Archon X1 Programming Thread (questions and experiences)

As an experiment, I set every assist level the same. Torque 30A, sensitivity torque level 1. To my surprise, the level of assist kept going up with each PAS level!

I conclude that the torque multiplier is actually set by the assist level. Sensitivity torque level is the blend of torque vs cadence control as mentioned in the programming video. 1 being pure torque control and 10 being pure cadence control.

Using sensitivity torque level 1 for all assist levels will provide better high power modulation for climbing technical sections of trail.
 
As an experiment, I set every assist level the same. Torque 30A, sensitivity torque level 1. To my surprise, the level of assist kept going up with each PAS level!

I conclude that the torque multiplier is actually set by the assist level. Sensitivity torque level is the blend of torque vs cadence control as mentioned in the programming video. 1 being pure torque control and 10 being pure cadence control.

Using sensitivity torque level 1 for all assist levels will provide better high power modulation for climbing technical sections of trail.
Interesting, I may have to modify the lower assist level sensitivity for climbs then!
 
I heard yes, but it would be nice to get a confirmation from them or WW and if so as to when it will be available.
 
New discovery day! Sensitivity torque level 0 is actually pure torque sensing mode. Startup crank angle is down to 1/16th of a turn! Motor overrun is down to about 1/4 second! The controller now works as I expected it should.

My Hydra is damn near perfect now. Next up is installing the 52V, 20 Ah battery I built, but that is for another discussion thread...
 
I'm surprised your Ultra with X1 has any overrun at all. I haven't noticed any using BarnBoy's settings. I'll pay closer attention to mine.
 
As an experiment, I set every assist level the same. Torque 30A, sensitivity torque level 1. To my surprise, the level of assist kept going up with each PAS level!

I conclude that the torque multiplier is actually set by the assist level. Sensitivity torque level is the blend of torque vs cadence control as mentioned in the programming video. 1 being pure torque control and 10 being pure cadence control.

Using sensitivity torque level 1 for all assist levels will provide better high power modulation for climbing technical sections of trail.
Very interesting, and unexpected, finding TDA78! Do you still feel this is correct after more time riding / adjusting? And when you watch the display, do you see the power/current output gauge confirming what you feel? I.e. more power at same pedal input as you increase PAS levels?
The reason I ask is that I'm having WattWagons add an Archon to my Ultra as we speak, but they are still waiting on programming cables to arrive, so they're going to program it to my specifications before shipping it back - thus I've only got one shot to get all the settings right! At least until the programming cables arrive...

If what you have found is correct - that the PAS level also controls torque multiplication - then it makes a big difference in how I want my settings set.

Wondering how folks with Archons have balanced the two variables that govern power output from pedaling:
1) Max Torque Amps: max Amps output when pedaling at max (determined by Sensitivity Torque)
2) Sensitivity Torque: how hard you have to pedal to reach Max Torque Amps

It seems like you could achieve the same 'pedal->power' output with various combinations of the above - i.e. high Max Torque Amps (say 30A) and low Sensitivity Torque values might give the same 'feel' as low Max Torque Amps (say 10A) and high Sensitivity Torque. In the first case, light pedaling gives a fraction of Max Torque Amps which is high, and in the second case light pedaling gives full Max Torque Amps, which is low. So they might both output 10A with the same pedal pressure? Am I understanding this correctly?

Of course, then if you factor in the 'torque multiplier' effect of the PAS level as you found it gets even more confusing!

TDA78 would you be willing to share the settings that worked well for you? And if other are willing to do the same that'd help too. I like to pedal and want torque sensor behavior as opposed to PAS (I've already got that on my two BBSHD bikes), and I'd like to be able to go from low power output (say 750W) with HARD pedaling when I want a good workout and long range, all the way to high power output (2000W+) with moderate pedaling when I'm trying to maintain 25mph on long uphills.

I've put together a best guess settings matrix here attached, but have NO idea if it's going to give me what I want. And definitely not sure how to factor in the 'torque multiplier' of the PAS level! Any input appreciated!
 

Attachments

  • Archon X1 settings v1.png
    Archon X1 settings v1.png
    19.5 KB · Views: 207
Very interesting, and unexpected, finding TDA78! Do you still feel this is correct after more time riding / adjusting? And when you watch the display, do you see the power/current output gauge confirming what you feel? I.e. more power at same pedal input as you increase PAS levels?
I currently have torque set to 30A for all assist levels (based on the cell rating for the Hydra 52V battery). Max speed torque is set to 32 km/h for eco and 99 for sport assist levels. All other settings are set to their minimum for all assist levels. While climbing a steep incline, Eco PAS 1 will draw about 10 A, PAS 2 about 20A, and PAS 3 about 30A. My input remains the same - full effort! Sport 1 will draw about 30A and I easily lift the front tire with light pedal pressure starting on an incline.

