• [mention=2387]Simon Knight[/mention] and [mention=36331]Bogmonster666[/mention] - you guys are getting great ranges out of your batteries which I'd think is mainly down to the max assist speed being set to 25kph (15.6mph) and you spending a good amount of time riding faster than that.
  
• I friend of mine has a Fazua fitted bike with a 250Wh battery and the battery still had charge after a 70 mile ride as he often rode above the 25kph cut-off.
  
• All testimony to using a lighter bike, more aerodynamic clothing/bike profile and being fit enough to peddle above 25kph.
  

Tbh, I suspect I am on the cusp of not really needing an ebike. I suspect if I had a decent road or gravel bike I would do ok for regular cycling. Tonight I cycled back from the pub after caving with 2 X 25l panniers crammed with totally mud caked cave equipment - the mud is quite something. The wind is ~40mph and I have just come down with a stinking cold and feel crap. It was great to put the bike on 'sport' and stick just below the assist level and let the motor take some of the strain. Likewise, with a supermarket run and 800 ft of ascent on the return, the motor will be a godsend.

 

On the utility front there is also a time aspect - it's 7 miles back from the supermarket with the big hill and in the ebike I can do that in 30 mins with shopping. There is no way I could do that on a non-ebike.

 

But I tend to think a bike either with a minimal motor or no motor will be better suited if I really want to do long rides. I've never really had a decent road bike so no idea what I could really achieve. I can see me ending up with an ebike workhorse and potentially a gravel bike with drop bars for leisure riding.

 

 

But the range thing is a useful exercise and a fun experiment.

 

C

Edited April 12, 20233 yr by Bogmonster666

My calculation says that you could have used a maximum of about 150wh until the first bar went on your display if everything is set correctly. For 60 miles at an average of 12 mph, that would be an average of 30 watts from the battery, or 21 watts of actual help.

Simon has 48V 15AH battery with Panasonic 217000 5,000mAH cells. The first bar would start to go as you said, after around (54-1.3)V*3AH=158W, so 150WH is very much ball park figure.

He managed 60 miles, that's an average 3WH per mile, which is excellent but there are two factors:

1) he still has strong legs - when you go from non-electric to electric, your pedalling is still excellent

2) he has a recumbent - which is aerodynamically much more efficient to slice through the air

I am pretty sure that overtime, the average consumption will go up but those early figures are still very pleasing to see.

The main weakness of the TSDZ2 is in the area of waterproofing. The water seal on the casing is a very thin (about 2.5mm-3mm thick) O-ring. It's fiddly to re-install the gasket correctly after you open the case. If I were to suggest an improvement, I would use a thicker O-ring or add a second O-ring.

I would recommend that owners a) never ride through a deep puddle, b) stick some deflector on the front of the motor or do like Vidtek, put some cling film over the casing.

The main weakness of the TSDZ2 is in the area of waterproofing. The water seal on the casing is a very thin (about 2.5mm-3mm thick) O-ring. It's fiddly to re-install the gasket correctly after you open the case. If I were to suggest an improvement, I would use a thicker O-ring or add a second O-ring.

 

• Yes, a weak area as I found out on a friend's TSDZ2 (my post on this is HERE).
  
• Personally I'd recommend a frequent inspection of this area if you are often going through deep puddles.
  

Riding home last night there was a lot of standing water - fortunately the roads were deserted so I could weave about.

 

As I understand it the main weak points are the seals under the chainring. These are not particularly easy to protect with a deflector (other than mudguards obviously). It almost needs something unsightly like an old fashioned static chain guard that covers top and bottom of chainring....let's hope toSeven have cracked this problem in their new motor. I am having a think on options and can imagine the shape of the deflector, struggling to work out how to mount it...

 

My guess is that tyre spray from below will pose more of an issue than rain from above?

 

Ultimately though, periodic checking as suggested by [mention=25387]Bikes4two[/mention] may still be required.

 

C

Edited April 13, 20233 yr by Bogmonster666

...let's hope toSeven have cracked this problem in their new motor. I am having a think on options and can imagine the shape of the deflector, struggling to work out how to mount it...

 

My guess is that tyre spray from below will pose more of an issue than rain from above?

let's hope so. Whenever a new motor is introduced, there is an opportunity to improve. Some areas that have been candidates for improvement:

1. Stronger motor spindle. It rarely breaks but it's still a weak point. As the motor spins twice as fast as the BBS01, it should be as strong as the BBS01's spindle, not smaller.

