line, the diode's voltage drop (Vf) is 8 volts and the current through the shunt is 5.8 A.
That's the most current and voltage drop I recorded, and it was only for about 10 seconds at 1500 rpm after a half hour leaving the headlights on. After that 9 or 10 seconds, the needle dropped below the C line, the Vf dropped to 4 or 5 volts and the shunt current was zero amps.
The other diode allowed the needle to go about 1/8 inch past the C line under similar conditions.

I'm guessing the needle has to travel further east before it literally pegs, but I would feel more comfortable if the ammeter didn't have to work so hard after the shunt kicks in.

The original active shunt by KBB_of_TMC used two 100BGQ030 Schottkeys in series on a 1975 Duster.
http://www.tidewatermoparclub.com/TECH/ ... shunt5.pdf

I cannot account for the apparent difference in voltage drop of the 1963 alt-to-starter relay path, but it seems to be quite less than the 1975 circuit.

I think a single 100BGQ030 diode would work fine on my 63 Valiant.
The needle would stall about halfway to the C line, but the shunt would turn on frequently.
Wouldn't that be desirable over my present setup?

I wonder how KBB's shunt is working on that 75 Duster.
Author:  KBB_of_TMC [ Thu May 10, 2007 10:17 am ]
Post subject:  Duster
I can't tell you how well the shunt is still working - the car was won and then a month or 2 later sold on Ebay. It worked at least for a few months OK, but that's all I know.
Author:  sixsignet [ Thu May 10, 2007 4:05 pm ]
Post subject:  NTE6094 Schottky
Now I'm trying a single NTE6094 and it works just right.

Its maximum forward voltage drop is 0.6v


The first drawing is at 0.4v drop.
The second drawing is at 0.5v and 0.56v (current almost doubled but needle didn't move)

diode_____________ shunt amps

0.35 v________________0 A

0.4 v ________________1.3 A

Image

0.45 v________________3 A

0.5 v ________________7 A

0.56 v_______________12 A

Image

The heatsink handled the heat generated by 20 seconds of 10-12 amps without the fan turning on.

I wonder why mine works at 0.6 volts, while the other one worked well at 1.2 volts.
It was very tricky finding a diode that turns on at the right point.
The 60 amp NTE6094 is not ideal for my 65 amp alternator. I need a 80-100 amp diode with the same characteristics as this http://www.nteinc.com/specs/6000to6099/pdf/nte6094.pdf

I wonder if a shunt with an adjustable turn-on point could be built around a 3-legged thyristor of some type.

Here's the unit mounted on a 1/4" hard neoprene sheet near the voltage regulator. The neoprene extends about 5/8" at the terminal block side and was trimmed to fit the other three sides.
The little grey cable that goes under the VR is for a temporary external fan LED and will be removed soon.

Image
Author:  supton [ Fri May 11, 2007 8:19 am ]
Post subject: 
I'd guess that the resistance of the ammeter isn't quite constant. That might explain why the difference.
Author:  SlantSixDan [ Fri May 11, 2007 9:53 am ]
Post subject: 
Quote:
I cannot account for the apparent difference in voltage drop of the 1963 alt-to-starter relay path, but it seems to be quite less than the 1975 circuit.
My guess: In '63, Chrysler were doing great. They were building high-quality cars and air conditioners everyone wanted, all over the world, they had the best warranty in the business, their suppliers were happy with them, there was no labour strife, and there were no money worries.

In '75, Chrysler were struggling to stay above water. They were building mediocre-quality cars nobody wanted, their suppliers and their workers were permanently pissed at them, and there were very serious money worries circling overhead.

Thinner wire and smaller contacts cost less money.
Author:  sixsignet [ Fri May 11, 2007 9:55 am ]
Post subject: 
Maybe the 1963 ammeter is slightly different electrically than the one used in 1975.

Maybe the sixties cars up to some point had an 0.025 ohm ammeter and later cars had a 0.05 ohm ammeter (0.03 ohms and 0.06 ohms, including the wiring)

That would explain why a 0.6 volt shunt works on a 63 and a 1.2 volt shunt works on a 75.

see K. Beard's calculations on pg 2 of the active_shunt5 article:

"The resistance of the stock path in good condition from the alternator to the starter relay
is around 0.06 ohms;that means 20A flowing though it will cause a ~1.2V drop"
Author:  KBB_of_TMC [ Fri May 11, 2007 3:45 pm ]
Post subject:  shunt
I'm very pleased that somebody doing a much better job than I did!

Please remember that my numbers for the '75 Duster's resistance was estimated from on the car measurements and could well reflect a less-than-optimal condition somewhere along the path - another car might give different numbers.
Author:  sixsignet [ Fri May 11, 2007 6:39 pm ]
Post subject: 
Thanks.

I think it could be much smaller using your diode, a 2-terminal block, and a small socket 7 heatsink-fan.

I wonder if the fan could be controlled by an inexpensive current-sensing circuit. (don't know if there is such a thing)

I wonder if you really need a fan for 25-30 watts max

Now I can't wait to give somebody a jump...

This project was based on stuff I read here:
http://www.slantsix.org/forum/viewtopic ... ing+wiring

************************

I think you could make a simpler, better shunt using one of these mounted on a big heatsink, with no fan and no terminal block.

http://rocky.digikey.com/scripts/Produc ... M=121NQ035

I like the D-67 Half-Pak package

---------------------------------------------------------------------------------------------------------------------
WARNING:
Your results may vary.

My ammeter is fed by the original wires *plus 12 ga wires that bypass the bulkhead connector.
I cleaned the ammeter and bulkhead connectors and treated them with contact cleaner. All of the other connectors are new.

~summary of other mods~
8 ga from alternator to shunt
10 ga from shunt to bulkhead connector*
10 ga between bulkhead connector* and starter relay
4 ga red battery cable
2 ga negative battery cable
two 8 ga block-body ground straps
12 ga ground loop from HEI to alternator to voltage regulator

note: the alternator-shunt-bulkhead connector path is 36 inches longer than the original alt-to-bulkhead wire.
Page 2 of 2 All times are UTC-08:00
Powered by phpBB® Forum Software © phpBB Limited
https://www.phpbb.com/