DIY Electric Car Forums banner

EVNetics Soliton1 - customer support thread

146K views 494 replies 75 participants last post by  dkubus 
#1 ·
Hi,

Not sure if this is the best place to ask but I know Tesseract and Qer frequent these forums and my question is for them.

I am a fair ways off installing my Soliton1 into my car, but I was wondering if it is OK to hook up a 12V car battery to the Soliton1 so that I can connect to it with the web page console. No motor or pot box would be connected.

The reason I want to do this is to take a look at the configuration stuff and get an idea of what it might take to integrate it with my in car computer/dash that I am planning.

The sort of thing I am thinking about is having a sport/economy setting for example on my dash and to be able to switch between those modes by making the necessary calls to change the controller settings (on the fly even?).

Cheers,
Pete.
 
#2 ·
Not sure if this is the best place to ask but I know Tesseract and Qer frequent these forums and my question is for them.
Well, I don't have a problem with it, but I'm not sure it's appropriate to use this website for such. Maybe start a poll asking if people mind if there is a thread for technical support on a specific controller?

I am a fair ways off installing my Soliton1 into my car, but I was wondering if it is OK to hook up a 12V car battery to the Soliton1 so that I can connect to it with the web page console. No motor or pot box would be connected.
Sure - it's perfectly fine, and often a lot more convenient, to power the Soliton1 from a wall wart or 12V battery to set everything up in the first place. Make sure the wall wart is rated for at least 1A and between 12 and 14VDC (regulated is preferred). This is how I test them immediately after assembly but before connecting them up to the dyno.

The sort of thing I am thinking about is having a sport/economy setting for example on my dash and to be able to switch between those modes by making the necessary calls to change the controller settings (on the fly even?).
This is an incredibly popular request but we haven't quite decided on the best way to implement it (without cluttering up the web interface page). We would really like to work with someone (hint, hint) to take the performance data the controller continuously streams to the ethernet port and turn it into pretty graphs and stuff for display in the car. We would like to make that whole spec open source and are willing to modify the data stream format as needed to facilitate such. Oops.. look at me, volunteering Qer (aka SPD - Swedish Programmer Dude; aka Martin) for even more work... :eek:
 
#4 ·
Well, I don't have a problem with it, but I'm not sure it's appropriate to use this website for such. Maybe start a poll asking if people mind if there is a thread for technical support on a specific controller?
More information is always good, I don't see any problem with sharing. I wish we had more product manufacturers discussing their product on this site.
 
#6 ·
Hi all!
I am still waiting for my soliton1 to come from Kostov with my motor. I am still trying to figure out my battery pack. I have a big motor and would like to use as high voltage as possible. Question is: How many 3.2 TS LiPoFe´s can i use? (If i can afford ) Number if i use 12v LiFePo´s? (nominal voltage 12V, maximal voltage 17V, minimal voltage 10V).
I have old Range Rover to convert so i need everything i can get.
I´ll start a build thread when motor arrives.
Thanks you all, Great forum!
 
#7 · (Edited)
Hi Glaurung.

The manual says 342V max and 300V for maximum current.

I think 90 LiFePO4 cells (288V nominal) are the optimal pack for the Soliton. If you charge them to 3,6V/cell (324V - battery pack), you'll still be able to drive the car although the cells didn't "settle" (LiFePO4 cells tend to reduce the voltage after a few hours - they come down to 3,3-3,4V after 5-6 hours)

What 12V LiFePO4 battery are you talking about? This can only be LiFeBatt or some e-bike battery pack. That are also just 4 LiFePO4 cells in series.

The e-bike packs are not the best choice for an EV and LiFeBatt packs are very expensive. Take a look at ThunderSky cells. I and many others have them in our cars. They are the best buy on the market if you build a "normal" car. A race car or dragster needs better (MUCH more expensive) batteries.

