High Power Mosfet Output Stage, More Power to your Track Options

[GregR]

Hey Rich, great work.

Am I to assume from the quick scematic that the intention is to drive the track direct from the daughterboard, hence eliminating the thin copper tracks on the existing PSU base?

I was a bit concerned as to whether the tracks on the PB after the output stage would be capable of handling the continuous power we will now be capable of drawing. Is this an issue at all? If so, is there any requirement to provide the 12VDC signal direct onto the daughterboard as well?




Greg

[RichG]

Hi Greg

Yes it will be necessary to provide a direct feed from the PSU to the daughter board. P1 & P2 will just be used to power the rest of the circuitry in the Powerbase.

The output stage will directly power the start track through the existing tracks, but it will be of benefit to take additional wires to the track directly from the output stage. A lot depends on how much current people want to take.

At 5 "real" Amps it will probably be ok, but at 10 or 20 Amps additional / better wiring will be necessary.

Rich

[pault86]

At 5 "real" Amps it will probably be ok, but at 10 or 20 Amps additional / better wiring will be necessary.

Rich

Woah! Rich are you hinting the new componenets you're testing could come anywhere close to providing 10 or 20 amps to the track?

Your previous charts were more modest.

For those - like me - trying to grok what's going on here the Wiki on MOSFETs helped a bit. Any other such links much appreciated.

Paul
Circuit TrustChrist

[RichG]

Good Morning power freaks, it's been a busy weekend chez RichG, and as ever nothing is ever simple.

During testing and making measurements it became clear that some of the numbers did not make sense. On investigation it transpired that with the increased currents I was testing my measurement setup as much as the mosfet output stages.

So I rebuilt the test setup, bigger wires everywhere, bigger diodes in the bridge feeding the electronic load, a separate bridge for the measurement system so that it is unnafected by the increasing vollt drop.

As an aside do not underestimate the amount of work you will need to put into the track continuity and power distibution if you don't want to throw away all of the good work that an improved Power Stage will bring

Of course this meant that all the numbers changed, so I had to go back and make them all again.




So what have we here? First column is a standard 1.5 Powerbase, numbers similar to before as the currents are relatively low. Next column is with L1 & L2, the output chokes removed, not that much difference with a standard PB, but they become significant as the currents rise, all subsequent measurements are with them out of circuit.

As before the 1.5 Powerbase Volts drop off very fast, at 3 Amps continuous the mosfets begin to smoke after 30 seconds and if left on would destroy themselves. A 10% drop comes at only 1 Amp, at a push you could take 2 Amps continuous with 3.5 Amps for short peaks.

Next column is the VNP20N07 Autoprotected Mosfets, these are driven by the HIP4081 driver chip. Performance is much improved, with a 10% drop in voltage at 9 Amps. With heatsinks this would be good for 8 Amps continuous and peaks to 15 Amps.

So final column is some Half bridges, 2 needed, intended for the Automotive industry windows, seats etc. These integrate the output stage and the drive electronics. They also have the benefit of being very well protected against short circuit, overtemperature etc. I won't say they are indestructable, because I am sure there will be a way.

So as you can see results are good. 10% drop in output voltage does not come until 18 Amps. I stopped the test at this point as was at the limit of my electronic load and despite my PSU being rated at 30Amps it was begining to make strange noises.

Anyway with a reasonable heatsink, yet to be worked out, these should be good for 15 Amps continuous with peaks to 25 Amps. Probably more, but let's be conservative for the moment.

For lower currents, say 5 Amps we can probably get away with no heatsink and 10 Amps with a small one.

The problem. is that none of the usual distributers in the UK stock them. However I am on the case and hope to come up with a solution.

So a long way to go, I need more testing, and to work out the best way to retrofit them into the Powerbase, however real and consistant power is on it's way.



Rich

[shaunbmx]

Not much to say appart from keep it up.... I am loving this.

Shaunbmx

[SinclairZX81]

OK, this is getting exciting! I will put a hold on my thoughts of defecting to the NINCO camp

I had to do some googling about this BTS7960 of course, and I found a company that will hand out FREE SAMPLES of the BTS7960B. Not sure if this is the same device - Rich? Just scroll to the bottom for the samples link.

http://www.arrowne.com/innov/in82/f_439.shtml

Maybe if we all sent off for samples and sent them off to Rich...?

[zipp]

Those figures look really impressive Rich.
If this mod is eventually simple enough to produce, it will mean i can discard the mss mod and all the associated wires/cables and pb's.
A few less items to trip over. That works for me.

[RikoRocket]

Zipp,
All my hard work for nothing?
I think currently Rich is going for the internal option, which is the best solution. But with MS you can add 5 slaves and have far more Amps on the track. Just keep adding....
Riko

Thanks Sinclair.... Sent off for mine

[RichG]

Yes thanks Sinclair, I have asked for some samples too. Arrow are going to wonder what's going on. They were on my list of people to contact, but I don't hold out too much hope as an order for 500 would normally be needed. Have asked Farnell if they will stock them, am waiting for a reply.


Rich

[gora]

Hi Rich,
maybe bad news for the BTS 7960...
The halfbridge is marked as 'not for new design ' in the productionlist from infineon.

Ralf

Sorry, have forgotten the line with the alternative type...

[RichG]

Hi Ralf

Not a problem it's been replaced by the BTN7960, which is virtually identical and has slightly better performance. Have passed this info to Farnell and asked if they will stock one of them. Either will be fine for us, as would the other members of the family.

Rich

[GregR]

This is sounding way promising now - if you have difficulties, let me know and I will chase some suppliers in OZ. Did a quick search and there looks like a couple of options but not sure on minimum purchase quantity.

I think I am well prepared for the track power distribution. Track is broken into roughly 2 - 3m sections and every joint is soldered within a section with a star power tap back to a main distribution from the PB's. I am also starting to trial copper tape as well.

This combined with PB Pro and Andy's latest S/W with pit lane sensing and speed control looks like finally delivering 2/3rds of what I want - reliable constant power and a workign pit lane. All I need now is a 3 lane start/finish straight that is 100% reliable.




Well done
Greg

[edropp]

Dear Richie,

I ahve aa question about your schemathic. I see that you use the two 0.1 ohm resistor. If you use that,
the PB will overload becuase with your new Power stage you are taking much more amps than a standard PB. Please explain.

Eduardo

[RichG]

Hi Eduardo

The two 0.1 Ohm resistors are the standard components in a Powerbase, and yes you are correct in a standard Powerbase it would overload, before additional current can be taken. I am assuming that this modification would always be used with the PB-Pro upgrade, this enables us to set the overload level in software.

If the Power Stage was fitted into a standard Powerbase then you would have to reduce the value of these resistors.

Rich

[RichG]

Bad news this morning. Account manager at Farnell has just go back to me. They will supply them but the minimum order quantity is 500. Now that is 250 units, which for me is too big a risk at £3.36 per device.

So I will go back to the people I got the samples from and see if they are happy to supply larger quantities. Trouble is they are in the USA and with larger quantities I will get hit with VAT & Duty even if they are happy to supply.

If anyone else can find these with more sensible MOQ's then please let me know.

Rich