2. Water proofing that we discussed. If you look at the underside of the motor, you can see the joint line. That is where water can get in, albeit very little. Water turns into steam and there is no escape route. A channel for the steam to escape would fix this.

3. A configurator tool for those who wants to tinker with their motor and not to have to rely on OSF firmware which will forever remain outside the warranty.

4. A better torque sensor. The torque sensor on the TSDZ2 is very clever but the friction between the rotating coil (chainring side) and the fixed coil (on the controller side) can detoriate the separator over time, forcing you to replace the whole thing. Bafang torque sensor is better built.

Tsdz2 main reported failures are air space within the motor unit and heat simply not being able to cool by passing to the exterior casing ( Air being a poor heat conductor).

The axle having a circlip groove that leads to the RHS especially snapping ( more so on the 100mm BB models).

Also lateral play on both cranks with insufficient bearing support on both sides (fortunately the fix is a simple cheap one).

Edited April 13, 20233 yr by Nealh

Not sure I want to know of all the failure modes...... head in sand! However, I will look at adding some extra water proofing / deflection before I venture out in the wet.

 

I have decided that the six bar battery display is not very useful so am in the process of adding a volt meter. I purchased these from Amazon https://www.amazon.co.uk/dp/B09XTMJTHW?psc=1&ref=ppx_yo2ov_dt_b_product_details. Beware, they are not waterproof as water can get in from the rear so I've potted mine with glue from a hot glue gun and mounted it above the display at the front of my trike in the hope that it will give me a better idea of the state of charge.

 

S

I would be interested to know how much use folks get out if a TSDZ2 before

Not sure I want to know of all the failure modes...... head in sand! However, I will look at adding some extra water proofing / deflection before I venture out in the wet.

 

I have decided that the six bar battery display is not very useful so am in the process of adding a volt meter. I purchased these from Amazon https://www.amazon.co.uk/dp/B09XTMJTHW?psc=1&ref=ppx_yo2ov_dt_b_product_details. Beware, they are not waterproof as water can get in from the rear so I've potted mine with glue from a hot glue gun and mounted it above the display at the front of my trike in the hope that it will give me a better idea of the state of charge.

 

S

I've bodged something with damp proof course wrapped around the housing and secured with cable ties. It ain't pretty but I think it will deflect the worst of the water.

 

C

Most of the issues are thsoe generally reported on ES where as usual they don't use the drives as intended, the obsession is to abuse drives with power and voltages well beyond the designed capabilities.

They do though, do most a favour by finding out the weaknesses and how far to push a drive.

 

The first thing I'd check if my battery was showing full on my LCD after 60 miles is whether it's set to 48v or 36v, and if it turns out that it's set correctly, I'd be looking to save weight by removing the electric kit. I know that when I ride my road bike, the effort is about the same for the same speed as on my electric bike on level one using about 50w from the battery or 35w of actul help. That's at an average speed of 14 mph (hilly).

 

The LCD should have automatic voltage recognition between 36v and 48v, but I know that if you fist connect a nearly flat battery to some LCDs, the software wrongly interprets the battery voltage, so you have to disonnect the battery and reset the LCD before reconnecting to a charged battery to get it right.

 

 

Hi, on reflection I should have chosen a better title for this thread, probably "how accurate is the battery charge indicator?" The short answer is "not very". This is confirmed by forum comments and a few reviews like this one.

Given there’s only six bars when full, the indicator is really only reliable to indicate “full”, “not full”, and “OMG flashing empty hope you’re nearly home”.

 

Initially I suspected that a voltage setting in the display was incorrectly set as you suggest, but I have been unable to find such a setting and if it were set to 36 volts then my 48 volt battery would register full charge throughout its 41v-54v cycle. Instead it has dropped one bar when the battery is just under 50% charge (76 miles). It is what it is so I am fitting a small digital volt meter to help me judge the state of the battery.

 

I can only dream of averaging 14 mph. Over the decades of recreational cycling when measured I average something like 10mph. I decided the simplest solution was to stop measuring and just enjoy the world at a slower speed. Nowadays if I am cycling faster than I can walk I call it a win.

