【convoy】Some items are unavailable during the 2024 New Year holiday.

It is normal that S21B XhP50 dont have spring from driver side?

I have sst40 and there is this spring but in XHP - i have 2 of this and there is no spring

I have the same


Calculating the (peak) sense voltage a driver uses is of help to determine how much current will go through a given resistor in it.

For the Convoy ∅17mm 5A linears with a 10mΩ sense resistor, V = I × R = 5A × 10mΩ = 50mV peak sense voltage.

So, since I = V / R you can calculate how much current for each resistor: R020 -> 2.5A, R025 -> 2A, R033 -> 1.51͡5A, R050 -> 1A, etc.

Power rating matters too, driver uses 1210 imperial size (3225 metric, 3.2 × 2.5mm), but if current is lower size can also be smaller: R020 or bigger value sense resistors (2.5A or less) could be 1206 imperial no problem guaranteed.

Thx Barkuti.

I just ordered a M21a with sst40. In the description it says the following:
Driver: 4 modes 0.1–3–30–100 / 12groups, max current output 6000mA,

Does this mean at 100% output the driver only allows the emitter and battery to run 6amps? I guess I’m asking, like a fw3a when on turbo runs a high amperage depending on your battery. Correct? Is this m21a constant current + FET or just constant current?

Artiet59, the drivers you mention are constant current linears, they use MOSFETs in triode mode or linear region. The principle of operation is like a gas valve (the MOSFET or MOSFETs) where the flow of gas (current) depends in how much the handle is turned open (MOSFET VGS or gate-source voltage). The driver MCU senses the current flow in the sense resistor, and tunes the MOSFETs VGS to limit the current flow accordingly. Of course, when the current flow is below the limit (battery too low, high emitter Vf, etc.) they limit nothing.

What you call “FET” mode is operating the MOSFETs in saturation mode, this is like opening the gas valve “all the way”. The current is then mainly limited by other reasons like spring, driver and battery resistances. If you can only operate the MOSFET blindly in saturation mode (open all the way), the only way to obtain lower modes is with PWM (quickly opening and closing the gas valve).

After checking out the thread CRX FW3A Mods & Teardown I see a driver with 8x 7135s plus MOSFET in saturation mode. Without wanting to be judgmental about the driver (great software I guess), it employs very outdated ways of doing the job. 7135s can only work at 350mA constant current, and so the only way to obtain lower modes is to PWM :facepalm: them. Honestly speaking, I don't see a reason to mix 7135s with a MOSFET onboard, as you can obtain lower modes by directly PWMing the MOSFET (and thus saving all the space occupied by the 7135s).

I prefer to always go regulated, but of course I know how to fine tune the sense resistor of a driver in case I need to lower or to increase the maximum current.

The FW3A has a very fine tuned driver, the Lume1 for FW3A/C/T - Constant Current Buck Boost + FET Driver with Anduril - Now for Sale! (not :X cheap, though).

Barkuti- thank you for this info. Fantastic job drawing comparison with an engine which I understand functionally better then a electronic driver. But I think I’m catching the drift of what you’re saying. One question- because I don’t know- what is the benefit of running PWM on the mosfet itself versus running PWM on the 7135’s? What exactly are the 7135’s, resistors?

Good explanation and simple-to-understand analogy for us who are not that well-versed in electronics.

However, regarding the PWM — based on my visual inspection (I’ve tried out the 4-mode and 12-mode driver [not the original Biscotti 12-mode]), both the 4-mode and 12-mode driver do not produce PWM (at least when I use the “shine light through a small fan” method). Does this mean it doesn’t use PWM for lower modes (in contrast, I can detect “fast” PWM with a flashlight that uses the Biscotti 12-mode driver, when set to lower brightness levels; but not in the lower brightness levels of the Convoy 4-mode (5A/6A) & 12-mode (5A/6A) driver… Or does it also use PWM but just too fast to detect? I remember Simon mention that these drivers are “constant current” without PWM (although I’m not that well-versed in electronics whether constant-current and PWM can co-exist?)

