Dissecting the Yi-Yun YK42-3 36v Controller

[Papa]

After receiving my YK42-3 36v Controller, I decided to crack its case, examine and photograph its innards. The first thing I noticed was the FETs. They are F1010E, not the 2SK3435 seen in other YK42-3 units. The pictures for mine, and the FET specs for both can be compared below. I originally planned to mod or eliminate the LVC (to allow 24v use), but decided instead to increase my supply voltage to 36v - providing my motor can safely handle the increased voltage. (edit - I modified image size so refresh your cache)

(FETs) IRF1010E (my YK42-3 - 36v shown above)

Absolute Maximum Ratings
Parameter Max. Units
ID @ TC = 25°C Continuous Drain Current, VGS @ 10V 84 A
ID @ TC = 100°C Continuous Drain Current, VGS @ 10V 59 A
IDM Pulsed Drain Current 330 A
PD @TC = 25°C Power Dissipation 200 W
Linear Derating Factor 1.4 W/°C
VGS Gate-to-Source Voltage ± 20 V
IAR Avalanche Current 50 A
EAR Repetitive Avalanche Energy 17 mJ
dv/dt Peak Diode Recovery dv/dt 4.0 V/ns
TJ Operating Junction and -55 to + 175
TSTG Storage Temperature Range
Soldering Temperature, for 10 seconds 300 (1.6mm from case )
°C
Mounting torque, 6-32 or M3 srew 10 lbf•in (1.1N•m)

-----------------------------------------
(FETs) 2SK3435 (other YK42-3 - 36v units)

ABSOLUTE MAXIMUM RATINGS (TA = 25°C)

Drain to Source Voltage (VGS = 0 V) VDSS 60 V
Gate to Source Voltage (VDS = 0 V) VGSS ±20 V
Drain Current (DC) (TC = 25°C) ID(DC) ±80 A
Drain Current (pulse) ±160 A
Total Power Dissipation (TC = 25°C) PT 84 W
Total Power Dissipation (TA = 25°C) PT 1.5 W
Channel Temperature Tch 150 °C
Storage Temperature Tstg −55 to +150 °C
Single Avalanche Current Note2 ISA 31 A
Single Avalanche Energy Note2 EAS 96 mJ
Notes 1. PW ≤ 10 µs, Duty cycle ≤ 1%
2. Starting Tch = 25°C, VDD = 30 V, RG = 25 Ω, VGS = 20 → 0 V

 

[RLT]

Interesting.

The 3435s seem to be a little harder to find, maybe they changed to something that was more widely available.

That copper wire down the back of the FET Bus: Did you add that, or did it come that way?

 

[Papa]

... That copper wire down the back of the FET Bus: Did you add that, or did it come that way?

Not my doing. I paid for new, so I'd like to think OEM... but suspect tampering fingers somewhere in between. I noticed the factory seal was broken when I unwrapped it, so who knows? Funny, I haven't a clue if it even works.

 

[HAL9000v2.0]

Is it boosted or is just "factory trimming"?

 

[fechter]

Great pics.
For as crude as those things are, they have a pretty good track record.

It looks like it would be easily modified to run anywhere from 24v to 48v.

I suspect the solder on the shunt is factory trimming, but I'd be real interested in seeing what the actual current limit is.

If your motor is unhappy at 36v, you could always limit the throttle to keep it lower.

 

[Papa]

Appreciate the PM suggestions, fechter. I may have to resort to 24v initially as I'm still one battery short. I'm not nearly as interested in speed as much as efficiency and reliability.

Also,..

I realize the circuits probably differ, but when I read through your Crystalyte R&M thread, I noticed that this YK42-3 (unlike the Crystalyte), had no insulators around the FET's and diode's mounting tab screws. There was thin waffers and thermal compound between components and the case, but essentially no insulation anywhere else, that I could see. Is this proper?

 

[fechter]

The mounting screws need to have insulators under the heads. There should be no connection between any of the FET or diode tabs and the heat sink.

In some cases, it might be possible that the power devices bolt directly to an aluminum bar and the entire bar is insulated from the case by a pad of some sort. I don't think they do this.

 

[Papa]

The mounting screws need to have insulators under the heads. There should be no connection between any of the FET or diode tabs and the heat sink.

That certainly isn't the case with this YK42-3. The only insulation I found, was the waffers and a small dab of thermal compound between the component's tabs and the heatsink/case. Also, no nuts are used on the tab side - just three (3) bare, steel pinch bars pressing all of the tabs to the heatsink.

[UPDATE] So far, I found no shorted FETs, so I'll be testing it tomorrow.

