Page 12 of 13

2 Speed Gearbox

Posted: Fri, 11 Jun 2010, 05:06
by T2
The purpose of a gearbox is to accelerate the mass quicker. When the change takes more time then the benefit of the gear change it results in a negative benefit (i.e better off leaving it in the same gear).

History repeats itself. Eaton Corporation's white paper published in EVS '83 described their EV powertrain which used a 2 stage hydraulically actuated gearbox. On this the technologist wrote that the best case of 2.2 secs to change gear during a 0 to 30 mph ramp gave up more time than the use of the extra gear had saved. It seems that even the use of microprocessor control could not overcome the magnitude of the problem.

Seven years later the Impact successfully outruns a Mazda Miata and Nissan 240Z to 60mph all with a single ratio.

Despite this Tesla began anew with their own two stage gearbox. It's now abandoned. The single ratio gearbox now ships in all their product.

Both the Tesla and EV1 used AC motors good for 13500rpm.

EV1 was successful,IMO, because they were prepared to limit top speed to 75-80mph. This of course lets them have a ratio > 10. In other words they chose a ratio which swapped fast for quick.

The Tesla team could not make that tradeoff. They absolutely needed 120mph for the roadster. (although they even trimmed that spec down from its original 130mph please note) They required the Tesla to have twice the acceleration of the 114Hp EV1. That was the easy part, they just selected a 200 Hp motor of the same rpm range which would give them double the torque. The intention was, at the completion of the 60mph ramp, to change gear and proceed on to 120mph.

The gearboxes were late in showing up. This was now to be a problem. Changing gear at 30mph involves spinning the motor down from 6000 rpm to 3000rpm. Not a trivial matter considering the solid rotor of an induction motor. It is not hard to imagine the same change at 60mph would produce an even more extreme situation. The gearboxes were found wanting.

Using the less favourable 120mph gear ratio would now require a doubling of the motor torque yet again in order to accomplish the 60mph benchmark. This now requires a motor of not just twice but four times larger than that of the EV1, which makes it equivalent to a motor of 400Hp peak power.   

Anyway for those who haven't yet been on the TESLAMOTORS site, the final solution was to upgrade the motor for more torque per amp and fit higher amperage transistors in the controller.

From this evidence it would appear that adopting the concept of the 2 speed gearbox is to go off on a blind alley.

So what is a good strategy ? To answer this I am going to assume ACIM usage as very few here seem to be using seperately excited shunt motors (SEPEX).
We could follow my lead and place a 15Hp inverter on a 325Vdc battery bus in order to power a 4-pole 15Hp 230Vac motor, as in my original intent, then we could gain full torque to 1500rpm followed by a field weakened region towards 3000rpm where torque decays inversely with speed. The so-called constant horsepower region.

What will this system with a direct drive to the diff yield ?
Well, if we are lucky. A top speed of 55mph. Sometime next week. Even if the drive is cranked to 150% max output current this system will continue to underwhelm the observer. This is the traditional approach used in industry where power peaks are rare. Though well matched for industry these two components are not matched for EV work.

The new rule then is that EV drives must be sized for 3:1 peaks and even 10:1 peaks when a more brisk response is desired. Perhaps someone has already tried this ?

Then I came across this site and read the activity here last summer.

ACMOTOR had spec'd an 11Kw (15Hp)415Vac motor for Red Suzi using a Danfoss VLT 4052 this 50Hp drive has the unusual charateristic of allowing 160% overload (64Hp) for one minute if set to nominal 40Hp.
He therefore achieved a 64/15 or 4.25:1 peak to continuous power ratio.

Figures show that motor power peaked at 30Kw arriving well before half of top speed from then on torque decayed rapidly. Too rapidly in fact. In order to counter this power drop off, the initial 10.5 gear ratio was changed on the fly to 6.5 as the vehicle gathered speed in a (desparate) attempt to slow down the motor and keep it out of saturation as much as possible. The motor did not appear to be in constant power mode perhaps because it wasn't getting even close to 415Vac input. Some would argue as low as 353Vac, my own calculations place it at 306Vac but whatever. What would be the solution if raising the voltage further with additional batteries was not on the table. ?