I’m also curious if PWM FET only drivers are more efficient than a 7135 based driver up to 3-5A.

I hesitated to buy some replacement drivers from Simon with the FET only layout that Artiet59 listed.

LEDs are more efficient at lower currents, so using PWM to lower the output of a 7135 is slightly more efficient than PWM on the FET in saturation. In actual practice, the difference in efficiency is probably negligible when compared to something like a buck converter.

FETs in the linear region will be more efficient than 7135 for lower modes (running the LED at lower current without PWM). For the high mode it will be the same efficiency.

Your 6A 4 mode driver will run at 60mA, 180mA, 1800mA and 6000mA. The 1800mA mode will be as efficient as a 5x7135 driver (1750mA), the 180mA mode will be a bit more efficient than a 1x7135 driver.


Simon might drop 3x21 soon.

US $34.16 | LUMINUS SBT90.2 LED 25mm 20mm

Just for information the T2 host does not come with an MCPCB, even though the AE listing clearly says “with MCPCB”. Simon confirmed this.

Guess I’ll have to file down a 16mm MCPCB or source a 14mm one from somewhere.

Buf! :facepalm:

You can read the article AMC7135 PWM LED control – works? by Pratik Panda. Basically, AMC7135 don't cope well with fast PWM. Pratik says “up to 6KHz”, although in my opinion the switch on time and distorted waveform period is already pretty significative at that frequency…

With all due respect, this is senseless. The efficiency arguing, I mean.

A MOSFET in saturation region is like a switch and it is meant to be very efficient in letting go of all the energy (electricity) which goes through. This is what is to be expected. If you blindly look at this like something “good” and cannot see the overall effects of it, you may be wrong.

A MOSFET in linear region is like a variable switch or valve, which is controlled via the gate-source voltage (VGS). When VGS is “maxed”, this is comparable as if someone “turns the lever all the way” (see below), entering into saturation (as above).

This is what I call a gas valve:

The car's gas pedal may also be a valid interpretation. What I call gas is not gasoline, but something like the air you breathe. Electricity is a “subtle fluid” which behaves sort of like a gas, but is in fact even more subtle than a gas, as it is a flow of electrons, tiny sub-atomic particles with charge and tiny mass.

A PWM FET driver employs a MOSFET like a gas valve where the turn on/off handle is fully set each time very quickly. Since you can only “turn it all the way” each time, now imagine how would you drive your car at cruise highway speed if you could only “press the pedal to the metal or nothing”. You would be forced to press it down for a small while, and let it go for another small while; rinse and repeat. Now imagine how would you do this in a very powerful vehicle, or if you had to cruise slowly through a dusty secondary road using a low gear… pedal down a tiny while, pedal up a longer while, pedal down a tiny while, pedal up a longer while… LMAO.

Does it make sense to compare linear devices in terms of efficiency? A linear driver is meant to burn all excess power (voltage times current), to avoid that excess power to reach the led. So, how can it be efficient? Efficiency @ Wikipedia. You tell me. I think they are efficient in accomplishing their goal.

People please stop with the @#$% about the newer regulated MOSFET drivers. They're very nice. Thanks to them we have ditched the PWM issue (so they are fully constant current) and can enjoy higher efficiency at the led emitter.

All the drivers dubbed “SST40” in the Convoy store are regulated MOSFET drivers.

The following is namely meant for Simon. Out of all the ∅22mm SST40 drivers:

4-modes, 12-groups and 2-modes; something seems bad with the blue, 2-modes driver in the picture. The sense resistor value is 20mΩ instead of 10mΩ, with maximum driving current being equal at 6A. This cannot be correct, as it would force the driver to employ a higher sense voltage (120mV instead of 60mV, and undesirable) for the same output current, or the maximum output current would be different (3A instead of 6A). I don't see a reason for the manufacturer to make such a senseless change besides board color or firmware… Just a side note.

That’s nicely said and well explained.

Have you got a plan for the use of it? What host?

C8+ reflector for 20mm sbt90 when

That’ll be one super heated C8+

Yolo and all that