 

You may ask so why have I fitted an electric motor?

 

My main reason is so that I don't use my car as much as I have in the past. I live out in the sticks where the amenities are between two and twelve miles away. Any one journey is quite doable on any cycle but several in one day can be more of a challenge especially if loaded with shopping and riding into a 30 kt headwind.

 

Last year I went to the York show and visited the ICE stand and tried out their newest trikes. Some were demonstrator machines and a number were personal machines that were in almost daily use. What was striking was with the exception of the VTX racing demonstrator I think all of them were fitted with motors. I was talking with one of the founding directors of ICE and asked why they had so many etrikes on their stand, one of the reasons stated was "because they are so much fun".

 

You mention system mass versus assistance used and that I might be better removing the motor system. You could be correct but I suspect that I have been using more power than the battery state indicator implies. Probably 2/3rds of the charge capacity over the 76 miles ridden to date which I think suggests that the motor and battery have been of use and not just dead weight.

 

As you imply the actual range figure is not that relevant as everyone's use case is different. When compared to my unassisted rides I see that the motor will allow me to achieve one of two things : 1) ride at higher speed over a similar distance or 2) ride a greater distance at a similar speed. I also hope that my motor will mean that I will ride more often and possibly get fitter but only time will tell.

 

S

Most of the issues are thsoe generally reported on ES where as usual they don't use the drives as intended, the obsession is to abuse drives with power and voltages well beyond the designed capabilities.

They do though, do most a favour by finding out the weaknesses and how far to push a drive.

I agree. Its a pity that there is not a similar obsession with efficiency.

 

If my Tongsheng lasts a couple of years I will be happy. After all it is not expensive to replace and unlike some of the big name systems spares are available for a reasonable cost.

So, I understand the calculation for wh per bar of the display - that's simple enough.

 

My question is about assistance at various cadences. I know the tsdz2 motor notionally like to spin at 4000 rpm ( which I am guessing is at 48v ). So presumably it is happy spinning somewhat faster at 54v on a fresh battery ~4500 rpm and as the battery depletes to 41v then ~3400 rpm?

 

I have only taken the battery down to 3 bars on the display but it really felt like everything was slowing down considerably and I was having to work a lot harder. I suspect it was because I was trying to maintain a higher cadence that worked ok'ish on the fresh battery but not on the half depleted battery. The issue not being the level of assistance the motor can give, rather the drop in cadence that the assistance can be delivered. Basically, needing to reduce my pedal cadence to a point that feels very unnatural - or just do all the work myself.

 

Edit: I have just done some calculations based on speed, wheel size and gearing. I think I was getting reasonable assistance at ~85 rpm on a full battery and ~75 rpm on a half depleted battery. This kind of makes sense I think and ties in with others who have said the tsdz2 peaks at about 75rpm at the pedals.

 

Edit 2:

 

4000 rpm is ~89% of 4500 rpm

75rpm is 88% of 85rpm

 

Close enough I think. Assuming I have not screwed up I think I can expect assistance at a miserable ~64rpm when the battery is almost fully depleted

 

 

 

C

Edited April 14, 20233 yr by Bogmonster666

In practice I think then that only the 1st half of the battery capacity is of practical use to me. The assisted cadence on the 1st couple of bars is ok but by bar 3 it was not nice. I think 80 to 90 rpm is comfortable for me. Less than 75 rpm just feels completely wrong, especially on hills - the amount of power I can deliver at lower cadences is greatly restricted and I find it much harder on my knees, I am an old codger after all.

 

In due course I think I will need to flash with OSF with its field weakening and support for higher cadence. Maybe I can adjust my cycling style but I doubt it. Fortunately the range of the 1st couple of bars is enough for now so not a big deal. I'll make the most of the warranty period before I flash the motor.

 

I'll do some more experimenting when my cold goes.

 

C

Thank you for the detailed remarks. I will look into how to optimise the motor for a specific cadence regardless of battery voltage.

Thank you for the detailed remarks. I will look into how to optimise the motor for a specific cadence regardless of battery voltage.

I know osf is not something that will ever be covered by warranty. For the tsdz2 I suspect getting new firmware from the manufacturer is unlikely.

 

Out of curiosity I had a little peek in the OSF code on my phone. Field weakening is started to be employed over 48 rpm and progressively up to 105 rpm.