The motor was pulled that August and REWOUND. Yep, a perfectly good machine was rewound and balanced at great expense ($900) with thermistor protection added for good measure.

Which comes to the point of my post. Rather than accept the manipulation of a two stage gearbox caused by an off the shelf motor, a more scientific proactive approach was taken towards the powertrain's deficiency.
The result is that the vehicle is now running in the 10.5 ratio full time with an effectively 200Vac winding in place of the original 415Vac.
This time on startup the lower impedance of the new motor winding keeps the drive in current limit beyond 10km/hr, unlike before, with the 10.5 ratio allowing the motor revs to rise rapidly. With the correct volts/Hz ratio applied the fully magnetized motor provides a more meaningful performance. Torque roll off will begin around 30km/hr and I believe a top speed in that gear is reached at 75km/hr.(6000rpm)
This is the end of the road for now as ACMOTOR has moved on to another project.
In summary - performance of the motor was enhanced by the ratio of 415/200 to 30Hp but that reduces the peak to continuous to 64/30. However a ~ 2 : 1 peak to continuous power ratio is not optimal.

On first sight the controller could use a current upgrade and the battery pack should go to 650Vdc also. However when I looked at this problem I noticed that an equivalent 230Vac industrial drive weighs in at 200 lbs against the current 106 lbs. Clearly the installation, if there was the room, of extra batteries with the original 460Vac drive would be the cheaper option, but the motor would still fall short of its potential.

I am particularly interested in this work since I happen to have aquired a MX8500 460Vac 15Hp controller. My enthusiasm to do a DD with this equipment has been curbed. And a 2 speed gearbox is not in my plans either. A 460Vac source won't be a problem here but using the original 460Vac 15 Hp motor with it seems to be. I hope I haven't hijacked the thread. I'll continue here later with my 2 SPEED GEARBOX NOT    solution.

2 Speed Gearbox

Posted: Fri, 11 Jun 2010, 13:47
by Tritium_James
T2, improving the motor/inverter system performance by oversizing the inverter is all well and good (in fact, it's great!) but an improved motor/inverter plus a two-speed gearbox is always going to work out even better.

2 Speed Gearbox

Posted: Fri, 11 Jun 2010, 14:08
by Squiggles
Tritium_James wrote: T2, improving the motor/inverter system performance by oversizing the inverter is all well and good (in fact, it's great!) but an improved motor/inverter plus a two-speed gearbox is always going to work out even better.

I agree. Not all of us want to drag race our cars, so the 0.5 seconds lost during a gear change is not a concern. Getting the most out of a motor is the issue.

2 Speed Gearbox

Posted: Fri, 11 Jun 2010, 20:07
by aegidius
EV2Go wrote:
Gearboxes without a clutch are a no go, drove one without a clutch and didn't like it at all. I want quick and hard changes so it doesn't loose momentum getting it into the next gear.

Has anyone tried using a low-mass armature motor, like a pancake motor? The armature wouldn't weight much more than a clutchplate - just cut off all power and change gear.

I agree entirely with a conventional motor - my last EV would change up OK, but not down :-)

2 Speed Gearbox

Posted: Sat, 12 Jun 2010, 03:05
by procrastination inc
How about a landrover transfer case?

the LT230 is a constant all wheel drive unit so it has an open center differential built in.

there are a few ratios available for high range roughly : 1.0:1, 1.2:1, 1.4:1, and 1.6:1

low range is 3.3:1

I think they are simple dog engagement, so shifting on the move would be um... noisey. If you got excited, you might be able to automate it so the motor speed matched the shafts.

2 Speed Gearbox

Posted: Fri, 18 Jun 2010, 15:32
by T2
Landrover transfer case ? Pancake motors ?

Look, all these alternative suggestions are diversions away from looking at the real problem.

The REAL problem being the lack of torque at the low end of the RPM range. To say that a two stage gear box is the better choice compared to the purchase of a higher output inverter is missing the point.

A power delivery system is needed that extends the constant power region back towards lower speeds. A system which is seamless and doesn't introduce problems of its own. There is only one way that does not impact the battery system nor require a spanner or wrench be taken to existing mechanics. That way is simply to fit a larger inverter.

