Anyone ever hear about a frazer-nash gokart

[Functional Artist]

I know how it IS set up (voided provided a schematic himself)
that one's easy enough to follow to not need any update(or clean up) on my behalf.

I wondered why it it setup that way and provided a schematic how I THINK it SHOULD be setup instead. 'sid

Alex, (Really?) I am simply trying to HELP figure THIS system out.

This is a professionally designed system
...& your basically suggesting to rewire it to how you "THINK it SHOULD be setup instead"?

IMO it just complicates the situation
...unless he starts all over from "scratch" & were NO where near that point yet.

 

[itsid]

Kevin, if it IS working correctly, then why isn't it working correctly?
So yes, I suggested to NOT RULE OUT a previous owners mistake

I made clear (more than once) that I cannot say the wiring (as it is atm) is incorrect.
I just said, that all controllers with a comparable setup are wired differently.
And yes that does imply it might be indeed incorrectly wired up as it sits.

Look at the wires.. see the incorrectly crimped connectors?
that's not been closed by a crimping tool, but pliers...
An added trimpot to adjust the potbox.. that's not original either I bet..
So professionally designed and then untouched my 455!

But frankly.. you're really pi55ing me off atm,
But this is not about you .. it's not even about me,
it's about giving voided as much information
(some only slightly relevant some maybe not relevant at all)
at hand, to bring his kart back to life.

You have a different approach to almost every electrically related thing,
I got it.. I do not agree on many things but as long as you're not about to kill yourself I don't give a shirt.
Please do me the favour and allow me to do things differently to what you think is correct as well.

I myself would've minimum wired the thing twice already
and checked if it moved or not..
BUT that's not sound advice, just because the controller might kick into safemode and could not be resurrected from there easily..
IDK so I don't advice to do so.

I never said voided SHOULD rewire the system... since that again would be ill advice at this time

But again, you do not read what I'm saying ...
you crossread every third word and make up the rest.
And that usually results in giving the worst possible advice aswell.

'sid

ps the kart maybe was big in malaysia??
http://ww1.utusan.com.my/utusan/inf...ub=Utusan_Express&sec=Front_Page&pg=fp_08.htm

 

[voided]

Never seen that article and I’ve looked all over the place.. gonna see if it’s possible to contact them. That article claims $3400 to build a kart. In the 90’s that seems way to cheap.. .

As for the crimped ends.. they were done with a crimper, I believe they used the wrong size for the wire and had to over crimp/ distort for it to hold..

 

[itsid]

There are plenty other articles regarding the company..
short story:
POEM built the karts for malaysain rental tracks..
and from the revenues start to finance their electric vehicle production..
which failed, they then got the national powercompany on board for help and still failed to get production up and running they then decided to buy in off the shelf parts instead of making them..
and it was essentially too late.. and if I understand it correctly, they're now no longer active (as frazer nash is nothing more than a dead website nowadays really)

BTW proper crimps (industrial not hobby) are done from four sides, not two.
look at the C and D connectors (microswitch potbox), those look to be crimped correctly
as are the waterproofed ones on the switch(O and P - I suspect).

Anyways
back to topic ..

have you checked the potbox microswitch yet?
if not, please do.. it needs to be connected in a NC (normally closed) setup, since you have a three terminal switch I suspect it to have both the NO and the NC terminal
it's counter intuitive since the potbox lever activates the switch when in rest...
So no continuity when the lever is in rest, and continuity when the lever is throttle is activated, still means normally closed.

remove the spade connectors and test for continuity of the switch in both throttle positions.
verify it working and we have another checkmark.
(reconnect)

Next: remove the green and yellow from the controller and test for resistance
if you can test before the potbox value is changed by the trimpot.
slowly increasing /decreasing throttle at least four cycles..
observe for hiccups in the change of resistance, and if there are none (no sudden jumps, no dead spots etc)
note down the min and max value.
and now measure the resistance of the trimpot as it sits (green lead both ends)
just note it down for now.
and yet another checkmark will be set.

Sorry, I forgot? the precharge resistor was checked, right?
The fuse and switch are okay as well IIRC..

what about the large battery disconnect, is that acting as it should?
if so I think the inside of that mystery box is now fully checked and confirmed to be working.

Which I'm afraid leaves us with not too many options, does it?
ideally incorrect trimpot settings and or potbox misvalues

What are these? and how does it change (0-5kOhms or 5k to 0 Ohms when throttle is applied)
And yes, continuity and resistance, do not check voltages!

