Kalkhoff Agattu C8 Impulse - one for the hills!

[flecc]

It probably wouldn't run without it, like the Bosch unit. It may work ok with the magnet and sensor transferred to a crank and the motor unit, which woud greatly increase the road cutoff speed, but I can't guarantee that. Bosch have made that trick more difficult now on their unit, so Kalkhoff may also have done that.

I don't think that power roll-off severe, it still has a high power level to over 13 mph and a roll off shouldn't be too abrupt for smooth riding.

 

[10mph]

Pending 10mph responding, the speed control is by a magnetic sensor on the rear wheel. No quoted figures for the power phase down, but this chart if accurate indicates it's from roughly 12 mph and gentle

Flecc is right about the sensor on the rear wheel. This will directly measure road speed irrespective of crank speed and gear settings. It is set up to cut off quite steeply between around 14 or 15 mph and a little over 17 mph.

The limit for assistance according to EN15194 is 25kph which corresponds to 15.53mph but I think in the setup they have taken full advantage of the 5% margin allowed in the test

4.2.6 Maximum speed for which the electric motor gives assistance
4.2.6.1 Requirements
The maximum speed for which the electric motor gives assistance may differ by ± 5% of the speed indicated on the label described within Clause 5 when determined according to the test method described in 4.2.6.2, from 25 km/h or lower values as specified by the manufacturer.

So that would take the speed to 16.31mph, but there is also a speed error of upto 2% is allowed in the test equipment.

4.2.6.2.1 Test conditions
a) The test may be performed either on a test track, a test bench or on a stand that keeps the motor driven wheel free of the ground.
b) The speed-measuring device shall have the following characteristics: Accuracy: ± 2%

So that could bring the cut off speed up to 16.64 mph. But the cutoff appears to me to be slightly higher.

I am sure there is still some assistance at 17mph, and that it has all gone before I reach 17.8mph. My impression that from how the ride feels on flattish terrain at 16.5 mph there is some really useful assistance. Doubtless it is not maximum power which you don't need anyway when on gentle slopes, but it is enough to make cruising at 16.5 mph definitely much more comfortable with the power and you notice the extra pedal force required if you hit the power off button.

We sometimes see here claims that a 10% margin is allowed on maximum assist speed, and if that is indeed legally allowed it would correspond to a speed of 17.08 mph. Even there I am still definitely getting a bit of assistance, because I can feel its removal when I hit the power off button.

I am not sure exactly where this 10% tolerance comes from and whether it is legal. It is probably from the 10% margin which I had believed applies to car and motorbike speedometers. However when I looked this up on Wikipedia I found,

European Union member states must also grant type approval to vehicles meeting similar EU standards. The ones covering speedometers are similar to the UNECE regulation in that they specify that:

• The indicated speed must never be less than the actual speed, i.e. it should not be possible to inadvertently speed because of an incorrect speedometer reading.
• The indicated speed must not be more than 110 percent of the true speed plus 4 km/h at specified test speeds. For example, at 80 km/h, the indicated speed must be no more than 92 km/h.

So motor vehicle speedometers appear to be required to read -0%+10% of the true speed. That would explain why when I check with my GPS all motor vehicles speedometers read slightly high.

But to return to the Agattu with the Impulse motor and the graph above. I suspect the cutoff is sharper than shown. It is rather similar to my Panasonic Agattu where when I measured current I found that maximum current could be drawn up to 14 mph and the cut off to zero current was a shade over 17 mph. I think the graph above of force may possibly be a salesman's illustration of the effect rather than an engineer's measurement. It does rather frustrate me that Kalkhoff don't put proper scales on their graphs and of course they never plot the engineers' actual data readings. I suspect we are just getting the salesman's simplified pitch rather than real data.

I was pleased to see that if the magnet on the rear wheel speed sensor is misaligned, you get an error reading and no assistance, so there is no simple way of removing the plumbed in cutoff speed. If you want to go fast get fitter and buy a road bike, or if you are too old for that buy a moped. However, I am quite happy tootling along at 10 mph in great comfort.

I have just downloaded the free copy of A to B April (issue 89) mentioned by jhruk earlier in this thread. It bring up a number of points about the Impulse Agattu which I have also noted, but I have a very slightly different take which I will expand later in another post.

Tangent also asked

One other question! How does the motor noise compare with the Panasonic based Agattu?

