SlotForum banner
1 - 20 of 44 Posts

·
Premium Member
Joined
·
4,882 Posts
Have been meaning to try this for a while. However first a word of warning. Have only modded one of these, seems to work, but there could be problems yet to find. This is not an easy mod to do physically.

This mod can be viewed either as a way of repairing a failed Saloon board or as an upgrade to the output stage. I suspect most failed boards have one or both of the output mosfets failed.

The device I have chosen to fit as a replacement is the one used by Scalextric in the early Powerbase's, the ZXMC3A16D. If two of them were nearly good enough to power 6 cars, one should be well over the top for one.


The output mosfets are the small black devices marked N02 & P03. First these need to be removed. Apply a soldering iron to the two connections on one side & when the solder melts flip up the device with a screwdriver, now desolder the final connection and remove the device.

To fit the device you need to slightly splay pins 1 to 4 in an outward direction to make them fit the four pads marked in red in a row on the picture. Be very careful not to bend the leads too close to the body of the IC and only tweak them just enough to reach the edges of the pads. The 4 Drain connections, pins 5 to 8 are then connected to a piece of tinned copper wire and the ends bent and soldered to the two remaining connections.



So thats it.


Have never been able to find a spec for the P03 & N02 devices. I assume they are mosfets, but have seen Beppe post that they are transistors?

However here is a comparison between this mosfet and the DF2C03 used by Slot.it in their new decoder.



As you can see it compares favourably in all respects.


The disappointing part is that in testing, I modified a Takara module for ease of swapping, the performance seems identical to the original. This is confirmed by on track performance and by measuring the voltage at the motor.

I therefore expect the gain to be with beefier motors and hopefully that it will not expire.

Ideally need a few people to try this and report back. Very little to loose if you have some failed Saloon boards. This will of course also work with the F1 board.

Farnell part nos for the ZXMC3A16D is 756-4953

Rich
 

·
Premium Member
Joined
·
4,882 Posts
Hi Bryan

Thanks for the info on P03 / PDTA114ET

Have had a look at the data sheet and am sceptical as to weather it is the correct device.

P03 is used to short the motor for Braking and N02 to feed the PWM signal to the motor.

I am concerned about the rating which is only 100mA & 150mW.
Also seems strange that the N02 transistor is not listed as they are likely to be a complimentary pair of either transistors or mosfets.

I think I found this data or similar a year ago when trying to track down P03 & N02 however my memory is failing me. Looking at my scribbles on the sheet that I roughed out the circuit diagram I think I came to the conclusion that the X reference was wrong and that the transistor is an A03. However as I say it is scribble.


Also seems odd to me that they would not use a mosfet in this application. Equally odd that I am getting identical performance after grafting in a mosfet.

Beppe if you are there, where did you get the info that these are transistors?

D. Not sure if this is what you mean but here are the Scalextric instructions for fitting a saloon ship to a dummy Takara module. As you have seen I have not followed it for the LED but have done that the same way as the ready made digital modules to put the LED further foreword.
 

Attachments

·
Premium Member
Joined
·
4,882 Posts
Discussion Starter · #10 ·
OK I stand corrected twas Mauro.


QUOTE (Xlot @ 9 May 2006, 13:38) <{POST_SNAPBACK}>If I understood Mauro correctly (something not to be taken for granted
), the performance of the Slot.it decoder should be - surprise, surprise - noticeably superior to that of the SSD original

This because the output stage uses low resistance MOSFETs instead of transistors, effectively eliminating the good 2 Ohm choke now present

Rich
 

·
Premium Member
Joined
·
4,882 Posts
Discussion Starter · #14 ·
Hi BT

I have scribbled out the circuit.

A8E are the driver transistors, one for each mosfet. I am going to assume they are mosfets until proved otherwise because I do not think it would make sense to use anything else.

The N02 device is in series with the motor & driven via an A8E driver with the PWM signal from the processor. The P03 device is across the motor & is driven again via an A8E driver with the brake signal.

All pretty simple, was just curious as to the spec of the original devices, but not important.

Rich
 

·
Premium Member
Joined
·
4,882 Posts
QUOTE (NASCARUSA @ 21 Jul 2006, 10:55) <{POST_SNAPBACK}>Although I do not know a thing about electronics,I have found the NASCARs to run well on the 30,000 rpm light blue scalextric motors...........except when using the braking buttons.The lap times increase with the dynamic brake,but the chip will fry very fast.I was going to mail you some chips once before Rich,never did.If you let me know where to mail them,I will send you a dozen or so.I am on vacation this week,so I will be sure to get them to the post office this time. Sounds like your on to a high performance upgrade!

Greg

Well the chips arrived from Greg this morning. If you think you have problems with fried chips, look at this.




So I thought I would share my findings as I go along.