 

[Papa]

What good is a thread without a little OT contamination? The image below is of the OEM controller that was originally paired with the motor i'm currently using. The YK42-3 is expected to replace this unit. It has 12 FETs (and one regulator?) attached to the heatsink. The heatsink measures 5.5" x 1.25" x .75". No idea what the specs are. It'll be my donor for the insulators (and anything else I need) for the YK42-3 controller.

 

[fdion]

After receiving my YK42-3 36v Controller, I decided to crack its case, examine and photograph its innards. The first thing I noticed was the FETs. They are F1010E, not the 2SK3435 seen in other YK42-3 units. The pictures for mine, and the FET specs for both can be compared below. I originally planned to mod or eliminate the LVC (to allow 24v use)

Anybody knows how one could mod the controller to run on 24 or 36v? In order to improve redundancy, I'd like to run on 36v but if I loose a pack, I can run on the other 2, at 24v. I can actually run with no controller using a bypass switch, but I really would like to be able to use the throttle.

 

[safe]

I've gone about 5,000 miles using this 36 volt controller and a significant portion was at 48 volts.

So what did I do today? Pushed my luck?

Yeah, well I had some wiring laying around and figured it would be fun to try rewiring my #001 bike as a 60 volt machine. I wired it up and since the capacitors are only rated as 63 volt peak I think that's what made one "pop" on me. The smell of burnt capacitors... ahhhh... like Castrol racing oil for gas powered machines... or napalm in the morning for a vietnam soldier. :lol:

Anyway...

Is there any problem with me replacing the burnt out capacitor with one of a higher voltage rating and higher capacity?

 

[317537]

Ok Im new to mods and learning electroniques but am very keen on working on these circuits.

That 7815 regulator is going to need some tinkering around.
http://www.farnell.com/datasheets/63524.pdf

You could use a 20v zenner or a larger resistor. Let me thinks about the resistor value.
Is this right.

Im getting Between 600 ohm at 5 watt.

Does this work? Dont quote me BTW.


At 41v subtract 15 volt output, leaves you with 26v with a 290ohm series resistor, you gets 2.2 watts at 80ma

At 65v subtract 15 volt output leaves you with 50v with a 600 ohm series resistor you gets 4.1 watts at 83ma.

At 65v subtract 15 volt output leaves you with 50v with a 1000 ohm series resistor you gets 2.5 watts at 50ma.


At 65v with a 600 ohm resistor the circuit can not generate anymore than 7 watts, subtract the 4.1 for the resistor and this leaves 3 watts on the reg. The reg can supply up to 15v @ 1 amp with 35v at the input and max at 1.2 amps short circuit so it will live at 65v (60v pack) with a 600 ohm 5 watt resistor just as it lived with a 48v pack installed. Its getting squishy for current if you go higher in resistance.

So maybe the larger ohm resistor is required at the risk of cutting current.

You will lose approx 4 watts an hour with a 65v 600 ohm conversion.

Or approx 2.5 of a watt with the 65v 1000 ohm conversion with the risk of starving the circuit of current.

I'm using 3x1 watt 10v zeners in parallel for a 12v overvolt now . Maybe 24v worth of zener would protect the circuit of a 65v system better, Just make sure you dont draw to much current through them.. Its easier with zeners as the voltage drop is almost constant and the supply is limited to the watage divided by the voltage i believe. So you may want to increment the voltage down by 12v at a time with a voltage divider. erm. With that 290 ohm resistor maybe 1 X 24v @ 5 watt zener would do the trick at a cost of 2.4 watt/hour.

If anyone can correct my equasions please do so.

 

[317537]

What good is a thread without a little OT contamination? The image below is of the OEM controller that was originally paired with the motor i'm currently using. The YK42-3 is expected to replace this unit. It has 12 FETs (and one regulator?) attached to the heatsink. The heatsink measures 5.5" x 1.25" x .75". No idea what the specs are. It'll be my donor for the insulators (and anything else I need) for the YK42-3 controller.

Do I read correct?

that controller was tested in 1992.


17yo controllers ?

Isnt this a relic by e-bike standards?

Oh BTW I think I've found lvc on the op's controller.

That 7815 reg has a min voltage of 18.8 volts. At 36v with that 290 ohm resistor and the voltage slump of a low battery wouldnt this trigger LVC, correct?

 

[bigbore]

I've been using a YK42-3 Controller on my e-scooter from one month. This controller does not have a switch to switch off when not using the e-scooter so I've been using a connector to power the controller when need it and to switch off when not in use to avoid battery discharging.
Can a power relay be used instead of a connector? Also I've noticed big sparks each time I connect the controller, is it normal? Could it cause any issue?

 

[texaspyro]

The sparks will quickly erode your connector and/or relay contacts. Most people add a pre-charge resistor to charge up the on-board caps before the full-power connection is made.