(in fact it's best to leave the battery disconnected when ever you test some static value,
that way even if you accidentally short something with the testlead you have a much better chance of not frying the controller )

Have you conducted the potbox test from earlier (post #24) already?
that I'm afraid needs to be done under battery voltage of course
(otherwise there's little chance the controller triggers the boot up sequence )
but make sure the rear is jacked up, since the wheels will spin in best case

And if I just assume that too has been done..
there is very little left that could cause the setup not working

a) the motors are shot
b) the controller is toast
c) the wiring is incorrect

Easiest to test at that point are the motors.
disconnect the motor and battery wiring from the controller (take notes what goes where)
Oh and they must be wired differently since I assume them to be identical and thus one must rotate clockwise the other one counterclockwise in order to not turn the kart in circles

now concentrate on the left motor first
Now prepare to get hit, then touch the motor wires to the battery terminal (I suggest you only test with 36V say the left batpack, just to lessen the jerk.
does it turn?
repeat with the right side motor.. and?

Where are we at?

'sid

 

[Functional Artist]

After cutting 100+ zip ties, removing a chain shield and a few other plates/panels.

Take note, I removed the rusted screw to get inside there. So I doubt anyone was in there before me..

These "facts" seems to indicate that it's still wired "as designed" & has NOT been tampered with.

---------- Post added at 12:46 PM ---------- Previous post was at 12:38 PM ----------

Kevin, if it IS working correctly, then why isn't it working correctly?

That's what "we" are tryin' to figure out.
IMO it's most likely a faulty component

But frankly.. you're really pi55ing me off atm,
...you crossread every third word and make up the rest.
'sid

How do you figure that?
...because I "respectfully" disagreed?

 

[itsid]

Kevin, you are currently in elmoto mode,
take a breath, re-read and come back when you want to reply to voided's problem

'sid

 

[voided]

Been busy...

I can say the motors are monsters... hit them with 72 volts. And let me tell you. Those bad boys spun..


I’ll try the other stuff and report back.

 

[itsid]

uhoh..

I mean great... but that removes yet another possible cause for failure

I'd rather replace a set of brushes in those motors than say trace and fix a broken MosFET inside the controller or -more precisely- talk you through tracing the controller internals.
Since that requires a ton of high resolution well lit images just to make out pcb traces and such...tedious work and you will need a few decent electronic tools as well.
And worse: from what the web archive was showing they used a propietary microcontroller to do all the 'mental' work.. not the slightest chance to see if that's still doing it's work without a full fledged Oscilloscope...if there is no 'smoking gun' to be found we'll be pretty much screwed.

but let's not get overly discouraged for now ..
let's just hope you find the small switch that isn't doing as it's supposed to.

And in case you don't..
I indeed will suggest you to rewire (minimum wiring I'd say) the controller before we crack the bad boy open.
in order to make sure we don't mess up I need yet a closer look at the screw terminals of the controller (hoping to find a faded terminal label or two)
But let's cross that bridge when we get there.
Just so that you know to take pics in case you haven't found a defective part once you tried everything else

'sid

 

[voided]

I’ll get out there Sunday and take some tests and pictures.. trying to get a friend over that knows more about the electric stuff than i do..

I could possibly just take the motor leads off and the battery leads off and take that to my friends house..

My battery is portable and we could bench test everything along side his electrical testing equipment ... just need to find if he has time.. I have been trying to get him over since I picked this thing up..

 

[Functional Artist]

The ports on the speed controller are
3 power connections .. ground to left motor / positive from contactor to both motors/ ground for right motor..

Harness wire with plug to control box as stated about with the red/blue/green/yellow.

I've been studding this system a bit more.
It indeed seems to be set up quite "differently" than what we have seen & discussed on this forum previously.

As Voided has described (& as it clearly shows in the pic) the contactor DOES seem to deliver power to the controller. (or at least to a junction terminal attached to the controller)

Notice, the (3) positive (+) cables connected to the "center" terminal on the controller.
This would the (B+) terminal
It looks like the cable (with red band) brings the positive (+) from the contactor to the controller & the other (2) cables deliver the positive (+) out to each of the motors.

Then, it looks like (& as Voided has verified) there are (2) negative "motor" terminals on this controller.
These would be the (M-) terminals
The left motors negative (-) cable goes to the left terminal on the controller (clearly visible) & the right motors negative (-) cable musty/probably goes to a right terminal on the controller. (not clearly visible)

But then, where/how is the negative (-) connecting the battery pack to the controller?
This would/should be the (B-) terminal
Is it grounded to the chassis?
...we'll need an answer to that one.

As for the small harness/connector, the small red wire seems to supply the (switchable) power to the positive (+) side of the contactor & controllers circuits. The negative (-) to complete the controllers "logic" circuit must be provided within the controller.

The small blue wire supplies the (switchable) ground to supply the negative (-) to the contactor (via the micro switch in the potbox) to complete the contactors "activation" circuit.

The small green & yellow wires are the throttles (potbox) input signal wires.