I did have all levels of Eco torque set to 10A to be street legal in these parts but I really dislike the sport power response for single track use.

Sensitivity torque 0 is a purely torque based response and 10 is purely cadence. Numbers in-between are a blend of the two. It sounds like you want sensitivity torque 0 across all settings like me. I haven't played around low torque amps and high sensitivity torque. The startup delay with cadence sensing is unacceptable to me.

I wouldn't count on receiving your motor programmed as you request. Mine didn't, but I'm glad it came with a programming cable, even though it wasn't supposed to.
 
Last edited:
@TDA78 can you comment about how differences in motor torque levels feel or seem to work? The description from the Innotrace documentation leads me to believe it's... just the initial startup torque to get things rolling/peak throttled load (not under throttle, but a temporary peak load that is throttled back by the system after some condition is met)

After going through the thread gain, sensitivity torque interpolated between torque/cadence inputs with 0 being pure torque and 1 being pure cadence, and then actual output is determined by torque and assist level, but assist level does not scale the throttle value at all?

The test you did confirms this, which in my opinion isn't ideal. Another test to confirm this is the case would be to try some math to assume this is just multiplied values and set your levels to aim for the same output given each assist level.

So something like
1 - torque 36%
2 - torque 18%
3 - torque 12%
4 - torque 9%
5 - torque 7% (unless it can run a float at 7.2%)
6 - torque 6%
7 - torque 5% (unless float @ 5.1%)
8 - torque 4% (unless float @ 4.5%)
9 - torque 4%
10 - torque 3% (unless float @ 3.6%)

There's also the possibility that each level's torque amount adds onto the previous assist level's amount, if this were the case, simple test to verify is setting level 1 at 5A and all the others at 0A, and see if assist still increases with each level. This is the situation that makes it easiest for users to determine the actual current output limit they will get from settings without having to mess with any files and potentially break stuff, and honestly the one I'm hoping for.

If display shows your current/power draw then you maybe could try setting sensitivity torque higher to make it easy to hit the max current given my the torque setting, keep all torque levels the same, then see if your current & power draw scale up with each assist level. Only issue here is motor torque setting - looks like there's 10 or 11 levels instead of a percentage like the torque setting? Might make it hard to match that value and we don't really know what conditions determine 'starting' - is it a timer after rpm>0, is it a timer once torque is exceeded, or a cutoff above a certain rpm? I skimmed through @BarnBoy 's thread and did see there was a text file which is a little more useful, but it looks like that file doesn't actually tell us what's being done beyond just what the controller sees as those input params for the actual operations. Does anyone have any screenshots of what files you have with the X1 tool and how much you can see or is that pretty much it?

I doubt Innotrace would grant any sort of access to that code since it's a little more uncopyable than the hardware itself (if protected properly). It may be possible to do a little work in the text file that has the input params to account for any sort of scaling by assist level done by the controller but if you don't know the exact operation it's a guessing game. If by some miracle there is a little more access into the code showing how these values are used then from there it shouldn't be difficult to get it to where the assist level is only an index to different groups of settings and has no affect on how any of those values are interpreted. Ideally in my situation I think I would want to know exactly what the constant output current limit (torque setting) for each assist level is, that way it's just easier to understand the actual assist amount between each level at a glance. No taking any unseen math into account. My only concern is early on in the thread I understood (hopefully misunderstood) that Pushkar said assist level itself is what blends the torque & cadence sensor? I need to rewatch the video too maybe that clears it up a bit...

as per Innotrace doc
1653807128304.png
 
I finally got around to registering with Archon, so I can some modifying.
This is a real noob question.....where does the programming harness plug in on the bike?
 
I finally got around to registering with Archon, so I can some modifying.
This is a real noob question.....where does the programming harness plug in on the bike?
Programming cable plugs in where the display plugs in, you unplug display and plug in programming cable. Like on the front of bike there is bundle of wires, find the wire that goes between bike and display, unplug those two and then connect programming cable to wire going to bike.
 
