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

Experimental Active Shunt using one Si diode
https://slantsix.org/forum/viewtopic.php?t=21081		 Page 1 of 2
Author:  sixsignet [ Tue Dec 12, 2006 12:47 pm ]
Post subject:  Experimental Active Shunt using one Si diode
My version of an Active Shunt is about ready for testing.
This is for 1963 wiring with a 65 amp alternator.
I added a small fan controlled by this circuit. The thermistor is strapped directly to the power diode.
Image

http://www.rason.org/Projects/fancont/fancont.htm

My silicon diode will turn on at 0.7 volts, whereas KBB's series diodes turn on at 1.2 volts.
That means it will turn on when the alternator output exceeds 12 amps or so.
That may be too soon and too often...
This is the diode, heatsink and terminal block that is the heart of the thing:
Image

I don't have a clamp-on current meter or anything that can accurately monitor current up to 65 amps, so I'm not sure how to test it to the alternators maximum output.
I'm going to put a remote LED on the dash that lights up when the fan turns on.

You can see the small leads of the thermistor and the red and black wires bring in Ground and switched 12v from Ignition.
Image

The fan control PCB is mounted in a box with the fan and there is a 4-pin connector that connects to the remote thermistor and the red
and black wires for PCB power.
Image

The box fits together with tabs and then squeezed next to the heatsink.
The fan is mounted with spacers so it is 1/4 inch away from the heatsink.

The whole thing fits in this Hammond 1554PGY box, which was the only small box I could find with the required 2.7 inch internal height for the heatsink-diode.
Its external dimensions are 4.7 x 4.7 x 3.15 inches and it will be mounted on the shelf near the master cylinder and voltage regulator, about five fingers from the firewall.
Image
Author:  KBB_of_TMC [ Thu Dec 14, 2006 11:58 am ]
Post subject: 
Very nice! I really like the addition of a fan - I'm working on a battery isolator I'm now building out of a broken & gutted PC supply - it has it's own thermal switch on the heatsink already. I have not measured its power draw when off; if it's too high, I may have to run it off IGN.

I've found that the inexpensive inductive ammeters that just sit on the wire work quite well; I got mine from Sears, but I don't think they sell them anymore and I got another from KD, I think it was. It is just some little magnets and plastic, but it was good to 20% or so.

Good luck, KBB
Author:  slantvaliant [ Thu Dec 14, 2006 12:07 pm ]
Post subject: 
Where is that emoticon with the little guy eating popcorn, watching intently?
Author:  sixsignet [ Thu Dec 14, 2006 12:19 pm ]
Post subject: 
Quote:
Very nice! I really like the addition of a fan - I'm working on a battery isolator I'm now building out of a broken & gutted PC supply - it has it's own thermal switch on the heatsink already. I have not measured its power draw when off; if it's too high, I may have to run it off IGN.

I've found that the inexpensive inductive ammeters that just sit on the wire work quite well; I got mine from Sears, but I don't think they sell them anymore and I got another from KD, I think it was. It is just some little magnets and plastic, but it was good to 20% or so.

Good luck, KBB
Thanks
You can get 100 amp terminal blocks at Boaters World or similar store that caters to boat owners.

If this thing turns on too often, I think I'll cut the heatsink in half and mount your two schottky diodes.

That amber LED in the round window will be relocated to the instrument panel for tests.
Image
Author:  SlantSixDan [ Thu Dec 14, 2006 12:19 pm ]
Post subject: 
Image
Author:  slantvaliant [ Thu Dec 14, 2006 3:24 pm ]
Post subject: 
That's the one, Dan!
It depicts me quite accurately on a number of topics on this forum. Watching, reading, thinking, and sometimes even learning!