It's been discussed ad infinitum on these pages in recent months that induction motors can take at least a 300% current overload before reaching breakdown torque, it's now time we got with the program.

2 Speed Gearbox

Posted: Fri, 18 Jun 2010, 18:23
by Squiggles
I thought this was a topic about gearboxes not AC motors :).

Also isn't torque highest at the low end of the rpm range?

2 Speed Gearbox

Posted: Fri, 18 Jun 2010, 20:12
by EV2Go
Correct Squiggles peak torque is from 0-2500 RPM, which basically means I can get to 100kph and still be right on the edge of peak torque.

I was looking towards gear reduction too simply provide more off the line oomph (can't accelerate fast enough in my world). Ok there is a slight element of making it easier on the motor, but mostly for the grunt.

The point many seem to be missing is that there is only so much power that can be pulled from the batteries, so while it is nice to think just lean on the motor/controller/batteries more, there are certain practical limits that just can't be pushed any further.

I already plan to squeeze every bit of life out of all three already, it then comes down to mechanical advantage of gears.

Mind you I have been entertaining another approach. If I use series / parallel switching instead of CW / CCW and use a low drag reversing unit (like a Lenco) for reverse. I can optimise the motor that fraction more.
Alternatively since the Soliton1 controller already has a beefy contactor built into it, if I use a reversing unit, I could straight wire it series or parallel, and do away with any extra contactors for motor reversing.

BTW got a price back today on a single Lenco forward ST1200 module. Looking at US$2400, unfortunately I can't reverse the motor because of the sprag in it. So if I wanted to go backwards I would have to push it, and I dont have the length for both it and a reverser unit.

Just thinking about T2's point of time it takes to spool down the motor from 6000rpm to 3000rpm to change gears...

While mine would be more like 2500 down to 1400 it did start me thinking... I wonder how long it takes a motor to spool down, and would jamming it in another gear unduely load up the motor shaft?

2 Speed Gearbox

Posted: Fri, 18 Jun 2010, 23:01
by Electrocycle
you could do reverse the way the motorcycle engined race cars do it - with a starter motor set up to drive either the drive shaft or a rear wheel.

I've seen ones where the edge of the rear brake rotor is a ring gear. It was a pretty neat system.
Still - since reverse on an electric motor is just a matter of reversing some connections it seems a bit crazy to add a second motor for reverse!

2 Speed Gearbox

Posted: Sat, 19 Jun 2010, 00:46
by EV2Go
Depends... If I was to do series / parallel switching which uses 5 contactors it is already getting expensive enough without adding a reversing contactor.

Series vs Parallel

A small hi-torque starter like a Tilton could run off the 12v system and probably be less parasitic than a gear box. A ring gear on the pinion or tailshaft would mean it doesn't need to be engaged unless reversing.


I figure 1.9hp should be enough to back up even without a full sized ring gear...

Edit: Jegs make one that is 3hp and draws 80-90 amps, with a 4.41:1 reduction surely that would be sufficent.

If I decided to use a transmisson that didn't use a sprag i.e convential gearbox perhaps I could drive the front motor shaft?

2 Speed Gearbox

Posted: Sat, 19 Jun 2010, 19:05
by Electrocycle
yep, should work.
You won't need massive power for reverse, unless you park in a very silly place :P
I still think reversing the main motor makes more sense than adding the extra weight and complexity of the starter motor reverse setup, but it depends on what'll work best as a whole package.

2 Speed Gearbox

Posted: Sun, 20 Jun 2010, 01:23
by EV2Go
Yeah well it's one to put on the backburner for now while I sort out a few other issues. Got a price back today, Lenco do a stand alone reversing unit for $1750, so would probably either reverse the motor, or if I go two speed use a starter motor, since I couldn't then run it backwards.

