Solution Finder

Linkswitch TNZ In 3 wire mode

Posted by: HenryRoe89 on 11/02/2021


I'm looking into implementing the Linkswitch TNZ in a transistor phase dimmer design and I've been looking at DER-865. I am looking for more information and guidance about how the device will operate with a neutral connection.

For context I am considering when the dimmer is fully on and current is only drawn at the input of the supply when live is negative, since the return for the supply will always be at live potential, i.e. a half wave rectified input and your design spreadsheets indicate a 2200uH inductor will be fine for a 70ma output, lovely stuff.

1. We will very likely be using either a +10dBm/+20dBm 433/868MHz transceiver and I'm concerned about the proper operation of the device with relatively high peak current draw when transmitting. I plan to increase the input capacitance to mitigate voltage drop at the input of the supply but if we are pulling 100-120ma for around 40-50ms when transmitting do you foresee any problems?

2.  If we were to operate the device with a neutral only, would D6 still be required for the zero cross? I assume not but I want make sure!

If you have any other gems of wisdom which aren't present on the datasheet or reference design it would be greatly appreciated.


Many thanks,



Submitted by PI-Crumb on 11/09/2021

hi Henry,

For item 1, you can definitely just increase the input capacitance to compensate for the half-wave rectification. However, you might need to redesign the buck converter to be able to deliver 120 mA. Please use a larger LinkSwitch-TNZ device.


For item 2, if it's neutral only, you still need to use D8 (I think you mean D8 instead of D6) to be able to accurately detect the zero-crossing on each phase of the AC line.


Submitted by HenryRoe89 on 11/10/2021

Hi PI-Crumb,

Thanks for getting back to me that's really useful information and yes I did mean D8 not D6.

I am still a bit confused about the inductor selection. So we will normally be drawing around 20-30ma or so and then around 120-130ma when transmitting, but only for 40-50ms and with a fixed interval between transmissions.

Ideally I would like to design the supply to be able to deliver the full load current continuously but I can mitigate for this by designing in a fail safe enable circuit for the transmitter to make sure it doesn't draw current continuously if it fails. I only really need to be able to deliver the peak current for a short time. 

Regarding Inductor selection: I can't find any design calculations for the inductor selection anywhere, but your design spreadsheets give very strange results. The inductance required seems to go up with load current to about 2.7mH @ 80ma and then drops back down to approx 1mH @ 90-100ma and the frequency of the converter remains constant(?). Is this to prevent overshoot at lower currents, something relating to the minimum on time of the converter? If so why such a drastic change in the required change between 80-90ma?

In our situation, should I go for the higher value inductance, or can I get away with a lower inductance in reality. I will test this in the lab, but some assistance would be great, if you could explain what's going on there that would be really helpful. A little more information about how to handle discontinuous currents when designing with your devices would be extremely reassuring :)

Failing this and as long as I limit the inrush current, do you foresee any issue if I were to put a massive capacitance at the output of the linear regulator?

Many thanks,