High Voltage DC/DC Converter Suggestions?

VietJr

10 mW
Joined
Aug 7, 2018
Messages
26
Hello there!

Just wondering if you have any recommendations for high voltage DC/DC converters. Looking to step down ~400V to 12V if possible to power some smaller electronics for my electric motorcycle design. Power consumption shouldn't be too heavy, I highly doubt it'll exceed 200W, but most of the high-voltage DC/DC converters I've found are rated for like >500W, which is a little overkill for my uses.

Another option would be to make an auxiliary 12V battery pack to power all the small stuff, so if anyone has any experience with implementing that option, if you could leave a comment that'd be great!

Cheers,
 
Are we talking under or over 400VDC?

Most common switch-mode power supplies will accept a wide input range of something like 100-240VAC but many will also run on DC.

240V RMS = ~340 VDC, a supply may take a little more depending on its circuitry (e.g. capacitors may be 400V rated).

A safer option might be to run two smaller PSUs in series I,.e ~200VDC to the input of each. You'd need to check the outputs can be safely paralleled or seriesed if they are non-isolated supplies.
 
Punx0r said:
Are we talking under or over 400VDC?

Just under 400V, about 384 V.

Punx0r said:
Most common switch-mode power supplies will accept a wide input range of something like 100-240VAC but many will also run on DC.

240V RMS = ~340 VDC, a supply may take a little more depending on its circuitry (e.g. capacitors may be 400V rated).

A safer option might be to run two smaller PSUs in series I,.e ~200VDC to the input of each. You'd need to check the outputs can be safely paralleled or seriesed if they are non-isolated supplies.

Interesting idea, kind of like putting chargers in series/parallel then? I guess the idea is pretty similar just in reverse to supply low voltage from high voltage supplies instead of using multiple low voltage supplies to charge a high voltage pack.
 
Punx0r said:
Are we talking under or over 400VDC?

Most common switch-mode power supplies will accept a wide input range of something like 100-240VAC but many will also run on DC.

240V RMS = ~340 VDC, a supply may take a little more depending on its circuitry (e.g. capacitors may be 400V rated).

A safer option might be to run two smaller PSUs in series I,.e ~200VDC to the input of each. You'd need to check the outputs can be safely paralleled or seriesed if they are non-isolated supplies.

A standard switching power supply made for 240vac will be close. Use a fuse. Fuses for 400vdc are very expensive.

Two supplies fed in series would work as long as they both share evenly. If one decided to take more, it would be uneven, but as long as they are both close to handling the full voltage, I don't see why that wouldn't work. Use a fuse.

I can't think of anything cheap that is really made for that kind of voltage.
 
VietJr said:
Just wondering if you have any recommendations for high voltage DC/DC converters. Looking to step down ~400V to 12V if possible to power some smaller electronics for my electric motorcycle design. Power consumption shouldn't be too heavy, I highly doubt it'll exceed 200W, but most of the high-voltage DC/DC converters I've found are rated for like >500W, which is a little overkill for my uses.
The CEN-60 series by Meanwell will operate up to 416 volts DC in. It is ~60 watts, and the output voltage is adjustable.
 
Thanks for the help billvon!
 
You're building a 400v electric motorcycle? That's pretty ambitious! Do you have a link to the project?
 
Mandres said:
You're building a 400v electric motorcycle? That's pretty ambitious! Do you have a link to the project?

Sorry for the (very) late reply. Never got a notification that you had commented on this thread. If you're still interested I will link my team's website. At the moment I'm updating my team's website to include this project. The goal is to race it in New Jersey at the eMoto Varsity Challenge. So far I believe our current design is 96S 5P of FEP90135 ~30C Lipo cells, a Rinehart 100DXR controller, and an Emrax 228 LV, which is supposed to output 120kW peak for a minute or two with a custom winding.

We're in the process of programming some BMS chips and trying to think of the biggest issues of managing 96S cells. We also have a NovAtel GNSS unit which we're trying to get recording position, velocity and acceleration data so we can hopefully optimize our performance on the track!
 
Why do you need a BMS in the first place ? Do you need those cells managed actively on the track ? Do you believe they will get significantly out of balance before the pack is depleted ? You can't just use more intelligent charging (perhaps in a different series configuration) before putting the bike on the on the track ?

Basically the way I see it, there is a major difference between a consumer product and a specialty-built, operated by engineers, narrowly tailored product. As an engineer, you don't always want to solve a complex problem. Instead sometimes you may want to redefine the problem.
 
cricketo said:
Why do you need a BMS in the first place ? Do you need those cells managed actively on the track ? Do you believe they will get significantly out of balance before the pack is depleted ? You can't just use more intelligent charging (perhaps in a different series configuration) before putting the bike on the on the track ?

The cells we're planning on using probably won't need to be actively balanced or managed on the track. To my knowledge, Lipo cells are quite good at staying balanced if they're balanced initially. Our BMS will do more monitoring than managing. We'd like to record basic parameters like temperature, cell voltages, and current over the race to see if we're using these cells optimally, or if we should pull more/less power from them. The BMS active management will probably be limited to over-temperature, under-voltage, and over-current, and may just display an LED warning light, leaving it to the pilot's discretion if they can push the cells a little more or bring the bike back gently. If you think recording these values wouldn't be of much use, let me know!

cricketo said:
Basically the way I see it, there is a major difference between a consumer product and a specialty-built, operated by engineers, narrowly tailored product. As an engineer, you don't always want to solve a complex problem. Instead sometimes you may want to redefine the problem.

I agree! Initially our BMS idea had involved several features that would have increased the complexity a ton more than needed. The students on our team are pretty ambitious so we appreciate the feedback we've been given from ES to help us keep our goals more realistic. Thanks!
 
VietJr said:
If you think recording these values wouldn't be of much use, let me know!

It may or may not yield any actionable data. I'd say perhaps you could just setup the wiring and temp sensors so any kind of monitoring can easily be added after the fact, but don't let this turn into "bike-shedding" for you - get the bike on the track first, then think about micro optimizations later :)
 
cricketo said:
It may or may not yield any actionable data. I'd say perhaps you could just setup the wiring and temp sensors so any kind of monitoring can easily be added after the fact, but don't let this turn into "bike-shedding" for you - get the bike on the track first, then think about micro optimizations later :)

Will do! We're mostly just establishing our goal priority for what needs to be on the bike versus what would be pretty cool/interesting to have on the bike. So micro-optimization is definitely a stretch goal for us. Thanks!
 
Using a AC PSU that also accepts DC is a good suggestion.
You should take a look at these ICs for cell monitoring: https://www.analog.com/en/parametricsearch/11532 if you haven't already found something
 
BotoXbz said:
Using a AC PSU that also accepts DC is a good suggestion.
You should take a look at these ICs for cell monitoring: https://www.analog.com/en/parametricsearch/11532 if you haven't already found something

Thanks! We are looking at a few of these right now.
 
Back
Top