Keep us informed on how this thing works out.
Author:  argentina-slantsixer [ Thu Dec 14, 2006 5:29 pm ]
Post subject: 
at the risk of making a fool out of myself... (well, been there so many times, here I go again)

what's that neat device for? I checked the site and the circuits are awesome, clean and easy to work. What's the use for an active shunt like this?
Author:  sixsignet [ Thu Dec 14, 2006 5:53 pm ]
Post subject: 
Quote:
at the risk of making a fool out of myself... (well, been there so many times, here I go again)

what's that neat device for? I checked the site and the circuits are awesome, clean and easy to work. What's the use for an active shunt like this?
Read "Upgrading the wiring, particularly the ammeter"
I borrowed ideas from KBB_of_TMC and 440_Magnum.
(and borrowed that thermal fan controller design from a HAM operator)

The picture with the lines drawn on paper (above) shows how it is hooked up
Author:  argentina-slantsixer [ Thu Dec 14, 2006 6:36 pm ]
Post subject: 
thanks, sorry
Author:  slantvaliant [ Thu Dec 14, 2006 7:13 pm ]
Post subject: 
It's used to protect the ammeter and associated wiring by bypassing it during heavy current-draw situations, when using an upgraded alternator. Really dead battery, jump starting, etc.
The charging system your car came with was sized for the expected usage back then. That includes the alternator, wiring, battery, and ammeter. Lots of folks use more juice these days. If you significantly increase the charging capacity and then actually use it, there can be problems with the old wiring. It can get ugly.
Author:  sixsignet [ Wed Dec 20, 2006 12:04 pm ]
Post subject: 
Quote:
Image
It's installed.
The thermal fan PCB and remote LED on steering column tested good with a hair dryer.
I added a pair of wires to the diode's terminals and routed them to a cheap multimeter at the driver's seat.
I taped the leads to the test probes and set it to DC VOLTS to look at voltage drop across the diode.

I started the car and idled at 1000 rpm: + 230 mV drop

I turned off the engine and turned on the radio+heater+highbeams for 20 minutes (this showed as negative 194 mV across diode)
I turned off everything and started the car @ about 1000 rpm:
+790 mV across the diode for about ten seconds.
At 790 mV, the ammeter's needle was slightly past the C line, but not pegged.
I expected to see the needle jump down a step when the diode turned on, but did not. Maybe it happened too fast for the gauge to indicate.
It tapered off quickly after that, hovering around 250-300 millivolts with the ammeter's needle halfway to the big C.

I think the shunt was turned on during that 10 second period, but the cooling fan and LED never turned on to confirm (likely due to 36 F air temperature)

I may have to buy a better meter than the pocket auto-ranging Radio Shack POS I am using.
Apparently 0.5 ohms is the bottom of its ohms range, so I can't use it to measure in the .01 range.
Do I have to get a high-end DVM to get accurate readings down to .01 ohms?

Next test - jump starting a dead car?

Notes: Unlike the original wiring, my 4 ga red battery cable goes directly to the mini-starter and a 6 ga jumper goes from it to the starter relay.

Then there is a 10 ga from it to the firewall.

There are added 12 ga wires parallel to the stock 14 ga wires connecting the 1963 ammeter to the firewall connector.

The alternator wire is 8 ga.

Would this possibly give me less resistance than the stock 70's alternator-ammeter-starter relay-batt path that KBB measured at .06 ohms?
Author:  supton [ Tue Dec 26, 2006 6:53 pm ]
Post subject: 
Quote:
Do I have to get a high-end DVM to get accurate readings down to .01 ohms?
Probably.

You could try doing something like adding a known load (either a resistor with a really good heatsink, or some other known load) and measuring voltage drop inbetween the two points that you are interested in. If you have a pair of "cheesy" meters, you could put one inline with the load on the current setting, and use the second in the appropriate voltage range to measure voltage drop. Just don't go over 10A (the upper limit on most of them), and make sure whatever it is won't mind a bit of voltage drop (probably 200mV, but don't quote me--that's a guess as to the resistance of the meter). You could try to parallel two meters up for 20A, I guess.

I bought some cheap cheap ones from Harbor Freight; Cummins Tools has them (they do traveling sales, real cheap stuff) also. In both cases I bought for $4.00. You can't buy DPM's that cheaply! I don't trust the AC range, and I suspect anything other than DC in (or any sort of electrical noise nearby) will cause false readings; but for ballpark measurements they seem to work ok. Walmart probably has a cheap (but not as cheap) version.
Author:  sixsignet [ Wed Jan 03, 2007 4:41 pm ]
Post subject: 
Thanks, supton.
I picked up a pretty good meter at H. Freight and will be doing tests soon.

The preliminary tests suggest that a single Schottky diode will be better than a silicon junction power diode (NTE 6050) for my wiring.