Very quiet just like the Panasonic Agattu. Detectable at slow speeds when accelerating hard. Wind noise in my ears dominates at high speeds. Overtaking a troop of soldiers who were marching up my 13% hill, I did get a shout of "cheating", but I am not sure whether they heard the very slight whine, saw the discreet battery, or just could not believe that an old git dressed in ordinary clothes, with white hair visible at the edge of his helmet, sitting bolt upright on a bike and not puffing at some 6 mph, could not somehow be cheating in comparison with their forced training march.

Added in edit- while I composing this Flecc already replied, and as usual his understanding absolutely bears out what I found in practice.

 

[flecc]

I am not sure exactly where this 10% tolerance comes from and whether it is legal. It is probably from the 10% margin which I had believed applies to car and motorbike speedometers. However when I looked this up on Wikipedia I found,

This was quoted by the DfT to e-bike suppliers not very long ago. I'm quite sure you are right in that they probably derived this from vehicle practice. However the 10% tolerance has also been quoted by police spokespersons at various times in the past as what they allow in measuring vehicle speed.

In any event, given the inaccuracies of measuring cyclists at low speed by any method, I've no doubt 17 mph is a safe speed to ride at power assisted.

 

[10mph]

The Start Up Boost

atob magazine have tested the Agattu C8 Impulse in their latest (April – No.89) edition - I’ve just read it as a subscriber to the exacteditions version. If you don’t already subscribe I see they are now offering a free pdf download of the whole magazine if you register on their site, although the April edition doesn’t appear on there yet.

Very many thanks to jhruk for pointing out that link for the free download. The April - No. 89 issue is now there. So I registered and downloaded it for free. The review is much more comprehensive than my report since as well as performance it covers some of the bike's components, and the reviewer is certainly more of an expert than me given his wide experience with different e-bikes. It makes a significant point about characteristic of the new Impulse drive which I noticed but did not mention in my report which already seemed rather long for a forum post.

A to B calls the effect "The start up boost".

This is a good description of the effect which manifests as a higher level of assistance as one starts pedalling from stationary or from free wheeling with the pedals stationary. It lasts for a few pedal rotations, roughly 2 to 4 seconds. Starting from stationary it provides a welcome boost as one accelerates. In what seems no time at all and with little effort one is over 10 mph and approaching 15 mph. But I really first picked up the effect when I was already moving and freewheeling, and then started to pedal hard. For example with the hub gear you have to momentarily (for less than 1 second) stop pedalling in order to change gear. When you start pedalling again there is much higher level of assist for 2 to 4 seconds.

In fact, when I was testing the Agattu Impulse by climbing my 13% hill in gear 8 (really too high a gear for that hill), I found that I could actually reduce the strain on my legs and get a brief but useful relief from the effort of using too high a gear by momentarily stopping pedalling (for perhaps about half a second) and then getting more relief for a few pedal turns with higher assistance. I probably took advantage of at least 5 such boosts during the 100 metre long section of the hill with 13% gradient. I gained more during the boost than I lost during the freewheel, and got some brief rest into the bargain!

As well as the good acceleration from a standing start this power boost thus nicely overcomes the disadvantage with the Panasonic system and a hub gear where gear changing means a short loss of propulsion while freewheeling. So once I got used to it, I regarded it as an improvement: more power briefly when you need it, without sacrificing the overall economy and long range of the system.

A to B take a somewhat different view, and even appear to call in their conclusion for Kalkhoff to recalibrate the control system perhaps to make the higher power boost to be continuous, so that one would be able to surmount very steep hills at high speed. They say "try before you buy if you live somewhere really hilly". I would not dispute that advice. Within a few pedal turns of 50 cycles shop in Richmond is the very steep Nightingale Lane which averages 12% and kicks up even steeper for the last few metres at the top. I tested my 3 speed Panasonic Agattu on that hill before purchasing it. Unless you have legs that a very very much weaker than mine the 8 speed Impulse Agattu will obviously surmount it by selecting a low enough gear.

 

[flecc]

Thanks for that extra information 10mph. I noticed on the power graph that the standstill start point has an elevated power level, rapidly falling away before rising again over most of the speed range.

 

[10mph]

Thanks for that extra information 10mph. I noticed on the power graph that the standstill start point has an elevated power level, rapidly falling away before rising again over most of the speed range.

Yes I agree, but I am sure that curve is something drawn by a salesman to try and illustrate the effect and not real data from Kalkhoff engineers.

If you assume a linear x axis, and assume that the cutoff is either 15.5mph ( the legal max) or the 17.1mph that I observed, then the graph would suggest that the start up boost only operates up to less than 1 mph.

My experience with the bike is that the boost does not seem to depend on your speed. It certainly operates to much higher than 1 mph when starting from rest, and you also experience it when at 4 mph on the hill in gear 8, and also at similar speeds when switching to lower gears on the hill. It seems much more to be a timed boost for around 2 to 4 seconds rather than something controlled by the bike's speed.