2 Speed Gearbox

Posted: Sun, 20 Jun 2010, 04:49
I have been working on a 2 speed planetry gearbox/diff from the start of this year, I have all the components for the geartrain in working order on the bench. Unfortunately I suffered a heart attack mid February which holted work on this unit.(required 6 by-passes) I am now back on deck and have now started on the gearbox/diff housing. This unit will be suitable for front wheel drive as well as rear wheel drive.
The gearbox has 2 forward ( 1:1 and 1.55:1) and one reverse gear. I am targeting 10 - 12kw continuous power range. I will need to increase the wet clutch size to cater for higher hp. If all works out as planned the gearbox will be approx 7 to 8 inches long with a diameter of 7.5 inches.

I decided to build this gearbox for my ev as I have been running a direct drive system for a year now, but 2 speeds (high and low) will make it more responsive at low speeds. I have tried the clutchless gear shifting, but not all of the family were happy with this. I have been caught out a couple of times at roundabouts because of the slow acceleration from a standing start.

Hopefully I will be able to test the completed unit in about 6 - 8 weeks. I will post a photo of the unit later.

2 Speed Gearbox

Posted: Sun, 20 Jun 2010, 05:42
by EV2Go
Sounds good keep us posted.

2 Speed Gearbox

Posted: Thu, 24 Jun 2010, 03:16
by procrastination inc
could this be a suitable replacement for a transaxle in FWD or mid mount applications?

2 Speed Gearbox

Posted: Sun, 27 Jun 2010, 16:22
by AMPrentice
EV2Go wrote: Thanks Joel and welcome to the forums. A Lenco was suggested previously in jest, but I don't know why it didn't dawn on me sooner that a Lenco was a planetary gearbox (being modular and all).

I just "assumed" to my detriment that because they handled such high horsepower that they would be some heavy gear driven box.

Hmmm wonder if it is too late for a redesign... Really like the look of that first one you posted, looks ideal for EVs.

The lenco unit is very expensive las I heard for a 5 speed
so maybe for a 2 speed we might be in luck with a group buy?

The pic of jim is with a volvo overdrive found readily at wreckers.
I often wondered about this unit as it will make it so much easier
being electronically controlled.

After all this talk what are the 2 best ratios for the street?

1:1 stays but what about take off/crawl 2:1, 1.5:1, 1.35:1 or 1.275:1?

Since most cars do well on 3rd and 4th gear Id guessed around
1.25 for take off on hills or crawling up driveways and 1:1 for
everything else.

Im thinking electric overdrives are better for RWD though meaning
from what I understand/dont is that it can change the ratio on the
fly with about 30amps of power for the solenoids. ... rivesjtype


jag, triumph and volvo

All about overdrives
overdrive info very detailed!!!!

Switch "on" - OD "in". Switch "off" - OD "out". No cables, no extra gear shift levers...just
one switch mounted under the gear shift knob! Sweet!


2 Speed Gearbox

Posted: Sun, 27 Jun 2010, 18:43
by EV2Go
A Lenco wouldn't be practical or cost efficient enough for most people to warrant a group buy. I probably have a more flexable budget than most, and even I am struggling to justify the price.

For a two forward and a reverse is over US$4k plus expenses of getting it here. I think the overdrive is a more more rational path to consider.

I was seriously considering the Gear Vendors under / over drive (~US$2.5k), but since they can't be bothered answering my emails they can jam it where the sun don't shine, I am no longer interested in dealing with them.

Never ceases to amaze me how people expect to sell products if they fail to respond to even an initial email.

While investigating the GV unit I actually came across similar units to the ones you posted which made me wonder if GV was merely purchasing the central unit part and adding their own input / output to it.

The ratio I think would depend on your current diff ratio. I using 3.1:1 diff, so a 2:1 first gear would give me 6.2:1, which is where I figure I would want to be.

For someone running a 4.5:1 diff maybe a 1.4:1 might be more practical.
Obviously the tyres selected will play a big part in the choice.

2 Speed Gearbox

Posted: Sun, 27 Jun 2010, 19:16
by Speedily
If you take the 2nd gear planetry gearset out of a turbo 400 G/box this will work and is capable of lasting behind the torque of EV's   Image

2 Speed Gearbox

Posted: Sun, 27 Jun 2010, 20:16
by EV2Go
Many of the different boxes we have discussed have been a planetary type gearbox i.e. Blackbox, Lenco, Gear Vendors, Volvo, Jag etc, without the resources to fabricate end plates, housings etc it really is a case of trying to adapt something off the shelf.