Hope it helps

Mate Rimac,
Croatia ;)
 
#8 ·
Thanks for your input Crodriver!
I have been watching your build very closely.
There is a Thunder Sky seller in Prague who has factory 12v batteries for all applications.And usual sizes too.They have used those in replacing VW Golf factory ev Varta batteries with good results. I dare not to put link here but pm if interested.
I was also supriced to see those batteries.If they are ok, wiring and many other things would be a lot simpler.
Plamen gave me your e-mail, but since i have not started building yet, i have not introduced myself to you(until now).
Thanks, Harri
 
#9 ·
If you purchase "unknown"/untested EV components you're risking a lot. A lot of companies are claiming specs they can't meet. Do you know if the 12V LiFePO4 has a integrated BMS?

If yes, how will it handle 300+V running trough it? The MOSFETs of the BMS must be oversized for that.

If not, how will you install a BMS in it? Don't believe the seller if he says you that the battery doesn't need a BMS ;) They do need it in the long term, especially if you can't check manually every cell voltage (if they are in a closed box) or don't want to bother with it.

I'm glad that Plamen gave you my mail. Are you building the components inside a BMW 7 series? He told me something about that a few months ago.
 
#12 ·
In the EVnetics manual it gives an example tach:

TACHOMETER – this input accepts pulses (amplitude of 4V to 15V) from a wide variety of transducers to read the motor RPM (overspeed protection and idle can't work without it). On our dyno we use a generic industrial inductive proximity (“prox”) sensor with a NPN open collector output."

I have come across a relatively cheap prox sensor that looks like the one in the manual but it is a PNP type - does that matter at all?????

Thanks,
Pete.
 
#14 ·
I have come across a relatively cheap prox sensor that looks like the one in the manual but it is a PNP type - does that matter at all?????

Doesn't matter at all - either type will work fine, you just need to use a pulldown resistor (from TACH to SGND) instead of a pullup (from TACH to S+12V).

The value of this resistor is not critical - anything from 150 ohms to 2.2k works fine which should accommodate just about any of this type of sensor. Even though very little power is dissipated in this resistor, the 1/2W size makes it easier to span from TACH to SGND, though I have managed to stretch a 1/4W resistor across those two terminals.

Any other questions just ask, and don't forget to update your code/logger to the latest version when you are ready to start driving around.
 
#13 ·
As long as the signal is clean it shouldn't matter. The controller doesn't care about the pulse length or anything like that, it trigs on the positive flank and measure the frequency. Just make sure the signal's as clean as possible, even though the noise filter seems to work fairly well it still degrades the accuracy.
 
#17 ·
I hear that 1.4 release will include a output to control the coolant pump....so that it doesn't have to run all the time, only when controller reaches a certain temperature. Can you elaborate on this nice feature at all?
Any hints of other new goodies to come with this release? :)

Gary
 
#18 ·
I hear that 1.4 release will include a output to control the coolant pump....so that it doesn't have to run all the time, only when controller reaches a certain temperature. Can you elaborate on this nice feature at all?
It's pretty straight forward. You select an output on the controller to switch on the pump and when the controller reach the temperature when it'll switch on the fans (~40C) it'll also switch on the pump. Not much more to elaborate about it...

Any hints of other new goodies to come with this release?

  • The special limits (for LVC, motor snap switch) has been unified to one simple limit (per customer requests) which will limit the throttle to a certain max percent. You can select all three inputs to different limits and lowest active wins. Can be both inverted and non-inverted.
  • SoC output is skipped since it never worked well (and only did a somewhat useful work for lead-acid). Instead there's a pack voltage gauge output.
  • The pack LVC now has a two point threshold so you define min pack voltage for no battery current and full battery current and the LVC will adjust linearly according to those two pints. For example, if you define LVC to 150 Volt at no load and 120 Volt at 400 Amps LVC will trig at 135 Volt at 200 Amps. If you define both LVC settings to the same voltage LVC will work like it does today.
  • There's a button for writing all settings to a text file. It's mainly implemented to make T's life easier since he simply can ask a customer to save and mail the settings to him, which will be a lot more convenient than to ask the user to do screen dumps in the browser, but I trust that you people will figure out a use for it as well (like you did with the logger which was mainly implemented for debugging and dyno runs :D ).