 

[flecc]

Thanks 10mph, I realised the graph wasn't accurate but that kickup near to the start line was unusual. The power pulse does seem an odd way to do things but I can see how it can be useful. I haven't read A to B's critique yet, just received the magazine this morning.

 

[flecc]

I've now read the A to B test of the Impulse and have to say I'm very disappointed with that high power mode profile. Although good for maximising range, I'd much rather have the high mode fully and constantly available for someone like me facing long uphill slogs. Titsey Hill in my area is a good example, one and a quarter miles over it's whole length with B269 extension, it contains sections that are very steep making it especially tough. The last thing I'd want on such a steep section is to have power reducing, nor do I want to twirl at high cadence in a low gear for an extended period. It is after all a 14 minute climb duration at the 5.5 mph you climbed at 10mph.

I can't help wondering if that high mode characteristic has anything to do with extending battery life, since Kalkhoff claim an 1100 charge life. I notice A to B on their Panasonic powered usage of the Kalklhoff 18 Ah battery (without that power reduction feature) have noted a loss of near 14% capacity in nine months. It's an isolated example so needs treating with care, but it's commensurate with getting down to 60% capacity by roughly 2.5 years, and few if any will get near to 1100 charges completed in that time.

 

[eddieo]

Have seen reports on German pedelec with disappointment with the power delivery, it was explained as akin to turbo lag in a car but without the kick in the pants......anyone thinking of a bike for hill climbing or whatever really needs to try a Bosch powered bike before parting with their monies......

 

[wissy]

In my travels around the country to measure races I have only ever encountered one hill I could not ride up - in Newborough Forest in Anglesey, where in one place the sandy trail was so steep I had to get help from a fit runner who was showing me the route to push the bike up.

And people always think Anglesey is flat!! I keep telling them......

Tempted again by Kalkhoff due to price decrease (after end of sale which I was so close to buying 'by midnight' tsch...) but........

 

[103Alex1]

And people always think Anglesey is flat!! I keep telling them......

Tempted again by Kalkhoff due to price decrease (after end of sale which I was so close to buying 'by midnight' tsch...) but........

Just to comment on the same point (...... again.......) those reports referred to in German forum were written before the firmware was upgraded. It is the controller which regulates this and not the motor.

The smooth start curve of the power delivery is something different to the issue people complained about on there and was reprogrammed. You would see if you tested the bike (rather than rely on reports based on testing of a superseded power management system programming which is not in use any more). I believe 50cycles sorted out upgrading the firmware of original bikes with programming and all the ones sold now have the new version.

As far as hills are concerned, I think you'd find that a Bosch was just the same as as any other 250W rated motor that can't deliver a high peak output for longer than a limited time - i.e. only of real help at low speeds or for a limited hill climb duration. Impulse is an extremely capable hill climber. However it's hard work at 8-10mph+ on prolonged very steep ascents taking several minutes to accomplish. As I explained in detail on a previous recent post. Getting a Bosch isn't going to give you something which is any different. Getting a moped might.

Having ridden an Impulse-powered bike every day since October, with regard to the 'power sag' referred to the only time I've experienced this are as follows to put things into balance a bit :

1) When riding over but near 15mph and dropping in and out of assistance

2) On long steep climbs after the power peak has been delivered and sustained for as long as it's likely allowed to be (this is down to the law on continuous assistance power output and not the Impulse motor)

If you're looking for very good assistance (i.e. nigh-on effortless riding so you barely notice you're climbing a hill) on short hill climbs the bike is fine. If you want a bike which delivers higher power when you start (more boy-racer stylee ) rather than delivers power in a very short but gradual upwards curve when you pull away (responsible mum & dad on the school run stylee) then the Impulse is probably a better choice. It all depends on what you prefer and the style you ride.

The "guts" point is likely mainly about how fast people feel they are getting helped to the maximum peak power delivery (which is way more than 250W in the case of both Bosch and Impulse bikes). I don't know Anglesey or how long / steep your climbs would be. But for me it's long climbs which are the killer and which slow you down rather than starts which I never have problems with even on very steep slopes.

Where I am there is only one road out of town which doesn't involve a couple of miles of steep climbing to get anywhere. I know now from experience that no 250W crank-drive eBike is going to take me up those climbs effortlessly at a good speed and leave me dry at the end. That's as objective as I can be about it. I've no agenda and remember steered you away from Agattu in the 1st place but it was to an E-motion Neo and not a generic Bosch bike.