Have only visually examined the chips at the moment, there are 3 obvious problems.

1) Fried mosfet. Always seems to be the N02 device, not suprising as this is the one that drives the motor. The P03 device is only used for braking.



This is usually visible as a bump on the surface of the chip, although some are more dramatic.

2) Failed Diode. These are the four larger block devices marked FS1A



Here you see a quite dramatic failure.

3) Desoldered Diode. This one was a surprise to me. What has happened here is that the Diode has become sufficiently hot that it has desoldered itself from the board.



This also seems to be the most common failure. Question for Greg, in the pack there were a lot of loose diodes, just want to check that these desoldered themselves, as opposed to you removed them?

So here is the final tally of the 32 failed boards.



Of course this is only a visual inspection, some boards may have multiple faults, it is also likely that a failing mosfet could cause the Diodes to overheat and desolder. Also the the boards without a visible failure will have something wrong with them.

However one thing is obvious. If we are going to get a beefier board, as well as the upgraded mosfet we are going to need uprated diodes and or a higher melting point solder to keep them in place.

Final note on the diodes. They are an FS1A rated at 1 Amp continuous and 30 Amps peak forward surge.
This should be ok for most cars. Greg do you know how many amps your motors take?

I note that the Slot.it board uses a device marked RA 63A but can find no data on this. Can anyone else help?

To be continued....

Rich
 

·
Premium Member
Joined
·
4,882 Posts
Hi Greg

Thanks for the extra info. Do you know how many amps the 35,000 rpm, red can motors take? Also does anyone know where I get one in the UK to do some tests with?

Have not investigated the failures any further, however have made on discovery.

Seeing that many of the bridge rectifier diodes had expired or desoldered themselves I dug out the spec for the FS1A.

I had assumed it would be a Schottky diode, however it turns out to be an ordinary silicon diode with a rating of 1 Amp, but more importantly a forward volt drop at 1 Amp of 1.3V. So this is not only losing us precious volts, but it is this power dissipation which is desoldering them.

So am now trying to source a suitable Schottky diode which should reduce the volt drop to a max of 0.5 Volts, this will also significantly reduce the power dissipation. A win
win.


Am not sure if I can find a suitable diode in the same package so solution may be to have the bridge remote from the board. This will mean that the wires from the guide go to the bridge and then to the board with it's bridge removed.

Another possibility is to resite the smoothing capacitor with the separate bridge, enabling the end centimeter of the board to be chopped off.

If it works this should be a big gain.
 

·
Premium Member
Joined
·
4,882 Posts
Hi Greg

Ok on the 2 Amps, no wonder the FS1A are not happy.
Will look for some 3 Amp plus diodes, these will not be the same size, so will have to be remote from the board.

If you could send me a motor that would be great, saves me having to look into and understand motors ( not my specialist subject) I assume these will just drop into something, ideally my slot.it Audi R8C or one of the Scaley Boxters?

When I have something that seems to work I will mend & mod four of the duff boards and send them back for test / destruction.


Sounds like you need some of Pete's throttles, my discard pile is growing all the time, and the ones that are ok are only just. My best throttle yesterday got a weird one when you press the LC button the car moves forward, still operating the LC's. Would not usually matter except that I use a 4 second push and hold on the LC with zero throttle to reset the display, which does not then work.


Rich
 

·
Premium Member
Joined
·
4,882 Posts
OK Guys you are going to love this.


Remember the power the lane changers separately mod,
remember the sort the throttles out mod,
well this mod is in the same league.


So a bit of a bodge but the only Schottky diodes I had in stock were wire ended SB130. Still 1 Amp but more importantly the forward volt drop at 1 Amp only 0.4 Volt.




So here they are soldered in in place of the FS1A, which remember has a 1.3 volt drop.


I have been doing my testing in a JGTC car. The modules make it very easy to change chips.
Had been running at 15 Volts. Swapped the chip and the car was uncontrollable.
Turned the volts down to get the same performance and ended up at 13.8 Volts. So only a quick test, but this is a big performance gain and has the added bonus of wasting less power.
Let's just hope the FIA don't ban Schottky diodes.


So not the ideal device. I will now try and find a better one, but has proven the point.

Scalextric please change to Schottky diodes asap, you will significantly improve performance and reduce failures.


Rich
 

·
Premium Member
Joined
·
4,882 Posts
Discussion Starter · #27 ·
Thanks Greg.

Look forward to receiving the car. Have just had a supplemental thought that we might be better off with two diodes in parallel on each leg of the bridge.

Trouble with diodes even schottky ones is that as the current goes up the volt drop goes up as well. Also the higher the rating of the diode in amps the higher the volt drop. So 2 in parallel could give the best of both worlds. Low volt drop at low and high current.
Just a thought.
 