However, the biggest change you won't really see. Most of the web server is rewritten to allow for the html to be compressed to use less flash memory so we can keep implementing features. Things started to get a bit cramped on the inside... :rolleyes:

The web server now uses a kind of template page system which makes it a lot easier to add new features too, like the writing settings to file-feature. Wouldn't have happened otherwise. The penalty for this is that the web server is a bit slower now, but since it's a motor controller rather than a search engine we're talking about I figured it was worth it. :p
 
#21 · (Edited)
Pete I have used the Soliton 1 powered up on 12 volts for checking out the web server before installed .

here is a picture of the sensor I installed for RPM
works well. I haven't traced down the wire I need to send the rpm from the Soltion 1 output to the dash tach yet so I am using your app to read RPM . best I could find laying around here was 225 ohm for the pullup/pulldown but it is working

a few days ago I threw a few pre charge errors . it would be great to get an error light on your app that signals an error was thrown on the Soliton 1.
 

Attachments

#22 · (Edited)
Just sent the new 1.4 to T for evaluation. If it survives his scrutinizing the release will be public during the coming week. If everything goes according to plan you will be able to upgrade tomorrow Monday already.

Soliton Junior owners that run their controllers at pack voltages above 250 Volt nominal should definitely upgrade! The same goes if you run software version 1.2 or older (for both Junior and Soliton 1). Those that already run 1.3 might want to upgrade for the additional features, but there's no safety risk in staying with version 1.3.

Please note! When you upgrade to 1.4 the inputs 1-3 and output 1-3 will be reset to off. Please make a note what your settings are before you upgrade and set up the controller in an appropriate way again after the upgrade is completed!

New features:


  • BMS LVC and Motor temp switch are now combined into a general purpose throttle limit. Any one of the three inputs can work as a limit and can be set to different values. Lowest active wins.
  • A limit of 0% will terminate idle, a limit that isn't 0% will not.
  • All programmable inputs now have a checkbox to invert their behavior (go from active high to active low).
  • SoC gauge output is converted to a pack voltage output with min and max voltage setting to better use the full gauge range.
  • Pack LVC now has two settings; LVC at no pack current and at max pack current. The controller then plots a linear limit between those points.
  • Added a pump output that makes it possible to let an external water pump run only when needed. An external relay will, however, be necessary.
  • Write settings to file-function.

Added safety:


  • Max motor current for Junior is now limited to 500 Ampere at 310 Volt pack voltage and above for the same reason the Soliton 1 limit max current to 900 Ampere; to protect the freewheeling diode.

So there.

EDIT: Forgot. Did some stuff with logger too to make it more readable and make it easier to import to for example Excel. Might be useful...
 
#27 ·
Here is a graph of a 95 cell pack, battery voltage versus battery current. Nominal voltage is slightly above 310 Volt but as you can see the pack voltage drops below 310 Volt already before 100 battery Amps.



WIth a pack of 100 cells the 310 Volt treshold will increase to about 300 Amps (depending on types of batteries etc too, of course) which might actually be possible to notice when you drive. On the other hand if you drop 1 or 2 cells from the pack and is content with 93 or 94 cells total the nominal voltage will be low enough for this to never be a problem.

And that's why we recommend 90 cells maximum, that gives enough margin for the initial pack voltage to fit within the 342 Volts and the 310 Volt threshold won't be a problem. ;)
 

Attachments

#28 ·
Hello, I'd like to chime in because I have the exact setup shown in the graph:

Soliton1 (v1.3 jan'11)
95* Winston 90AH
Kostov 250V 11"

Looking at the graph it seems resting voltage after charge is (315/95) 3.3~V, charging to (332.5/95) 3.5V maybe?
Lower down the graph there seems to me a sweet spot at 270V/600A, this would be (270/95) 2.84V for a cell which has 2.8 as discharge voltage on the datasheet (WB LYP90AHA).

