# TVS Diode selection

I am designing a motor controller for a 24V 2.2KW d.c motor but I have a problem. I have been unable to get the 1.5KE39CA TVS Diode spcified in the design guide i am following. But i have two alternatives:

1. A Z2 033 TVS diode (say D1) found here and

2. A 1.5KE18CA TVS diode (say D2) found here

D1 is rated 33V (Breakdown voltage Vbr), but has Peak Pulse Power of 500W. D2 is rated 17.1V Vbr and has Peak Pulse Power of 1500W. (same power but different Vbr as the 1.5KE39CA)

Can i combine two D2's in series to get a total of 34.2V Vbr and 3000W Ppp (as described here) or should i use D1 the way it is (because this material claims the Ppp is not of utmost importance )? What is the implication of either method?

• why do you need TVS's for a 2kW motor-drive? – JonRB Apr 16 '15 at 21:31
• "Voltage spikes coming from motors are also problems in DC motor drives. The OSMC handles these with devices called Transient Voltage Suppressors (TVS). These devices can be considered “super Zener diodes”. They are optimized for handling high-current voltage spikes safely. They are connected in the OSMC to clip spikes across the battery leads and to protect the FETs from voltages exceeding their Drain to Source breakdown limit." - From the project documentation – Jimi Apr 16 '15 at 21:42
• and this TVS is to be fitted where? you generally do not need a TVS, you need a freewheel diode for sure, but not a TVS – JonRB Apr 16 '15 at 22:02
• Do you have an answer to the question? – Jimi Apr 16 '15 at 22:15
• I do (as in how to correctly take TVS datasheet figures for series stack) but I am still questioning where this TVS is to go? Actually finding the link (which you could have posted: robotpower.com/downloads/OSMC_project_documentation_V4_21.pdf ) goes a bit to explain why they think they need it. Correct decoupling would solve the issue negating the need for TVS – JonRB Apr 16 '15 at 22:26

Both solutions are not fully correct. First of all, look at the datasheet of 1.5KE series. It claims that 1.5KE39 has nominal breakdown voltage $V_{br}$ of 39V, but don't forget about accuracy and possible breakdown voltage in 37V to 42V range (7% accuracy). In practice breakdown voltage of TVS never equals to its nominal, I've checked that many times.
First diode (Z2033) has nominal $V_{br}$ of 33V (29.7 to 36.3V), so less than Your original one. The best choice from VRD series is Z2039 with 39V nominal breakdown voltage.
If you combine two 1.5KE18 (nominal $V_{br}=18V$) in series, theoretically you will have 36V breakdown voltage. But due to accuracy of every diode, the breakdown voltage will be somewhere between 34V and 38V - it's still too little. You should combine two 1.5KE20, then range of the breakdown voltage will be 39V to 42V.