·
Premium Member
Joined
·
4,882 Posts
Morning Guys

Just a few thoughts and snippets of information in no particular order.

The two mods can either be done separately or together.

The one with the greatest benefit is the Diode mod, as this makes the board able to drive beefier engines, improves the max volts you can get, and reduces power dissipation in the diodes.

The mosfet mod did not seem to improve performance, but should improve reliability. Remember we do not have a spec for the N02 & P03 devices so do not know how the ZXMC3A16 compares.

However bear in mind that the diode mod will put a greater load on the output mosfet and it may then fail.
However you have nothing to loose, so probably do the diode mod first, run the car, then if the mosfet fails change it.

Have not checked the plate of chips yet but visual examination only showed the N02 motor drive mosfet failing. An alternative mod may be to just change that one.

So have had a look around for alternative mosfets and diodes, this is what I have found and ordered for test.



So the first one is the device used in the early PB's and that I have put in the first modded board.
Second one is the device Slot.it is using, slightly lower spec, but cheaper.

Third one is the best dual mosfet I can find, slightly better but more expensive.

Finally I have looked for direct replacements in SOT23 packages

The PMV31XN looks a good bet for the N02 device and a nice price at 25P


Am not sure if the P03 device (motor brake) fails but the IRLML26402 is the best I can find at the moment.

Details on diodes to follow when I write it up.

Rich
 

·
Premium Member
Joined
·
4,882 Posts
So here's the best Schottky diodes I can find at the moment. Open to any other suggestions.



The FS1A is the original silicon diode with it's massive 1.3V drop, and attendant power dissipation.

Next the SB130 that I happened to have. Wire ended but much improved forward volt drop.
The SMA device is the same device size as the original, 3 Amp rating and a low forward volt drop. Probably the pick of the bunch,
but not tested yet.
The SMB device is larger, has a lower volt drop and is cheap. Not quite sure how you would fit it.

So thats it for now. Sax will be very interested to see how you get on.

Greg I have high hopes that the two mods combined will cope with the 35K motors. However I do not think the PB will take more than 3 cars at a time. However remember you should get similar performance at a volt less than you used to. So at 14V you might just squeeze 4 cars.

Have just received a Scalextric Light Blue can 30K motor, so will be able to see how the mod copes with that.

QUOTE (NASCARUSA @ 21 Jul 2006, 10:55) <{POST_SNAPBACK}>I have found the NASCARs to run well on the 30,000 rpm light blue scalextric motors...........except when using the braking buttons.The lap times increase with the dynamic brake,but the chip will fry very fast.
Greg can you expand on this, are you saying that using the Braking button fry's chips faster?

To be continiued.....

Rich
 

·
Premium Member
Joined
·
4,882 Posts
Started making a few measurements, many more to do, but these are interesting.

Standard Scalextric Boxter.

12V to the Powerbase
Max throttle at motor 10.1 Volts.
Temperature of diodes at start of test 30 deg C
Temperature of diodes max throttle wheels in the air settles at 40 deg C.
Temperature of diodes max throttle virtually stalled motor 100 deg C

Fitted 4 off the 15MQ040 Diodes

12V to the Powerbase
Max throttle at motor 11.1 Volts.

Temperature of diodes at start of test 30 deg C
Temperature of diodes max throttle wheels in the air settles at 36 deg C.
Temperature of diodes max throttle virtually stalled motor 80 deg C

So far so good.


Just as an aside if Digital ever catches on for club racing.
It's going to be really hard to scrutineer the electronics.


Rich
 

·
Premium Member
Joined
·
4,882 Posts
Beppe

Slot.it board uses a device marked RA 63A but can find no data on it. Would be nice to know the spec if anyone can find it.

Yes maximum throttle , car held with its wheels in the air.

Virtually stalled is wheels rubbing on the track and "virtually stalling" the motor. The temperature of the diodes rises very fast when you do this.


Not very scientific, but you have got to start somewhere.

Rich
 

·
Premium Member
Joined
·
4,882 Posts
A few more measurements, same tests as above except this time with the Scalextric Light Blue can 30K motor fitted.

Standard Scalextric Boxter with 30K motor.

12V to the Powerbase
Max throttle at motor 9.55 Volts.
Temperature of diodes at start of test 30 deg C
Temperature of diodes max throttle wheels in the air settles at 60 deg C.
Temperature of diodes max throttle virtually stalled motor 107 deg C

My temp meter at max and at that point a puff of smoke from the mosfet. However board survived.


Fitted 4 off the 30BQ030 Diodes

These are the larger SMB devices. They fill the available space but were actually easier to fit than the SMA type.

12V to the Powerbase
Max throttle at motor 10.4 Volts.
Temperature of diodes at start of test 30 deg C
Temperature of diodes max throttle wheels in the air settles at 43 deg C.
Temperature of diodes max throttle virtually stalled motor 65 deg C.