 

At my cursory glance it takes no account of voltage when setting the field weakening angle - but might be easier to read on a proper computer.

 

Assuming osf field weakening really works (and anecdotally it appears to), it's a great shame Tonsheng haven't baked this into the stock firmware.

 

I understand that machining and material costs to fix physical weaknesses can be high but a software change should be relatively cheap to implement. Of course flashing the tsdz2 is a faff with nasty error prone cables...so even if Tongsheg created a new firmware, it would not be easy for existing owners to flash. Of course, I think this is all somewhat academic as I get the impression the tsdz2 isn't being onwardly developed?

 

C

as you said before, it's not a big issue because you would go out with a full battery most of the time. Still, manufacturers are naturally reticent to change the firmware for good reason, one mistake could sink the ship!

If you buy your kit from abroad, then OSF is the way to go because you are on your own anyway but most of my customers prefer to stick with the original firmware until the warranty runs out.

PS: I just took a peek at the add field_weakening_angle code and got lost in their math or logic.

Will look again over the weekend. Maybe possible to change the formula to add an extra factor for battery voltage.

 

/* // add field_weakening_angle

ui16_temp = (uint16_t) ui8_g_duty_cycle;

ui16_temp = (ui16_temp * 100) / PWM_DUTY_CYCLE_MAX;

if (ui8_g_field_weakening_enable_state)

{

ui16_temp += (((uint16_t) ui8_g_field_weakening_angle) * 14) / 10;

}

ui8_tx_buffer[15] = (uint8_t) ui16_temp;

*/

Edited April 14, 20233 yr by Woosh

Hi, on reflection I should have chosen a better title for this thread, probably "how accurate is the battery charge indicator?" The short answer is "not very". This is confirmed by forum comments and a few reviews like this one.

 

 

Initially I suspected that a voltage setting in the display was incorrectly set as you suggest, but I have been unable to find such a setting and if it were set to 36 volts then my 48 volt battery would register full charge throughout its 41v-54v cycle. Instead it has dropped one bar when the battery is just under 50% charge (76 miles). It is what it is so I am fitting a small digital volt meter to help me judge the state of the battery.

 

The battery inicator is very accurate. Only people that don't understand how it works say otherwise. It's basically a voltmeter that tells you when the voltage crosses various boundaries. Tht was explained by Woosh above. It uses a voltage divider to indicate to the LCD's CPU an actual voltage. The CPU uses software to send a repeatble instruction from its software to the LCD so that each bar on the display will change at exactly the time the battery voltage reaches the condition for each one. You can check it with a voltmeter, and you will find that it's consistent and accurate. It's up to you how you interpret and use the information, which is where the inconsistency comes in.

 

Battery voltage is not a good indication of range because not only is the voltage non-linear as it discharges at a constant current, but it also varies according to how much current you take fom the battery, which is varying all the time when you have a torque sensor system. When you pedal hard, the current will increase and the voltage will go down, then the voltage will bounce back up when you stop pedalling or pedal with less effort. Also, like Woosh explained, the motor's power decreases as the battery voltage goes down, so riders tend to demand more current from the system by pedalling harder, which causes increased voltage sag. In practical terms, that all means that the battery goes down slowly at first, then accellerates downwards during its discharge cycle.

 

To stop bars flashing on and off as you pedal, the software includes a damping factor, which is basically a moving average of the voltage that has the effect of averaging out and delaying the response of the battery level indication in the LCD.

 

To summarise, the LCD is not a range-meter. It's a voltmeter. To estimate your range at any time you need to understand how hard you're likely to be pedalling for the rest of your journey and how much hill-climbing you will be doing, and whether you expect to change your speed. No computer can know those things. What you will always know is exactly what voltage is in your battery when any segment goes out.

 

As suggested above, if you want an accurate indication of how much battery charge you have left, you can use a wattmeter in conjunction with the LCD. You can see at how many watt-hours you've used when each segment goes out, which will be consistent as long as it doesn't happen at a ime of excess voltage sag, like when hill-climbing. You can then interpret each bar on the display as a known percentage of charge remaining, and, as others have mentioned, there won't be much charge left when three of five are out.