If nothing else in this 12 page thread I think we have narrowed the selection somewhat...

Just need someone to pick up the ball and go into production.

2 Speed Gearbox

Posted: Tue, 29 Jun 2010, 12:02
by T2
Hi EV2Go,

I see where you're headed. I have a cheap electrical solution that you may not have thought of.

How did this get to 12 pages you ponder ? Well the brain is a strange thing. In general when we set out on a course of action we believe in, we are really not looking for advice. We are in fact searching for corroboration. We can, of course, kid ourselves that we are of an open mind and if we see a better idea or a cogent argument we may exercise free will and jump ship. Almost never happens, for good reason. At this time the subconcious brain sets up filters such that we consider as the truth only that which we want to hear. Everything else is some kind of dissent to be disregarded.

Case in point- the idea of a larger current output inverter was dropped because of cost and I'm fine with that. But You wrote :

The point many seem to be missing is that there is only so much power that can be pulled from the batteries, so while it is nice to think just lean on the motor/controller/batteries more, there are certain practical limits that just can't be pushed any further.

Well my answer to that is -it's quite in order to set up the battery limit to 150Amps while at the same time set up the motor current to 300 Amps. The controller will output 300Amps continuous current to the motor(s) at the slow speeds in question until the vehicle is at half speed for that connection, at which point the 300Amps will be progressively reduced until the motor voltage reaches the battery voltage at which point the motor current will be down to 150 amps. Sure the torque will have been diluted just like changing from first to second gear except that it is seamless in operation. And no batteries will have been harmed. I've already run this argument elsewhere and so have others, but you're just not having a piece of that....

As TJ has said a 2-speed gearbox will help you optimise however small the output comes from your controller ! Can't argue with that but a gearbox in an EV is an anachronism and makes the vehicle almost undriveable to others. Is that the image you want to put out there ?

I am going to give this one more kick at the can. This time for a DC system.

If you are using two motors, and you go from series to parallel but can only get 2500rpm top speed. Then your system would be improved if the motors could run faster. Placing the series fields in parallel would provide field weakening that would have the desired effect of increasing the powerband. The series to parallel switch would made at higher rpms this time. When the armatures are placed in parallel each motor has to see only its own field as before, but each field could now be bridged with an inductor or resistor allowing a fraction of the armature current to bypass each field, giving the field weakening effect all over again.
The thing is that the series motor is its own field controller. Not for nothing have they been used on streetcars since the 1880's. We just need to to tweak its top end so we can get by with a higher ratio gear without that higher ratio impinging on our top speed.
That's the way I see it anyway.

There will be enough inductance in the armature circuit providing the   motor is rated at least 100 volts - to keep the controller happy. BTW I tried this with a 15Hp Baldor motor.

2 Speed Gearbox

Posted: Thu, 01 Jul 2010, 20:05
by EV2Go
Hi T2, I see where your coming from but perhaps you have have missed the point I was trying to make.

When I said "The point many seem to be missing is that there is only so much power that can be pulled from the batteries, so while it is nice to think just lean on the motor/controller/batteries more, there are certain practical limits that just can't be pushed any further."

It has taken a while but it has finally sunk in for me, there is only so much power in a battery pack that can be utilised.

Let me explain further... When I owned a car last I put bigger injectors, bigger fuel pump, a reasonably large turbo, big camshafts, ported the head etc, all to make it go harder, if the engine ran lean I just altered the fuel map in the aftermarket ECU to add more fuel.

In a car I have almost unlimited power potential available (including changing to other fuels of a higher octain), if I want more power I just pump more fuel and air in its that simple.

The point I was trying to make is that there is a very hard ceiling on the battery pack. Once the battery pack is installed and fully charged, there are only so many volts and so many amps that can be drawn from the pack.

I can push it even harder and pull or exceed the C rating of those batteries, but at that point I can't draw any more power from the pack.

Once I have that determined that limit, I can put that current into the motor, because of mechanical losses I get even less power to the wheels.

What I was trying to say is no matter how I play the shell game and tweak this or pull that, there is a very real hard and fast limit to what the motor can produce.