So, it looks like, the positive (+) power is only connected/supplied to the controller/motors when the throttle is activated. (when the microswitch turns the contactor on)
…& the negative (-) is only connected to each of the motors when (& for as long/duration) as the throttle (potbox) "signals" to the controller.


It seems to be set up like them old golf carts where when you press the accelerator, you hear a click (contactor switching on/closing the circuit) & you go.

Then, when you let off of accelerator, you hear another click (the contactor switching off/opening the circuit) & you just coast.

 

[itsid]

I....
But then, where/how is the negative (-) connecting the battery pack to the controller?
This would/should be the (B-) terminal
Is it grounded to the chassis?
...we'll need an answer to that one.
...

*cough* you are again NOT reading what has been said already

After cutting 100+ zip ties, removing a chain shield and a few other plates/panels.. I’ve found a 4th lead at the speed controller for the negative that goes directly to a large battery disconnect then to the battery.
....

Anyways, we already have settled on what "seems" to be going on..
the puzzle we need to solve is what NEEDS to be going on.

you know why the throttle needs a microswitch?
it's to give the controller a split second to actually activate the throttle
(hot boot sequence).. switch toggled 'on'...
and it starts cycling:
*read throttle pos as 'zero' reference
*engage powercircuitry
_cycle start:
*check for brake signal
*read throttle pos against previous zero
*adjust speed
_cycle restart until zero or switch toggled 'off'...

you know what happens if the throttle switch triggeres the wrong pin on the microcontroller
(pulling pin 19 instead of pin 12 high let's say) ?
NOTHING!
absofrickinglutely not a single thing.
controllerwise it's the same thing as you trying to switch on the bedroom light by flicking the kitchen switch.
in this case it could as well be a timing issue
(depends on if the blue wire is low or high when the microswitch is disconnected)

Sure.. there's a good chance the controller is indeed dead
and that's why the throttle doesn't show any effect on the motors.
but it's not the only reason left to think about

'sid

 

[voided]

I’ll be out there Sunday.
There is a battery to controller ground lug under the left motor ground that can’t be seen in the pictures.. it is not grounded to the frame..

As for the way functional artist explained about the electric golf cart throttle... it sounds exactly like that. As I was on a golf cart today and i wanted to tear into it and see how it was wired after giving it throttle and hearing it click... I did try on the golf kart giving slight throttle, heard the click and was able to not give it enough throttle to move it. I also looked at the whole brake system and don’t see any fail safes there.

 

[itsid]

that's the point..
electronics are not always that simple..
what was described as ground might not be one at all..
in fact the two lines (red and blue) only need to be at different potentials
to activate the contactor.
(say for example one at +72V one at +60V) leaves us with a 12 volts potential difference.
Think of your batpacks.. what's the plus side for one battery is the negative for the next in order to step up the potential.
and if you want 12V to power an automotive lightbulb
you could take negative from the third battery's plus side
and positive from the fifth battery's negative
to see the 12V from the fourth battery in line;
+ and - are 'mislabelled' and none of that is actually ground

Don't get me wrong I hope he's correct,
that blue is indeed ground and no "inbetween" potential.
In fact I hope it's as simple as fixing the resistance from the potbox or finding the dead external switch.
but if you're out of luck with that tomorrow,
blue being ground raises the chance it's wired incorrectly
(an inbetween potential would be an indicator for more shenanigans going on than just toggling the potbox readout )

What I would like to ask you for -and the reason for this post-
is to take the wiring diagram you drew and note
down all voltage potentials on it
(mostly interested in the mystery box contents; but if you are willing to include the motor and controller terminals we'd have some information that might come in handy later on )
in three states I'm afraid:

1) main switch off (please verify 0V on all lines but the contactors power line (72 and 68 IIRC))
2) main switch on, potbox at rest
3) main switch on, potbox at full throttle

probing all terminals
if at all possible against battery ground (chassis ground when the first battery is out of reach)

jack up the rear,
if the potbox is out of sync or misvalued, the probing can upset the resistance,
and in best case the motors start turning in the third round (verifying a resistance issue).
Do not get hurt

'sid

 

[voided]

I might have some help later with the testing, here is the whole system in ms paint...

 

[voided]

Contactor , pot box, trim pot, switch, microswitch, resistor, and everything has tested correct according to my electrical engineer friend and the guides from sid and functional artist...

The controller does have some fail safes that we found .. without motors connected , controller will not signal power to the motor terminals. It does have to see resistance.

We are gonna test some more tomorrow and get some more feedback once We hook the controller / control box back up to the kart ...

We opened the speed controller to test the leads for the potbox and the red and blue wires at the source on the board.... blue is ground and red is positive. The microswitch does not work if the wires are swapped around.. or on the center contact.