@TDA78 can you comment about how differences in motor torque levels feel or seem to work? The description from the Innotrace documentation leads me to believe it's... just the initial startup torque to get things rolling/peak throttled load (not under throttle, but a temporary peak load that is throttled back by the system after some condition is met)

After going through the thread gain, sensitivity torque interpolated between torque/cadence inputs with 0 being pure torque and 1 being pure cadence, and then actual output is determined by torque and assist level, but assist level does not scale the throttle value at all?

The test you did confirms this, which in my opinion isn't ideal. Another test to confirm this is the case would be to try some math to assume this is just multiplied values and set your levels to aim for the same output given each assist level.

So something like
1 - torque 36%
2 - torque 18%
3 - torque 12%
4 - torque 9%
5 - torque 7% (unless it can run a float at 7.2%)
6 - torque 6%
7 - torque 5% (unless float @ 5.1%)
8 - torque 4% (unless float @ 4.5%)
9 - torque 4%
10 - torque 3% (unless float @ 3.6%)

There's also the possibility that each level's torque amount adds onto the previous assist level's amount, if this were the case, simple test to verify is setting level 1 at 5A and all the others at 0A, and see if assist still increases with each level. This is the situation that makes it easiest for users to determine the actual current output limit they will get from settings without having to mess with any files and potentially break stuff, and honestly the one I'm hoping for.

If display shows your current/power draw then you maybe could try setting sensitivity torque higher to make it easy to hit the max current given my the torque setting, keep all torque levels the same, then see if your current & power draw scale up with each assist level. Only issue here is motor torque setting - looks like there's 10 or 11 levels instead of a percentage like the torque setting? Might make it hard to match that value and we don't really know what conditions determine 'starting' - is it a timer after rpm>0, is it a timer once torque is exceeded, or a cutoff above a certain rpm? I skimmed through @BarnBoy 's thread and did see there was a text file which is a little more useful, but it looks like that file doesn't actually tell us what's being done beyond just what the controller sees as those input params for the actual operations. Does anyone have any screenshots of what files you have with the X1 tool and how much you can see or is that pretty much it?

I doubt Innotrace would grant any sort of access to that code since it's a little more uncopyable than the hardware itself (if protected properly). It may be possible to do a little work in the text file that has the input params to account for any sort of scaling by assist level done by the controller but if you don't know the exact operation it's a guessing game. If by some miracle there is a little more access into the code showing how these values are used then from there it shouldn't be difficult to get it to where the assist level is only an index to different groups of settings and has no affect on how any of those values are interpreted. Ideally in my situation I think I would want to know exactly what the constant output current limit (torque setting) for each assist level is, that way it's just easier to understand the actual assist amount between each level at a glance. No taking any unseen math into account. My only concern is early on in the thread I understood (hopefully misunderstood) that Pushkar said assist level itself is what blends the torque & cadence sensor? I need to rewatch the video too maybe that clears it up a bit...

as per Innotrace doc
View attachment 124641
I hope you receive your programming cable soon so you can experiment to your heart's desire. I'm done with that after 10 iterations to find settings I like.
 
I kinda thought that, but wasn't sure. Thanks.
Rocking the display connector while pulling it apart makes it much easier to separate the waterproof seal. Be very careful aligning the connector when plugging it back in. The pins are easily damaged.
 
Rocking the display connector while pulling it apart makes it much easier to separate the waterproof seal. Be very careful aligning the connector when plugging it back in. The pins are easily damaged.
Thanks, I will. Still waiting for activation from Archon.
 
Good stuff guys! I found some pretty good settings for myself but have been experimenting a bit more with varying the amounts of pedal torque between assist levels instead of changing the torque sensor sensitivity levels. I'm finding anything above about 350 watts is excessive for singletrack, especially the technical rooty stuff around here.

Thanks, I will. Still waiting for activation from Archon.
Just a warning, Innotrace ties your registration to the MAC (hardware) address of your computer, so you cannot login to the utility on any other devices. I registered without knowing this and now I'm stuck using the 7 year old Surface Book I have that is on its death bed, I need to contact WW and see if they can change that lol.
 
Just a warning, Innotrace ties your registration to the MAC (hardware) address of your computer, so you cannot login to the utility on any other devices. I registered without knowing this and now I'm stuck using the 7 year old Surface Book I have that is on its death bed, I need to contact WW and see if they can change that lol.
Well, that's annoying. Mine is on a 10 year old laptop.
 
Well, that's annoying. Mine is on a 10 year old laptop.
Yeah I thought something was wrong when I tried to log in on my desktop. WW told me it was tied to the device MAC address. I know you can spoof the MAC address of a device, but that would be annoying to have to do every time.
 
Back