Reality is as much as we would like to believe we can build this you-beaut car that will go like a rocket, once all the limitations / power draws are factored in, it always comes down to a compromise and slower vehicle than most would expect.

This is the sole reason for investigating a 2 speed gearbox. I plan to optimise everything best I can, of course you want the most you can get through tweaks etc.

The single motor I am buying is already bi-wired for series and parallel to optimise the top end (still undicided if I will use it though).

But this post was as much for the EV community as it was for me, I could see a opportunity for someone with the right tools to replace what has up to now been using a "make shift" part which is the ICE gearbox.

2 Speed Gearbox

Posted: Mon, 19 Jul 2010, 18:10
by AMPrentice
Im convinced Id rather spend all this effort on a second motor and a decent controller!!!!!
Look at this conversion its direct drive its out of the box literally

Suzuki Vitara twin engined EV



2 Speed Gearbox

Posted: Mon, 19 Jul 2010, 23:08
by antiscab
Its tricky to stay on topic when the various parts are so tightly inter related.

heres my 2 cents
EV2Go wrote:
The point many seem to be missing is that there is only so much power that can be pulled from the batteries, so while it is nice to think just lean on the motor/controller/batteries more, there are certain practical limits that just can't be pushed any further.

I would have to agree with T2, the battery practical limits are the same in both cases.
the motor controller limits are different (higher) when you use a bigger motor and controller.

a bigger motor and controller are also more efficient than a smaller set, which matters alot when you want big acceleration at low speeds.

case in point:
say you have a battery that will give you at max power point 57.6kw at the battery. say its a headway pack, so its is rearrangeable electrically when your setup changes.

say you are thinking of using a 144v 500A controller with a kostov R17.
initially you single ratio so 100kmh occurs at peak power point at 135Nm @ 3150rpm or 44.4kw.

that gives you 0-50kmh of 3.8secs and 0-100 of 7.8secs.

thats too slow, so you decide to spend US$2500 to make it faster:
you could use a two speed gearbox.
if you do that 0-100 would be reduced to 5.9 secs + shifting time with 0-50kmh being 1.9 secs.

alternatively, you could spend the same money on a 300V 1000A controller with a kostov R20.
max power would be 95Nm @ 4800rpm or47.7kw mech (remember only 57.6kw is being drawn from the battery)

start of power plateu (battery current limit) is 400Nm @ 825rpm
its ratioed so 100kmh = 4800rpm, giving 0-50kmh 1.5 secs and 0-100kmh 4.5 secs.

the question is more a matter of whether its easier to fit a bigger motor and controller, or even whether it is cheaper.
usually its cheaper to put the money in the motor and controller and not use a gearbox, but YMMV.

back to deciding on a gearbox for when you have to use one:
you could have two gearsets always in gear with two clutches.
that would avoid having to actually shift, though a little harsh on the clutches.

I had the opportunity to drive a EV with a gearbox the other day.
so used to having to shift often with an ICE, the owner was forever saying "just leave it in 2nd".

when i did shift up, it was tricky as the motor takes so long to spin down. the motor doesn't have alotta inertia, but you do still feel it give the back wheels a kick.


Edit: added shifting time, though from what i gather the 0-100 wasn't as important as say the 0-50

2 Speed Gearbox

Posted: Sat, 24 Jul 2010, 03:53
by EV2Go
While I originally started the thread more as a thought provoking excersize than anything else, I have decided not to run a gearbox myself.

Being that I am going to be running a Kostov R20 with close to 300v and a controller capable of 1000+ Amps the reality is I think I would loose more time in shifting then I would gain.

While this may not hold true for everyone, I believe it is the case for me, and adding a gearbox would be a counter productive excersize.

The thread was started with people who are running a 9" motor and Curtis controller mostly in mind, not so much myself and I believe a cheap simple two speed would still be of benifit, but it is a case of horses for courses.

2 Speed Gearbox

Posted: Sun, 25 Jul 2010, 01:07
by AMPrentice
I would prefer a 3 speed jatco auto if I didnt need the torque converter
2.4,1.4,1:1 should be plenty for a 9 inch