A puff of smoke from the mosfet on this board, so stoped the test, but again it survived to tell the tale. Looks like they have two lives. I suspect if I had kept the motor nearly stalled for much longer the mosfet would have blown.

On the standard board I think it would have been a close run thing between the mosfet blowing and a diode desoldering. Perhaps they were designed to be virtual fuses.


So the diodes are doing well, next thing is to start swapping mosfets.

Rich
 

·
Premium Member
Joined
·
4,882 Posts
No more tests yet, but a thought for the morning, and something to look out for if trawling through spec sheets looking for mosfets.

Further work needed, but my gut feeling is one of the schottky diodes is going to be fine. Going through the boards I have even the silicon diodes have very few failures, but because they get so hot they desolder. The schotty diodes are running much cooler and will probably fix that, with the added bonus of more volts for the motor.

So that leaves the mosfet. Have yet to test the ones in the chart above, but I have come across is a version called a protected mosfet. It's just like an ordinary one but has the ability to shut down. It has overvoltage, current limiting, overtemperature and short circuit protection. In other words they are bomb proof.


I think this type of mosfet combined with the Schottky diodes is the full answer.


Lower volt drop.
Ability to drive 2 Amps plus.
But then shut down, not destruction if the device is pushed too hard for too long.

You then have to return to the pits to cool down, but then everything springs back into life.


Here's a link to the data sheet on a VNS3NV04D, which is a dual N channel device if you want to see the sort of thing I am talking about. They of course come in all sizes and ratings.

http://www.st.com/stonline/products/literature/ds/7396.pdf

Problem is I can only find them in N channel not P, and cannot find any complimentary versions with P & N in one package.

Will separately have a look see if the brake circuit can be reconfigured to work with an N channel device.

Still probably the way forward as the N channel device is the key motor drive one, however if anyone comes across a suitably rated P channel protected mosfet let us know.

To be continued....

Rich
 

·
Premium Member
Joined
·
4,882 Posts
Discussion Starter · #47 ·
Well I eventually have found time to get back to this. Did a few more tests and have come to the conclusion that the best value for money mod is.

ZXMC3A16D to replace the N02 & P03 output mosfets
20BQ3A16D to replace the diodes.

Look back earlier in the thread if you want to see the spec and prices.

The mosfet is tricky to fit in particular bending the legs.

The diodes are larger than the original ones but actually fit quite well.

I have repaired and modded 6 of the failed boards that Greg sent me, they will be on the post on Monday.
Will be interesting to see if they survive. I have only lightly tested them.



Here they are ready to go. In practice the only faults I found were desoldered diodes, failed diodes and failed output mosfets, usually N02. The boards that visually looked ok all had N02 failed.

Matt before you ask.
Your F1 board will be done over the weekend and will also be in the post on Monday.

Rich
 

·
Premium Member
Joined
·
4,882 Posts
Ok here is the modded F1 Board. had to use the 15MQ040N Schottky diodes as the ones used on the Saloon board were too big. Should not affect performance, they are just a bit more expensive.



The mosfet was slightly easier to fit as the pads were closer together and the legs needed less tweaking.

These are going off to Matt who will hopefully report back on how they perform.

Rich
 

·
Premium Member
Joined
·
4,882 Posts
Discussion Starter · #52 ·
It was, the board you sent was burnt through. I "borrowed"
a couple of Gregs failed ones from the big pile, so you have a spare. You need to thank Greg.

Rich
 

·
Premium Member
Joined
·
4,882 Posts
Discussion Starter · #55 ·
Hi Matt

Glad the arrived OK. I suggest you give these two a try and see how they perform relative to the standard chip before worrying about any more.

It is not viable to modify on an ongoing basis. The Saloon board components are £1.90 and the F1 £3.66 because the cheaper diodes do not fit. Each board takes at least half an hour to modify and test. It may be possible to shoehorn the cheaper diodes into the F1 Board.

So even with the Saloon board costs you would have to charge at least £5 to make it worthwhile. I do not think that people would be willing to pay a total of £15 for a board.

Anyway we are getting ahead of ourselves here. We need to know how they perform. If they do very well with higher RPM motors and give the performance improvment I expect then perhaps people will be willing to pay £15 for an uprated board?

Rich
 

·
Premium Member
Joined
·
4,882 Posts
Hi Greg

Glad the chips got there.

Sounds good so far, look forward to the results of further testing.


All the data on a comparison between these upgraded boards and the Slot.it ones is contained in this thread and the Slot.it one.

Read them both carefully, in particular there are some tables earlier in this thread, and look at Maurizio's comments in the Slot.it thread. Would be wrong of me to draw any conclusions as have not tested a Slot.it & my modded ones are not yet fully tested.

Rich
 
1 - 20 of 44 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top