 

I say again from my vast experience that if you do 60 miles on any electric bike and the LCD still shows all bars, either something is wrong or you did something exceptional that a normal e-bike rider wouldn't do, like riding with the power switched off or an exceptionally flat ride with the wind behind you. It can't just happen without explanation and you confirmed that you average 10 mph, so it wasn't that you were riding over the 15 mph cut-off, which would explain how that can happen with a fit rider on a fast bike, and your recumbent trike gives no advantage at that speed. In fact its additional mass and roolling resistnce would be a disadvantage. You should check the function of your battery indicator with a voltmeter.

 

When you said that the first indicator went out at 50% charge, that doesn't make sense to me. How did you figure out how much charge was left without a wattmeter? Also, the indicator levels are fixed by the hardware and software. The only way it could happen is if a wrong resistor were soldered to the PCB, but I never heard of that. I'm losing sleep over this. Please give an explanation so that I can make sense of it all.

Edited April 14, 20233 yr by saneagle

as you said before, it's not a big issue because you would go out with a full battery most of the time. Still, manufacturers are naturally reticent to change the firmware for good reason, one mistake could sink the ship!

If you buy your kit from abroad, then OSF is the way to go because you are on your own anyway but most of my customers prefer to stick with the original firmware until the warranty runs out.

I made my choice to buy from a UK company with CS and a warranty that may actually be usable rather than importing. Tbh, it was mainly the lack of availability of batteries of a suitable quality and difficulty sourcing a 48v 250w motor.

 

I don't regret the decision to buy from the UK.

 

C

as you said before, it's not a big issue because you would go out with a full battery most of the time. Still, manufacturers are naturally reticent to change the firmware for good reason, one mistake could sink the ship!

If you buy your kit from abroad, then OSF is the way to go because you are on your own anyway but most of my customers prefer to stick with the original firmware until the warranty runs out.

PS: I just took a peek at the add field_weakening_angle code and got lost in their math or logic.

Will look again over the weekend. Maybe possible to change the formula to add an extra factor for battery voltage.

 

/* // add field_weakening_angle

ui16_temp = (uint16_t) ui8_g_duty_cycle;

ui16_temp = (ui16_temp * 100) / PWM_DUTY_CYCLE_MAX;

if (ui8_g_field_weakening_enable_state)

{

ui16_temp += (((uint16_t) ui8_g_field_weakening_angle) * 14) / 10;

}

ui8_tx_buffer[15] = (uint8_t) ui16_temp;

*/

 

 

This is the bit of code that works out the angle:

 

// Calculate Field Weakening max Angle based on current motor speed

// Field weakening max angle is 0 when cadence is below 48rpm

// Field weakening max angle is 6 when cadence is above 105rmp

// between 48 and 105rpm goes from 1 to 5

#define FW_COUNTER_MIN 65U

#define FW_DIVISOR 16U

#define FW_COUNTER_MAX (FW_COUNTER_MIN + (FW_DIVISOR*5))

 

if (ui16_PWM_cycles_counter_total >= FW_COUNTER_MAX)

ui8_fw_angle_max = 0;

else if ((uint8_t)ui16_PWM_cycles_counter_total <= FW_COUNTER_MIN)

ui8_fw_angle_max = 6;

else

ui8_fw_angle_max = 5 - ((uint8_t)((uint8_t)ui16_PWM_cycles_counter_total - FW_COUNTER_MIN) / (uint8_t)FW_DIVISOR);

 

I think the code you pasted above applies this. There may well be compensation for voltage elsewhere but I need a bigger screen... If not, I am sure it's possible to adjust the angle based on rpm and voltage, its only code, but I suspect it isn't important. Possibly tweaking the angle based on voltage might improve efficiency slightly but I don't know enough about motors to say either way.

 

My overall aim would be to get an elongated cadence range, I suspect the OSF field weakening logic is quite sufficient as it is.

 

I am guessing the Whoosh motor has the older controller supported by osf, not the V2? Like I say I'm, not in a hurry, but maybe when out of warranty...

 

C

Edited April 14, 20233 yr by Bogmonster666

Which version are you looking at?

Emmebrusa - TSDZ2-v20.1C.2-update-2 for VLCD5-VLCD6-XH18, I think...

 

Motor.c line ~768...

I think the code you pasted above applies this. There may well be compensation for voltage elsewhere but I need a bigger screen... If not, I am sure it's possible to adjust the angle based on rpm and voltage, its only code, but I suspect it isn't important. Possibly tweaking the angle based on voltage might improve efficiency slightly but I don't know enough about motors to say either way.