In the controller interface I'd like to set: Maximum battery current 600A. Maximum motor voltage 240V.
With that setup I believe output would also automatically be limited motor side to: 270V * 600A = 162000 kW / 240V = 675 Amps motor side and producing a peak of ((162k / 746) /* 0.8) ~170HP.

If a slight current derating is really the only thing I should be worried about then I'd like to go ahead and use the full 95 pack. It wil help me stay above 2.8V per cell and give me just that bit longer before 80% DoD.

Am I okay with those settings, or missing a beat here? Also very curious what the further setup of the graphed vehicle is!

Thank you,
AK
 
#33 ·
...Looking at the graph it seems resting voltage after charge is (315/95) 3.3~V, charging to (332.5/95) 3.5V maybe?
Lower down the graph there seems to me a sweet spot at 270V/600A, this would be (270/95) 2.84V for a cell which has 2.8 as discharge voltage on the datasheet (WB LYP90AHA).
Sort of. The voltage your cells rest at immediately after charging depends on whether you subscribe to the top, bottom or no balancing schools of LFP use and abuse. If you top balance and charge each cell to, e.g., 3.65V then a 95 cell pack will rest at 347V and the controller will refuse to start up until you bleed off that surface charge. If you bottom balance (my personal preference) and stop the charger as soon as the first cell hits 3.45V then a 95 cell pack will, at most, reach 328V, which is well within the operating range of the Soliton1/Jr.

In the controller interface I'd like to set: Maximum battery current 600A. Maximum motor voltage 240V.
With that setup I believe output would also automatically be limited motor side to: 270V * 600A = 162000 kW / 240V = 675 Amps motor side and producing a peak of ((162k / 746) /* 0.8) ~170HP.
You didn't specify what you intend to set the motor current limit to. I highly recommend you not try to push 1000A through a 250V Kostov motor because it is wound with smaller "wire" to allow it to tolerate a higher voltage with neutral brush timing. What you set motor current is up to you, of course, but I'd take it easy on the Kostov for the first few hundred miles.

Otherwise, your theoretical maximum power is determined by multiplying voltage and current on the same side of the controller - you multiply battery current and battery voltage, not motor voltage. Thus your maximum power would be your minimum battery voltage (including sag) times maximum battery current, or 270V x 600A = 162kW. Your motor power will be the same, minus a kW or so in losses.

http://www.diyelectriccar.com/forum...lla-31655p5.html?highlight=soliton+thermistor

Tess said they are a pain, and you mentioned paypal at least once :D

But maybe something has changed since 2009?
Much has changed in the controller code since 2009, but we still don't read thermistors directly... ;)

Logger records the voltage at each aux input so you can hook up the thermistor to one of them - as part of a voltage divider - and then import the data into a spreadsheet to convert the voltage reading into a temperature.
 
#29 ·
I'm not nearly enough of a battery expert so I can't say anything about if this will be ok for your batteries or not, but I can at least say that the controller will be quite ok as long as the pack voltage doesn't pass 342 Volt (at which point the precharge cycle will simply interrupt and the controller won't start).

I don't think the data for this graph covers when the car's taken directly from the charger. If it were I'd expected a higher pack voltage at the beginning since you usually charge to a higher cell Voltage than 3.3 Volt, so probably the car was driven for a few minutes before logger was started. On the other hand, the pack voltage usually drop to nominal voltage pretty quick when the pack's loaded so it wouldn't take much driving to "shave off" the excess voltage.

Again; I'm not a battery expert, but it seems that it's safe to drop below minimum cell voltage at load which makes sense since the cells internal resistance will make the Voltage at the terminals drop lower than the "actual" cell voltage, and that's why we've added a two point LVC-curve rather than a fixed minimum pack voltage in 1.4.

...which btw will be released real soon now. Initially it was supposed to be released in the beginning of this week, but then we decided we weren't happy with the temp deration of Junior so we've been tinkering a bit with that. So, well, soon. Real soon... :D
 
#30 ·
Thank you Qer,
As I am not looking to get max Amp I figured a Amp derating would not really affect me, I'll be sure to stay under 340V though.