 

[itsid]

uhhhhhhh
I see something
might be the image me tricking into false assumptions...
but check the zenner diode @ D15 it looks darker than D10 below,
might be flash glare on D10 fooling my eyes or a burned D15
Also R59 .. looks odd on the pic with one glossy end
(might be again a lighting issue and my old eyes failing..)

And lots of 20 year old electrolytic capacitors...

I only have one or two that old that not yet failed.. and I swapped about three dozen that weren't even half that old.

when they heat up for one reason or another (I mean they're thermocoupled to the case for a reason )
they could crack their seal and dry out if unused, or worse they cook off and leak
they do not appear to have leaked (the thermal tape would have turned yellow)
but I'd suspect a good chance of at least one being dry by now.

So keep this in mind
one of the first electronic check routines to be the canned caps
(all incl the smaller smd versions.. they are not as prone to failure as the larger ones.. but wet caps and 20 years of age... you want them tested )


Btw:
I see a nice LED right there.. usually manufacturers are 'smart' enough to include a debug function (dem blinkin lights and such if it's the only control light they include)
looks like monochromatic to me (likely red.. maybe green.. doesn't matter much)
does it fire up and stay lit, or does it blink?
if so how?
blinking a few times staying off for a green light usually means failed testroutine
blinking then staying ON for a red light could (doesn't need to) mean the same.
and/or does the blinking speed change when booting up the controller?
(red led in fail mode usually blink slower than before if "permanently on" means all good)
Since we have no manual we don't know what any of this means ...
but we have something to base some assumptions of

no reaction of a LED at all is never a good sign of course.

Have you traced the wires /pcb traces back to ground & +72V or have you just poked it with a multimeter?

the microswith like yours is commonly an OR gate
(leftmost in the pics would be IN, middle one A and right one B.. should be drawn on the case really)
A and B never connect, and the switch toggles IN to either A OR B
So if you swap the wires on the microswitch (still using the outermost two terminals) nothing would change electrically.. IN and B would still connect at the same switch state they do now.

So that has me baffled... you sure the switch is still good not just intermittent?
(oh that would be perfect.. tiny microswitch $2 incl shipping.. fixed! I'll have my fingers crossed )

'sid

 

[voided]

LED is permanently on once switch is on.

All the diodes and other soldering joints are uniform/ same color . Flash / lighting does make it look different in the pictures. Some of them do have black rtv? On them ..

Nothing smells burnt, looks burnt, or anything odd / out of place..

My friend and i are stumped and I will be bringing the whole kart to a local maker /hacker space that has a few more testing instruments that we don’t have..


Hopefully I can report back with some positive findings.. without throwing new parts at this other than some diodes, resisters , or whatever else inside that controller my need.

 

[itsid]

Yeah that's the bad thing about microelectronics,
you often do not have a smoking gun to see what went wrong.
especially not with electrolytic caps that went dry

luckily you can almost verify their failure with simple tools without desoldering them first.
(unfortunately you cannot always verify their fitness )

Best of luckfor that! (keep us posted!)..

Aaaanyways, while we're so comfy at this camp fire atm..
why don't you give a minimal wire a chance to see how that goes?

skip all switches except for the contactor
(basically set all switches to the "on" position by interconnecting their respective leads directly)

something like this:


(the full setup for that alltrax btw looks like this

the "on/off" switch in the diagram is actually the microswitch on the potbox in your case and the "key switch" in the diagram is your on/off switch ..
as you can see yet another mfg that toggles the switch on the positive lead
just saying with microcontrollers, timing can be important...

once you skipped all switches it doesn't matter anymore
so stick with your setup

it's a rather quick test and if it moves you don't need to tediously test the controller
since frankly if the LED lights up and stays lit it's not too common to see a failed testroutine from within the microcontroller... then again who knows how common "common routines" are to uncommon controllers
sure it could still be that the MosFets are no longer working
(a transistor tester for ~10 bucks can check them rather reliably [like the "M328 AVR Transistor Tester" chinese copy of a german design.. cool stuff! I have one myself])
MakerSpace should at least have such (or quality made equivalent).
it's also able to test the caps rather reliably.. albeit it might have troubles with some when they're soldered in (misreading of close by components)...

R59 btw has a non perfect solder joint (you can see it mushrooming peeking in from below)
but if that's the fault the LED should stay off so.. yeah...


'sid

 

[Quinc]

Ok you have had plenty of time! When we going to see some video of this thing in action?!

 

[itsid]

yeah.. I was afraid to ask myself tbh..
I feel like the hackerspace guys did their best to measure and test..
and found a dead brick controller..
And now voided is in search for a properly sized dual motor replacement
which I'm afraid will be a unicorn hunt

IDK I soo hope I'm wrong and he just forgot to let us know once the kart was running
(excitement overwhelmed eating, drinking, sleeping on the driver seat ever since)

curious minds need to know

'sid