The car's ammeter is almost full scale when the shunt turns on.
If I understand NTE 6050's specs correctly, the voltage drop across it increases as the current increases, up to a maximum of 1.25 volts.

The car's ammeter points to the far edge of the letter C when the shunt turns on (when the shunt and the ammeter are each dropping about 0.7 volts). It will peg badly if the drop across the diode becomes 0.9 volts at 30 amps or so.

I want the shunt to turn on when the ammeter is line halfway to the C.

:oops:
Am I totally wrong about the diode? Does the voltage drop across the diode creep up above the initial .7 volts as the current increases?
Author:  supton [ Thu Jan 04, 2007 10:05 am ]
Post subject: 
Quote:
I want the shunt to turn on when the ammeter is line halfway to the C.

:oops:
Am I totally wrong about the diode? Does the voltage drop across the diode creep up above the initial .7 volts as the current increases?
Yes, it does. I'm not sure which diode you are using, but in general, they all follow the same characteristic: the diode voltage drop increases linearly as the current increases exponentially. For example, at 1A the drop might be 500mV; at 10A 700mV, and at 100A 900mV. Give or take. It's best to go by the curves in the datasheet (assuming they were given).

I took a quick look, and NTE6050 looks horrible for this application: at 20 *milliamps* it is rated at 1.25V drop. At 20A it is probably on the order of 1.5 to 2V, if I had to guess. I'm not too familar with such large devices, and the datasheet that I found isn't very helpful.

http://www.nteinc.com/specs/6000to6099/ ... 048_73.pdf

I'd use something else. I can't find what the ammeter resistance is, but IIRC from the other threads it was something like 60 milliohms? That'd be 1.2V at 20A. You want most of the current through the ammeter, up to 10A or so, if I had to guess--that gives pretty good indication of charging/discharging. The Schotkeys I'm finding are around 0.5 to 0.6V at 20A or more, that should work I'd think. Maybe an International Rectifier MBR7535. IIRC, your heatsink is set up for DO-5 cases, so that may work.

If I were doing this, I might think about of using a bridge rectifier, like the 36MB series:
http://www.irf.com/product-info/datashe ... a/36mb.pdf
using just two of the diodes in parallel, and using the fender as the heatsink. Not sure if it would work, haven't tried it, usual disclaimers: but perhaps something along those lines.

I took a quick look at Digikey, and only found a couple of Schotkey diodes, and they are each $20. SBR6045 came up:
http://www.microsemi.com/datasheets/msc1522.pdf
and it looks good. At the least, it has nice graphs, showing how the forward drop varies with current.

On Edit:
I'm not sure what to make of that NTE6050 datasheet--maximum forward drop at 20mA? maybe that was supposed to be 20A? [Typo's are very easy to make, and there's alot of them out there in datasheets, unfortunately... that or I'm misreading it.]
Author:  sixsignet [ Thu Jan 04, 2007 12:24 pm ]
Post subject: 
Quote:
I took a quick look, and NTE6050 looks horrible for this application: at 20 *milliamps* it is rated at 1.25V drop. At 20A it is probably on the order of 1.5 to 2V, if I had to guess.

http://www.nteinc.com/specs/6000to6099/ ... 048_73.pdf
I was unsure whether ALL silicon diodes had the same turn on voltage or not.
My picking one with a high one explains why the ammeter is allowed to go so high before the diode turns on. This is what it looked like:
Image

Quote:
I'd use something else. I can't find what the ammeter resistance is, but IIRC from the other threads it was something like 60 milliohms? That'd be 1.2V at 20A. You want most of the current through the ammeter, up to 10A or so, if I had to guess--that gives pretty good indication of charging/discharging. The Schotkeys I'm finding are around 0.5 to 0.6V at 20A or more, that should work I'd think. Maybe an International Rectifier MBR7535. IIRC, your heatsink is set up for DO-5 cases, so that may work.
Indications are that the resistance of my 1963 ammeter and upgraded wires is less than 60 milliohms - the other thread I was looking at referred to 1975 wiring.

That MBR7535 you suggested looks good for my modified 63 wiring.

I disabled the diode by removing the blue fusible link prior to taking a road trip. I have to take care of a few mechanical things before I start playing with the shunt again.
Page 1 of 2 All times are UTC-08:00
Powered by phpBB® Forum Software © phpBB Limited
https://www.phpbb.com/