Setup with current version 1.3 has been real smooth. Getting the logging info through, and the android ev speedo is picking up Volt, Amp and RPM in realtime! working on temp hookup now..

Just read an interesting convo between yourself, Tess and Plamenator (~Re: Ideal motor for the Zilla...?(2009)), where you were also talking about the hookup for an KTY84-130 thermistor.
Do you guys have a suggested/ best practice method for this input?
AK
 
#31 ·
Just read an interesting convo between yourself, Tess and Plamenator (~Re: Ideal motor for the Zilla...?(2009)), where you were also talking about the hookup for an KTY84-130 thermistor.
Hooo, that's ancient and I don't remember that. You have a link to that thread?
 
#34 ·
Thank you Tesseract,

I meant to say I'd limit motor current to 675-700 Amps. The 162kW output figure is also the top of what I'd like this drivetrain to produce, and most often much less.
We are replacing a Volvo B20 (or AQ130 as it is the marine version), 4 cilinder 130HP engine.

We are working on the homebrew 0-5 Volt output for the thermistor, as we really want motor temp as part of the digital dash.
AK
 
#35 ·
I meant to say I'd limit motor current to 675-700 Amps.
That sounds reasonable for a 250V 11" Kostov motor.

We are working on the homebrew 0-5 Volt output for the thermistor, as we really want motor temp as part of the digital dash.
AK
If you are concerned about your motor's health then monitor the commutator temperature with an industrial optical pyrometer (e.g. - from omega.com). The commutator heats up and cools down much more rapidly than the field or the armature, even, and it is usually the first thing to get damaged so monitoring it makes the most sense.
 
#36 ·
I have a support question.
I get a slow blinking error light when I enable input 2 to be motor temp.
I have 12v going through the N/C switch in the motor.
If I disable input 2 the error light stays off.
Do I need to wire it a different way?

It's a Soliton1 connected to a Advanced DC FB1-4001A motor.
 
#37 · (Edited)
This motor uses a Normally Open temperature switch, not a Normally Closed. See the motor's product description PDF: http://www.evparts.com/cms/picts/products/mt2120productinfo.PDF

You can leave it wired the way it is now after you upgrade to release 1.4 of the controller software as it has the option to toggle the polarity of any of the aux inputs (if - change them from "active high" to "active low"). In the mean time, you need to wire the switch in between the input and ground then use a pullup resistor (value of 1k-10k is fine) between the input and 12V.

The full release package for 1.4 should be posted on our website any day now. I still need to finish editing the manual first.
 
#40 ·
These questions were asked in another thread, I moved my answer here to not further clutter that thread.

Is your built in contactor something that is field replaceable if it fails?
No. There's no need. If the contactor ever blows up it's a sign that something else has blown up as well and just replacing the contactor won't help one bit. It'll probably just mean you'll blow up a second one.

Have no idea what type contactor you're using or if it's a semiconductor switch but I've blown a few, probably due to not having a precharge.
And that's the reason. We've blown one contactor in total and that was because the software didn't bother to check for negative pack voltage but in one early experiment the polarity was swapped by mistake. When the precharge code decided that precharge was finished (ie when it couldn't detect an increase in capacitor voltage anymore which it couldn't since it can't measure negative voltage) it enabled the contactor and the ~-150 Volt was connected directly to the only partly charged capacitor and the resulting current made the contactor pretty much evaporate instantly. This is, as far as I know, the only contactor we managed to blow up though.

This was a very early beta controller though and nowadays (since the first ones was shipped actually) the Soliton abort the precharge and goes into fault mode if the voltage doesn't follow a normal precharge curve. This is also the reason why an external contactor (yes, I know that at least one customer has that, "just to be safe") has to engage within a second from when the Soliton is powered up. If it's slower than that the precharge sequence will be aborted immediately. If the precharge sequence finish according to the required parameters the contactor will be engaged and by that time the voltage over it will be well within specs for it to safely do so without risk for damage or excessive wear.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top