You have already seen that the cutoff speed of the motor is directly proportional to the battery voltage, the coefficient is known as motor constant Kv. It follows that you would want to increase Kv to compensate for the drop in voltage. The way to do that is by field weakening, So I imagine that you would want to apply a FW angle that depends on voltage instead of cadence, something like FW_angle = 0.5 * (54- battery voltage)

The battery indicator is very accurate. Only people that don't understand how it works say otherwise. It's basically a voltmeter that tells you when the voltage crosses various boundaries.

• When I first read the above, I was not in agreement with it at all but I then read further (and more than once) and the bit that says:

  
  

Battery voltage is not a good indication of range

• had me re-thinking it all and coming round to agreeing with what you say.
  
• Within OSF you can change the voltage thresholds for the bar indicators which I guess is useful for those who understand the implications of doing so(not me - well, just a little bit anyway).
  
• In my case I have over time and a good many charges/discharges for varying rides/miles/conditions, noted through the use of a watt meter the amount of power consumed and have formed a reasonably accurate 'achievable range' for my batteries vs the sort of riding and conditions I undertake.
  

So thanks for the above [mention=3847]saneagle[/mention] - every day is a school day.

The battery inicator is very accurate. Only people that don't understand how it works say otherwise. It's basically a voltmeter that tells you when the voltage crosses various boundaries.

 

The point of my original post was to gain knowledge of how the display operates by comparing what I see with what other users experience. "Very accurate" is an unfortunate term perhaps you would care to specify how accurate and how you know.

 

Also I have two volt meters on my system, the one we are talking about on the display and the one on the battery. The four segment display on the battery appears to be the more useful of the two as it has dropped two bars whereas the display is still showing 5 with the sixth bar pulsing. To be blunt I trust the meter that is showing the lower value.

 

As suggested above, if you want an accurate indication of how much battery charge you have left, you can use a wattmeter in conjunction with the LCD.

 

The issue I have with a watt meter is that they are high current devices so need to be well insulated and protected from the weather. I have one and I don't think it is of higher enough quality to have in circuit. I will however use it on the charging side to get some idea of how much energy is being used in charging the battery.

 

It can't just happen without explanation and you confirmed that you average 10 mph,

 

What I wrote was that over many years my average speed when recreational cycling was 10 mph i.e. on non electric cycles.

 

When you said that the first indicator went out at 50% charge, that doesn't make sense to me.

How did you figure out how much charge was left without a wattmeter?

Quite, it does not make a great deal of sense to me which is rather where this thread started and no, I have not fitted a wattmeter. Half charge is an estimate based on the second volt meter which is now showing 2 bars from 4. The battery off load voltage is 47.1 volts +/- 0.5v. In a perfect world both volt meters would change at a similar rate.

 

Also, the indicator levels are fixed by the hardware and software. The only way it could happen is if a wrong resistor were soldered to the PCB, but I never heard of that. I'm losing sleep over this. Please give an explanation so that I can make sense of it all.

 

No need to lose any sleep it is what it is. The explanation is that the measuring device is cheaply made with little or no quality control and no calibration. None of this is surprising and it does not cause me to lose any sleep. The equally cheap volt meter I have just fitted will give me a better idea how the system is performing and when I start measuring joules in I will be to have some idea how the volt meters perform.

 

Moving on, so far I have not noticed any issues with regards to cadence but I probably have some way to go before the issues reported may become apparent.

 

Placing all of this post to one side I am happy with both the kit and the support from Woosh. To put it simply, fitting the kit has made my trike a better car replacement than when it was unassisted. The only negative is that in early spring a greater number of layers need to be worn owing to the wind blast from higher speed in conjunction with lower physical input from me.

 

S

• had me re-thinking it all and coming round to agreeing with what you say.
  
• In my case I have over time and a good many charges/discharges for varying rides/miles/conditions, noted through the use of a watt meter the amount of power consumed and have formed a reasonably accurate 'achievable range' for my batteries vs the sort of riding and conditions I undertake.
  

 

So how does the six segment display of voltage operate on your rides? Does the rate of change of the display accelerate as the number of segments decrease and is there any correlation with what you see on your watt meter?

 

S